3828
S. Orlandi et al. / Tetrahedron: Asymmetry 14 (2003) 3827–3830
A breakthrough in the field was obtained by Enders in
1995, via the use of the chiral 1,2,4-triazolium salt 1 as
a catalyst for the dimerisation of benzaldehyde, to
afford the corresponding benzoin in 66% yield and 75%
ee (Fig. 2).12,13 Comparable enantioselectivities were
later obtained by Knight and Leeper, by employing the
bicyclic triazolium salt 2.14 The comparison between the
enantiomeric excess obtained with 2 (80% ee) and a
similar, bicyclic thiazolium salt 3 (20.5% ee) seems a
clear indication of the importance of the N-aryl ring in
determining the stereochemical outcome of the reac-
tion. In 2 it is possible that the benzyl group at the
stereocenter preferably shields one side of the interme-
diate C in Figure 1, with the N-phenyl playing a
decisive role, in pre-orienting the approach and the
attack of the second molecule of aldehyde.
enantiomerically pure (1S,2S)-2-amino-1-phenyl-1,3-
propanediol 6 was reacted with ortho-fluoro nitroben-
zene, in the presence of potassium carbonate in
aqueous ethanol for 60 h to give 7 in 57% non opti-
mized yield; reduction of the nitro group followed by
the reaction with formic acid gave the benzimidazole 8
in 65% overall yield.17 Protection of the diol with
2,2-dimethoxypropane afforded 9. Alkylation with
iodomethane in acetonitrile gave the enantiomerically
pure N-methyl-benzimidazolium salt 10 in 95% yield.18
The ability of these chiral salts to act as a catalyst in
asymmetric benzoin reaction of benzaldehyde was then
investigated. Optimal reaction conditions were estab-
lished by preliminary experiments carried out with the
achiral salt 11 (Scheme 1). Selected data are shown in
Table 1. The N-methyl salt 11 promoted the reaction as
well in THF with DBU (entry 1), or NaH (entry 2), as
in methanol with DBU (entry 3) or NaOH 50% (entry
4). Having established that N-methyl benzimidazolium
salts were efficient catalysts, the activity of the chiral
salt 10 was investigated.19 Unfortunately 10 proved to
be less reactive than 11, as it did not catalyze the
reaction in the previously explored conditions (see for
example entry 5 versus 1). The best result obtained was
by using 5 mol% of the catalyst, in the presence of NaH
as base in THF; in these conditions the benzoin was
obtained in 31% yield after 2 h reaction at 60°C, as a
racemic mixture.
However catalyst 1 was unstable because of degrada-
tion due to competing deprotonation at the 3-position,
to irreversibly give an N-cyanobenzamidine, via triazole
ring opening.13 In order to solve the problem, we
decided to investigate the catalytic activity of chiral
N-alkyl benzimidazolium salts, structurally inspired by
catalyst 1 and in which degradation is not possible.
Recently Enders has developed the new chiral tri-
azolium salts 415 which are able to promote asymmetric
benzoin condensation with enantiomeric excesses up to
93%. A similar bicyclic triazolium salt 516 has also been
described as very efficient catalyst for intramolecular
Stetter reactions (ee up to 97%) (Fig. 2). Prompted by
these works, we report herein the synthesis of C1 and
C2 symmetric, conceptually new, chiral benzimida-
zolium and thiazolium salts and preliminary studies on
their catalytic activity.
Disappointed by the modest reactivity and enantiose-
lectivity shown by these catalysts, we decided to explore
the behavior of completely new chiral salts. Thus, the
synthesis of C2-symmetric enantiomerically pure salts
containing a stereogenic axis as stereochemical element
able to control the outcome of the reaction was
attempted (Scheme 2).
The synthesis of a chiral benzimidazolium salt is
described in Scheme 1. The commercially available
Figure 2. Chiral thiazolium and triazolium salts.