Angewandte
Chemie
DOI: 10.1002/anie.201200248
Asymmetric Catalysis
Trienamine Catalysis with 2,4-Dienones: Development and Applica-
tion in Asymmetric Diels–Alder Reactions**
Xiao-Feng Xiong, Quan Zhou, Jing Gu, Lin Dong, Tian-Yu Liu, and Ying-Chun Chen*
Since the seminal papers at the beginning of this century,[1]
asymmetric organocatalysis based on chiral amines has been
extensively and actively explored. A variety of catalytic
modes, which includes enamine, iminium, dienamine, and
SOMO pathways, have been elegantly applied to a large
number of asymmetric reactions, especially for aliphatic and
a,b-unsaturated aldehyde substrates.[2] Very recently, our
group as well as Jørgensen and co-workers have reported
a new reactive model, trienamine catalysis, for 2,4-dienals.
This method relies on the generation of an electron-rich
triene system in situ, which can perform as a diene counter-
part in the Diels–Alder (DA) reaction with an electron-
deficient dienophile, and gives exclusive b,e-regioselectivity
along with excellent stereoselectivity.[3]
It would be reasonable that such a trienamine catalytic
protocol should also be applicable to 2,4-dienone analogues
Scheme 1. Logical development of trienamine catalysis for 2,4-dienone
as the common enamine or iminium catalysis for a-enolizable
substrates. Conditions: i) 1a (20 mol%), TFA (40 mol%), 608C, tolu-
ketones or a,b-unsaturated ketones, respectively.[2a,b]
ene. X=9-quinyl.
Unfortunately, as outlined in Scheme 1, the expected trien-
amine catalysis was unsuccessful and almost no reaction
occurred between hepta-3,5-dien-2-one (2a) and the dieno-
phile N-phenylmaleimide (3a) when catalyzed by a primary
amine, such as 9-amino-9-deoxyepiquinine (1a, Table 1).[4] As
the two different trienamine intermediates I and II could be
formed, we reasoned that trienamine I might be preferred,
which would not promote the desired cycloaddition. As
a result, dienone 2b, which does not contain an a-enolizable
alkyl group, was used under the same conditions. However,
the a,d-regioselective DA cycloadduct 5a was isolated as the
sole product, albeit in low yield. It was later confirmed that
this reaction proceeded in a noncatalyzed manner, and
substrate 2b, rather than the trienamine III, directly acted
as the diene.[5] We envisaged that the introduction of another
methyl group at the d position of the dienone, as in the
structure of the 2,4-dienone 2c, would significantly inhibit the
reaction that proceeds through the undesired DA pathway, as
a quaternary center shown in the proposed product 5b has to
be formed. Moreover, the electron-donating effect of the d-
methyl group would also increase the HOMO energy of the
resulting trienamine intermediate IV,[6] which would further
facilitate the amine-catalyzed DA process.
To our gratification, it was found that the expected DA
reaction of 2c and 3a catalyzed by amine 1a and trifluoro-
acetic acid (TFA) in toluene at 608C was efficient.[7] Cyclo-
adduct 6a was isolated as a single endo diastereomer and,
more pleasingly, with excellent enantioselectivity (Table 1,
entry 1). It should be noted that only a trace amount of the
noncatalyzed DA product 5b was detected, as proposed
before. Subsequently, a number of reaction parameters were
explored. A much lower yield was obtained when the weaker
o-fluorobenzoic acid (OFBA) was used (Table 1, entry 2), but
good results were achieved in the presence of salicylic acid
(SA, Table 1, entry 3).[8] Both the yield of the reaction and the
ee value were decreased by applying sulfonic acids as the
additives (Table 1, entries 4 and 5). A lower yield and ee value
were obtained in chloroform (Table 1, entry 6), and almost no
reaction occurred in THF (Table 1, entry 7). The yield also
decreased upon lowering the loading of TFA, but without
effect on the enantioselectivity (Table 1, entry 8). Other chiral
primary amines derived from cinchona alkaloids were also
tested. 9-Amino-9-deoxy-epicinchonidine (1b) provided the
same enantioselectivity as 1a but with a lower yield (Table 1,
entry 9), whereas the demethylated derivative 1c[4f] could not
catalyze this reaction (Table 1, entry 10). Furthermore, 9-
amino-9-deoxyepiquinidine (1d) and 9-amino-9-deoxyepicin-
chonine (1e) delivered unsatisfying results and the product
[*] X.-F. Xiong, Q. Zhou, J. Gu, Dr. L. Dong, Prof. Dr. Y.-C. Chen
Key laboratory of Drug Targeting and Drug Delivery System of the
Education Ministry, Department of Medicinal Chemistry
West China School of Pharmacy, Sichuan University
Chengdu, 610041 (China)
E-mail: ycchenhuaxi@yahoo.com.cn
Dr. T.-Y. Liu, Prof. Dr. Y.-C. Chen
College of Pharmacy, The Third Military Medical University
Chongqing, 40038 (China)
[**] We are grateful for financial support from the NSFC (21125206 and
21021001) and the National Basic Research Program of China
(973 Program, 2010CB833300).
Supporting information for this article is available on the WWW
Angew. Chem. Int. Ed. 2012, 51, 4401 –4404
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4401