Acceleration of the Baylis-Hillman reaction in the presence of ionic liquids

Article abstract of DOI:10.1002/hlca.200390087

The Baylis-Hillman reaction is accelerated in the presence of ionic liquids. Of various 1-butyl-3-methylimidazolium (bmim)-based ionic liquids tested, [bmim][PF₆] has been found to result in the highest rate increase. In the company of Lewis ac

Full text of DOI:10.1002/hlca.200390087

894
Helvetica Chimica Acta Vol. 86 (2003)
Acceleration of the Baylis Hillman Reaction in the Presence
of Ionic Liquids
by Eun Jin Kim and Soo Y. Ko*
Department of Chemistry and Division of Molecular and Life Sciences, Ewha Womans University,
Seoul 120-750, Korea (phone: (‡ 822) 3277-3283; fax: (‡ 822)3277-2384; e-mail: sooyko@mm.ewha.ac.kr)
and
Choong Eui Song
Life Sciences Division, Korea Institute of Science and Technology, Seoul 130-650, Korea
The Baylis Hillman reaction is accelerated in the presence of ionic liquids. Of various 1-butyl-3-
methylimidazolium (bmim)-based ionic liquids tested, [bmim][PF6] has been found to result in the highest rate
increase. In the company of Lewis acid and H-bond-donor additives, the reaction rates further improve, albeit
only modestly. A preparatively useful Baylis Hillman procedure prescribes the use of [bmim][PF6] with
La(OTf)3 and 2,2'2''-nitrilotris[ethanol], in which the net effect of the ionic liquid is to bring about a more than
twofold rate increase over the otherwise same reaction in MeCN.
Introduction. The Baylis Hillman reaction, a coupling of activated alkenes with
aldehydes promoted by tertiary amines, is emerging as a useful CÀC bond-forming
reaction (for reviews, see [1]). The adduct possesses rich functionalities, including an
allylic alcohol and an activated alkene, that can be utilized in further transformations. A
major drawback of the Baylis Hillman reaction is the low reaction rate; reaction times
as long as several weeks are sometimes required. To overcome this problem, a number
of additives have been reported to speed up the reaction, including Lewis acids [2],
tertiary phosphines [3], quaternary ammonium salts [4], H-bond donors [5], and
combinations thereof. In addition, high concentrations (often without solvent) and
pressure [6], and ultrasound [7] and microwave irradiations [8] have been tried.
Room-temperature ionic liquids are gaining much attention in organic chemistry as
reaction solvents and catalysts (for reviews, see [9]). The advantages of using ionic
liquids are diverse. Due to their low volatility, they are regarded as environmentally
friendly −green solvents×. Product isolation or catalyst recycling can be very easy in ionic
liquids [10]. In some cases, great rate accelerations or selectivity improvements are
observed [11]. An example of the reactions wherein this last advantage of using ionic
liquids has been realized is a Sc(OTf)3-catalyzed Diels Alder reaction [11c]. Noting
that Sc(OTf)3 was also one of the effective additives for accelerating the Baylis
Hillman reaction [2b], we speculated that the Lewis acid-catalyzed Baylis Hillman
reaction might be further accelerated in the presence of ionic liquids. This led us to
investigate the effects of ionic liquids on the rate of Baylis Hillman reaction. In a
recent publication, a rate-accelerating effect of ionic liquids in Baylis Hillman
reactions has been reported [12]. Disclosed herein are our own findings.

