Rate acceleration of the Baylis-Hillman reaction in polar solvents (water and formamide). Dominant role of hydrogen bonding, not hydrophobic effects, is implicated

Article abstract of DOI:10.1021/jo016073y

A substantial acceleration of the Baylis-Hillman reaction between cyclohexenone and benzaldehyde has been observed when the reaction is conducted in water. Several different amine catalysts were tested, and as with reactions conducted in the absence of so

Full text of DOI:10.1021/jo016073y

510
J . Org. Chem. 2002, 67, 510-514
Ra te Acceler a tion of th e Ba ylis-Hillm a n Rea ction in P ola r
Solven ts (Wa ter a n d F or m a m id e). Dom in a n t Role of Hyd r ogen
Bon d in g, Not Hyd r op h obic Effects, Is Im p lica ted
Varinder K. Aggarwal,*,† David K. Dean,‡ Andrea Mereu,§ and Richard Williams§
School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, U.K., GSK,
New Frontiers Science Park (North), Third Avenue Harlow, Essex, CM19 5AW, U.K., and
University of Sheffield, Department of Chemistry, Dainton Building, Sheffield S3 7HF, U.K.
V.Aggarwal@bristol.ac.uk
Received August 30, 2001
A substantial acceleration of the Baylis-Hillman reaction between cyclohexenone and benzaldehyde
has been observed when the reaction is conducted in water. Several different amine catalysts were
tested, and as with reactions conducted in the absence of solvent, 3-hydroxyquinuclidine was found
to be the optimum catalyst in terms of rate. The reaction has been extended to other aldehyde
electrophiles including pivaldehyde. Attempts to extend this work to acrylates was only partially
successful as rapid hydrolysis of methyl and ethyl acrylates occurred under the base-catalyzed
and water-promoted conditions. However, tert-butyl acrylates were sufficiently stable to couple
with relatively reactive electrophiles. Further studies on the use of polar solvents revealed that
formamide also provided significant acceleration and the use of 5 equiv of formamide (optimum
amount) gave faster rates than reactions conducted in water. Using formamide, further acceleration
was achieved in the presence of Yb(OTf)3 (5 mol %). The scope of the new conditions was tested
with a range of Michael acceptors and benzaldehyde and with a range of electrophiles and ethyl
acrylate. The origin of the rate acceleration is discussed.
A number of physical and chemical methods have been
through either stabilization of the enolate by hydrogen
bonding or by activation of the aldehyde again through
hydrogen bonding or indeed both.2,4 As the initial addition
of the tertiary amine to the Michael acceptor is reversible,
a solvent that is able to solvate both the enolate and the
ammonium cation should provide a further acceleration
as it should increase the concentration of intermediate
1 (Scheme 1). Strongly polar solvents, especially water,
should be able to provide such solvation and thereby
enhance rates.6 Furthermore, there is the additional
possibility of achieving acceleration through hydrophobic
effects7 as the Baylis-Hillman reaction shows a large
negative volume of activation (∆Vq ) -70 cm3 mol-1).8
Indeed, Auge´ has shown that the Baylis-Hillman reac-
tion between acrylonitrile and various aldehydes can be
accelerated in water.9 However, it was not clear whether
the origin of the acceleration was due to hydrophobic
effects as the key test results involving salting in and
out agents were ambiguous: generally, both salting-in
(CsI) and salting-out (LiCl) agents caused a small reduc-
tion in rate, although LiI and NaI caused a small increase
in rate. It was curious that they had chosen to study
reactions involving acrylonitrile as this is already a
relatively fast-reacting substrate in the Baylis-Hillman
reaction. We have focused on slower reacting substrates,
which are very much in need of acceleration: â-substi-
tuted enones and acrylates. In this paper, we describe
developed to accelerate the notoriously slow Baylis-
Hillman reaction.1 These efforts not only allow reactions
to proceed more rapidly but also allow previously unre-
active partners to couple. Of these methods, the chemical
methods have the advantage of not requiring specialized
equipment, and so are especially attractive. We previ-
ously reported that the combination of a metal catalyst
and co-ligand (5 mol % La(OTf)3 and 80 mol % trietha-
nolamine) provided up to a 40-fold increase in rate,2 and
recently, Kobayashi showed that lithium perchlorate in
ether can lead to increased rates.3 We have also discov-
ered that DBU is a better catalyst than 3-hydroxyqui-
nuclidine (which was previously regarded as the optimum
catalyst4) and provides rate accelerations of up to 50-fold
over the standard amine catalyst (DABCO).5 We dem-
onstrated that these new conditions provided signifi-
cantly increased rates and also increased the scope of the
Baylis-Hillman reaction. However, there is still scope
for further improvements.
It is well-known that protic solvents (e.g., methanol,
ethylene glycol) accelerate the Baylis-Hillman reaction,
† University of Bristol.
‡ GSK.
§ University of Sheffield.
(1) For reviews, see: (a) Basavaiah, D.; Rao, P. D.; Hyma, R. S.
Tetrahedron 1996, 52, 8001-8062. (b) Drewes, S. E.; Roos, G. H. P.
