56
Chemistry Letters Vol.38, No.1 (2009)
New Ytterbium Complex-catalyzed Multicomponent Reactions for Synthesis
of Dihydropyrimidines: [4 þ 2] Cycloaddition vs. Biginelli Type Reaction
Jie Zhu, Mingjie Zhang,ꢀ Bo Liu, and Xiaojuan Li
Department of Chemistry, School of Science, Tianjin University, Tianjin 300072, P. R. China
(Received October 14, 2008; CL-080985; E-mail: zhujie@tju.edu.cn)
A new schiff base ytterbium complex was synthesized and
Ar
Ar
RO2C
Me
X
RO2C
Me
Me
used as catalyst for a three-component, one-pot reaction involv-
ing 1,3-keto ester with urea or thiourea and aldehyde. The reac-
tions resulted in the formation of two different dihydropyrimi-
dines, the Biginelli type 3,4-dihydropyrimidin-2-(1H)-ones and
non-Biginelli type. A new mechanism based on [4 þ 2] cycload-
dition was proposed which serves as a reasonable explanation for
the two different products.
Yb(pic)3· L
NH
NH
+
+
ArCHO
N
+
H2N
NH2
O
THF, RT
N
H
X
RO2C
N
H
X
N
O
a
b
HO
HO
OH
OH
L=
RO2C
Me
CHO
Y
Yb(pic)3·L
+
HO
+
H2N
NH2
O
THF, RT
RO2C
Me
NH
Multicomponent reactions (MCRs) have received increasing
interest in organic and medicinal chemistry1 during the past two
decades mainly owing to the high demands of drug discovery.2
The venerable Biginelli reaction, one-pot cyclocondensation of
aldehyde, ꢀ-keto ester, and urea or thiourea, is a classical way
to produce multifunctionalized 3,4-dihydropyrimidin-2(1H)-
ones (DHPMs).3 DHPMs show a wide scope of important phar-
macological properties and make up a large family of medicinal-
ly relevant compounds. The present attention on Biginelli type
DHPMs is mainly due to their close structural relationship to
the pharmacologically important as calcium channel modulators
of the nifedipine type, and several marine natural products con-
taining the dihydropyrimidine-5-carboxylate core exhibit anti-
cancer properties.4,5
Pharmacological studies concerning the absolute configura-
tion have demonstrated individual enantiomers perform oppos-
ing biological activities in most cases.4 Nevertheless, only a
few examples of asymmetric synthesis of this heterocyclic target
have been reported.6,7 Although the reaction was described over
a century ago, and more and more attention are focused on
Biginelli DHPMs, the mechanism of the classical three-compo-
nent Biginelli condensation has not been elucidated with cer-
tainty and remains controversial.8 The first so-called ‘‘ureido-
crotonate mechanism’’ was proposed by Folkers and Johnson9
in 1933. They suggested that the primary bimolecular condensa-
tion product N,N0-benzylidenebisurea is the first intermediate in
this reaction. 40 years later in 1973, the ‘‘carbenium ion mecha-
nism,’’ proposed by Sweet and Fissekis,10 suggested that an acid-
catalyzed aldol condensation is the first and limiting step of the
Biginelli reaction. The widely accepted mechanism proposed by
Kappe11 is the ‘‘N-acyliminium ion intermediate’’ path, which is
a confirmation of Folkers and Johnson’s. According to this
mechanism, only a single product can be obtained. Herein, we
proposed a new mechanism for the Biginelli reaction via
[4 þ 2] cycloaddition,12 which serves as a reasonable explana-
tion for the observation of two different dihydropyrimidine prod-
ucts during our experiment.
O
1. Y=O, R=Et
2. Y=S, R=Et
3. Y=S, R=t-Bu 17a
15a
16a
N
H
Y
Scheme 1. The reaction resulted in different products.
The Biginelli condensation of ethyl acetoacetate, benzaldehyde,
and urea was catalyzed by ytterbium picrate [Yb(pic)3] with the
ligand (Scheme 1, Table 1, Entry 1). The result was not satisfac-
tory for the reaction showing high conversion but the yield of ex-
pected product 1a is very low (37%). After screening the reac-
tion mixture carefully, we surprisingly found a new product 1b
yield in 54% which is also a dihydropyrimidine and is different
in the methyl and ester group positions from the expected one.
The differences are demonstrated by 1H NMR characterizations.
Because this intriguing finding, we proceeded to examine
the scope of this side product with various substituted aromatic
aldehydes, two 1,3-keto esters, and urea/thiourea.13 Two differ-
ent dihydropyrimidine products were found for most substrates
(Table 1, Entries 1–14), and for salicylaldehyde (Table 1, Entries
15–17), because of the Biginelli type precursors were much easi-
er to conduct intramolecular etherification, three sole diazatricy-
clic products were produced. The X-ray diffraction analysis of
compound 4a, 7a, 11a, and 16a were then performed to confirm
that the molecular structures are indeed as shown in Supporting
Information.15
The non-Biginelli products could not be obtained according
to the mechanism of Kappe’s.11 Thus, we proposed a new mech-
anism based on a [4 þ 2] cycloaddition (Scheme 2). Condensa-
tion of urea and benzaldehyde affords imine, which is then
enolized to form a diene, 1-benzylideneisourea. The Yb(pic)3ꢁL
complex is assumed to contribute to the stabilization of diene
intermediate. This is the key step which is different from the
N-acyliminium intermediate. Subsequent cycloaddition with the
enol generates the ꢀ-carbonyl compound and then gives the
corresponding dihydropyrimidine 1a and 1b after elimination
of a water molecule.
Methyl acrylate, instead of 1,3-keto ester, serving as a di-
enophile, was added to the mixture of benzaldehyde, urea and
the ytterbium catalyst to validate the existence of diene inter-
mediate. After 24 h stirring, two new compounds were observed
(Scheme 3). The formation of A and B demonstrated generation
At the beginning, we focused on the asymmetric synthesis of
DHPMs via the Biginelli reaction. A new BINOL-salen type
chiral ligand was developed from the corresponding amine and
BINOL-derived aldehyde in alcohol under ambient atmosphere.
Copyright Ó 2009 The Chemical Society of Japan