Helvetica Chimica Acta Vol. 86 (2003)
895
Results and Discussion. With methyl acrylate, benzaldehyde, and 1,4-diazabicy-
clo[2,2,2]octane (DABCO) (1:1:1 molar ratio) as the standard components for the
Baylis Hillman reaction, a series of ionic liquids were screened first without any other
additives in order to assess the rate-enhancing abilities of ionic liquids alone (Scheme).
‡
Six ionic liquids of the general structure ([bmim][X], [bmim] ˆ 1-butyl-3-methyl-
imidazolium; [X]À ˆ OAc, OTf, N(Tf)2, BF4, SbF6, PF6) were used. To achieve a
meaningful comparison with the reported data, each reaction was conducted under the
set of conditions resembling, as closely as possible, those in the literature. Thus, when
an ionic liquid was added to a reaction, the amount of solvent (MeCN) was reduced
accordingly so that the concentration would be maintained at a constant level1). Also,
the amount of ionic liquid added to each reaction was set constant by volume2). Each
reaction was monitored by GC during the first 50min, and the rate was then compared
against the standard Baylis Hillman reaction performed in MeCN at the same
concentration. The results are summarized in Table 1. Most of the ionic liquids studied
brought about modest (6 7 times) rate accelerations. Two of them, [bmim][OTf] and
[bmim][PF6], seemed more effective than the rest (ca. 9 10times rate accelerations;
Entries 6 and 7)3). These two ionic liquids were, thus, selected for further studies.
Scheme
Next, we evaluated the rate-enhancing effects of the selected ionic liquids in the
presence of a series of additives, including Sc(OTf)3, La(OTf)3, and 2,2',2''-nitrilo-
tris[ethanol] (Table 2)4). These additives have been reported to be effective in
accelerating the Baylis Hillman reaction under various reaction conditions [2b], and
we confirmed their efficacy when the coupling reactions were performed in MeCN
alone (Entries 2 6). It is interesting to note that, while each of these additives brought
ca. 4 15 times rate increase, a combination of the best two, La(OTf)3 and 2,2'2''-
nitrilotris[ethanol], resulted in only a modest further rate increase and far below the
1
)
)
The combined volume of MeCN and a given ionic liquid was 20 ml per mmol of each reactant.
The amount of ionic liquid was set initially at 5.0mol-% in the case of [bmim][PF 6], which corresponded to
10 ml of the ionic liquid per mmol of each reactant (Entry 6 in Table 1). For other ionic liquids, this volume
was equivalent to 3.4 6.0mol-% (see Table 1).
The results with [bmim][BF4] and [bmim][PF6] are qualitatively in agreement with those reported in [12];
direct comparison is not meaningful as the two studies have been conducted under different sets of
conditions from each other.
2
3
4
)
)
The amount of each additive was adopted from a literature report [2b] as having given the maximum rate
acceleration.

896
Helvetica Chimica Acta Vol. 86 (2003)
Table 1. The Effects of Ionic Liquids (IL) on the Baylis Hillman Reaction Ratea)
Entry
IL
Amount of IL
Relative rate
1
2
3
4
5
6
7
1
b
[bmim][BF4]
[bmim][SbF6]
[bmim][OAc]
[bmim][N(Tf)2]
[bmim][PF6]
[bmim][OTf]
6.0mol-%
)
6.4
6.4
6.9
6.9
8.6
10.1
4.6 mol-%b)
5.5 mol-%b)
3.4 mol-%b)
b
5.0mol-%
)
4.5 mol-%b)
a) The reaction was performed with 5 mmol each of methyl acrylate, PhCHO, and 1,4-diazabicyclo[2.2.2]octane
(DABCO) in MeCN. The combined volume of MeCN and ionic liquid was 100 ml in each case. b) 50 ml of a given
ionic liquid for 5 mmol of substrate.
a
Table 2. The Combined Effects of Ionic Liquids (IL) and Additives on the Baylis Hillman Reaction Rate
)
Entry
ILb)
Additives
-
Relative rate
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
4
12
14
9
17
10.1
8
12
14
17
8.6
8
c
Sc(OTf)3
La(OTf)3
)
)
c
2,2',2''-nitrilotris[ethanol]d)
Sc(OTf)3, 2,2',2''-nitrilotris[ethanol]e)
La(OTf)3, 2,2',2''-nitrilotris[ethanol]e)
[bmim][OTf]f)
[bmim][OTf]f)
[bmim][OTf]f)
[bmim][OTf]f)
[bmim][OTf]f)
[bmim][PF6]g)
[bmim][PF6]g)
[bmim][PF6]g)
[bmim][PF6]g)
[bmim][PF6]g)
c
Sc(OTf)3
La(OTf)3
)
)
c
2,2',2''-nitrilotris[ethanol]d)
La(OTf)3, 2,2',2''-nitrilotris[ethanol]e)
c
Sc(OTf)3
La(OTf)3
)
)
c
19
22
25
2,2',2''-nitrilotris[ethanol]d)
La(OTf)3, 2,2',2''-nitrilotris[ethanol]e)
a) The reaction was performed with 5 mmol each of methyl acrylate, PhCHO, and DABCO in MeCN. The
combined volume of MeCN and ionic liquid was 100 ml in each case. b) 50 ml of a given ionic liquid for 5 mmol of
substrate (for Entries 7 16). c) 5 mol-%. d) 80mol-%. e) 5 mol-% Lewis acid 50mol-%. f) 4.5 mol-%.
g) 5.0mol-%.
level that would be expected if the Lewis acid and H-bond donor behaved
independently (Entry 6). These findings are consistent with the reported results [2b].
Then, in the presence of a selected ionic liquid, the Baylis Hillman reactions were
performed with these additives. Sc(OTf)3, if anything, slowed the reaction when
combined with [bmim][OTf] (Entry 8 vs. 7). This is in contrast to what has been
observed when the same reactions were performed in MeCN without any ionic liquid
(Entry 2 vs. 1) [2b], but consistent with the results reported for ionic liquids [12].
Unlike Sc(OTf)3, and contrary to the published results [12], La(OTf)3, 2,2'2''-
nitrilotris[ethanol], and the combination thereof did bring about the Baylis Hillman
rate enhancements in the company of [bmim][OTf], albeit only slightly (1.2 1.7 times,
Entries 9 11 vs. 7), and in much lower ratios than observed in MeCN without any ionic