Tetrahedron 1988, 44, 4653-4670. (c) Ciganek, E. Org. React. 1997,
51, 201-350. (d) Langer, P. Angew. Chem., Int. Ed. 2000, 39, 3049-
3052.
(6) For a review, see: Lubineau, A.; Auge´, J .; Queneau, Y. Synthesis
1994, 741-760.
(7) Breslow, R. Acc. Chem. Res. 1991, 24, 159-164.
(8) (a) Hill, J . S.; Isaacs, N. S. Tetrahedron Lett. 1986, 27, 5007-
5010. (b) Hill, J . S.; Isaacs, N. S. J . Phys. Org. Chem. 1990, 3, 285-
288.
(2) Aggarwal, V. K.; Mereu, A.; Tarver, G. J .; McCague, R. J . Org.
Chem. 1998, 63, 7183-7189.
(3) Kawamura, M.; Kobayashi, S. Tetrahedron Lett. 1999, 40, 1539-
1542.
(4) Ameer, F.; Drewes, S. E.; Freese, S.; Kaye, P. T. Synth. Commun.
1988, 18, 495-500.
(5) Aggarwal, V. K.; Mereu, A. Chem. Commun. 1999, 2311-2312.
(9) Auge´, J .; Lubin, N.; Lubineau, A. Tetrahedron Lett. 1994, 35,
7947-7948.
10.1021/jo016073y CCC: $22.00 © 2002 American Chemical Society
Published on Web 12/27/2001

Baylis-Hillman Reaction in Polar Solvents
Sch em e 1. Ba ylis-Hillm a n Rea ction
J . Org. Chem., Vol. 67, No. 2, 2002 511
Ta ble 2. Cou p lin g of Cycloh exen on e w ith Differ en t
Ca r bon yl Com p ou n d s in Wa ter a
entry
carbonyl compound
time (h)
yield (%)
1
2
3
4
5
benzaldehyde
o-anisaldehyde
isobutyraldehyde
pivaldehyde
4
24
17
10 d
4
82
74
59
36
99
formalinb
a
Reaction conditions: 2.0 mmol of 2-cyclohexen-1-one, 1.0 mmol
of carbonyl compound, 1.0 mmol of 3-hydroxyquinuclidine, and 1.0
b
mL of water. 5 equiv of formalin was used with 1.0 mmol of
2-cyclohexen-1-one.
Ta ble 1. Effect of Differ en t Am in e Ca ta lysts on th e Ra te
of th e Ba ylis-Hillm a n Rea ction in Wa ter a
Ta ble 3. Cou p lin g of ter t-Bu tyl Acr yla te w ith Ca r bon yl
Com p ou n d s in Wa ter a
entry
catalyst
rate (% min-1
)
krel (DABCO)
entry
carbonyl compound
time (h)
yield (%)
1
2
3
4
5
DABCO
4-DMAP
0.17
0.35
0.44
1.62
0.86
1.0
2.1
2.6
9.5
5.1
1
2
3
4
5
6
7
p-nitrobenzaldehyde
o-nitrobenzaldehyde
o-chlorobenzaldehyde
o-anisaldehyde
4
4
97
80
53
33
40
88
56
3-HDQ
3-HDQ with GnClb
3-HDQ with LiClb
48
17
72
10 d
4
a
2,2,2-trifluoroacetophenone
Reaction conditions: 2.0 mmol of cyclohexenone, 1.0 mmol of
PhCHO, 1.0 mmol of catalyst, and 1.0 mL of water were stirred
formalinb
glyoxylic acidc
b
at rt. 2.0 mmol of cyclohexenone, 1.0 mmol of PhCHO, 1.0 mmol
of 3-HDQ, and 1.0 mL of 4 M salt solution. GnCl ) guanadinium
chloride. The rate of each reaction was determined at low conver-
sion (<10%) by 1H NMR.
a
Reaction conditions: 2.0 mmol of tert-butyl acrylate, 1.0 mmol
of carbonyl compound, 1.0 mmol of 3-hydroxyquinuclidine, and 1.0
b
mL of water. 1.0 mmol of tert-butyl acrylate, 3.0 mmol of
formalin, 1.0 mmol of 3-HQD, and 1.0 mL of water. c 1.0 mmol of
NEt3 was also added to neutralize the acid.
the use of polar solvents to accelerate a broad range of
Baylis-Hillman reactions.
We began our studies with the reaction between
cyclohexenone10 and benzaldehyde in water and inves-
tigated the effect of different amine catalysts on reaction
rates (Table 1). DMAP has been compared with DABCO
and found to be superior in the reaction between cyclo-
hexenone and formaldehyde in THF/water.11 We found
that DMAP was indeed better than DABCO, but as with
reactions conducted neat, 3-hydroxyquinuclidine gave the
fastest rates (Table 1).12
To test whether the origin of the acceleration in water
was due to hydrophobic effects, salting-in [guanidinium
chloride (GnCl)] and salting-out (LiCl) agents were tested
(Table 1, entries 4 and 5).7,13 If hydrophobic effects were
operative, salting-in agents should decelerate the reaction
while salting-out agents should accelerate the reaction.
