To obtain a preliminary indication of the cumulative effects
of the cis-acetate and the methyl substituent on the diene,
the hydroxyl groups in 13 were reacted simultaneously with
chloromethyl methyl ether in the presence of diethylisopro-
pylamine to provide the bis-methoxymethyloxy ether 14. The
silyl group was removed with tetrabutylammonium fluoride
to afford the allylic alcohol 16. Oxidation of this secondary
alcohol 16 with Dess-Martin periodinane in dichloro-
methane generated decatriene ketone 17 in situ, which
cyclized spontaneously to 23 at room temperature (21 °C)
over 60 h or more efficiently in 4 h at 40 °C during the
oxidation reaction (87%) (Scheme 3). The mild conditions
Table 1. Solvent Effects: Cycloaddition Diastereoselectivitiy
Ratios
Scheme 3. Cycloadditions of Substituted Dienes
However, the most intriguing result is reflected in the data
for entry 5, in which the ratio of 24/27 increased to 23:1
with acetonitrile as the solvent at 40 °C. In comparison, entry
2 in chloroform under similar conditions afforded a 5.3:1
ratio. Generally, the rates and stereoselectivities of Diels-
Alder reactions are immune to significant solvent effects.12
Experiments conducted in water, in which substantial rate
enhancements are observed, are an exception.13 Normally
the diastereofacial selectivity is not affected in aqueous
conditions unless the components are partially water-
soluble.14 In these examples the adduct ratios are influenced
by the hydrogen-bond donating ability of the solvent.
It seems likely that this “acetonitrile effect” is a conse-
quence of the preferred association of the solvent dipole with
the oxygen atoms on the top face of the molecule. This
interaction may resemble the type of charge-transfer and
π-π interactions encountered in aromatic systems.
If hydrogen bonding is important, then a free hydroxyl
on the underside of the molecule should reduce the diaster-
eomeric ratio. To examine this possibility and provide an
additional, differentially functionalized adduct, alcohol 15
was treated with tetrabutylammonium fluoride and oxidized
with manganese dioxide (Scheme 2). The expected ketone
20 was not isolated but rather hemiacetal 19 was formed
preferentially. In refluxing chloroform the lactol and keto
forms are in equilibrium and the cis diastereomer 25 was
generated in a ratio of 5:1. This value increased to 11:1 in
for this cycloaddition of a sterically encumbered system are
particularly expedient.
For natural product syntheses, differentiation of the
hydroxyl groups is required. Consequently iododiene 1210
was prepared from 10 via 11, and the derived lithium salt
condensed with 9 to afford alcohol 15 in which the
R-hydroxy diastereomer predominated.11
Protection of the secondary alcohol as its acetate, removal
of the silyl ether, and Dess-Martin periodinane oxidation
afforded the highly substituted cycloaddition precursor 18.
In this instance the ketone could be isolated, but as
summarized in Table 1 (entry 1) at room temperature (21
°C) it cyclized slowly over 2 weeks to the decalin. Treatment
of the dienes with Lewis acid did not facilitate cycloaddition
but led to decomposition. As expected, higher temperatures
reduced the level of selectivity.
(12) (a) Sauer, J.; Sustmann, R. Angew. Chem., Int. Ed. Engl. 1980, 19,
779. (b) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry;
VCH: Cambridge, 1990.
(13) (a) Rideout, D.; C.; Breslow, R. J. Am. Chem. Soc. 1980, 103, 7817.
(b) Engberts, J. B. F. N. Pure Appl. Chem. 1995, 67, 823. (c) Meijer, A.;
Sijbreu, O.; Engberts, J. B. F. N. J. Org. Chem. 1998, 63, 8989.
(14) Lubineau, A.; Auge´, J.; Lubin, N. J. Chem. Soc., Perkin Trans. 1
1990, 3011. (b) Cativiela, C.; Garcia, J. I.; Mayoral, J. A.; Royo, A. J.;
Salvatella, L. J. Phys. Org. Chem. 1992, 5, 230.
(10) Wong, T.; Tjepkema, M. W.; Audrain, H.; Wilson, P. D.; Fallis, A.
G. Tetrahedron Lett. 1996, 37, 755.
(11) Tjepkema, M. W.; Wilson, P. D.; Audrain, H.; Fallis, A. G. Can. J.
Chem. 1997, 75, 1543.
Org. Lett., Vol. 2, No. 18, 2000
2795