Helvetica Chimica Acta Vol. 86 (2003)
897
liquid (12 17 times, Entries 3 6 vs. 1). Thus, it appears that the rate-enhancing effects
of [bmim][OTf] and of Lewis acid/H-bond donors are not quantitatively additive; in
fact, the effect of the ionic liquid seems virtually wiped out when the Baylis Hillman
reaction is conducted in the presence of Lewis acid or H-bond-donor additives
(Entries 9 11 vs. 3 6).
When [bmim][PF6] was present in the reaction, the effects of the other additives
generally followed the pattern observed in the previous systems, but the magnitude was
different. Thus, Sc(OTf)3 now had little effect on the Baylis Hillman reaction rate
(Entry 13 vs. 12), while La(OTf)3 and 2,2'2''-nitrilotris[ethanol] accelerated the raction
(Entries 14 15 vs. 12). The effects of the latter two additives were greater with
[bmim][PF6] (2.2 2.6 times rate increases) than in the previous system with
[bmim][OTf] (1.2 1.4 times, Entries 9 and 10), but still fell short of the proportions
observed when the reactions were performed in MeCN without any ionic liquid (12
14 times, Entries 3 and 4). A combination of La(OTf)3 and 2,2'2''-nitrilotris[ethanol]
produced a modest further rate increase in the presence of [bmim][PF6] (Entry 16).
The rate-accelerating effects of La(OTf)3, 2,2'2''-nitrilotris[ethanol], and [bmim][PF6]
are thus combined to result in a 25-fold rate increase from the standard Baylis Hill-
man reaction of methyl acrylate, PhCHO, and DABCO performed without any
additives in MeCN at the same concentration.
Having established the rate-accelerating effects of [bmim][PF6] in the company of
La(OTf)3 and 2,2'2''-nitrilotris[ethanol], we turned our attention to finding the
optimum amount of the ionic liquid to effect the maximum Baylis Hillman rate
increase, bearing in mind the possibility of using it as the reaction solvent (Table 3). In
the reactions described so far, a catalytic amount (5.0mol-%) of [bmim][PF ] had been
6
employed (Entry 1). At the very high concentration that these reactions had been
performed (20 ml of MeCN ‡ [bmim][PF6] per mmol of each reactant), this meant the
reactions having been run in a 1:1 (v/v) mixture of the solvent and ionic liquid5). The
amount of [bmim][PF6] could not be increased while maintaining the concentration
constant, as the reaction mixture became heterogeneous when the ionic liquid
outbalanced MeCN at that concentration. When, however, the concentration
constraint was lifted, and the amount of the ionic liquid was increased beyond 20 ml
Table 3. The Effects of the Ionic Liquid (IL) Amounts on the Baylis Hillman Reaction Ratea)
Entry
IL/Solvent (composition v/v)
Amount of IL
Concentrationb)
Relative ratec)
1
2
3
4
5
6
[bmim][PF6]/MeCN (1 : 1)
[bmim][PF6]
[bmim][PF6]/MeCN (9 : 1)
[bmim][PF6]
[bmim][PF6]
MeCN
5 mol-%
50mol-%
45 mol-%
200 mol-%
600 mol-%
20 ml/mmol
100 ml/mmol
100 ml/mmol
400 ml/mmol
1200 ml/mmol
100 ml/mmol
25
26
19
17
2
12
a) Each reaction was performed with 5 mmol each of methyl acrylate, PhCHO, and DABCO in the presence of
La(OTf)3 (5 mol-%) and 2,2',2''-nitrilotris[ethanol] (50mol-%). b) Combined volume of [bmim][PF6] and
MeCN per mmol of methyl acrylate. c) % Conversion/min. The rate of the standard Baylis Hillman reaction in
MeCH in the absence of any additives (Entry 1, Table 1) is calculated to be 1.
5
)
See Footnote 2.

