tion, starting from two rather simple starting materials
(aldehydes and N-acylamido acrylate) and resulting in the
formation of valuable highly enantioenriched a-amino acid
derivatives [Eq. (1)].
we observed that the use of two equivalents of the Michael
acceptor had a positive effect on conversion (Table 1,
entries 6 and 7). Switching to stronger bases such as
KHMDS and KOtBu restored the enantioselectivity to
95%, and the conversion increased substantially (67% and
65%; Table 1, entries 8 and 9, respectively).[16] We surmise
that these bases are strong enough to be fully protonated by
the triazolium salt, so that free base, which might racemize the
product, is no longer present in the reaction mixture. Low-
ering the reaction temperature to 08C led to full conversion
and high levels of enantioinduction after only 4 h (Table 1,
entry 10). Clearly, racemization is not a problem under these
mild reaction conditions, even with a prolonged reaction time
(Table 1, entry 11).
Two equivalents of 2 were essential for full conversion
under these conditions (Table 1, entry 12). Nevertheless, the
catalyst and base loadings could be further reduced without
affecting the yield or enantioselectivity; this led to the
optimized reaction conditions (Table 1, entry 13; highlighted
in bold). The reaction has also been applied successfully on a
gram scale (5 mmol) under these optimized conditions to
yield enantiomerically enriched a-amino acid derivative 4a in
excellent yield and stereoinduction (Table 1, entry 13). How-
ever, lowering the amount of catalyst significantly below
10 mol% resulted in an incomplete reaction (not shown).
Furthermore, we tried other Michael acceptors in the
reaction to evaluate their reactivity. Other amino protecting
groups such as tert-butoxycarbonyl (Boc) or phthalimido
failed to provide the desired product. In addition, the N-H
group of the amide seems to be crucial for the reactivity, since
the tertiary N-methylated variant of Michael acceptor 2 did
not react.[17] Similarly, the b-substituted acrylate (Z)-methyl-
acetamido cinnamate (MAC) also did not react under our
standard conditions.
In the course of our research on NHC organocatalysis[14]
we decided to explore an enantioselective Stetter reaction in
which a stereoselective protonation was the key step by using
aromatic or aliphatic aldehydes and a dehydroamino ester as
the Michael acceptor. Our study commenced with the
observation that the reaction between aldehyde 1a and
Michael acceptor
2 was catalyzed by chiral NHC 3
(Table 1).[15] The Stetter product 4a was obtained with an
excellent enantioselectivity of 97% but with a poor yield of
only 10% (Table 1, entry 1). An extensive screening of the
reaction conditions revealed several crucial parameters. We
first showed that the use of TBD as the base and toluene as a
solvent gave an improved yield compared to dioxane or THF
(Table 1, entries 2–4); however, the enantioselectivity
decreased to 84%. Similar results were obtained with DBU
or K2CO3 as the base (Table 1, entries 5 and 6). Additionally,
Table 1: Optimization of the reaction conditions.[a]
Thereafter, we studied the scope and the generality of this
reaction. First, we screened aromatic aldehydes bearing an
electron-withdrawing group (Scheme 3). Methyloxycarbonyl
(4b), trifluoromethyl (4c), and cyano (4d) groups were
compatible with the reaction conditions. In all cases, the
reaction led to the desired products in good yields and with
excellent ee values over 90%. It is noteworthy that halides
such as bromide in position 4 or 3 (products 4e and 4 f,
respectively) or chloride (product 4g) were also tolerated,
and the corresponding products could undergo further
functionalization by cross-coupling reactions for the con-
struction of more elaborate molecules. Challenging aldehydes
with a substituent in the ortho position were then evaluated.
With 2-fluorobenzaldehyde, we obtained the desired product
4h with an exceptional enantioselectivity (99% ee). More
sterically hindered aldehydes such as 2-methylbenzaldehyde
or 2-chlorobenzaldehyde failed to participate in the reaction
(not shown). Notably, heteroaromatic aldehydes such as
furfuraldehyde were also successful in yielding the expected
product 4i with high stereocontrol (98% ee). The highly
reactive 2-naphthalenecarboxaldehyde can also participate in
the reaction to give the desired product in excellent yield and
selectivity. As often reported, electron-rich aromatic alde-
hydes are known to be less reactive in NHC-catalyzed
processes. For example, attempts to carry out the reaction
Entry
Base
Solvent
T
[8C]
t
[h]
Yield
[%][b]
ee
[%]
1[c]
KHMDS
TBD
TBD
TBD
DBU
K2CO3
K2CO3
KHMDS
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
dioxane
dioxane
THF
25
25
25
25
25
25
25
25
25
0
24
24
24
24
24
24
24
24
24
4
(10)
22
17
41
49
23
31
67
65
97
2[d]
3[d]
4[d]
5[d]
6[d]
7
n.d.
n.d.
84
85
n.d.
90
95
95
95
95
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
8[e]
9[e]
10[e]
11[e]
12[e,f]
13[g]
99 (98)
99
72
0
0
0
65
24
3
n.d.
95
95 (93)
[a] General reaction conditions: 1a (0.2 mmol, 1.0 equiv), 2 (0.4 mmol,
2.0 equiv), NHC·HCl 3 (10 mol%), base (10 mol%), 0.67 mL solvent
(0.3m). [b] Yield determined by 1H NMR spectroscopy using 1,3,5-
trimethoxybenzene as an internal standard. Yields of isolated products
are given in parentheses. [c] 1a (0.4 mmol, 2.0 equiv), 2 (0.2 mmol,
1.0 equiv). [d] 2 (0.22 mmol, 1.1 equiv). [e] NHC·HCl 3 (15 mol%). [f] 2
(0.3 mmol, 1.5 equiv). [g] Base (8 mol%). On a 5 mmol scale, 95% yield
and 94% ee were obtained. n.d.=not determined. DBU=1,8-
diazabicyclo[5.4.0]undec-7-ene, KHMDS=potassium hexamethyldisil-
azide, TBD=1,5,7-Triazabicyclo[4.4.0]dec-5-ene.
Angew. Chem. Int. Ed. 2011, 50, 1410 –1414
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1411