Bifunctional acid-base ionic liquid organocatalysts with a controlled distance between acid and base sites

Article abstract of DOI:10.1002/chem.200901519

Bifunctional acid-base ionic liquid organocatalysts with different distances between the two sites have been synthesised, and their activity for the Knoevenagel condensation has been tested. As has been found to be the case with enzymes, the distance betw

Full text of DOI:10.1002/chem.200901519

FULL PAPER
DOI: 10.1002/chem.200901519
Bifunctional Acid–Base Ionic Liquid Organocatalysts with a Controlled
Distance Between Acid and Base Sites
Mercedes Boronat, Maria J. Climent, Avelino Corma,* Sara Iborra, Raquel Montꢀn, and
Maria J. Sabater[a]
Abstract: Bifunctional acid–base ionic
liquid organocatalysts with different
distances between the two sites have
been synthesised, and their activity for
the Knoevenagel condensation has
been tested. As has been found to be
the case with enzymes, the distance be-
tween the acidic and basic sites deter-
mines the activity of the bifunctional
organocatalyst, and at the optimal dis-
tance the reaction rate increases by
two orders of magnitude with respect
to the purely acidic or basic counter-
part organocatalysts. The experimental
results have been rationalised through
the study of the reaction mechanism of
the Knoevenagel condensation be-
tween malononitrile and benzaldehyde
by means of DFT calculations. It has
been found that it consists of two con-
secutive steps. First, deprotonation of
malononitrile on the basic site to
obtain a methylene carbanion inter-
mediate takes place, and second, co-ad-
sorption and activation of benzalde-
hyde on the acid centre of this inter-
ꢀ
mediate followed by the C C bond-for-
mation reaction. The calculations and
the kinetic study indicate that there is
an inversion of the rate-controlling
step when the distance between the
acidic and the basic sites is modified,
with a direct implication on the reac-
tion rate.
Keywords: acid–base interactions ·
density functional calculations
ionic liquids Knoevenagel
condensation · organocatalysis
·
·
Introduction
tion,[3] esterification,[4] acrylation,[5] dimerisations[6] and
Diels–Alder.[7] However, the sensitivity of these ILs to
water can be a limitation. The first non-chloroaluminate
acidic ionic liquid[8] was based on dialkylimidazolium salts
with a side chain containing a -SO3H group, and were used
as catalysts and/or solvent in many conventional acid-cata-
lysed reactions such as olefin oligomerisation,[9] etherifica-
tion,[8] esterification[10] and Friedel–Crafts alkylation.[11]
Other types of Brønsted acid ionic liquids refer to those ob-
tained through the combination of a Brønsted acid and a
Brønsted base, such as N-protonated 1-alkylimidazolium
salts,[12] N-protonated lactams[13] and N-protonated pyridini-
um salts.[14] These protic ionic liquids have been used in a
variety of organic reactions such as esterification,[15] carbon-
yl protection[16] and Mannich.[17] Recently, an interesting
review concerning the properties and applications—includ-
ing organic reactions—of protic ionic liquids has been re-
ported by Greaves and Drummond.[18] The weak acidity of
the C-2 proton in the dialkyl imidazolium ring is also well
known and has been anticipated as the cause of some acid
catalytic effects.[19]
Ionic liquids (ILs) are highly relevant in organic synthesis as
environmentally benign solvents.[1] They possess favourable
properties such as negligible vapour pressure, high polarity,
non-flammability, easy handling and good solvating ability
for a wide range of substrates and catalysts. The number of
ionic liquids that can be produced is in essence limitless, and
the variation in the properties as a consequence of small
changes in the molecular structure greatly extend the scope
of applications. Besides their interesting physical properties,
the fact that some ionic liquids can show acidity or basicity
allows them to be also used as solvent and catalysts in sever-
al acid- or base-catalysed processes.[2] For instance, over the
past decade, chloroaluminate ionic liquids have gained in-
creasing attention as Lewis acid catalysts and they have
been used to catalyse a variety of reactions such as alkyla-
[a] Dr. M. Boronat, Dr. M. J. Climent, Prof. A. Corma, Dr. S. Iborra,
R. Montꢀn, Dr. M. J. Sabater
Instituto de Tecnologꢁa Quꢁmica (UPV-CSIC)
Av. de los Naranjos s/n, 46022 Valencia (Spain)
Fax : (+34)96-387-7809
Basic ionic liquids that form anions of the carboxylate
type (such as lactate,[20] formate[21] and acetate), hydroxy
anion[22] and dicyanamide[23] have also been prepared. How-
E-mail: acorma@itq.upv.es
Chem. Eur. J. 2010, 16, 1221 – 1231
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1221

ever, there is an alternative in designing ILs with basic sites
that incorporate the basic moiety in the cation. These ILs
have the advantage of being more thermally stable than
those containing basic anions. Thus, a range of salts using a
ties of ILs. We will show that bifunctional molecules act as
active, selective and recyclable catalysts for Knoevenagel
condensation reactions. When an optimum distance between
the acidic and basic sites exits, the reaction rate increases by
two orders of magnitude with respect to the counterpart
monofunctional basic catalyst.
On the basis of this observation, we thought that it could
be of much interest to prepare a bifunctional acid–base or-
ganocatalyst with the optimum distance and configuration
between the two sites that simultaneously incorporates the
interesting solvent extraction and recyclable properties of
ILs. In the present work we have synthesised organocata-
lysts with IL properties that present acidic and basic sites in
the same molecule and in which the distance between the
acidic and the basic sites has been changed. This bifunction-
al molecule acts as very active, selective and recyclable cata-
lyst for condensation reactions, thereby increasing the reac-
tion rate by almost two orders of magnitude with respect to
the counterpart basic catalyst.
