Superior amine catalysts for the Baylis-Hillman reaction: The use of DBU and its implications

Article abstract of DOI:10.1039/a907754e

DBU, which is normally regarded as a hindered and non-nucleophilic base, is in fact the optimum catalyst for the Baylis-Hillman reaction, providing adducts at much faster rates than using DABCO or 3HQD; the scope of the Baylis-Hillman reaction is enhanced

Full text of DOI:10.1039/a907754e

Superior amine catalysts for the Baylis–Hillman reaction: the use of DBU and
its implications†
Varinder K. Aggarwal* and Andrea Mereu
Department of Chemistry, University of Sheffield, Brook Hill, Sheffield, UK S3 7HF.
E-mail: v.aggarwal@sheffield.ac.uk
Received (in Cambridge, UK) 27th September 1999, Accepted 8th October 1999
DBU, which is normally regarded as a hindered and non-
nucleophilic base, is in fact the optimum catalyst for the
Baylis–Hillman reaction, providing adducts at much faster
rates than using DABCO or 3HQD; the scope of the Baylis–
Hillman reaction is enhanced using this catalyst and
implications of this finding are discussed.
amidine catalysts (entries 4, 5) were stable under the reaction
conditions except for DBU 7, which not only gave a clean
reaction but also the fastest rate (Table 1, entry 6; Table 2, entry
6). The substituted guanidine gave a lower yield due to a
competing side reaction involving the acrylate (entry 7).
In comparison with other commonly used catalysts (Table 2),
DBU was over an order of magnitude faster than the current best
catalyst (3-HQD, entry 4) and superior to the DABCO–
The Baylis–Hillman reaction has great synthetic utility as it
converts simple starting materials into densely functionalised
3
La(OTf) –triethanolamine system that we have developed
1
2
products. However, it suffers from low reaction rates (espe-
(entry 2). Even at 10 mol% loading, DBU was superior to both
cially for acrylates) and this often limits the range of substrates
that are tolerated. Whilst acceleration can be achieved by
physical methods this usually requires specialised apparatus.
We recently described a chemical method for accelerating the
stoichiometric DABCO or 3-HQD (Table 2, entry 5).5
We investigated the scope of the DBU-catalysed Baylis–
Hillman reaction by reacting benzaldehyde with a range of
Michael acceptors (Table 3), and methyl acrylate (Table 4) and
cyclohex-2-en-1-one (Table 5) with a range of electrophiles.
Notable examples from Table 3 include a fast reaction with
tert-butyl acrylate (entry 3; DABCO gives 65% yield after 28
3
reaction: the use of La(OTf) and triethanolamine as cocata-
lysts, which provided up to 40 fold rate increase over the use of
DABCO alone.2
6
Structural variations of the amine catalyst have been probed
and DABCO and 3-hydroxyquinuclidine (3-HQD) provide the
highest rates. Indeed, if the optimum features of the catalyst are
that it should be nucleophilic and unhindered, it is hard to
d ), and a very rapid reaction with cyclohex-2-en-1-one (entry
7
5). Methyl vinyl ketone (MVK) is not a suitable substrate with
DBU (vide infra) but, in any case, is a fast reacting substrate
with other catalysts (e.g. DABCO).
3
imagine superior ones. Here we describe our study on the
Notable examples from Table 4 include reaction with
4-anisaldehyde which gave a good yield after just 2 days (entry
nature of the catalyst and the discovery that much higher rates
can be achieved with alternative structures.
8
5; DABCO gives 90% after 20 d ), reaction with the notoriously
The rate-determining step of the Baylis–Hillman reaction is
difficult, hindered and deactivated 2-anisaldehyde (entry 4) and
the reaction of the aldehyde 4 with the ammonium enolate 3
3
(
Scheme 1). This enolate is formed by conjugate addition of the
Table 1 Amine catalysts in the Baylis–Hillman reactiona
nucleophilic amine 1 to the Michael acceptor 2 (a reversible
process) and therefore to obtain faster rates, higher concentra-
tions of the enolate are required. Amines which can shift the
equilibrium towards the enolate 3 by stabilising this species
should achieve this. We therefore screened a range of amines
which all had the potential for the positive charge on nitrogen to
be stabilised through conjugation with another heteroatom
Entry
Nucleophilic compound
t/h
96
120
120
1
2 min
6
48
Yield (%)
1
2
3
4
5
6
7
a
Dimethylaminopyridine
1-Methylimidazole
2-Methyl-2-oxazoline
N-Methyl-4,5-dihydroimidazole
DBN
87
0
0
c
10
d
13
(
Table 1). Of the aromatic heterocyclic catalysts (entries 1, 2)
DBU
89
30
4
only DMAP gave any Baylis–Hillman adduct, but at a rate only
slightly higher than DABCO (Table 2, entries 1, 3). None of the
Substituted guanidined
Reactions conducted on 2 mmol scale using 1+1+1 ratio of benzaldehy-
b
c
de+methyl acrylate+catalyst. Complex mixture of products. A fast
reaction occurred but the catalyst decomposed rapidly. d 1,3,4,6,7,8-Hexa-
hydro-1-methyl-2H-pyrimido[1,2-a]pyrimidine.
Table 2 Comparison of DBU with the other catalytic systemsa
Reaction
rate
Yield
(%)
b
Entry Catalytic system
k
rel
t/h
1
2
DABCO
DABCO–La(OTf) –
3
0.016
0.511
1
120
91
c
N(CH
2
2
CH OH)
3
31.9
2.4
4.8
10.2
49.5
12
96
30
24
6
83
87
91
75
89
3
4
5
6
a
Dimethylaminopyridine 0.038
3-Hydroxyquinuclidine 0.076
DBUd
DBU
0.163
0.762
Scheme 1
Reactions conducted on 2 mmol scale using 1+1+1 ratio of benzalde-
b c
hyde+methyl acrylate+catalyst. % Product per minute. 1 equiv. of
DABCO, 0.05 equiv. La(OTf)
equiv.
d
3
and 0.5 equiv. N(CH
2
CH
2
OH)
3
.
0.1
†
Experimental and spectroscopic data, and further references, are available
from the RSC web site, see http://www.rsc.org/suppdata/cc/1999/2311/
Chem. Commun., 1999, 2311–2312
This journal is © The Royal Society of Chemistry 1999
2311