898
Helvetica Chimica Acta Vol. 86 (2003)
per mmol of each reactant (in consequence, the overall concentration was no longer
constant), the reaction mixture became homogeneous. Thus, with 50mol-% of
[bmim][PF6] (100 ml per mmol reactant), the entire mixture, containing La(OTf)3 and
2,2',2''-nitrilotris[ethanol] additives as well as methyl acrylate/PhCHO/DABCO,
remained homogeneous without addition of MeCN (Entry 2). The rate observed from
this reaction was, despite the apparent dilution, comparable to that at a higher
concentration with a smaller (a tenth) amount of the ionic liquid (i.e., in a 1:1 (v/v)
mixture of MeCN and [bmim][PF6]; Entry 1). When a fraction of the ionic liquid was
replaced by MeCN while maintaining the same dilute concentration (100 ml per mmol),
a decrease in the rate was noted (Entry 3), confirming the positive effects of the ionic
liquid on the Baylis Hillman rate. When the amount of [bmim][PF6] was further
increased so as to be regarded as bona fide solvent, a gradual drop in the rate was
observed, probably due to dilution (Entries 4 and 5). Thus, the highest rate increase
observed so far in our study was when the DABCO-promoted Baylis Hillman reaction
was performed in the presence of La(OTf)3 and 2,2',2''-nitrilotris[ethanol] in
[bmim][PF6] (100 ml per mmol). The net effect of the ionic liquid under these
conditions was to bring about more than two-fold rate increase over the otherwise same
reaction in MeCN (Entry 2 vs. 6)6).
The results so far obtained pointed toward interesting, but also perplexing, effects
of ionic liquids in the Baylis Hillman reaction rate. Whereas many ionic liquids
accelerated the uncatalyzed Baylis Hillman reaction significantly, the effect was
diminished in the company of Lewis acid/H-bond donor additives almost to the point of
complete elimination in some cases. The incompatible effects of ionic liquids/Lewis
acid/H-bond donors will be a subject for further studies, but may lead to speculations as
to the roles of ionic liquids in accelerating the Baylis Hillman reactions. While a
favorable shift in the Baylis Hillman equilibria has been suggested to be caused by
ionic liquids [12], the incoherent results obtained by the combinations of ionic liquids/
Lewis acid/H-bond donor may hint a weakly Lewis acidic role of ionic liquids, which
compete unfavorably against more conventional catalysts in the Baylis Hillman reaction7).
From a synthetic point of view, the rate-enhancing effects of [bmim][PF6] are
translated into a practically useful Baylis Hillman procedure. In particular, the
conditions of Entry 2, Table 3, allow a use of organic liquid (up to 100 ml [bmim][PF6]
per mmol substrate, which is a relatively large amount for the Baylis Hillman process),
while still producing the highest reaction rate reported at this concentration8). This
may prove useful for Baylis Hillman reactions with solid reactants wherein even
such a small amount of organic liquid will help maintain the reaction mixture
homogeneous.
In conclusion, a catalytic amount of ionic liquids has been found to accelerate the
Baylis Hillman reaction. In the company of Lewis acid and H-bond donor additives,
6
)
Under these conditions, the combined effects of La(OTf)3, 2,2'2''-nitrilotris[ethanol], and [bmim][PF6]
amount to a 43-fold rate increase over a DABCO-promoted, uncatalyzed Baylis Hillman reaction in
MeCN.
7
8
)
A similar, weakly Lewis acidic role of ionic liquids has been suggested for Diels Alder reaction in ionic
liquids (see [13]).
Under these conditions, the Baylis Hillman adduct was produced in 63% yield after 14 h (see Exper.
Part). With acrylonitrile as the activated olefin component, the yield was 80%.
)

