Table 1. Preparation of N-Substituted Sulfoximines
Figure 1.
dition, and of most other asymmetric reactions involving
chiral auxiliaries, is that access to the enantiomeric series of
products requires the assembly of precursors bearing the
enantiomeric partner of the chiral auxiliary.3 In our case this
would need a seven-step synthesis starting from com-
mercially available (-)-(Ss)-menthyl p-tolylsulfoxide. We
envisaged that it might be possible to overcome this
drawback by altering the priorities of the substituents at the
sulfur in a way that reversed the diastereofacial selectivity
of the cycloaddition. Although this tactic has few precedents,
it has been reported that bulky aluminum-based Lewis acids
are capable of changing the steric priority of the sulfinyl
oxygen and thereby induce an inversion of topicity in
sulfinyl-directed radical reactions.4 This type of Lewis acid
cannot be used in our system because of the basicity of the
pyrone, but we reasoned that transforming the sulfoxide
moiety of 1 into a sulfimide or a sulfoximine might be a
practical strategy to alter the sterochemical outcome of the
cycloaddition.
Since our initial attempts to prepare the N-tosylsulfimide
from the sulfoxide 1a gave low yields and seemed to be
flawed by partial racemization, we turned our attention to
the configurationally more robust sulfoximine derivatives.
Sulfoximines have received rather little attention as chiral
auxiliaries in asymmetric synthesis even though the presence
of a nitrogen substituent on the sulfur would seem to offer
interesting possibilities for modulating the steric and elec-
tronic characteristics of the chiral unit.5 Among the various
possible routes to prepare the required optically active
sulfoximines, reaction of the alkenylsulfoxide 1a with MSH
(0-mesitylsulfonylhydroxylamine)6 appeared to be the best
choice since the resulting “free” aminated product is
amenable to divergent N-substitution. Unfortunately, all
attempts to carry out this transformation failed, most of them
giving the desilylated pyrone as the only product. Nonethe-
less, the sulfoximine 7a, an immediate precursor of the
desired cycloaddition substrate, was efficiently prepared from
optically active 51 by sequential alkylation with diethylma-
lonate and subsequent amination with MSH in CH3CN (Table
entry
conditions (c)a
R
7 (yieldb)
1
2
3
4
5
H
7 (79%)
CH3COCl, Et3N, 0 °C
(CF3CO)2O, Et3N, rt
PhCOCl, Et3N, DMAP, rt
p-NO2C6H5COOH, EDC,
DMAP, rt
CH3SO2Cl, Et3N, 0 °C
(CF3SO2)2O, Py, 0 °C
p-TolSO2Cl, Py, 0 °C
MPA, EDC, DMAP, rt
COCH3
COCF3
COPh
7b (82%)
7c (72%)
7d (60%)
7e (81%)
COpNO2Ph
6
7
8
9
SO2CH3
SO2CF3
SO2pTol
MPA
7f (75%)
7g (65%)
7h (71%)
7i (74%)
a All of these reactions were carried out in CH2Cl2, except entry 8.
b Overall yield for steps b and c.
1).7 That the amination reaction proceeded with complete
stereoselectivity was confirmed by analysis and comparison
of the H NMR spectra of the Mosher [(+)-MPA and (()-
1
MPA] amide derivatives.6b This analysis revealed an optical
purity of at least 96%, similar to that of the sulfoxide
precursor 5. To our best knowledge this is the first reported
case of the formation of an optically active N-unsubstituted
alkenylsulfoximine from an R,â-unsaturated sulfoxide.5,6
After a brief, small-scale test of the viability of the coupling
between the N-acetyl derivative 7b and the bromopyrone 8
and of the cycloaddition of the resulting product, sulfoximine
7a was transformed into a variety of N-substituted derivatives
(Table 1).
The couplings between the alkenylsulfoximines 7 and the
bromopyrone 8 all proceeded cleanly to give the expected
cycloaddition precursors 9 in yields of between 77% for the
N-acetyl derivative 9b and 93% for 9c.
Remarkably, the [5C + 2C] cycloaddition reactions of
alkenylsulfoximines 9a-h took place over three times faster
than those of their sulfinyl analogues, an acceleration that
must be related to the stronger electron-withdrawing character
of the sulfonimidoyl group.5 As illustrated in Table 2, except
for the N-unsubstituted derivative 9a, the cycloadditions of
the acyl or sulfonylsulfoximines 9b-h gave a reasonable
degree of facial diastereoselectivity, which ranged from a
modest 58:42 for the p-nitrobenzoyl derivative 9e to a notable
90:10 for the N-tosyl compound 9h. In all cases, except those
of the N-tosyl, N-mesyl, and N-acetyl derivatives, it was
possible to separate the two diastereoisomeric cycloadducts
by flash chromatography.
(3) Stereochemistry of Organic Compounds; Eliel, E., Wilen, S. H., Eds.;
John Wiley and Sons: New York, 1995.
(4) (a) Renaud, P.; Bourquard, T.; Gerster, M.; Moufid, N. Angew.
Chem., Int. Ed. Eng. 1994, 33, 1601. (b) Lacoˆte, E.; Delouvrie´, B.;
Fensterbank, E.; Malacria, M. Angew. Chem., Int. Ed. 1998, 37, 2116. (c)
Delouvrie´, B.; Fensterbank, E.; Lacoˆte, E.; Malacria, M. J. Am. Chem. Soc.
1999, 121, 11395.
(5) A recent comprehensive review on sulfoximines: Reggelin, M.; Zur,
C. Synthesis 2000, 1.
(6) It has been shown that amination of sulfoxides with MSH proceeds
with complete retention of chirality at the sulfur atom: (a) Johnson, C. R.;
Kirchhof, R. A.; Corkins, H. G. J. Org. Chem. 1974, 39, 2458. See also:
(b) Yabuuci, T.; Kusumi, T. J. Am. Chem. Soc. 1999, 121, 10646.
The structures of the major diastereoisomers were un-
equivocally established as the oxabicycles 10b-h by the
(7) Use of CH3CN was critical for the success of these aminations; see:
Johnson, C. R.; Corkins, H. G. J. Org. Chem. 1978, 43, 4136.
624
Org. Lett., Vol. 3, No. 4, 2001