The results obtained showed that both GnCl and LiCl
caused an increase in the rate of these water-promoted
reactions. As both salting-in and salting-out agents
caused an increase in rate, this means that hydrophobic
effects are not the primary cause for the acceleration
observed. These results concur with Auge´’s observations.9
The optimum catalyst, 3-hydroxyquinuclidine, was
screened with a range of electrophiles (Table 2). As
benzaldehyde worked so well, providing high yields in
short reaction times (Table 2, entry 1), we only investi-
gated relatively difficult/unusual substrates. o-Anisalde-
hyde is a hindered and deactivated aldehyde but still
gave good yields after only 1 day (Table 2, entry 2).
Hindered aliphatic aldehydes such as isobutyraldehyde
(Table 2, entry 3) worked well although in the case of
pivaldehyde (Table 2, entry 4) required longer reaction
times. The use of pivaldehyde in the Baylis-Hillman
reaction is rare, and this is only the second report on its
use.5 Aqueous formaldehyde (formalin) also coupled
extremely efficiently under these conditions (Table 2,
entry 5).
The use of acrylates with 3-hydroxyquinuclidine in
water was less successful due to rapid hydrolysis par-
ticularly of the methyl and ethyl esters.14 The problem
of ester hydrolysis or transesterification is a general
problem when acrylates are employed with alcohol
solvents to enhance rates, and this is clearly accentuated
if the acrylate is attached to a solid support for combi-
natorial synthesis.15 However, we considered the use of
tert-butyl esters, which are much less rapidly hydrolyzed,
but this benefit has to be balanced against the fact that
tert-butyl acrylate is a notoriously slow partner in the
Baylis-Hillman reaction.1c In the event, tert-butyl acry-
late was sufficiently stable at short reaction times to
allow coupling to occur with reactive electrophiles (Table
3). It is notable that in most cases high yields were
obtained with this highly unreactive substrate after only
a few hours. However, aliphatic aldehydes were not
(10) â-Substituted enones were previously regarded as unreactive
substrates in Baylis-Hillman reaction (see ref 1c), but we have found
that cyclohexenone displays moderate reactivity and certainly greater
the tert-butyl acrylate (see ref 5).
(11) (a) Rezgui, F.; El Gaied, M. M. Tetrahedron Lett. 1998, 39,
5965-5966. (b) Sugahara, T.; Ogasawara, K. Synlett 1999, 4, 419-
420.
(12) This finding surprised us as the rate acceleration of 3-HDQ over
DABCO has been attributed to hydrogen bonding either intra- or
intermolecularly (ref 4). Clearly, hydrogen bonding cannot be respon-
sible for the increase in rate of 3-HDQ relative to DABCO in water
and this prompted us to investigate the origin of the rate acceleration
further. We have found that in the absence of steric effects there is a
correlation between the pKa of the base and its rate in the Baylis-
Hillman reaction, and this can account for the enhanced rate of 3-HDQ
[pKa 9.9 (see: Grob, C. A. Helv. Chim. Acta, 1985, 68, 882)] relative to
DABCO [pKa 8.7 (see: Hine, J .; Chen, Y.-J . J . Org. Chem., 1987, 52,
2091)]. Full details of these investigations will be reported in due
course.
(14) During the course of this work, Hu reported the use of water
to promote Baylis-Hillman reactions involving methyl acrylate and
also noted that partial hydrolysis of the acrylate occurred. He used
the less basic amine, DABCO, in which case the rate of hydrolysis will
be slower compared to the use of 3-HQD. Yu, C.; Liu, B.; Hu, L. J .
Org. Chem. 2001, 66, 5413-5418.
(13) (a) Breslow, R.; Guo, T. J . Am. Chem. Soc. 1988, 110, 5613-
5617. (b) Schmitt, M.; Bourguignon, J . J .; Wermuth, C. G. Tetrahedron
Lett. 1990, 31, 2145-2148. (c) Dack, M. R. J . Chem. Soc. Rev. 1975, 4,
211-220.
(15) Ra¨cker, R.; Do¨ring, K.; Reiser, O. J . Org. Chem. 2000, 65, 6932-
6939.

512 J . Org. Chem., Vol. 67, No. 2, 2002
Aggarwal et al.