Helvetica Chimica Acta Vol. 86 (2003)
899
the DABCO-promoted reaction rates further improve, albeit only modestly. A
preparatively useful Baylis Hillman procedure prescribes a use of [bmim][PF6] with
La(OTf)3 and 2,2',2''-nitrilotris[ethanol].
Experimental Part
General Procedure for the Rate Determination of the Baylis Hillman Reaction. A mixture of methyl
acrylate, PhCHO, and DABCO (5 mmol each) was dissolved in solvent as specified. The additives were added as
specified, and the mixture was stirred under N2. At 10-min intervals for the initial 50 min, a 100 -ml aliquot of the
reaction mixture was taken out and placed on a short pad of silica, which was then washed with AcOEt. The
filtrate was evaporated. The residue was combined with a sample of accurately weighed GC standard (dodecane,
10mg). The mixture was analyzed on GC, and the conversion was calculated based on the data generated from
authentic pure samples.
The Baylis Hillman Reaction Rrocedure in the Presence of Ionic Liquid: PhCHO (0.533 g, 5.0 mmol) was
added to [bmim][PF6] (500 ml). DABCO (0.56 g, 5.0 mmol), La(OTf)3 (0.15 g, 0.25 mmol), and 2,2',2''-
nitrilotris[ethanol] (0.33 ml, 2.5 mmol) were successively added. Finally, methyl acrylate (0.45 ml, 5.0 mmol)
was added, and the mixture was stirred at r. t. for 14 h. It was diluted with AcOEt and washed with 2% aq. HCl,
then with H2O. The aq. phase was extracted with portions of AcOEt. The combined org. phases were dried
(Na2SO4) and concentrated. Flash-chromatographic purification (hexane/AcOEt 3 :1) yielded the Baylis
Hillman adduct methyl 3-hydroxy-2-methylidene-3-phenylpropanoate (1; 0.60 g, 63%). IR 3465 (br.), 3060m,
2955s, 2312s, 1700m, 1631s, 1494s, 1441m, 1268m, 1151m, 1043m. 1H-NMR (CDCl3): 3.03 (br., 1 H), 3.73 (s,
‡
3 H), 5.51 (s, 1 H), 5.84 (s, 1 H), 6.34 (s, 1 H), 7.28 7.47 (m, 5 H). MS: 192 (M ).
This work was supported by the KOSEF (R01-2002-000-00274-0). E. J. K. is grateful to the Brain Korea 21
project for a graduate fellowship.
REFERENCES
[1] a) E. Ciganek, Org. React. 1997, 51, 201; b) D. Basavaiah, P. D. Rao, R. S. Hyma, Tetrahedron 1996, 52, 80 0 1.
[2] a) V. K. Aggarwal, G. J. Tarver, R. McCagur, Chem. Commun. 1996, 2713; b) V. K. Aggarwal, A. Mereu,
G. J. Tarver, R. McCagur, J. Org. Chem. 1998, 63, 7183; c) T. Kataoka, T. Iwama, S. Tsuhiyama, Chem.
Commun. 1998, 197.
¬
[3] a) F. Roth, P. Gygax, G. Frater, Tetrahedron Lett. 1992, 33, 1045; b) S. Rafel, J. W. Leahy, J. Org. Chem.
1997, 62, 1521.
[4] M. Shi, Y.-S. Feng, J. Org. Chem. 2001, 66, 406.
[5] a) F. Ameer, S. E. Drews, S. Freese, P. T. Kaye, Synth. Commun. 1988, 18, 495; b) S. E. Drews, S. D. Freese,
N. D. Emslie, G. H. P. Roos, Synth. Commun. 1988, 18, 1565; c) D. Basavaiah, P. K. S. Sarma, Synth.
Commun. 1990, 20, 1611; d) Y. M. A. Yamada, S. Ikegami, Tetrahedron Lett. 2000, 41, 2165.
[6] a) J. S. Hill, N. S. Issacs, Tetrahedron Lett. 1986, 27, 5007; b) J. S. Hill, N. S. Issacs, J. Chem. Res. 1988, 330.
[7] G. H. P. Roos, P. Ramoersadh, Synth. Commun. 1993, 23, 1261.
[8] M. K. Kundu, S. B. Mukherjee, N. Balu, R. Padmakumar, S. V. Bhat, Synlett 1994, 444.
[9] a) J. S. Wilkes, Green Chemistry 2002, 4, 73; b) R. Sheldon, Chem. Commun. 2001, 2399; c) P. Wasserscheid,
W. Keim, Angew. Chem., Int. Ed. 2000, 39, 3772; d) T. Welton, Chem. Rev. 1999, 99, 2071; e) Y. Chauvin, H.
Olivier, CHEMTECH 1995, 26.
[10] a) C. E. Song, E. J. Roh, Chem. Commun. 2000, 837; b) C. E. Song, C. R. Oh, E. J. Roh, D. J. Choo, Chem.
Commun. 2000, 1743.
[11] a) C. E. Song, W. H. Shim, E. J. Roh, J. H. Choi, Chem. Commun. 2000, 1695; b) T. Tsuchimoto, T. Maeda,
E. Shirakawa, Y. Kawakami, Chem. Commun. 2000, 1573; c) C. E. Song, W. H. Shim, E. J. Roh, S.-g. Lee,
J. H. Choi, Chem. Commun. 2001, 1122.
[12] J. N. Rosa, C. A. M. Afonso, A. G. Santos, Tetrahedron 2001, 57, 4189.
[13] a) J. Howarth, K. Hanlon, D. Fayne, P. McCormac, Tetrahedron Lett. 1997, 38, 3097; b) T. Oh, I. Meracz,
presented at the 224th ACS National Meeting, Boston, Aug. 2002, Abstract No. ORGN 621.
Received October 20, 2002