J.-J. Filippi et al. / Tetrahedron Letters 47 (2006) 6067–6070
6069
BF3
F3B
O
S
BF3
O
R
R
O
S
R
O
R
S
S
S
R
S
O
O
R
BF3
6
5
BF3
Scheme 2. Proposed mechanism for the isomerisation of c-thionolactones.
When the cis-thionolactone derivative 1e was submitted
to isomerisation (entry 5), a clean reaction took place
with the formation of trans-thiolactone 2e in 93% selec-
tivity. No cis-2e was observed. Interestingly, the iso-
meric trans-thionolactone 1f (entry 6) underwent a
slow isomerisation to cis-thiolactone 2f, though with a
selectivity of 50% (due to the presence of lactone 3f in
41% selectivity). Compound 2f was exclusively formed
as the cis isomer. These results are in agreement with a
previous study by Schmarr et al. dealing with P4S10 sul-
furisation of cis- and trans-whisky lactone diastereo-
mers, in which an inversion of configuration was
observed upon thiolactone formation.2b
highly slowed down and only 11% of the starting thiono-
lactone was converted after 10 h of reaction.
When In(OTf)3 was used to perform the reaction with
(R)-1a (ee = 98%), (S)-2a was obtained in 66% yield
after 24 h, with a 87% enantiomeric excess, indicating
that a minor pathway of SN1-type process could occur
with this catalytic system during the isomerisation pro-
cess. These results indicate that the nature of the Lewis
acid strongly influences the isomerisation mechanism.
In conclusion, we describe here a novel boron trifluo-
ride-catalysed thiono- to thiolactone isomerisation,
which resulted to be highly regio- and stereoselective.
This unexpected high isomerisation stereospecificity,
with inversion of configuration at the a position to oxy-
gen in starting 1 (and a to sulfur in compounds 2)
prompted us to further examine the possibility of an
enantioselective isomerisation process.
Acknowledgements
is made to the French Ministry of Education for a doc-
toral scholarship to J.J.F. and to Pr. Dr. U. J. Meierh-
enrich for his advice on chiral separations.
When applied to the optically active (R)-c-thionodeca-
lactone, (R)-1a (98% ee, entry 7), the isomerisation
reaction carried out with BF3ÆEt2O (10 mol %) led to
the formation of (S)-c-thiodecalactone 2a with a total
inversion of configuration at C-5 (97% ee) (Scheme 1).
Chiral GC was used for ee analysis; the elution of (R)
and (S)-2a have already been reported on the same col-
umn.2c Additionally, it was checked that no racemisa-
tion occurred once (S)-2a was formed, when running
the reaction of (R)-1a with BF3ÆEt2O for more than four
days.
References and notes
1. For reviews on volatile organosulfur compounds see: (a)
Boelens, M.; Van Gemert, L. Perfum. Flavor. 1993, 18, 29–
39; (b) Goeke, A. Sulfur Rep. 2002, 23, 243–278; (c)
Jameson, S. B. In Chemistry and Technology of Flavors and
Fragrances; Rowe, D. J., Ed.; Blackwell Publishing:
Oxford, 2005; p 116.
2. (a) Roling, I.; Schmarr, H. G.; Eisenreich, W.; Engel, K.
H. J. Agric. Food. Chem. 1998, 46, 668–672; (b) Schmarr,
H. G.; Engel, K. H.; Eisenreich, W. J. Agric. Food Chem.
2001, 49, 5923–5928; (c) Schellenberg, A.; Schmarr, H.-G.;
Eisenreich, W.; Engel, K.-H. In Frontiers of Flavour
Science; Schieberle, P.; Engel, K.-H. Eds., Deutsche
The complete inversion of configuration observed for
(R)-1a tends to exclude the possibility of a classical cat-
ionic-type mechanism to explain the results of the
isomerisation.8
Forschungsanstalt fur Lebensmittelchemie, Garching,
¨
Germany, 2000.
The results seem to indicate an SN2-type mechanism
involving at least two molecules of thionolactone. The
possibility to consider the formation of several dimer
intermediates can be considered. For example, an ylide
intermediate of type 5, followed by the formation of
dimer 6, as presented in Scheme 2, may explain the
observed enantioselectivity.
3. Levine, J. A.; Ferrendelli, J. A.; Covey, D. F. J. Med.
Chem. 1986, 29, 1996–1999.
4. (a) Scheibye, S.; Kristensen, J.; Lawesson, S. O. Tetra-
hedron 1979, 35, 1339–1343; (b) Curphey, T. J. J. Org.
Chem. 2002, 67, 6461–6473; (c) Filippi, J.-J.; Fernandez,
X.; Lizzani-Cuvelier, L.; Loiseau, A.-M. Tetrahedron Lett.
2003, 44, 6647–6650; (d) Filippi, J.-J.; Fernandez, X.;
Lizzani-Cuvelier, L.; Loiseau, A.-M. Flavour Fragr. J.
2006, 21, 175–184.
5. Taguchi, Y.; Suhara, Y. Bull. Chem. Soc. Jpn. 1986, 59,
2321–2324.
6. Furuya, M.; Tsutuminai, S.; Nagasawa, H.; Komine, N.;
Hirano, M.; Komiya, S. Chem. Commun. 2003, 2046–2047.
In order to confirm the possibility of a bimolecular
mechanism, the isomerisation of (R)-1a was performed
at a 10-fold lower concentration (0.05 M). As expected,
(S)-2a was obtained with a complete inversion of config-
uration (ee = 97%), but the kinetics of the process were