Ta ble 4. Rea ction betw een ter t-Bu tyl Acr yla te a n d
Ta ble 5. Rea ction of Differ en t Mich a el Accep tor s w ith
Ben za ld eh yd e u n d er Op tim ized Con d ition s
Ben za ld eh yd e in Differ en t Solven ts/Con d ition sa
entry
solvent
acetonitrile
rate (% min-1
)
krel (neat)
stoichiometrica
catalyticb
1
2
3
4
5
6
7
8
9
5 × 10-6
5 × 10-3
0.11
0.23
0.18
0.001
1
22
46
37
timec
(h)
yield
(%)
timec
(h)
yield
(%)
neat
entry
alkene
formamide
formamide (5 equiv)b
water
water/Sc(OTf)3
water/Yb(OTf)3
N-methylformamide
formamide/Ti(OiPr)4
formamide/Al(OiPr)3
formamide/Yb(OTf)3
formamide/La(OTf)3
1
2
3
4
5
6
7
methyl acrylate
ethyl acrylate
tert-butyl acrylate
acrylonitrile
6
3.75
14
4
74
80
96
95
10
63
70
17
17
48 (72) 59 (71)
78
84
c
0.059
12
4
97
c
methyl vinyl ketoned 10 min
0.022
0.31
0.32
0.33
0.24
4
62
64
66
48
2-cyclohexen-1-one
2-cyclopenten-1-one
24
24
96
24
63
10
d
d
10
11
12
d
a
Reaction conditions: 1.2 mmol of Michael acceptor, 1.0 mmol
of PhCHO, 5.0 mmol of H2NCHO, 1.0 mmol of 3-hydroxyquinu-
clidine, and 0.05 mmol of Yb(OTf)3. b 1.2 mmol of Michael acceptor,
1.0 mmol of PhCHO, 1.25 mmol of H2NCHO, and 0.1 mmol of
3-hydroxyquinuclidine were stirred at rt under an argon atmo-
sphere. c No further increase in yield after the times reported.
d
a
Reaction conditions: 2.0 mmol of tert-butyl acrylate, 1.0 mmol
of PhCHO, 1.0 mmol of 3-hydroxyquinuclidine, and 1.0 mL of
solvent. 2.0 mmol of tert-butyl acrylate, 1.0 mmol of PhCHO, 1.0
mmol of 3-hydroxyquinuclidine, and 5.0 mmol (220 µL) of forma-
mide. c 2.0 mmol of tert-butyl acrylate, 1.0 mmol of PhCHO, 1.0
mmol of 3-hydroxyquinuclidine, 1 mL of water, and 0.05 mmol of
Lewis acid were stirred at rt. 2.0 mmol of tert-butyl acrylate, 1.0
mmol of PhCHO, 1.0 mmol of 3-hydroxyquinuclidine, 5.0 mmol of
formamide, and 0.05 mmol of Lewis acid were stirred at rt.
b
d
Rapid decomposition of methyl vinyl ketone took place under
reaction conditions.
d
Ta ble 6. Rea ction of Differ en t Ca r bon yl Com p ou n d s
w ith Eth yl Acr yla te u n d er Op tim ized Con d ition sa ,b
stoichiometrica
catalyticb
compatible with these conditions, and this limitation
prompted us to investigate other polar solvents (Table
4).
timec
(h)
yield
(%)
timec
(h)
yield
(%)
entry
carbonyl compound
1
2
3
4
5
6
7
8
o-chlorobenzaldehyde
o-anisaldehyde
p-anisaldehyde
o-nitrobenzaldehyde
isobutyraldehyde
cyclohexylcarboxaldehyde
trifluoroacetophenone
pivaldehyde
4
12
28
2
72
72
20
8 d
99
90
18
95
71
88
37
31
6
99
96 (120) 66 (75)
96
2
42
42
72
We were particularly interested in the use of forma-
mide as, like water but to a lesser extent, it is known to
contribute to hydrophobic acceleration7 and because it
has an even higher dielectric constant (ꢀr ) 111 vs 78.5
for water)16 will be better able to stabilize the charged
intermediate. A further advantage of using formamide
was the fact that there was no possibility of ester
hydrolysis. We were pleased to find that the use of
formamide did indeed provide an acceleration over reac-
tions conducted neat (Table 4, entries 2 and 3). N-
Methylformamide, which has an even higher dielectric
constant than formamide (ꢀr ) 182 vs 111),16 was also
tested but gave a lower17 rate indicating that the rate
acceleration is not dominated by the solvents ability to
solvate/stabilize the charged intermediates.
It seems that the origin of the rate acceleration
observed with water, formamide, and N-methylforma-
mide is dominated by hydrogen bonding but with smaller
contributions from hydrophobic effects (with water) and
solvent polarity. The small increase in rate observed
when salts (GnCl or LiCl) are added to water could be
due to the solvation of the cation, which will increase the
hydrogen bond donor ability of water.
Further studies on concentration revealed that 5 equiv
of formamide was optimum (Table 4, entry 4) and
provided a rate acceleration that was similar to water.
