Angewandte
Chemie
[
2]
after in catalytic asymmetric intramolecular aldol reactions.
l-Proline and other chiral amines have recently been shown
to be efficient catalysts of asymmetric intermolecular aldol
[
3]
reactions and a variety of other imine- and enamine-based
[
4–7]
reactions.
Although impressive, the synthetic scope of l-
proline is not sufficient to address all aspects of the aldol
[
3b]
reaction.
For example, the synthesis of compounds with
quaternary carbon atoms is currently one of the most
challenging topics in asymmetric organic chemistry that is
[
8,9]
not addressed efficiently with l-proline catalysis.
l-Pro-
line-catalyzed aldol reactions have focused on the use of a-
monoalkyl-substituted or a-heteroatom-substituted carbonyl
compounds as donors. The use of a,a-dialkyl aldehyde donors
should provide direct access to enantiomerically enriched
products with a quaternary carbon atom. However, the
application of this approach to reactions of a,a-dialkyl
aldehyde donors has not provided satisfactory results.
Since an amine-catalyzed aldol reaction proceeds via an
enamine intermediate, acceleration of the formation of the
enamine intermediate can be key to improving the construc-
tion of a,a-dialkyl aldol products. Recently we demonstrated
[
10]
the utility of a fluorescence detection system for monitor-
ing the progress of CÀC bond formation in the reaction of the
Figure 1. Initial rates of reactions of chiral-amine/acidcombinations
with the fluorogenic substrate 1 andacetone ( 2). Catalysts andruns:
l-proline (4, runs 1–16), l-prolinol (5, runs 17–32), (S)-(À)-a,a-
diphenyl-2-pyrrolidinemethanol (6, runs 33–48), (S)-(+)-2-(methoxyme-
thyl)pyrrolidine (7, runs 49–64), and( S)-(+)-1-(2-pyrrolidinylmethyl)-
pyrrolidine (8, runs 65–80). Each catalyst was evaluatedwith each of
the following acid additives (from left for each catalyst): none,
Sc(OTf) , Cu(OTf) , Zn(OTf) , Y(OTf) , La(OTf) , Eu(OTf) , Yb(OTf) ,
maleimide 1 and acetone (2). This Michael-type reaction can
then be used as a surrogate-reporter reaction for other
enamine-based reactions. By monitoring the formation of the
fluorescent product 3 (Figure 1), catalysts of enamine for-
mation were evaluated, and an effective pyrrolidine/acetic
acid bifunctional catalyst was identified for the use of a,a-
3
2
2
3
3
3
3
H SO , CF SO H, p-TsOH, d-(+)-10-camphorsulfonic acid, HNO ,
2
4
3
3
3
[
11]
dialkyl aldehydes as aldol donors. This study provided us
with the incentive to find asymmetric catalysts for this
important class of aldol reactions. Herein we present the
results of our investigation of direct asymmetric intermolec-
ular aldol reactions of a,a-dialkyl aldehydes with aryl
aldehydes through high-throughput fluorescence-based
screening.
CF CO H, CH CO H, H PO . Reactions were carriedout with the chiral
3
2
3
2
3
4
amine (3 mm), acid(3 m m), and 1 (6 mm) in 20% acetone/80%
DMSO, andthe initial reaction rates were de terminedby monitoring
the fluorescence (l =315 nm, l =365 nm) over 20 min. DMSO=
ex
em
dimethyl sulfoxide, RFU=relative fluorescence unit, Tf=trifluorome-
thanesulfonyl, Ts=p-toluenesulfonyl.
To evaluate the catalytic efficiency of the chiral amines 4–
8
in the presence of various acid additives, such as Lewis,
MeCN, THF, PhMe, iPrOH, and MeOH. DMSO was
determined to be the most effective solvent.
Brønsted, and organic acids, the reaction of 1 with acetone
was performed in the presence of each of these catalysts, and
the increase in fluorescence was monitored (Figure 1). The
best results were observed in the reactions with the catalyst 8
and the acid additive trifluorosulfonic acid (run 74, RFU =
1
trifluoroacetic acid (run 78, RFU = 152.5 s ). The addition of
these acids significantly improved the reaction rate relative to
that with the catalyst 8 in the absence of an acid (run 65,
RFU = 35.0 s ). The initial rate of reactions catalyzed by l-
proline (4, run 1, RFU = 79.2 s ) and l-prolinol (5, run 17,
The utility of the chiral amine catalysts was evaluated in
intermolecular aldol reactions between a,a-dialkyl aldehyde
donors 9 and aryl aldehyde acceptors 10 (Table 1). The
catalyst 8 and CF CO H (0.3 equiv; Table 1, entry 7) provided
3
2
À1 [12]
60.0 s ),
and with the catalyst 8 and the acid additive
11a (X = NO ) in excellent yield with 94% ee. A lower
2
À1
catalyst loading (0.05 equiv; Table 1, entry 9) also led to good
results (92% yield, 96% ee). The addition of an acid, in an
amount equimolar to the amine 8, improved the reactivity
(Table 1, entry 6 versus entry 7) as well as the enantioselec-
tivity (Table 1, entry 6 versus entries 7–9). The data in Table 1
are consistent with the results of the fluorescence assay. The
combination of 8 and CF SO H was also effective (Table 1,
À1
À1
À1
RFU = 73.9 s ) were not increased by the addition of any of
the acids. The reaction with the catalyst 6, which has a bulky
3
3
diphenylhydroxymethyl substituent, had a low rate in the
absence of an acid (run 33, RFU = 1.6 s ), and the rates
entry 10). However, the use of the catalysts l-proline (4;
Table 1, entry 1), l-prolinol (5; Table 1, entry 2), and 6/
CH CO H (Table 1, entry 3) provided the aldol product 11a
À1
remained low even when acids were added. For the catalyst 7,
the rate was enhanced by the addition of acetic acid (RFU =
3
2
in low yields. Although a reaction with the combination 7/
CH CO H (0.3 equiv) provided 11a in better yield in much
À1
À1
1
4.8 s without acid, run 49 and RFU = 85.5 s with acetic
3
2
acid, run 63). The chiral-amine/acid combinations were also
evaluated in different solvents, such as dimethyl sulfoxide
less time (Table 1, entry 5) compared to the reaction with 7
alone (Table 1, entry 4), both reactions proceeded with low
enantioselectivities. The catalyst 8/CF CO H was effective in
(
DMSO), N,N-dimethylformamide, 1,4-dioxane, acetone,
3
2
Angew. Chem. Int. Ed. 2004, 43, 2420 –2423
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2421
Mase, Nobuyuki
