give dihydrofuran 4. This is followed by elimination of acetic
acid to furnish the desired furan 5 in excellent yield.
The above method was discovered in the course of an
attempted synthesis of several 2-thiomethyl-5-alkoxy or
5-dialkylamino cyclic furans such as 17 or 24 (vide infra).
In planning an approach to these compounds, we relied on
earlier studies in our laboratory which showed that the acid-
catalyzed reaction of â-oxo ketene thioacetals 8 and 9 gave
furans 12 and 13 in 70% and 83% yields, respectively. This
reaction most probably proceeds by a protonation/cyclization/
elimination sequence as outlined in Scheme 2.
Scheme 4
Scheme 2
step leading to the formation of 19 involves an initial 1,3-
hydrogen shift of 15 to give 18 which is followed by
conjugate addition of the neighboring thiomethyl group onto
the activated π-bond. Once formed, episulfonium ion 19 can
undergo a subsequent ring opening in the opposite direction
to produce thionium ion 16 by deprotonation and double
bond isomerization. A related set of reactions also takes place
when 16 is subjected to the acidic conditions, eventually
giving thioketene acetal 15 as the minor product (20%) from
the equilibrating mixture.
The series of steps shown in Scheme 4, although credible
in retrospect, was totally unexpected. Our inability to prepare
furan 17 from the acid-catalyzed reaction of thioketene acetal
15 led us to consider some alternate ways to synthesize this
compound.14 It was known from earlier work in the literature
that treatment of thioketals with DMTSF causes the carbon-
sulfur bond to become labile upon methylthiolation.15 The
initially formed alkylthiosulfonium ion easily dissociates to
produce a thionium ion and methyl sulfide. We reasoned
that by converting the hydroxyl group present in lactone 14
into the corresponding acetate, it should be possible to
promote cyclization of the lactone carbonyl group onto the
resulting thionium ion formed from the DMTSF-induced
reaction. Once the dihydrofuran ring has been forged,
elimination of acetic acid should proceed readily to furnish
However, when this reaction sequence was applied to
thioketene acetal 15, prepared by dehydration of the aldol
condensation product 14, none of the desired furan 17 could
be detected. Instead, the only product isolated (80%) cor-
responded to the rearranged thio-conjugated lactone 16.
Subjection of a sample of 16 to the acidic conditions gave
thioketene acetal 15 (20%) in addition to recovered starting
material (80%). This same equilibrium mixture (i.e., 15:16
) 1:4) was observed in the acid-catalyzed reaction of 15
(Scheme 3).
Scheme 3
(10) Alvarez-Ibarra, C.; Quiroga, M. L.; Toledano, E. Tetrahedron 1996,
52, 4065.
(11) Annunziata, R.; Cinquini, M.; Cozzi, F.; Raimondi, L. Tetrahedron
Lett. 1989, 30, 5013.
(12) Padwa, A.; Rudiger, H.; Kappe, C. O.; Reger, T. S. J. Org. Chem.
1998, 63, 1144. Kappe, C. O.; Padwa, A. J. Org. Chem. 1996, 61, 6166.
Padwa, A.; Kappe, C. O.; Cochran, J. E.; Snyder, J. P. J. Org. Chem. 1997,
62, 2786.
(13) Trost, B. M.; Murayama, E. J. Am. Chem. Soc. 1981, 103, 6529.
Trost, B. M.; Sato, T. J. Am. Chem. Soc. 1985, 107, 719. Commercially
available from Aldrich Chemical Co.
(14) The difference in the acid-catalyzed behavior of 14 from that
encountered with ketene thioacetals 8 and 9 may possibly be related to the
lower concentration of the enol tautomer of the lactone, thereby permitting
thiomethyl group migration to compete with cyclization. Further studies
with other lactones are currently underway to help clarify this issue.
(15) Rice, J. L.; Favstritsky, N. A. J. Am. Chem. Soc. 1969, 91, 1751.
Smallcombe, S. H.; Caserio, M. C. J. Am. Chem. Soc. 1971, 93, 5826. Kim,
J. K.; Pau, J. K.; Caserio, M. C. J. Org. Chem. 1979, 44, 1544.
The interconversion of 15 and 16 is believed to occur via
the transient episulfonium ion 19 (Scheme 4). The critical
1560
Org. Lett., Vol. 1, No. 10, 1999