1-alkyl-4-aza-1-azoniabicycloACHTNUGRTENUNG[2.2.2]octane cation (Cndabco)
in combination with bis(trifuoromethanesulfonyl)amide
have been synthesised.[24] Seddon and co-workers have also
recently reported a series of basic ILs of formula [Cat+-Z-
Bas][Xꢀ] in which Cat+ is a positively charged moiety con-
sisting of a heterocyclic ring structure, Bas is a hindered
basic moiety that comprises at least one nitrogen, phospho-
rous, sulphur, oxygen or boron atom, and Z is a carbon
chain joining Cat+ with Bas. The authors claim that these
ILs can be used as promoters or catalysts for a variety of re-
actions such as esterifications, transesterifications, aldol re-
actions, polymerisations, Robinson annulation, Heck and
Suzuki coupling and so on.[25] Finally, proton-conducting
non-aqueous electrolytes in fuel cells have been prepared
based on bis(trifluoromethanesulfonyl)amide.[26]
gem-Diamines present high pKa values and have been
named as proton sponges.[27] On account of their strong ba-
sicity they have been used as organocatalysts for reactions
such as Knoevenagel and aldol condensations.[28,29] We have
recently shown that by anchoring the gem-diamine proton
sponge 1,8-bis(dimethylamino)naphthalene (DMAN) on a
carrier containing mildly acidic sites, a bifunctional catalyst
containing acid and base sites is formed that enhances the
rate of the Knoevenagel condensation between benzalde-
hyde and methylene active compounds.[30] This was shown to
occur through a mechanism in which the mildly acidic site
of the support activates the benzaldehyde through the car-
bonyl group, whereas the basic site of the proton sponge
was responsible for the activation of the methylene active
compound (diethyl malonate) by proton abstraction. In the
case of enzymes with acid–base sites, key factors for the en-
hanced activity are related to the distance between the acid
and basic sites and their geometrical configuration. In the
case of organic bases supported on acidic carriers, it is diffi-
cult to control the above-mentioned factors. However, this
should be easier in the case of organocatalysts, in which the
distance between the acidic and basic sites can be controlled
by the synthesis of the molecule.
Results and Discussion
Catalytic activity: To check the catalytic activity of the
newly synthesised [diamine-A]BF4 IL in which the acid and
basic sites are separated by a -CH2- group, the Knoevenagel
condensation between benzaldehyde and several compounds
with active methylene groups and different pKa values was
selected as a test reaction. The Knoevenagel condensation
(Scheme 1) is an important carbon–carbon bond-forming re-
Scheme 1. The Knoevenagel condensation between benzaldehyde and
several compounds with active methylene groups (X=CN, CH3, COOEt;
Y=CN, COOEt).
action that is typically catalysed by base,[31] acid,[32–34] or
acid–base catalysts.[35] Then, reactions between benzalde-
hyde and malononitrile, ethyl cyanoacetate, ethyl acetoace-
tate and diethyl malonate were performed using [diamine-
A]BF4 IL as catalyst (1 mol%) in the absence of solvent,
since the reactants and catalyst form a homogeneous mix-
ture. The results obtained (Figure 1) show that the order of
reactivity of the different reactants is in accordance with the
decrease in the pKa value of the methylene active com-
pounds, that is, malononitrile>ethyl cyanoacetate>ethyl
acetoacetate>diethyl malonate. The selectivity to the con-
densation product is 100% in all cases, and the ionic liquid
characteristics of [diamine-A]BF4 allow the extraction and
recycling of the catalyst. Indeed, three reaction–extraction
cycles using diethyl ether have been performed, and the
mass balance was always above 98% without any loss of ac-
tivity.
In the present work we have synthesised bifunctional
acid–base IL organocatalysts with different distances and
configuration between the two sites, and which also incorpo-
rate the interesting solvent extraction and recyclable proper-
1222
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 1221 – 1231

Ionic Liquid Organocatalysts
FULL PAPER
plied to a variety of carbonyl compounds (aldehydes and
ketone) as well as different methylene compounds and the
results are given in Table 1. In general, the reactions were
Table 1. Results obtained in the Knoevenagel condensation of different
substrates at room temperature using [diamine-A]BF4 as catalyst.[a]
Figure 1. Knoevenagel reactions between benzaldehyde (32 mmol) and
different methylene active compounds (28 mmol) in the presence of [dia-
Entry R1
R2
E1
E2
t [min] Yield [%][b]
^
mine-A]BF4 (0.28 mmol): (ꢃ) malononitrile at 258C, ( ) ethyl cyanoace-
1
2
3
4
5
6
7
Ph
H
H
H
H
H
H
CO2CH2CH3 CO2CH2CH3 20
CO2CH2CH3 CO2CH2CH3 15
CO2CH2CH3 CO2CH2CH3 15
99
99
*
tate at 258C, (&) methyl acetoacetate at 608C, ( ) diethyl malonate at
i-C3H7
n-C7H15
Ph
Ph
Ph
808C.
100
100
100
100
100
CN
CN
CN
CO2CH2CH3
1
1
COCH3
CO2CH2CH3 10
CO2CH2CH3 10
For purposes of comparison, the reaction between benzal-
dehyde and malononitrile was also performed using only
[N-methyl piperidinium]BF4 as acid catalyst, or only the
highly active gem-diamine-A precursor as base catalyst, and
the results obtained are presented in Figure 2. According to
CH3
CH3 CN
[a] Reaction conditions: [diamine-A]BF4 (20 mol%), methylene com-
pound (5 mmol), aldehyde or ketone (5 mmol), room temperature, under
nitrogen. [b] Calculated by GC.
completed within a few minutes and proceeded very cleanly
at room temperature. Indeed the Knoevenagel products
were obtained selectively with very high yields.
In general, the aliphatic aldehydes reacted faster in the
Knoevenagel condensation with diethyl malonate than the
aromatic benzaldehyde (see entries 1–3 in Table 1), although
the yields of the final condensation products were always
quantitative. Notably, the reaction rate improved significant-
ly when benzaldehyde was allowed to react with malononi-
trile, ethyl cyanoacetate and ethyl acetoacetate (see en-
tries 4–6 in Table 1). In all cases studied, the performance of
the [diamine-A]BF4 was clearly superior to that of the 1-
butyl-3-methylimidazolium hydroxide ([BmIm]OH) catalyst
reported in the literature since the condensation products
were obtained with very high yields and shorter times than
with [BmIm]OH.[22a]
Figure 2. Knoevenagel condensation between benzaldehyde (32 mmol)
and malononitrile (28 mmol) in the presence of [N-methyl piperidine]BF4
&
&
^
(0.28 mmol) ( ), diamine-A ( ), [diamine-A]BF4 ( ) and [diamine-B]BF4
~
).
