Enantiomerically Enriched Thiiranium Ions
FULL PAPER
phene. These results are strong indicators that diphenyl di-
sulfide attacks the thiiranium ion 5 at the sulfur atom at
which point a dithiosulfonium salt 15 and (E)-4-octene are
reversibly formed (Scheme 13). In fact, attack of disulfides
at the sulfur atom of thiiranium ions is also supported by
calculations and kinetic experiments.[17,36] In contrast, attack
at the carbon atom has been shown with dimethyl disul-
fide.[37] These experimental results support Pasquatoꢁs pro-
posal, that chiral, nonracemic thiiranium ions generated by
enantiomerically enriched dithiosulfonium salts might race-
mize after formation.[23] However, racemization depends, of
course, on the relative rates of attack at the sulfur or carbon
atom of the thiiranium ion which led to the next mechanistic
experiments.
readily occur (Scheme 10). An associative sulfenyl group
transfer to an olefin seemed likely for this reason. In addi-
tion, computational studies by Radom and co-workers sug-
gested an associative pathway.[38] Because the initially
formed thiiranium ion is chiral and enantiomerically en-
riched, it can distinguish the enantiotopic faces of the added
olefin thus leading to an enantiomerically enriched transfer
product. The magnitude of the asymmetric induction, how-
ever, is not important.
Scheme 14. Results from the transfer experiment.
In this study, sulfenyl group transfer from an enantiomeri-
cally enriched thiiranium ion to an olefin is carried out fol-
lowed by methanolysis and analysis of enantiomeric compo-
sition of the products. This experiment had to be designed
and optimized carefully because two chiral, nonracemic
thiiranium ions and two olefins would be present in the re-
action mixture. In principle each of the two thiiranium ions
could deliver sulfenyl cations to the olefins, hence four sulfe-
nyl group transfers are possible. All of these transfers could
ultimately lead to the racemization of the present thiiranium
ions. The challenge was to find substrates and conditions,
where only one of the four pathways occurred irreversibly.
For this demanding experiment (2S,3S)-2,3-dibenzyl-1-phe-
nylthiiranium hexafluoroantimonate (13) and (E)-4-octene
were chosen as substrates for several reasons. The inverse,
thermodynamically unfavorable sulfenyl group transfer from
thiiranium ion 5 to (E)-1,4-diphenyl-2-butene does not lead
to fast racemization even at 08C (Table 1, entry 4) and
transfer from 5 to (E)-4-octene can be suppressed thermally
(Table 1, entries 1–3).[7] Sulfenyl group transfer from 13 to
its parent olefin (E)-1,4-diphenyl-2-butene was expected to
have a similar temperature profile. The mandatory preserva-
tion of enantiomeric purity of both thiiranium ions before
methanolysis seemed possible for these substrates.
The synthesis of (2S,3S)-2,3-dibenzyl-1-phenylthiiranium
hexafluoroantimonate (13) was executed uneventfully in
good overall yield. Critical to the success of the experiment
was the development of conditions where the sulfenyl group
transfers of 5 and 13 to their parent olefins and hence race-
mization of the thiiranium ions was thermally suppressed
but wherein the transfer from 13 to (E)-4-octene could still
occur occurred. Additionally, the reaction time was limited
because of the competing decomposition of the thiiranium
ions 5 and 13 above ꢀ208C. The experiment was performed
at ꢀ108C, where 5 and 13 were still sufficiently stable for
1 h. Gratifyingly, the transfer product was isolated not as a
racemate, but with slight enantiomeric enrichment in ent-6
(Scheme 9). Thus, the olefin-to-olefin transfer afforded
modest stereochemical induction which provides compelling
Scheme 13. Racemization of thiiranium ion 5 catalyzed by diphenyl disul-
fide.
Competition experiments: Competition reactions were per-
formed to investigate whether attack on a thiiranium ion at
sulfur by diphenyl disulfide or at carbon by nucleophiles is
faster. This factor is mechanistically relevant, since once
formed, a chiral, nonracemic thiiranium ion must be inter-
cepted faster than racemization occurs to observe asymmet-
ric induction. In view of our ongoing efforts to develop
enantioselective Lewis base catalysis of chalcogeno function-
alizations, a Lewis base would always be present. These re-
actions were conducted as competition experiments wherein
enantiomerically enriched 5 was treated with a preformed
mixture of diphenyl disulfide, (E)-4-octene and tetrabutyl-
ammonium acetate or methanol (Scheme 6). Attenuated
stereospecificities were anticipated in these inherently
biased intermolecular cases. However, no erosion of configu-
rational purity occurred! This striking result suggests that
the intramolecular trapping of an in situ formed, enantio-
merically enriched thiiranium ion might not be necessary to
preserve enantioenrichment. Therefore, one could ultimate-
ly also envision intermolecular chalocogeno functionaliza-
tions.
Investigation of the “olefin-to-olefin” transfer mechanism:
Because the “olefin-to-olefin” transfer process represents a
viable racemization pathway, a mechanistic investigation
was undertaken to gain more insight into its mechanism. No
experimental precedent addresses the possibility of either a
dissociative or an associative mechanism (Scheme 14). The
preceding studies had already shown that enantiomerically
enriched thiiranium ions are inherently configurationally
stable at low temperatures and racemization via carbocation
intermediates or dissociation of a sulfenyl cation does not
Chem. Eur. J. 2009, 15, 11737 – 11745
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11743