Similar experiments on concentration with water showed
that the rate increased up to 0.2 M (Table 4, entry 5)
and then plateaued.18 For comparison, rates of reaction
conducted neat and in acetonitrile (2 M) are given (Table
4, entries 1 and 2) and serve to demonstrate the high
level of acceleration that can be achieved in the polar
protic solvents (>104 fold increase in rate). Further
30
95
21
42
46
a
Reaction conditions: 1.2 mmol of ethyl acrylate, 1.0 mmol of
carbonyl compound, 5.0 mmol of H2NCHO, 1.0 mmol of 3-hy-
doxyquinuclidine, and 0.05 mol of Yb(OTf)3. 1.2 mmol of ethyl
acrylate, 1.0 mmol of carbonyl compound, 1.25 mmol of H2NCHO,
and 0.1 mmol of 3-hydroxyquinuclidine were stirred at rt under
an argon atmosphere. c No further increase in yield after the times
reported.
b
studies with the optimized concentration of formamide
revealed that additional acceleration could be achieved
using Lewis acid (Table 4, entries 9-12).2 We believe that
formamide coordinates to the Lewis acid, which results
in even more polarized NH bonds leading to increased
hydrogen bond donor ability and therefore increased
rates. These new conditions (Table 4, entry 11) provided
significantly higher rates than using water alone. In
contrast to the use of formamide, no additional rate
acceleration was observed with Sc(OTf)3 or Yb(OTf)3
(Table 4, entries 6 and 7) in water. As these conditions
are now the best to date, we tested a range of different
Michael acceptors with benzaldehyde (Table 5) and a
range of different electrophiles with ethyl acrylate (Table
6) using both stoichiometric and catalytic quantities of
3-hydroxyquinuclidine.
All the acrylates and acrylonitrile reacted efficiently
(Table 5, entries 1-4) under both stoichiometric and
catalytic conditions. Less reactive enones coupled smoothly
(Table 5, entries 6 and 7) but more reactive enones led
to polymerization (Table 5, entry 5). This is typical of
reactive enones, which are usually reacted in the pres-
ence of solvent to slow the reaction to prevent polymer-
ization. Indeed, with reactive enones more conventional
catalyst systems can be employed for efficient coupling.
Ethyl acrylate coupled efficiently with most aldehydes
(Table 6). Particularly noteworthy is the unusually high
reactivity of o-anisaldehyde relative to p-anisaldehyde.
(16) Solvents and Solvent Effects in Organic Chemistry, 2nd ed.;
Reichardt, C., Ed.; VCH Verlagsgesellschaft: Weinheim, 1988.
(17) These results parallel Auge´’s observations who observed that
rate of Baylis-Hillman reaction between acrylonitrile and benzalde-
hyde in different solvents followed the order H2O ≈ H2NCHO >
MeNHCHO (ref 9).
(18) Reactants are not soluble in water, which is why there is no
peak in the rate of reaction with amount of water used. The reaction
is zero order in water (ref 9).

Baylis-Hillman Reaction in Polar Solvents
J . Org. Chem., Vol. 67, No. 2, 2002 513
droxyquinuclidine (127 mg, 1.0 mmol) were reacted in water
(1.0 mL). Purification by column chromatography using pe-
troleum ether/diethyl ether (2:1) as eluant gave the adduct as
a colorless oil (120 mg, 40%): Rf 0.27 (petroleum ether/diethyl
ether, 2:1); IR (film) 3420, 2980, 1720 cm-1; 1H NMR (250 MHz;
CDCl3) 1.3 (s, 9H), 5.5 (s, 1H), 5.95 (s, 1H), 6.45 (s, 1H), 7.25-
7.4 (m, 3H), 7.5-7.65 (m, 2H); 13C NMR (101 MHz, CDCl3)
27.7, 79.0 (q, J CF ) 29 Hz), 83.5, 124.5 (q, J CF ) 280 Hz),
126.9, 127.8, 128.2, 128.8, 129.1, 130.1, 135.5, 137.4, 166.2;
MS (EI) (m/z) 302 (M+, 12), 177 (100), 159 (59); HRMS (EI)
(m/z) calcd for C15H17O3F3 302.1130, found 302.1125.
2-[Hyd r oxy(2-ch lor op h en yl)m eth yl]a cr ylic Acid ter t-
Bu tyl Ester . tert-Butyl acrylate (256 mg, 2.0 mmol), o-
chlorobenzaldehyde (141 mg, 1.0 mmol), and 3-hydroxyquinu-
clidine (127 mg, 1.0 mmol) were reacted in water (1.0 mL).
Purification by column chromatography using 5% ethyl acetate
in hexane as eluant gave the adduct as a yellow oil (143 mg,
53%): Rf 0.32 (10% ethyl acetate in hexane); IR (film) 3426,
2933, 1710 cm-1; 1H NMR (250 MHz, CDCl3) 1.45 (s, 9H), 3.2
(d, J ) 4.5 Hz, 1H), 5.54 (t, J ) 1.1 Hz, 1H), 5.95 (bd, J ) 4.5
Hz, 1H), 6.26 (d, J ) 1.1 Hz, 1H), 7.2-7.4 (m, 4H); 13C NMR
(63 MHz, CDCl3) 28.0, 69.5, 87.1, 125.7, 127.0, 128.1, 129.0,
129.4, 132.9, 136.8, 138.7, 165.8; MS (EI) (m/z) 268 (M+, 6),
154 (76), 139 (74), 57 (100). Anal. Calcd for C14H17ClO3: C,
62.6; H 6.4. Found: C, 62.3; H 6.4.