(
our previous results, the basic gem-diamine-A is more active
than the acidic [N-methyl piperidinium]BF4 catalyst, but less
active than the acid–base [diamine-A]BF4 IL. The enhanced
activity of the [diamine-A]BF4 IL can not be attributed to a
higher basicity of this molecule, because gem-diamine-A is
more basic than [diamine-A]BF4. A key difference between
the two organocatalysts is the presence in [diamine-A]BF4
of a mildly acidic protonated amino group together with the
basic site. Indeed, we have seen that the protonated amino
groups of [N-methyl piperidinium]BF4 and [diamine-A]BF4
are acidic enough to catalyse the acetalisation of benzalde-
hyde with trimethyl orthoformate (TMOF), a reaction typi-
cally catalysed by mildly acidic sites, but are not acidic
enough to catalyse the Knoevenagel condensation, as dem-
onstrated by the null conversion obtained when performing
the Knoevenagel condensation between the different meth-
ylene active compounds described above and benzaldehyde
in the presence of [N-methyl piperidinium]BF4.
Since the excellent catalytic results for the Knoevenagel
condensation obtained with the bifunctional [diamine-A]BF4
are difficult to explain through the independent action of
the acidic and basic sites present in the organocatalyst, a
combined acid–base-catalysed mechanism could be pro-
posed in which the weak Brønsted acid site associated with
the N-protonated amino group interacts with the benzalde-
hyde carbonyl group, thereby polarising the C=O bond and
increasing the net positive charge on the C atom. This polar-
isation makes this C atom more likely to be attacked by the
methylene carbanion intermediate formed on the vicinal
Lewis basic site associated with the neighbour amine group
(Scheme 2).
If the bi-site mechanism is operative, one can expect that
the distance between the acidic and basic sites should be a
critical parameter in achieving the stabilisation of the transi-
tion state involved in the mechanism, just as it occurs in
many enzymatic processes. We have studied this effect by
synthesising an analogous diamine IL, [diamine-B]BF4, in
The activity of the protonated compound [diamine-A]BF4
in the Knoevenagel condensation has been successfully ap-
ꢀ
which the two nitrogen atoms are separated by a -CH2
Chem. Eur. J. 2010, 16, 1221 – 1231
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1223

A. Corma et al.
Computational study of the
reaction mechanism: The mech-
anism of the Knoevenagel con-
densation between benzalde-
hyde and malononitrile cata-
lysed by [diamine-A]BF4 and
[diamine-B]BF4 was investigat-
ed theoretically by means of
DFT calculations. According to
the
bifunctional
acid–base
Scheme 2. The bifunctional acid–base mechanism proposed for the Knoevenagel condensation shown in
Scheme 1.
mechanism proposed above and
depicted in Scheme 2, it was ini-
tially assumed that the carbonyl
CH2- chain instead of one -CH2- bridge. Table 2 offers an
overview of the physical properties of these organic salts.
group of benzaldehyde would
interact with the N-protonated
acidic site, and malononitrile
would adsorb on the basic
neighbouring N atom. The opti-
mised geometry of the complex
formed by co-adsorption of ben-
zaldehyde and malononitrile on
[diamine-A]BF4 is depicted in
Figure 3. The carbonyl group of
benzaldehyde interacts with the
proton of the N-protonated site,
and as a result the carbonyl
group is polarised and the C=O
bond length increases from
1.214 ꢄ in the gas phase to
Table 2. Properties of organic salts [diamine-A]BF4, [diamine-B]BF4 and
[N-methyl piperidinium]BF4.
Organic salt
T (dec)
[8C][a]
1
Physical M.p.
s
[b]
[d]
[gcmꢀ3
]
state[c]
[8C]
[mScmꢀ1
]
[diamine-A]BF4
[diamine-B]BF4
120, 307,
392, 437
303, 344,
387
1.195
1.196
1.215
solid
108.8 167.0
solid
72
–
88.4
[e]
[N-methyl piperi-
–
liquid
137.0
dinium]BF4
[a] Decomposition temperatures were obtained from thermogravimetric
Figure 3. Optimised geometry
of benzaldehyde and malono-
nitrile co-adsorbed on [dia-
mine-A]BF4: O red, C orange,
N blue, H white, B pink, F
yellow.
analysis under an N2 flow (rate: 108Cminꢀ1). [b] Calculated at 308C.
[c] Physical state at 208C. [d] Ionic conductivity calculated on 10ꢀ3
aqueous solutions at 208C. [e] Not determined.
m
ꢀ
1.228 ꢄ. The calculated O(CO)
H(NH)
distance
is
short
(1.856 ꢄ), whereas the distance
between the H atom of malono-
nitrile and the basic N atom of
The acidity of the [diamine-B]BF4 salt was first tested by
performing the acetalisation reaction of benzaldehyde with
TMOF. The activity obtained with [diamine-B]BF4 was simi-
lar to that measured for [diamine-A]BF4, thus indicating a
similar acidity for both catalysts. However, when the Knoe-
venagel condensation between malononitrile and benzalde-
hyde was performed at 258C with [diamine-B]BF4, the yield
was much lower than that with the [diamine-A]BF4 catalyst
(see Figure 2). Moreover, when diethyl malonate was used
as the reagent, only 1% yield of the Knoevenagel adduct
was observed after a 6 h reaction time with [diamine-B]BF4,
whereas 85% yield was obtained with [diamine-A]BF4,
which results in a catalytic activity more than two orders of
magnitude larger.