View Article Online
Table 3 Reactions of activated alkenes with benzaldehydea
Entry
Alkenes
t/h
Yield (%)
1
2
3
4
5
Methyl acrylate
Ethyl acrylate
tert-Butyl acrylate
Acrylonitrile
6
24
72
3
89
80
74
92
60
b
Fig. 1
c
Cyclohex-2-en-1-one 0.5
a
Reactions conducted on 2 mmol scale using 1+1+1 ratio of benzalde-
acrylates work extremely well with DBU as a catalyst indicates
that enolisation is not a requirement and that the reaction must
proceed via the b-ammonium enolate derived from DBU and
the acrylate. Furthermore we have not been able to reproduce
some of their results: when we repeated the reaction between
MVK and acetaldehyde we obtained polymeric material
instead, no MVK was returned.
b
hyde: alkene+DBU. No side products; 20–25% benzaldehyde and acrylate
recovered. No benzaldehyde remained, 30% cyclohex-2-en-1-one re-
covered.
c
Table 4 Reactions of carbonyl compounds with methyl acrylatea
Entry
Aldehyde
t/h
Yield (%)
We were surprised that DBU worked so well as it is
considered to be a non-nucleophilic hindered base; features that
are diametrically opposite to what is normally required of amine
catalysts for the Baylis–Hillman reaction. DABCO, for exam-
ple, is one of the best amine catalysts and is an unhindered,
nucleophilic base. The incorporation of substituents a to
nitrogen result in substantial reduction in rates and this has been
ascribed to the lower nucleophilicity of the base due to steric
hindrance.13 In fact, a more likely explanation is that the
additional substituent contributes substantial steric hindrance to
the b-ammonium enolate intermediate 3 which results in a shift
in the equilibrium back to starting materials. In contrast, with
DBU the intermediate b-ammonium enolate is stabilised
through conjugation (Fig. 1), which increases its equilibrium
concentration, and this results in significantly enhanced rates.
These studies reveal that to achieve high rates in the Baylis–
Hillman reaction the nucleophilicity of the amine is much less
important than factors which stabilise the intermediate b-
ammonium enolate.
1
2
3
4
5
6
7
8
9
Benzaldehyde
6
1.5
89
95
95
2-Nitrobenzaldehyde
4-Nitrobenzaldehyde
2-Anisaldehyde
1
4
48
24
70
2
b
25
4-Anisaldehyde
62
b
Propionaldehyde
17
Trimethylacetaldehydec
b
20
60
78
d
2,2,2-Trifluoroacetophenone
2,2,2-Trifluoroacetophenonee
48
a
Reactions conducted on 2 mmol scale using 1+1+1 ratio of carbonyl
compound: acrylate: DBU. Low yield due to decomposition of aldehyde.
Reaction performed in presence of 0.05 equiv. La(OTf) . Product found
3
to be unstable in presence of high concentration of catalyst. 0.1 equiv.
DBU used.
b
c
d
e
Table 5 Reactions of aldehydes with cyclohex-2-en-1-onea
Entry
Aldehyde
t/h
0.5
50 min
7
Yield (%)
We thank the EPSRC for financial support for this work.
1
2
3
4
Benzaldehyde
2-Anisaldehyde
Cyclohexanecarbaldehyde
Trimethylacetaldehyde
60
70
73
75
Notes and references
b
21
1
For reviews see: D. Basavaiah, P. D. Rao and R. S. Hyma, Tetrahedron,
1996, 52, 8001; E. Ciganek, Org. React., 1997, 51, 201.
a
Reactions conducted on 2 mmol scale using 1+1+1 ratio of aldehyde-
b
2 V. K. Aggarwal, A. Mereu, G. J. Tarver and R. McCague, J. Org.
Chem., 1998, 63, 7183; V. K. Aggarwal, G. J. Tarver and R. J.
McCague, Chem. Commun., 1996, 2713.
+
cyclohex-2-en-1-one+DBU. 1.2 equiv. cyclohex-2-en-1-one and 0.05%