This is presumably due to the superior ability of the ortho
isomer to bind the Lewis acid and thus activate the
aldehyde. Aliphatic aldehydes (Table 6, entries 5 and 6)
coupled efficiently under stoichiometric conditions, but
at the longer reaction times required under the catalytic
conditions, product decomposition was observed. Acti-
vated ketones worked moderately well, but simple ke-
tones (acetone) were not effective (Table 6, entry 7).
Pivaldehyde (Table 6, entry 8) worked moderately well,
and indeed, this is only the second use of pivaldehyde in
the Baylis-Hillman reaction with any Michael acceptor.5
In conclusion, we have developed new conditions for
accelerating the Baylis-Hillman reaction, which we
believe are now the best to date. These improved condi-
tions employ either water as solvent or small amounts
of formamide and Yb(OTf)3. Using the formamide and
Yb(OTf)3, rate enhancements of up to 120-fold have been
achieved in the coupling of methyl acrylate with benzal-
dehyde over standard DABCO-catalyzed reaction. These
new conditions will be particularly appropriate for solid-
phase Baylis-Hillman reactions of acrylates as high
rates without premature cleavage of the ester from the
resin should be achieved. Hindered and deactivated
aldehydes can now be coupled efficiently, and attempts
to encourage unactivated ketones to couple is underway.
2
1
2-Hyd r oxy-3-m eth ylen esu ccin ic Acid 4-ter t-Bu tyl Es-
ter . tert-Butyl acrylate (256 mg, 2.0 mmol), glyoxylic acid
monohydrate (92 mg, 1.0 mmol), triethylamine (101 mg, 1.0
mmol), and 3-hydroxyquinuclidine (127 mg, 1.0 mmol) were
reacted in water (1.0 mL) to give the adduct as a colorless oil
(114 mg, 56%): IR (film) 3580-2350, 2980, 1721 cm-1 1H
;
Exp er im en ta l Section
NMR (250 MHz, CDCl3) 1.45 (s, 9H), 3.1 (bs, 1H), 4.8 (s, 1H),
5.8 (s, 1H), 6.2 (s, 1H), 7.5 (bs, 1H); 13C NMR (63 MHz, CDCl3)
27.9, 70.2, 83.6, 128.9, 137.9, 166.8, 174.2; MS (CI) (m/z) 220
(MNH4+, 83), 203 (MH+, 12), 164 (100), 147 (34); HRMS (CI)
(m/z) calcd for C9H15O5 203.0919, found 203.0919.
Rea gen ts. Methyl, ethyl, and tert-butyl acrylate were used
as purchased. Cyclohexenone, cyclopentenone, formamide, and
all aldehydes were distilled prior to use. DABCO was purified
by sublimation under vacuum prior to use. DMAP and 3-hy-
droxyquinuclidine were used as purchased without further
purification. All Lewis acids (except Ti(OiPr)4) were stored in
a vacuum desiccator and used without further purification.
Ti(OiPr)4 was distilled under reduced pressure prior to use.
Gen er a l P r oced u r e for th e Ba ylis-Hillm a n Rea ction
(Ta bles 1 a n d 2). To a stirred mixture of 2-cyclohexen-1-one
(2.0 mmol) and carbonyl compound (1.0 mmol) in water (1.0
mL) at room temperature was added the amine catalyst (1.0
mmol). The reaction was stopped by dilution with diethyl ether
and washed with 2 M HCl, followed by water. After drying
over sodium sulfate, filtration, and evaporation, the crude
Mon itor in g of Wa ter -Ba sed Rea ction s (Ta ble 4). To a
stirred mixture of tert-butyl acrylate (256 mg, 2.0 mmol),
benzaldehyde (106 mg, 1.0 mmol), and water (1.0 mL) at room
temperature was added 3-hydroxyquinuclidine (127 mg, 1.0
mmol). At selected time points, an aliquot (yield between 0
and 10%) of the reaction mixture was quenched with 2 M HCl/
brine (1:1, 0.5 mL), extracted with CDCl3 (0.5 mL), passed
through a hydrophobic filter (Whatman IPS filter media
1
catalog no. 6987-1299), and analyzed by H NMR.
Mon itor in g of F or m a m id e-Ba sed Rea ction s (Ta ble 4,
En tr y 11). To a stirred mixture of tert-butyl acrylate (256 mg,
2.0 mmol), benzaldehyde (106 mg, 1.0 mmol), and formamide
(225 mg, 5.0 mmol) at room temperature were added ytterbium
triflate (32 mg, 0.05 mmol) and 3-hydroxyquinuclidine (127
mg, 1.0 mmol). At selected time points (yield between 0 and
10%), an aliquot of the reaction mixture was removed, diluted
1
mixture was analyzed by H NMR.
Gen er a l P r oced u r e for th e Ba ylis-Hillm a n Rea ction
(Ta ble 3). To a stirred mixture of tert-butyl acrylate (256 mg,
2.0 mmol) and carbonyl compound (1.0 mmol) in water (1.0
mL) at room temperature was added 3-hydroxyquinuclidine
(127 mg, 1.0 mmol). The reaction was stopped by dilution with
diethyl ether and washed with 2 M HCl, followed by water.