the diamine is considerably larger (2.286 ꢄ). The different
strength of the two interactions involved, namely, the ben-
zaldehyde acidic site and malononitrile basic site, is also re-
flected in the calculated adsorption energies: ꢀ8.8 kcalmolꢀ1
for benzaldehyde on the acidic site and only ꢀ2.6 kcalmolꢀ1
for co-adsorption of malononitrile on the basic N of the
benzaldehyde–[diamine-A]BF4 complex. Benzaldehyde ad-
sorption on the acidic proton of [diamine-B]BF4 is exother-
mic by 5.3 kcalmolꢀ1, but interaction of malononitrile with
the basic site of the benzaldehyde–[diamine-B]BF4 complex
is negligible, and no minimum corresponding to such a
system could be obtained. Moreover, despite all our efforts,
it was not possible to find a reaction path that, starting from
the benzaldehyde–malononitrile–[diamine-A]BF4 complex
(Figure 3) and through a simultaneous deprotonation of ma-
lononitrile and attack of the negatively charged carbon
atom to the carbon atom of the carbonyl group, yields the
condensation alcohol intermediate shown in Scheme 2. In-
stead, it was found that the mechanism consists of two
steps: 1) deprotonation of adsorbed malononitrile to yield a
methylene carbanion intermediate that remains adsorbed on
a positively charged diprotonated diamine, and 2) co-adsorp-
tion of benzaldehyde on this system and reaction with the
Several conclusions can be extracted from these results.
ꢀ
First, it appears that the BF4 anion, which in principle
could act as a weak base, does not play any key role as basic
catalytic site in the reaction. Secondly, the results corrobo-
rate that a purely acidic mechanism is not acting in the pro-
cess. Finally, and most importantly, it appears that there is
an optimal distance between the acidic and basic sites to
achieve their cooperative effect. In an attempt to check this
hypothesis and find the optimal distance between both
active centres, the reaction mechanism was studied by
means of density functional calculations.
1224
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 1221 – 1231

Ionic Liquid Organocatalysts
FULL PAPER
methylene carbanion to form a condensation alcohol inter-
mediate.
The species involved in the first step of the mechanism
and the calculated energy profiles are depicted in Figures 4
and 5, respectively, and the adsorption, activation and reac-
anion intermediate (I1-DA) is formed (Figure 4c). In the
ꢀ
ꢀ
transition state, the C(methylene) H and H N(diamine) distances
involving the hydrogen atom that is being transferred are
1.463 and 1.275 ꢄ, respectively, and the proton of the N-pro-
tonated amino group continues to interact with one of the
ꢁ
C N groups of malononitrile. In the methylene carbanion
ꢁ
intermediate I1-DA, the N atom of this C N group is hydro-
gen bonded to the two protons of the diprotonated [dia-
ꢀ
mine-A]BF4, with calculated N(CN) H(diamine) distances of
1.591 and 1.601 ꢄ, and with a concomitant increase in the
CN bond length from 1.159 ꢄ in the gas phase to 1.196 ꢄ.
Moreover, an important electronic reorganisation occurs
spontaneously as the geometry of the system is optimised.
The negative charge that should be localised on the tri-coor-
dinated carbon atom, which is not involved in any interac-
tion, moves to the N atom that is interacting with the dia-
ꢁ
mine protons, the triple C N bond is converted into a
double bond, and a new C=C double bond is formed. The
calculated activation energy for this step is 10.9 kcalmolꢀ1,
and the reaction intermediate is only 0.7 kcalmolꢀ1 more
stable than the transition state. The low stability of the reac-
tion intermediate can be explained by the strong repulsive
interaction that exists between the two positive charges in
the diprotonated diamine, which are localised on the N-pro-
tonated atoms and are separated by only 2.469 ꢄ.
The process is not completely equivalent when [diamine-
B]BF4 is used as catalyst. In the initial adsorption complex
(M-DB), the acidic proton of [diamine-B]BF4 interacts with
ꢁ
one of the C N groups, whereas one of the hydrogen atoms
of malononitrile adsorbs on the basic N centre of the dia-
mine. Both interactions are weak, so that the calculated ad-
sorption energy is only 0.5 kcalmolꢀ1. However, the molecu-
lar conformation is very favourable for the hydrogen-trans-
fer process, and therefore the calculated activation energy is
low (1.9 kcalmolꢀ1). As in the methylene carbanion inter-
mediate I1-DA, the negative charge in the reaction inter-
mediate I1-DB is not localised on the tri-coordinated C
atom, but on the N atom that interacts with the two protons
of the diprotonated [diamine-B]BF4. The optimised geome-
try of the methylene carbanion fragment in I1-DA and I1-
ꢀ
DB is similar, and the N(CN) H(diamine) distances are slightly
longer in the case of [diamine-B]BF4 (1.629 and 1.663 ꢄ).
The main difference between the two intermediate com-
ꢀ
plexes is found in the N(diamine) N(diamine) distance (2.469 in I1-
Figure 4. a) Adsorbed malononitrile, b) transition state for deprotonation
and c) adsorbed methylene carbanion intermediate on [diamine-A]BF4
(left) and [diamine-B]BF4 (right: C orange, N blue, H white, B pink, F
yellow.
DA and 3.300 ꢄ in I1-DB). This is the distance between the
two N atoms bearing the positive charge in the diprotonated
diamines, and explains the higher stability of the reaction in-
termediate with a larger charge separation, namely, I1-DB.
Thus, whereas step 1 is clearly endothermic when catalysed
by [diamine-A]BF4 (see Figure 5 and Table 3), it is energeti-
cally favourable with [diamine-B]BF4.
tion energies are summarised in Table 3. Initially, malononi-
trile interacts with the acid proton of [diamine-A]BF4
through the lone pair of electrons on the N atom of one of
The second step in the reaction mechanism involves ad-
sorption of benzaldehyde on the adsorbed methylene carb-
anion intermediate complex formed in the first elementary
step, and attack of the methylene carbanion onto the car-
bonyl group to form a condensation alcohol intermediate.
