.
Angewandte
Communications
DOI: 10.1002/anie.201107317
Organocatalysis
Enantiodivergent and g-Selective Asymmetric Allylic Amination**
Jianmin Wang, Jie Chen, Choon Wee Kee, and Choon-Hong Tan*
There are many readily available methods for the preparation
of enantiopure carbonyl compounds containing a- and b-
chiral centers. However, the selective functionalization of the
g position has been met with more difficulties and less
progress. Asymmetric vinylogous aldol reactions,[1] Michael[2]
reactions, and phosphine-catalyzed nucleophilic addition to
alkynes and allenes [Eq. (1)][3] are some of the successful
attempts made to address this difficult problem. In particular,
the asymmetric vinylogous reactions seem to be most
promising at delivering the desired results, despite the fact
that the high electron density at the a position makes it
kinetically favorable relative to the g position. A cinchona-
alkaloid-catalyzed direct enantioselective g amination was
reported using activated alkylidene cyanoacetates and malo-
nonitriles [Eq. (2); EWG = electron-withdrawing group].[4]
Proline derivatives were also employed to catalyze the
direct asymmetric g functionalization of a,b-unsaturated
aldehydes by transforming the electron-poor alkene into an
electron-rich one [Eq. (3)].[5]
achieved by tuning various parameters, including additive,
counter ion, temperature, pressure, and solvent, used in the
reactions.[7a] Few organocatalytic enantiodivergent reactions
have been observed; however, several were reported for an
asymmetric Baylis–Hilman reaction.[7b–d] In a recent example,
a
guanidine/bisthiourea-catalyzed Mannich-type reaction
achieved a reversal of enantioselectivity by utilizing different
solvent conditions.[8] Authors frequently include the following
statements in their reports, “unexpected inversion of enan-
tioselectivity”, “change/dramatic change in the sense/direc-
tion of enantioselectivity”, “switch of the expected chiral
sense”, and “unexpected reversal of the enantioselectivity”.
These statements imply that these enantiodivergent syntheses
were not planned and it is difficult to design such reactions. It
was reported that E and Z enolates exhibit different enantio-
facial selectivities in enantioselective protonation reactions
because the two diastereomeric transition states for the
protonation of the E enolate are different from those for the
Z enolate.[9] Accordingly, the four possible transition-state
structures, resulting from the relative positioning of the
substitutents, drive diastereomeric differentiation which leads
to enantiodivergent selectivities of the protonation. We, thus,
envisioned a reaction that leverages the E/Z geometry of
a double bond to develop enantiodivergent asymmetric
synthesis.
Guanidine derivatives were reported as efficient catalysts
in asymmetric reactions.[10] Previously, we reported that
a bicyclic guanidine effectively activates dithiomalonates in
asymmetric Michael reactions.[11] The thioesters not only
increased the acidity of a-hydrogen atoms of the nucleophiles,
but also provided the substrate with a handle for modifica-
tion. We hypothesized that with the help of an adjacent vinyl
group, the acidity of the a-hydrogen atoms of unactivated
thioesters might be additionally enhanced and facilitate
deprotonation under mild reaction conditions (Scheme 1).[12]
Both a-amination and g-amination adducts are possible.
Enantiodivergent syntheses enable the preparation of
both enantiomers utilizing the same chiral catalyst.[6] Metal-
catalyzed enantiodivergent asymmetric catalysis can be
[*] J. Wang,[+] J. Chen,[+] C. W. Kee,[+] Prof. Dr. C.-H. Tan
Department of Chemistry, National University of Singapore
3 Science Drive 3, Singapore 117543 (Singapore)
E-mail: chmtanch@nus.edu.sg
Scheme 1. Brønsted base catalyzed allylic addition.
[+] These authors contributed equally to this work.
We prepared substrate 1a by a simple dicyclohexylcarbo-
diimide (DCC) coupling reaction of commercially available 3-
hexenoic acid and 2-methyl-2-propanethiol. The reaction
between the (E)-b,g-unsaturated thioester 1a and di-tert-
butyl azodicarboxylate (2c) was slow and was not complete
within 24 hours (Table 1, entry 1). However, we were pleased
[**] This work was supported by ARF grant R-143-000-461-112 and
scholarships (to J.M. and C.W.K) from the National University of
Singapore.
Supporting information for this article is available on the WWW
2382
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 2382 –2386