After drying over sodium sulfate, filtration, and evaporation,
the crude mixture was purified by column chromatography.
2-[Hyd r oxy(2-n itr op h en yl)m eth yl]a cr ylic Acid ter t-
Bu tyl Ester . tert-Butyl acrylate (256 mg, 2.0 mmol), o-
nitrobenzaldehyde (151 mg, 1.0 mmol), and 3-hydroxyquinu-
clidine (127 mg, 1.0 mmol) were reacted in water (1.0 mL).
The mixture was purified by column chromatography using
petroleum ether/diethyl ether (1:1) as eluant to produce the
adduct as a yellow oil (233 mg, 80%): Rf 0.3 (petroleum ether/
1
with CDCl3, and analyzed by H NMR.
Gen er a l P r oced u r e for th e Ba ylis-Hillm a n Rea ction
u n d er Op tim u m Ca ta lytic Con d ition s (Ta bles 5 a n d 6).
To a stirred mixture of carbonyl compound (1.0 mmol), Michael
acceptor (1.2 mmol), and formamide (1.25 mmol) at room
temperature under argon was added 3-hydroxyquinuclidine
(0.1 mmol). After the time indicated, the reaction was stopped
by dilution with ether and washed with 2 M HCl, followed by
water. After drying over sodium sulfate, filtration, and evapo-
ration, the crude mixture was purified by column chromatog-
raphy.
Gen er a l P r oced u r e for th e Ba ylis-Hillm a n Rea ction
u n d er Op tim u m Stoich iom etr ic Con d ition s (Ta bles 5
a n d 6). To a stirred mixture of carbonyl compound (1.0 mmol),
Michael acceptor (1.2 mmol), and formamide (5.0 mmol) at
room temperature under argon were added 3-hydroxyquinu-
clidine (1.0 mmol) and Yb(OTf)3 (0.05 mmol). After the time
indicated, the reaction was stopped by dilution with ether and
washed with 2 M HCl, followed by water. After drying over
sodium sulfate, filtration, and evaporation, the crude mixture
was purified by column chromatography.
1
diethyl ether, 1:1); IR (film) 3452, 2932, 1709, 1528 cm-1; H
NMR (250 MHz; CDCl3) 1.3 (s, 9H), 3.7 (bs, 1H), 5.6 (s, 1H),
6.0 (s, 1H), 6.2 (s, 1H), 7.4 (td, J ) 7.5 Hz, 1.5 Hz, 1H), 7.5-
7.65 (m, 2H), 7.85 (d, J ) 7.5 Hz, 1H); 13C NMR (101 MHz,
CDCl3) 27.8, 67.6, 81.7, 124.5, 125.5, 128.6, 128.8, 133.4, 136.6,
+
142.1, 148.5, 165.1; MS (CI) m/z (rel intensity) 297 (MNH4
,
29), 241 (100), 206 (19); HRMS (CI) (m/z) calcd for C14H21O5N2
297.1450, found C14H21O5N2 297.1436. Anal. Calcd for C14H17
NO5: C, 60.2; H 6.1; N 5.0. Found: C, 60.1; H 6.1; N 5.0.
-
2-(2,2,2-Tr iflu or o-1-h ydr oxy-1-ph en yleth yl)acr ylic Acid
ter t-Bu tyl Ester . tert-Butyl acrylate (256 mg, 2.0 mmol),
2,2,2-trifluoroacetophenone (174 mg, 1.0 mmol), and 3-hy-
2-[Hydr oxy(2-m eth oxyph en yl)m eth yl]acr ylic Acid Eth -
yl Ester . Ethyl acrylate (120 mg, 1.2 mmol), o-anisaldehyde

514 J . Org. Chem., Vol. 67, No. 2, 2002
Aggarwal et al.
(136 mg, 1.0 mmol), formamide (225 mg, 5.0 mmol), ytterbium
triflate (31 mg, 0.05 mmol), and 3-hydroxyquinuclidine (127
mg, 1.0 mmol) were reacted. Purification by column chroma-
tography using 10% ethyl acetate in hexane as eluant gave
the adduct as a pale yellow oil (212 mg, 90%): Rf 0.11 (10%
135638-64-1;25 2-[hydroxy(2-cyclohexyl)methyl]acrylic acid eth-
yl ester, 145316-19-4;26 2-[hydroxy(2-tert-butyl)methyl]acrylic
acid ethyl ester, 252756-39-1;27 2-(hydroxyphenylmethyl)-2-
cyclohexen-1-one, 94348-71-7;28 2-[hydroxy(2-methoxyphenyl)-
methyl]-2-cyclohexen-1-one, 254729-46-9;5 2-(hydroxymethyl)-
2-cyclohexen-1-one, 105956-40-9;11a 2-(1-hydroxy-2-methyl-
propyl)cyclohexen-1-one, 94348-72-8;29 2-(1-hydroxy-2,2-di-
methylpropyl)cyclohexen-1-one, 189870-31-3;30 2-(1-hydroxy-
phenylmethyl)cyclopenten-2-one, 122617-89-4;31 3-(hydroxy-
phenylmethyl)but-3-en-2-one, 73255-39-7;32 2-(hydroxyphenyl-
methyl)acrylonitrile, 19362-96-0,33 2-(2,2,2-trifluoro-1-hydroxy-
1-phenylethyl)acrylic acid ethyl ester, 8063518.34
1
ethyl acetate in hexane); IR (film) 3473, 2938, 1712 cm-1; H
NMR (250 MHz, CDCl3) 1.2 (t, J ) 7 Hz, 3H), 3.5 (d, J ) 6
Hz, 1H), 3.75 (s, 3H), 4.1 (q, J ) 7 Hz, 2H), 5.65 (s, 1H), 5.75
(d, J ) 6 Hz, 1H) 6.2 (s, 1H), 6.75-7.0 (m, 2H), 7.15-7.25 (m,
2H); 13C NMR (63 MHz, CDCl3) 14.1, 55.4, 60.8, 68.3, 110.5,
120.7, 125.4, 127.7, 128.9, 129.2, 141.6, 156.6, 166.7; MS (EI)
(m/z) 236 (M+, 36), 190 (33), 135 (100); HRMS (EI) (m/z) calcd
for
C13H16O4 236.1049, found 236.1048. Anal. Calcd for
Ack n ow led gm en t. We thank EPSRC and GSK for
support of this work.