The co-adsorbed reactants (R-DA and R-DB), the transition
ꢁ
the C N groups, thereby forming an adsorption complex
(M-DA) 3.6 kcalmolꢀ1 more stable than the separated reac-
tants. Then, the basic N of [diamine-A]BF4 abstracts one of
the hydrogen atoms of malononitrile through the transition
state depicted in Figure 4b (TS1-DA) and a methylene carb-
Chem. Eur. J. 2010, 16, 1221 – 1231
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1225

A. Corma et al.
sition state (TS2-DA) is quite
similar, with the carbonyl group
highly activated (rCO =1.270 ꢄ),
the proton beginning to be
ꢀ
transferred (N(diamine) H(diamine)
=
=
ꢀ
1.115 ꢄ,
H(diamine) O(CO)
1.447 ꢄ) and the carbonyl and
methylene carbon atoms being
separated by only 2.235 ꢄ. The
product obtained in this process
(I2-DA) is an alcohol adsorbed
on the initial [diamine-A]BF4
catalyst and stabilised by two
acid–base interactions: one be-
tween the N-protonated acidic
centre of the diamine and one
ꢁ
of the C N groups, which also
exist in R-DA and TS2-DA,
and another one between the H
atom of the hydroxyl group and
the basic nitrogen of the cata-
lyst. The activation energy ob-
tained
for
this
step
is
6.7 kcalmolꢀ1, and the alcohol
intermediate
formed
is
7.2 kcalmolꢀ1 more stable than
adsorbed reactants.
Figure 5. Calculated energy profile for the condensation of malononitrile and benzaldehyde catalysed by [dia-
mine-A]BF4 (full line) and [diamine-B]BF4 (dashed line).
Adsorption of benzaldehyde
on the I1-DB intermediate is
weaker than on I1-DA, and dif-
Table 3. Calculated adsorption, reaction and activation energies (in kcal
molꢀ1) for the two steps of the mechanism catalysed by [diamine-A]BF4
and [diamine-B]BF4.
ferent in nature. Benzaldehyde is not able to displace the
methylene carbanion from the acidic protons of the diproto-
nated [diamine-B]BF4, and the carbonyl group can only in-
teract with the hydrogen atoms of the cyclohexane ring,
with a calculated adsorption energy of 6.6 kcalmolꢀ1. How-
ever, the optimised geometry of the transition state TS2-DB
is similar to that of TS2-DA. The carbonyl group is strongly
[Diamine-A]BF4
[Diamine-B]BF4
Step 1
Step 2
Step 1
Step 2
adsorption energy
activation energy
reaction energy
ꢀ3.62
10.92
10.23
ꢀ10.83
6.68
ꢀ7.18
ꢀ0.50
1.92
ꢀ2.90
ꢀ6.64
8.64
ꢀ5.51
ꢀ
interacting with the acidic centre (rCO =1.270 ꢄ, H(diamine)
O(CO) =1.536 ꢄ) and the methylene and carbonyl carbon
atoms are separated only by 2.174 ꢄ. The activation barrier
is 8.6 kcalmolꢀ1, and the intermediate formed is
5.5 kcalmolꢀ1 more stable than the separated reactants. In
this case, the reaction intermediate is not an adsorbed alco-
hol, but an alcoholate, since the proton transfer from the di-
amine has not occurred yet.
The complete energy profile for the condensation of ben-
zaldehyde and malononitrile catalysed by [diamine-A]BF4
and [diamine-B]BF4 is depicted in Figure 5. Malononitrile
adsorption on [diamine-A]BF4 is exothermic, but the activa-
tion energy necessary for its deprotonation is high, and the
methylene carbanion intermediate is a very unstable species.
Adsorption of benzaldehyde on this system is energetically
states for the process (TS2-DA and TS2-DB) and the con-
densation intermediates (I2-DA and I2-DB) are depicted in
Figure 6, and the calculated adsorption, activation and reac-
tion energies are summarised in Table 3. In the case of [dia-
mine-A]BF4, benzaldehyde shows a strong tendency to
adsorb on one of the acidic protons of the intermediate
complex I1-DA obtained in the first step, thereby partially
displacing the methylene carbanion. Irrespectively of the ini-
tial distance and orientation of benzaldehyde with respect to
the intermediate complex, geometry optimisation always
leads to a very stable structure (R-DA, Figure 6a) in which
the carbonyl group is hydrogen bonded to one of the pro-
tons of the diamine, and deprotonated malononitrile is inter-
ꢀ
very favourable, and the activation energy for the C C
ꢁ
acting with the other acidic proton through one of the C N
bond formation is lower than the backward desorption of
benzaldehyde. This, together with the favourable orientation
of the reactants in the R-DA complex, suggests that the
second step of the mechanism will be fast, and step 1 will be
ꢀ
groups. The C(CO) C(methylene) distance is 4.518 ꢄ and the rela-
tive orientation of both molecules is very favourable for the
condensation reaction. The molecular geometry of the tran-
1226
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 1221 – 1231

Ionic Liquid Organocatalysts
FULL PAPER
with the acidic site in the R-DB complex, thereby suggesting
that entropy effects could have a strong influence on the ob-
served reaction rate. To confirm the proposed mechanism
and to check the influence of entropy effects on the catalytic
behaviour of the two diamines, a kinetic study of the reac-
tion including the experimental determination of the activa-
tion energies was carried out.
Kinetic study: The kinetics of the Knoevenagel condensa-
tion between benzaldehyde and malononitrile catalysed by
[diamine-A]BF4 and [diamine-A]BF4 were studied at differ-
ent temperatures to obtain experimental activation energies.
According to the mechanism proposed by the theoretical
study, the first step is adsorption of malononitrile (M) on
the diamine catalyst (cat.) [Eq. (1)] and deprotonation of
the adsorbed malononitrile ([M–cat.]) to obtain the methyl-
ene carbanion intermediate ([I1–cat.]) with a rate constant
k1 [Eq. (2)]; and the second step is co-adsorption of benzal-
dehyde (B) on the [I1–cat.] intermediate to form a reactant
complex ([R–cat.]) [Eq. (3)] that is converted into the con-
densation intermediate ([I2–cat.]) with a rate constant k2
[Eq. (4)]:
K1
M þ cat:
½Mꢀcat:ꢂ
½Mꢀcat:ꢂ
½I1ꢀcat:ꢂ
ð1Þ
ð2Þ
ð3Þ
ð4Þ
ƒ!ƒ
k1
ƒ!
