should produce enol IV. After tautomerization of IV to
activated ester V, the attack of a suitable nucleophile
completes the catalytic cycle.10,11
Scheme 1. General Umpolung Strategy
We began our investigations of this intriguing strategy by
employing the simplest electrophile, a proton (Table 1). The
Table 1. Investigation of Homoenolate Reactivitya
pattern has significant utility in organic synthesis, establishing
the requisite nucleophilic character at the â-position of a
carbonyl compound is challenging.
Inspired by the proposed catalytic cycle for the thiazolium-
catalyzed benzoin reaction,8 we hypothesized that the exten-
sion of carbonyl anion reactivity to a â-position could be
achieved by use of an alkene in the form of R,â-unsaturation
(Scheme 2). The addition of a nucleophilic carbene catalyst9
entry
ROHb
catalyst (mol %)
additivec
yield (%)
1
2
3
4
5
EtOH
PhOH
BnOH
BnOH
BnOH
A (30 mol %)
A (30 mol %)
B (10 mol %)
C (20 mol %)
C (5 mol %)
57 (2)
55 (3)
47 (4)
82 (4)
82 (4)
PhOH
PhOH
PhOH
a Reactions performed with a 1:1 molar ratio of catalyst to DBU at reflux
temperature. b Performed with 5 equiv of nuclephilic alcohol. c Performed
with 2 equiv of proton additive.
Scheme 2. Proposed Mechanism to Generate Homoenolate
Equivalents from R,â-Unsaturated Aldehydes
use of the optimal alcohol would presumably serve a double
purpose as the electrophile source and subsequent nucleo-
phile. To probe this possibility, cinnamaldehyde (1) and
various alcohols were heated in toluene in the presence of
catalytic quantities of ammonium salts A-C and DBU. With
ethanol and ammonium salt A, we were pleased to isolate
the desired ester (57%), but clearly the system required
improvement. Initially, we pursued a moderate-yielding
approach that utilized a single reagent as the potential proton
donor and subsequent nucleophile required for catalyst
turnover (entries 1 and 2).
Fortuitously, while using PhOH as the alcohol in chloro-
form subjected to Al2O3 filtration, but not distillation, a
substantial amount of ethyl ester was observed. Our initial
hypothesis for this result was that the putative activated
acylation agent (i.e., V) was being trapped by the ethanol
typically added to stabilize chloroform (not shown). Armed
with the knowledge that the electrophile and nucleophile can
be decoupled in this manner, we discovered that the use of
imidazolium salt C (5 mol %) with phenol as the proton
source and a second alcohol as the nucleophile affords the
best yields for the reaction. This process is noteworthy since
no self-condensation (benzoin) products are observed, and
combinations of acids and bases in the reaction still generate
effective concentrations of catalyst I. The combination of
DBU and imidazolium salts in the presence of phenolic
functionality seems to be uniquely suited for this reaction.
Encouraged by these interesting results, we proceeded to
examine the ability of other alcohols to act as nucleophiles
to the carbonyl compound should generate tetrahedral
intermediate II, and subsequent hydrogen migration would
generate the reactive dienamine III. The normal nucleophilic
character of the carbonyl carbon is then extended to the
â-position and, in the presence of an electrophile (E-X),
(7) For reviews on homoenolate equivalents, see: (a) Hoppe, D.;
Kraemer, T.; Schwark, J.; Zschage, O. Pure Appl. Chem. 1990, 62, 1999-
2006. (b) Ahlbrecht, H.; Beyer, U. Synthesis 1999, 365-390. For recent
examples, see: Seppi, M.; Kalkofen, R.; Reupohl, J.; Frohlich, R.; Hoppe,
D. Angew. Chem., Int. Ed. 2004, 43, 1423-1427. (c) Reuber, J.; Frohlich,
R.; Hoppe, D. Org. Lett. 2004, 6, 783-786. (d) Hilgenkamp, R.; Zercher,
C. K. Org. Lett. 2001, 3, 3037-3040. (e) Whisler, M. C.; Beak, P. J. Org.
Chem. 2002, 68, 1207-1215.
(8) (a) Breslow, R.; Schmuck, C. Tetrahedron Lett. 1996, 37, 8241-
8242. (b) White, M. J.; Leeper, F. J. J. Org. Chem. 2001, 66, 5124-5131.
(9) (a) Bourissou, D.; Guerret, O.; Gabbai, F. P.; Bertrand, G. Chem.
ReV. 2000, 100, 39-91. (b) Herrmann, W. A. Angew. Chem., Int. Ed. 2002,
41, 1291-1309.
(10) Similar activated esters have been implicated in the reactions
between N-heterocyclic carbenes and R-halo aldehydes (Reynolds, N. T.;
de Alaniz, J. R.; Rovis, T. J. Am. Chem. Soc. 2004, 126, 9518-9519) or
epoxy-aldehydes (Chow, K. Y. K.; Bode, J. W. J. Am. Chem. Soc. 2004,
126, 8126-8127).
(11) For recent and related approaches accessing homoenolates from R,â-
unsaturated aldehydes for the synthesis of γ-butyrolactones, see: (a) Sohn,
S. S.; Rosen, J. W.; Bode, J. J. Am. Chem. Soc. 2004, 126, 14370-14371.
(b) Glorius, F., Burstein, C. Angew. Chem., Int. Ed. 2004, 43, 6205-6208.
906
Org. Lett., Vol. 7, No. 5, 2005