C
13H16O4: C, 66.1; H 6.8. Found: C, 66.2; H 6.8.
Registr y Nu m ber s (P r ovid ed by th e Au th or s). The
following known compounds showed identical data with the
literature: 2-(hydroxyphenylmethyl)acrylic acid methyl ester,
29540-54-3;19 2-(hydroxyphenylmethyl)acrylic acid ethyl ester,
37442-45-8;19 2-(hydroxyphenylmethyl)acrylic acid tert-butyl
ester, 135513-98-3;19 2-[hydroxy(2-nitrophenyl)methyl]acrylic
acid ethyl ester, 198902-99-7;20 2-[hydroxy(4-nitrophenyl)-
methyl]acrylic acid ethyl ester, 88039-47-8;21 2-[hydroxy(2-
chlorophenyl)methyl]acrylic acid ethyl ester, 88039-46-7;22
2-[hydroxy(4-methoxyphenyl)methyl]acrylic acid ethyl ester,
88039-45-6;23 2-hydroxymethylacrylic acid ethyl ester, 121065-
74-5;24 2-[hydroxy(2-isopropyl)methyl]acrylic acid ethyl ester,
Su p p or tin g In for m a tion Ava ila ble: Spectral data for all
new compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
J O016073Y
(25) Adam, W.; Albert, R.; Hasemann, L.; Nava Salgado, V. O.;
Nester, B. J . Org. Chem. 1991, 56, 5782-5785.
(26) Masuyama, Y.; Nimura, Y.; Kurusu, Y. Tetrahedron Lett. 1992,
33, 6477-6478.
(27) Ramachandron, V. P.; Reddy, V. R. M.; Rudd, M. T. Chem.
Commun. 1999, 1979-1980.
(28) Kataoka, T.; Iwama, T.; Tsujiyama, S.; Iwamura, T.; Watanabe,
S. Tetrahedron 1998, 54, 11813-11824.
(19) Fort, Y.; Berthe, M. C.; Caubere, P. Tetrahedron 1992, 48,
6371-6384.
(29) Hwu, J . R.; Hakimelahi, G. H.; Chou, C.-T. Tetrahedron Lett.
1992, 33, 6469-6472.
(20) Familoni, O. B.; Kaye, P. T.; Klaas, P. J . Chem. Commun. 1998,
2563-2564.
(30) Bachki, A.; Forbela, F.; Yus, M. Tetrahedron 1997, 53, 4921-
4934.
(21) Brand, M.; Drewes, S. E.; Loizou, G.; Roos, G. H. P. Synth.
Commun. 1987, 17, 795-802.
(31) Kusuda, S.; Veno, Y.; Turu, T. Bull. Chem. Soc. J pn. 1993, 66,
2720-2724.
(22) McFadden, H. G.; Harris, R. L. N.; J enkins, C. L. D. Aust. J .
Chem. 1989, 42, 301-314.
(32) Basavaiah, D.; Gowriwari, V. V. L.; Dharma-Rao, P.; Bharthi,
T. K. J . Chem. Res. Miniprint 1995, 7, 1656-1673.
(33) Hill, J . S.; Isaacs, N. S. J . Chem. Res., Miniprint 1988, 10,
2641-2676.
(23) Ameer, F.; Drewes, S. E.; Emslie, N. D.; Kaye, P. T.; Mann, R.
L. J . Chem. Soc., Perkin Trans. 1 1983, 10, 2293-2295.
(24) Basavaiah, D.; Krishnamacharyulu, M.; Rao, A. J . Synth.
Commun. 2000, 30, 2061-2070.
(34) Ramachandron, V. P.; Reddy, V. R. M.; Rudd, M. T.; Alaniz, J .
R. Tetrahedron Lett. 1998, 39, 8791-8794.