K2
½I1ꢀcat:ꢂ þ B
½Rꢀcat:ꢂ
ƒ!ƒ
k2
½Rꢀcat:ꢂ
½I2ꢀcatꢂ:
ƒ!
Owing to the extremely large reaction rates obtained even
at the lowest temperature and catalyst concentration al-
lowed by the experimental setup, a number of approxima-
tions were used to simplify the kinetic rate equations and to
obtain useful information. When the catalyst used is [dia-
mine-A]BF4, the rate-determining step is malononitrile de-
protonation [Eq. (2)], which is directly dependent upon the
concentration of adsorbed malononitrile. Assuming that the
adsorption equilibrium [Eq. (1)] is fast and complete, and
since the catalyst concentration is constant, the reaction rate
(r) could be simplified to give Equation (5):
Figure 6. a) Co-adsorbed benzaldehyde and methylene carbanion inter-
ꢀ
mediate, b) transition state for the C C bond formation and c) condensa-
tion alcohol intermediate on [diamine-A]BF4 (left) and [diamine-B]BF4
(right): O red, C orange, N blue, H white, B pink, F yellow.
ꢀd½Mꢂ
ð5Þ
r ¼
¼ k1½Mꢂ
dt
the rate-determining step. On the other hand, malononitrile
adsorption on [diamine-B]BF4 is not very favourable, but
the deprotonation step involves a low energy barrier and
the reaction intermediate obtained is more stable than the
initial reactants, thus suggesting that step 1 is fast with this
catalyst. Subsequent adsorption of benzaldehyde on the ad-
sorbed methylene carbanion intermediate is exothermic, but
not enough so as to overcome the activation barrier in-
and the experimental results should be fitted to Equa-
tion (6)]:
ln½Mꢂ ¼ ln½Mꢂ0ꢀk1t
ð6Þ
On the other hand, with [diamine-B]BF4 catalyst, malono-
nitrile deprotonation is fast and therefore it can be assumed
that the concentration of the intermediate [I1–cat.] is con-
stant and equal to the initial concentration of the catalyst.
The rate-determining step [Eq. (4)] depends on the concen-
tration of the reactant complex [R–cat.], which, according to
Equation (3), is directly proportional to the concentration of
ꢀ
volved in the C C bond-forming process. More energy is
needed for the reaction to proceed, and therefore step 2 is
the rate-determining step on this catalyst. Moreover, as pre-
viously discussed, benzaldehyde does not directly interact
Chem. Eur. J. 2010, 16, 1221 – 1231
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1227

A. Corma et al.
benzaldehyde. Thus, the reaction rate on [diamine-B]BF4
could be simplified to give Equation (7):
study. The situation is not so clear when the reaction is cata-
lysed by [diamine-B]BF4. In this case, the fitting of the ex-
perimental data to the two models is of a similar quality,
with r2 =0.989 for the first-order reaction, and r2 =0.977 for
the second-order reaction. This suggests that competitive ad-
sorption on the active centres or entropy effects related with
co-adsorption of the reactants could be quite relevant when
the reaction is catalysed by [diamine-B]BF4, as will be seen
later.
Activation energies (Eact) were calculated from the mea-
sured rate constants at different temperatures (T) according
to Equations (11) and (12) (in which A is the frequency
factor and R is the gas constant):
ꢀd½Mꢂ ꢀd½Bꢂ
ð7Þ
r ¼
¼
¼ k2½Bꢂ
dt
dt
and the experimental data should fit a first-order equation
[Eq. (8)]:
ln½Bꢂ ¼ ln½Bꢂ0ꢀk2t
ð8Þ
Finally, we have also considered the one-step mechanism
initially proposed (Scheme 2), in which malononitrile and
benzaldehyde co-adsorb on the diamine with the simultane-
ous occurrence of malononitrile deprotonation and attack
onto benzaldehyde to give the condensation intermediate.
In this case, the reaction rate would depend on the concen-
tration of both reactants [Eq. (9)]:
ꢀEact
=
ð11Þ
RT
k ¼ Ae
ꢀ ꢁ
Eact
R
1
T
ln k ¼ ln A ꢀ
ð12Þ
ꢀd½Mꢂ ꢀd½Bꢂ
ð9Þ
r ¼
¼
¼ k3½Mꢂ½Bꢂ
dt
dt
as depicted in Figure 8. The quality of the fitting is good in
all cases, and the activation energies obtained from the
slope of the plot are 8.4 kcalmolꢀ1 for [diamine-A]BF4,
7.1 kcalmolꢀ1 for [diamine-B]BF4 using the first-order rate
constants and 10.2 kcalmolꢀ1 for [diamine-B]BF4 using the
second-order rate constants. At this point, it is still not pos-
sible to definitively state which is the reaction mechanism
followed with [diamine-B]BF4. However, there is an impor-
tant difference when considering the first- and second-order
mechanisms with [diamine-B]BF4. In the case of a second-
order mechanism, a isokinetic point must exist at a reaction
temperature of 388C in which the reaction rates on [dia-
mine-A]BF4 and [diamine-
and using initial concentrations [M]0 =[B]0, the experimental
data should fit a second-order equation [Eq. (10)]:
1
1
1
1
ꢀ
¼
ꢀ
¼ k3t
ð10Þ
½Mꢂ ½Mꢂ0 ½Bꢂ ½Bꢂ0
The plots of ln[M] and 1/[M] versus time depicted in
Figure 7 for both diamines indicate that when [diamine-
A]BF4 is used as catalyst, the reaction rate is first-order with
respect to the concentration of malononitrile, thus support-
ing the two-step mechanism that arises from the theoretical
B]BF4 are the same (see
Figure 8) and at temperatures
higher than 388C the reaction
should be faster on [diamine-
B]BF4. On the other hand,
when the first-order mechanism
is considered, the isokinetic
point between the two catalysts
should occur at
a
lower
(ꢀ118C) reaction temperature.