La(OTf)
3
used.
3
J. S. Hill and N. S. Isaacs, J. Phys. Org. Chem., 1990, 3, 285.
9
4 DMAP has been used as a catalyst for the Baylis–Hillman reaction
between cyclohex-2-en-1-one and formaldehyde but no comment was
made on its effectiveness or relative rate compared to DABCO. F.
Rezgui and M. M. El Gaied, Tetrahedron Lett., 1998, 39, 5965.
for the first time reactions with pivaldehyde (entry 7) and
trifluoroacetophenone (entries 8, 9). Pivaldehyde required
La(OTf)
without it. La(OTf)
3
to promote the reaction as no adduct was obtained
3
often enhances the rates and gives higher
5
At 10 mol% loading of DBU the reaction is somewhat capricious and
more reliable results are obtained using a stoichiometric amount of the
base. There is a variable amount of catalyst decomposition during the
course of the reaction.
yields of adducts (Table 4, entry 7; Table 5, entry 4) but when
used with benzaldehyde a less clean reaction was observed and
the additional acceleration was minimal. Aldehydes with
enolisable protons gave low yields of Baylis–Hillman adducts
6 Y. Fort, M. C. Berthe and P. Caubere, Tetrahedron, 1992, 48, 6371.
7 Cyclopent-2-en-1-one and cyclohept-2-en-1-one were both ineffec-
tive.
because of competing aldol reactions.10
Notable examples from Table 5 include high yielding
reactions with all the difficult aldehydes (2-anisaldehyde, entry
8
9
A. Foucaud and F. El Guemmout, Bull. Chim. Soc. Fr., 1989, 3, 403.
Reaction using DABCO reportedly failed. See: M. C. Berthe, P.
Caubere and Y. Fort, Eur. Pat. Appl., 1992, EP 465,293 (Chem. Abstr.,
2
; pivaldehyde, entry 4; even the aliphatic enolisable aldehyde
cyclohexanecarbaldehyde, entry 3), demonstrating the effec-
tiveness and superiority of this catalyst. Again La(OTf) (5
1
992, 116, 152605c); US Pat., 1994, 5,332, 836.
3
1
0 Low yields of adducts with enolisable aldehydes using DBU as a
catalyst have been reported. See: D. Basavaiah and V. V. L. Gowriswari,
Synth. Commun., 1987, 17, 587; P. Auvray, P. Knochel and J. F.
Normant, Tetrahedron, 1988, 44, 6095.
mol%) was required in the reaction with pivaldehyde (entry 4);
in its absence 44% yield of adduct was obtained after 4 days.
Evidently with faster reacting enones compared to acrylates,
enolisable aldehydes are better tolerated.
11 J. R. Hwu, G. H. Hakimalahi and C. T. Chou, Tetrahedron Lett., 1992,
3, 6469.
3
There is one example of the use of DBU as a catalyst for the
1
2 The use of DBU in the presence of 2 equiv. of 4-methoxyphenol has also
been reported in a patent, although without noting its superior
properties. See: K. Oohashi and S. Ido, Jpn. Kokai Koho JP, 1993, 05,
Baylis–Hillman reaction although no comment was made on its
rate.1
1,12
The focus of the paper was on a-alkylation of enones
with acrylates as acceptors and it was proposed that DBU acted
as a base rather than a nucleophile, with the reaction occurring
via a diene enolate rather than a b-enolate. This was supported
by their observation that reaction between MVK and acet-
aldehyde only returned starting material. Our observation that
1
7,375 9Cl. C07B41/02 (Chem. Abstr., 1993, 118, 254552s).
1
3 R. J. W. Schuurman, A. Vanderlinden, R. P. F. Grimbergen, R. J. M.
Nolte and H. W. Scheeren, Tetrahedron, 1996, 52, 8307.
Communication 9/07754E
2312
Chem. Commun., 1999, 2311–2312