C O M M U N I C A T I O N S
excess of aluminum chloride. In the case of the Diels-Alder
products derived from 2a and 2b, very effective ring expansion to
the [4 + 3] cycloadducts was observed. The reaction of aluminum
chloride with the exo and endo isomers 9a and 9b derived from 2d
was especially interesting as the major exo isomer 9a underwent
an essentially quantitative rearrangement to the [4 + 3] cycloadduct
3d (eq 4), while the minor endo isomer 9b preferentially underwent
a retro-Diels-Alder reaction (eq 5).6 The retro-Diels-Alder reaction
was a major side reaction for the Diels-Alder cycloadducts derived
from 2c and 2e, which may explain why the yields for the formation
of 3c and 3e (Table 1) are relatively low.
Due to the fact that a stoichiometric amount of Lewis acid is
required to efficiently induce the formation of the formal [4 + 3]
cycloadducts, the use of chiral Lewis acids for the direct transfor-
mation would not be practical. The asymmetric synthesis of the
formal [4 + 3] cycloadducts, however, can be readily achieved in
a two-step process beginning with a chiral Lewis acid-catalyzed
asymmetric Diels-Alder reaction followed by aluminum chloride
induced rearrangement (eq 8). For example, using Faller’s ruthenium-
based chiral Lewis acid,2c the Diels-Alder product 13 was obtained
in 85% ee. Aluminum chloride induced rearrangement of 13
generated 14 in 74% yield with retention of the enantioselectivity.
This is consistent with the formation of 14 from the direct
rearrangement of 13 rather than a retro-Diels-Alder followed by
a [4 + 3] cycloaddition, where loss of enantioselectivity would be
expected.
On the basis of these observations, it became of interest to
reevaluate the recently published studies by Harmata on the
scandium triflate-catalyzed reaction of the 2-siloxyacrolein 10 with
cyclopentadiene, which results in the formation of [4 + 3]
cycloadducts 11a,b (eq 6).3a Although the mechanism of this
reaction has not been fully elucidated, the working hypothesis has
been a stepwise mechanism involving a zwitterionic intermediate,
which preferentially forms the [4 + 3] cycloadduct rather than the
[4 + 2] cycloadduct. A recent computational study on the reaction
of 2-siloxyacrolein with furan supported a stepwise mechanism.7
An alternative mechanistic possibility would be the tandem Diels-
Alder reaction/ring expansion.
In conclusion, we have discovered a remarkably simple method
to achieve a formal [4 + 3] cycloaddition between cyclopentadiene
and R,â-unsaturated aldehydes, which proceeds by a tandem Diels-
Alder reaction/ring expansion. Further studies are in progress to
determine the full scope of this chemistry.
Acknowledgment. This work was supported by the National
Science Foundation (CHE-0350536). We thank Oksana O. Gerlits
for the X-ray crystallographic analysis and Dr. Jaemoon Yang for
helpful discussions.
Supporting Information Available: Full experimental data for the
compounds described in this paper (PDF). This material is available
References
(1) For recent reviews, see: (a) Corey, E. J. Angew. Chem., Int. Ed. 2002,
41, 1650. (b) Evans, D. A.; Johnson, J. S. In ComprehensiVe Asymmetric
Catalysis; Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer:
Berlin, 1999; Vol. 3, p 1177. (c) Corey, E. J.; Guzman-Perez, A. Angew.
Chem., Int. Ed. 1998, 37, 488. (d) Ishihara, K.; Yamamoto, H. Eur. J.
Org. Chem. 1999, 527, 7.
To test this possibility, attempts were made to isolate the Diels-
Alder cycloadduct from the reaction of 2-siloxyacrolein 10 with
cyclopentadiene. The scandium triflate-catalyzed reaction at -78
°C gave a mixture of 11a,b and other unidentified products. The
Diels-Alder product was eventually formed by a microwave-
induced cycloaddition between 10 and cyclopentadiene (eq 7), but
attempted purification of the product by conventional silica gel
chromatography led to its rearrangement to a mixture of bicyclo-
[3.2.1]octenones. Purification of the Diels-Alder product was
possible using silica gel deactivated by triethylamine. Furthermore,
the isolated exo isomer 12 was shown to readily undergo a scandium
triflate-catalyzed rearrangement under Harmata’s conditions3a to the
endo product 11a. Considering the facility of the rearrangement of
12 to 11a and of the known ketol rearrangement of related Diels-
Alder products,8 it is conceivable that the Harmata [4 + 3]
cycloaddition3a and related reactions3b-d are also examples of the
tandem Diels-Alder reaction/ring expansion. The scandium triflate-
catalyzed reaction of 2-methylacrolein with cyclopentadiene also
generates [4 + 3] cycloadducts, but the yield is low and the
diastereoselectivity is poor.
(2) For recent examples, see: (a) Sprott, K. T.; Corey, E. J. Org. Lett. 2003,
5, 2465-2467. (b) Corey E. J.; Shibata, T.; Lee, T. W. J. Am. Chem.
Soc. 2002, 124, 3808. (c) Faller, J. W.; Grimmond, B. J.; D’Alliessi, D.
G. J. Am. Chem. Soc. 2001, 123, 2525. (d) Ishihara, K.; Kurihara, H.;
Matsumoto, M.; Yamamoto, H. J. Am. Chem. Soc. 1998, 120, 6920.
(3) (a) Harmata, M.; Sharma, U. Org. Lett. 2000, 2, 2703. (b) Aungst, R. A.;
Funk, R. L. Org. Lett. 2001, 3, 3553. (c) Sasaki, T.; Ishibashi, Y.; Ohno,
M. Tetrahedron Lett. 1982, 23, 1693. (d) Blackburn, C.; Childs, R. F.;
Kennedy, R. A. Can. J. Chem. 1983, 61, 1981.
(4) Davies, H. M. L.; Xiang, B.; Kong, N.; Stafford, D. G. J. Am. Chem.
Soc. 2001, 123, 7461.
(5) NOe studies support the relative configurations of compounds 3a-e, 5a,
and 8a-c. The relative configuration of 3d was confirmed by the X-ray
analysis of its 2,4-dinitrophenylhydrozone derivative. The X-ray crystal-
lographic data have been submitted to the Cambridge Structural Database
[Gerlits, O. O.; Coppens, P. Private Communication (1078) 2003, CCDC
219152].
(6) For a review, see: Rickborn, B. Org. React. 1998, 52, 1.
(7) Saez, J. A.; Arno, M.; Domingo, L. R. Org. Lett. 2003, 5, 4117.
(8) Creary, X.; Inocencio, P. A.; Underiner, T. L.; Kostromin, R. J. Org.
Chem. 1985, 50, 1932.
JA039908Q
9
J. AM. CHEM. SOC. VOL. 126, NO. 9, 2004 2693