To check this hypothesis, the in-
itial reaction rates with both di-
amines were measured at differ-
ent temperatures between ꢀ10
and 608C. The results from
Table 4 clearly indicate that the
reaction is faster on [diamine-
A]BF4 over the whole tempera-
ture range, and that at ꢀ108C
the reaction rates for the two
catalysts are very close, thus
confirming the existence of an
isokinetic point at the tempera-
ture predicted by the first-order
Figure 7. Plots of ln[A] and 1/[A] versus time for the Knoevenagel condensation of malononitrile and benzal-
dehyde catalyzed by [diamine-A]BF4 (left) and [diamine-B]BF4 (right) at 258C. [A] (in mmolLꢀ1) is the malo-
nonitrile concentration for [diamine-A]BF4 and the benzaldehyde concentration for [diamine-B]BF4.
1228
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 1221 – 1231

Ionic Liquid Organocatalysts
FULL PAPER
¼
=R
DS
kBT
h
ð13Þ
A ¼
e
Indeed, the frequency factor A obtained for [diamine-A]BF4
from the plot in Figure 8 is an order of magnitude larger
than that obtained for [diamine-B]BF4, and this difference
must be related to the entropy effects associated with the
adsorption and reaction of benzaldehyde on the methylene
carbanion intermediates I1-DA and I1-DB. As previously
discussed, benzaldehyde adsorbs strongly on the acidic site
of intermediate I1-DA, thereby adopting a very favourable
orientation for the condensation reaction, so that the entro-
py effects in the second step of the mechanism are negligible
and the reaction rate on [diamine-A]BF4 only depends on
the rate of step 1, which is determined by the activation
energy. Adsorption of benzaldehyde on I1-DB is weaker
and the relative orientation of the two molecules in the re-
actant complex R-DB is not so favourable (see Figure 6).
When passing from the reactant complex R-DB to the tran-
sition state TS2-DB, benzaldehyde has to get very close to
the diamine acidic site and to the methylene fragment, with
a consequent entropy loss that is reflected in a smaller fre-
quency factor and therefore a lower reaction rate.
Figure 8. Relationship between rate constants and temperature for the
Knoevenagel condensation of malononitrile and benzaldehyde catalysed
&
^
by [diamine-A]BF4
( ) and [diamine-B]BF4 assuming first-order ( ) or
~
second-order ( ) reaction rates.
Table 4. Conversion [%] at different temperatures.
T [8C]
[Diamine-A]BF4
[Diamine-B]BF4
ꢀ10
0
25
40
50
60
7
28
41
68
82
90
9
16
32
54
65
74
model. It can then be concluded that on both catalysts the
reaction follows the two-step mechanism proposed by the
theoretical study, and that the different catalytic behaviour
arises from a change in the rate-determining step when the
distance between the acidic and basic sites is modified.
Comparison of the experimental activation energies with
those calculated above is not straightforward, since the
mechanism consists of several elementary steps including
adsorption processes and consecutive reactions with differ-
ent activation barriers. But it can be seen in Table 5 that the
Conclusion
Two different bifunctional organocatalysts with ionic liquid
properties that contain an acidic and a basic site in the same
molecule but with different distances between the two sites
have been synthesised, and their activity for the Knoevena-
gel condensation between benzaldehyde and methylene
active compounds has been investigated and compared with
that of the purely acidic or basic counterpart organocata-
lysts. It has been found that the best bifunctional acid–base
catalyst is more active than the corresponding monofunc-
tional acid or base catalyst. As has been found to be the
case with enzymes, the distance between the acidic and
basic sites is a determinant of the activity of the bifunctional
organocatalyst. The optimum distance between the two sites
and molecular configuration have been rationalised through
the study of the molecular mechanism of the Knoevenagel
condensation between malononitrile and benzaldehyde by
means of DFT calculations. It has been found that it consists
of two consecutive steps: 1) deprotonation of malononitrile
on the basic site to obtain a methylene carbanion intermedi-
ate, and 2) co-adsorption and activation of benzaldehyde on
Table 5. Calculated and experimental activation energies (in kcalmolꢀ1
)
for the two steps involved in the mechanism of the Knoevenagel conden-
sation of malononitrile and benzaldehyde catalysed by [diamine-A]BF4
and [diamine-A]BF4.
[Diamine-A]BF4
Step 1 Step 2
[Diamine-B]BF4
Step 1 Step 2
calculated
experimental
10.92
8.4
6.68
fast
1.92
fast
8.64
7.1
experimental activation energies correlate well with those
calculated for the rate-determining step on each catalyst
(step 1 on [diamine-A]BF4 and step 2 on [diamine-B]BF4)
and are approximately 2 kcalmolꢀ1 higher for [diamine-
A]BF4 than for [diamine-B]BF4. The experimental observa-
tion that the reaction proceeds faster on the catalyst that in-
volves a higher activation energy is explained by a larger
frequency factor [A in Equation (11)] with this catalyst that
must be related to entropy (S) effects according to Equa-
tion (13) in which kB is the Boltzmann constant and h is the
Planck constant:
ꢀ
the acidic centre of this intermediate followed by the C C
bond-formation reaction. The calculations indicate that
there is an inversion of the rate-controlling step when the
distance between the acidic and the basic sites is modified,
with a direct effect on the reaction rate. Thus, with [dia-
mine-A]BF4, in which the acidic and basic sites are separat-
ed by a -CH2- group, malononitrile deprotonation is the
most difficult step of the mechanism and requires an activa-
tion energy of approximately 11 kcalmolꢀ1. But once depro-
tonation occurs, benzaldehyde co-adsorption on the highly
Chem. Eur. J. 2010, 16, 1221 – 1231
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1229

A. Corma et al.
Experimental techniques: NMR spectra were recorded using a Bruker
Avance 300 spectrometer at 300.13 (1H), 75.47 (13C) and 128 MHz (11B),
in deuterated solvents with TMS as an internal standard. FTIR spectra
were recorded using a Nicolet 710 FT spectrophotometer in the range of
300–4000 cmꢀ1. Solutions of the respective organic salts in acetonitrile
were deposited onto a flat disk of monocrystalline Ge (provided by
Sorem-France). After slow solvent evaporation under a nitrogen flow,
the spectra were recorded at room temperature. Fast-atom-bombardment
(FAB) mass spectra of organic salts were performed with a ditranol
matrix using VG-Autospec equipment.
