Scheme 1. Azomethine Ylid Strategy for the Synthesis of
Scheme 3. Dithioformates as 1,3-Dipolarophilesa
Penams and Penems
a Thermal vs microwave conditions studied: (a) (thermal) 8a,
MeCN, 2 days, 80 °C (19%); (b) (microwave) screw cap pressure
vessel, 8a, PhMe, 55 W, 1 h (76%) or 8a, emimPF6 (10 mol %),
55 W, PhMe, 1 h (56%); (c) (microwave) open vessel, 8a, PhMe,
200 W, 5 h (62%) or 8a, emimPF6 (10 mol %), 55 W, PhMe, 4 h
(40%); (d) (microwave) open vessel, 8b, PhMe, 200 W, 4 h, (45%);
(e) (microwave) screw cap pressure vessel, 8c, PhMe, emimPF6
(10 mol %), 55 W, 1 h, (45%).
substituted penems 6 (R1 ) alkyl, aryl, S-alkyl) (Scheme
1).5 The extension of the azomethine ylid strategy to the
synthesis of the C(2)-unsubstituted penem moiety (i.e., 6,
R1 ) H) that is associated with 1 and 2 is the focus of this
paper.
Achieving this objective required access to dithioformates
7 and an evaluation of the ability of these units to function
as effective 1,3-dipolarophiles. Only a very limited range of
dithioformates have been reported to date,6,7 and various
approaches to S-alkyl dithioformates 7 were evaluated
(Scheme 2).
7b and 7c were best obtained by direct thionation of the
corresponding thioformates 99 and 109 using Lawesson’s
reagent.10 In both cases, the target dithioformates 7b and 7c
were not isolated but were trapped in situ with cyclopenta-
diene to give cycloadducts 8b and 8c, respectively, in
moderate yields for this two-step sequence.11,12
Scheme 2. Generation and Trapping of Dithioformates
Cycloadducts 8a-c were especially attractive for our
purposes, representing a potentially controlled supply of the
requisite dithioformate (via 4 + 2 cycloreversion); the retro
Diels-Alder reaction provides an in situ source of dipo-
larophile that is compatible with release of the key azome-
thine ylid intermediate 4 from oxazolidinone 3.13
This strategy was validated, and thermolysis of 8a in the
presence of oxazolidinone 3 provided the racemic cycload-
duct 11a as a 2.5:1 mixture of exo and endo isomers (Scheme
3). The structure of exo-11a was confirmed by X-ray
crystallography (see Supporting Information).
However, this thermal process (conditions a) did require
2 days to go to completion and this only achieved a very
S-Methyl dithioformate 7a is available by reduction of CS2
with LiBEt3H followed by S-methylation.7 We found it most
convenient (see below) to trap 7a with cyclopentadiene to
give the corresponding cycloadduct 8a as a 1.5:1 mixture of
exo and endo isomers in 35% yield.8 The S-benzyl variants
(9) Bax, P. C.; Holsboer, D. H.; Van der Veek, A. P. M. Recl. TraV.
Chim. Pays-Bas 1971, 90, 562-567.
(10) S-Benzyl dithioformate (PhCH2SC(S)H) could not be prepared using
the reduction/S-alkylation strategy associated with 7a but was obtained by
direct thionation of the corresponding thioformate (PhCH2SC(O)H) using
Lawesson’s reagent. NMR (CDCl3) for PhCH2SC(S)H: δH 11.28; δC 216.8).
However, we were unable to isolate this product in an efficient manner,
but this served as a model for the preparation of 7b and 7c.
(11) 8b and 8c were obtained as a 0.6:1 and 1.7:1 mixture of exo and
endo isomers, respectively.
(5) (a) Planchenault, D.; Wisedale, R.; Gallagher, T.; Hales, N. J. J. Org.
Chem. 1997, 62, 3438. (b) Highly reactive thioaldehydes have also been
generated and trapped in situ: Brown, G. A.; Anderson, K. M.; Large, J.
M.; Planchenault, D.; Urban, D.; Hales, N. J.; Gallagher, T. J. Chem. Soc.,
Perkin Trans. 1 2001, 1897-1900.
(6) Block, E.; Aslam, M. Tetrahedron Lett. 1985, 26, 2259-2262.
(7) (a) Seyferth, D.; Womack, G. B. Organometallics 1984, 3, 1891-
1897. (b) Gandhi, T.; Nethaji, M.; Jagirdar, B. R. Inorg. Chem. 2003, 42,
4798-4800. Jagirdar et al.7b were able to isolate 7a by distillation. Block6
and Seyferth7a did not isolate this volatile component but trapped it in situ
as a Diels Alder cycloadduct and as an Fe-based coordination complex,
respectively.
(8) Similar exo and endo cycloadducts based on thioaldehydes have been
characterized previously. Kirby, G. W.; Lochead, A. W. J. Chem. Soc.,
Chem. Commun. 1983, 1325-1327. Vedejs, E.; Stults, J. S.; Wilde, R. G.
J. Am. Chem. Soc. 1988, 110, 5452-5460.
(12) One of the issues contributing to the yields obtained for 8b/c was
purification of the desired product from the residues associated with the
thionation step. Ley (Ley, S. V.; Leach, A. G.; Storer, R. I. J. Chem. Soc.,
Perkin Trans. 1 2001, 358-361) has reported a solid-phase variant of
Lawesson’s reagent. In our hands, this reagent worked well for the thionation
of amides, but we were unsuccessful in our attempts to thionate S-benzyl
thioformate (PhCH2SC(O)H); see ref 10.
(13) Mechanistic studies provide evidence that oxazolidinone 3 is in
equilibrium with the carboxylated azomethine ylid 4. This pathway provides
an equilibrium concentration of 4, and we have exploited this to trap highly
reactive and short-lived 1,3-dipolarophiles.5b
2782
Org. Lett., Vol. 6, No. 16, 2004