unstable methylene carbanion intermediate complex is spon-
taneous and the molecular disposition in the reactant com-
ꢀ
plex formed is so favourable that the C C bond formation
occurs readily. On the other hand, with [diamine-B]BF4, in
which the two positively charged NH centres in the methyl-
ene carbanion intermediate are separated by increasing the
hydrocarbon chain between the two nitrogen atoms, both
the transition state and the intermediate of the first step are
stabilised and the activation barrier decreases to approxi-
mately 2 kcalmolꢀ1. However, co-adsorption of benzalde-
hyde on the methylene carbanion intermediate complex and
its activation to react and form the condensation product is
energetically and entropically disfavoured, so that step 2 be-
comes the rate-determining step. The kinetic study and the
agreement between the experimental and calculated activa-
tion energies and the isokinetic point confirm the proposed
bi-site two-step mechanism, and explain the different behav-
iour of the two catalysts in terms of the change in the rate-
determining step that occurs when the distance between the
acidic and basic sites is modified.
The density of compounds [diamine-A]BF4, [diamine-B]BF4 and [N-
methyl piperidinium]BF4 was determined using a helium Accupyc 1330
picnometer (cell capacity: 1 cm3; measurements were taken at 308C).
Ionic conductivity (s) measurements were determined using an Orion-
160 conductivity meter (conductivity cell model 016010; K (cell constant)
= 0.609 cmꢀ1); reference values: Milli-Q-water s
(208C): 1 mScmꢀ1; etha-
ACHTUNGTRENNUNG
nol s
AHCTUNGTRENNUNG
means of thermogravimetric analysis with a SETARAM Setsys Evolution
16/18 thermobalance (108Cminꢀ1; N2: 30 mLminꢀ1).
The water content was determined by using the Karl Fischer method and
METROHM 702 SM Titrino equipment. The boron content was deter-
mined by means of an inductively coupled plasma optical emission spec-
trometer (ICP-OES) Varian 715-ES in combination with 11B NMR spec-
troscopy.
General procedure for the Knoevenagel reactions: In a typical experi-
ment, the basic catalysts (0.28 mmol, previously activated for 2 h at 808C
under vacuum) were added to a solvent-free solution of the methylene
compound (4.5 g, 28 mmol) while stirring under an inert atmosphere.
After temperature adjustment (808C), the aldehydes (3.4 g, 32 mmol)
were added and the reaction was periodically monitored by GC.
Experimental Section
General procedure for acetalisation of benzaldehyde: In a two-necked
10 mL round-bottomed flask, the catalyst (0.84 mmol) was weighed.
Triethyl orthoformate (2.73 mL, 25 mmol) and benzaldehyde (1.05 mL,
10 mmol) were added to the flask with a syringe. The mixture was heated
at 1308C with vigorous stirring for 6 h. During this time several samples
were taken, extracted with diethyl ether and analysed by GC.
[Diamine-A]BF4: A 100 mL round-bottomed flask was charged with dipi-
peridinomethane (10 g, 54.85 mmol) and diethyl ether (30 mL). Then
equimolar amounts of tetrafluoroboric acid/diethyl ether were added
dropwise into an ice bath. The mixture was stirred at room temperature
for 1 h. A solid was formed. It was recovered by filtration and was
washed exhaustively with diethyl ether. The solid was dried under
vacuum to give the organic salt as a yellow solid (8.1 g, 55%). 1H NMR
(300 MHz, CDCl3): d=1.5 (brs), 1.65 (brs), 2.85 (brs), 3.65 ppm (s);
13C NMR (300 MHz, CDCl3): d=22.5, 23.9, 51.0, 79.8 ppm; MS (FAB+):
General procedure for reuse reactions: A normal Knoevenagel reaction
was developed with larger amounts of every reactant and catalyst
(64 mmol of benzaldehyde, 56 mmol of diethylmalonate and 0.56 mmol
of [Amine 1]BF4). After being left to react for 6 h, products and reactants
were extracted with diethyl ether and the viscous yellow catalyst was col-
lected and dissolved in dichloromethane, dried with magnesium sulfate,
filtered and evaporated. Finally, the yellow-brown catalyst was dried
under vacuum at 408C for 2 h.
m/z: 98 [Mꢀ
T
(C6H11N)]+; FTIR: n˜ =3153 (s), 2937
ACHTUNGTRENNUNG
(s), 2853 (s), 2800 (s), 1479 (m), 1453 (m), 1311 (w), 1116 (s), 1069 (vs),
1006 (vs), 511cmꢀ1 (w); water content (<0.3%); fluorine composition
(%): 27.8 [C11H23N2BF4 requires 28.0%].
[Diamine-B]BF4: A 100 mL round-bottomed flask was charged with 1, 2-
di(N-piperidine)ethane (5.0 g, 25.47 mmol) and diethyl ether (30 mL).
Then equimolar amounts of tetrafluoroboric acid/diethyl ether were
added dropwise into an ice bath. The mixture was stirred at room tem-
perature for 1 h. A solid was formed. It was recovered by filtration and
was washed exhaustively with diethyl ether. The solid was dried under
Dried catalyst was weighed and used for the next reaction (reactants
were added in appropriate amount to the catalyst). Five consecutive runs
were carried out with the same catalyst using this method.
Computational details: Calculations were carried out by means of the
Gaussian 03 program package[36] using the density functional B3PW91
ACTHNUTRGNEUNG
method[37] and the standard 6-31G(d,p) basis set.[38] The geometries of all
vacuum to give the organic salt as
a yellow solid (4.25 g, 58.7%).
1H NMR (300 MHz, CD3CN): d=1.65 (brs), 1.75 (brs), 2.85 (brs), 2.9
(s), 5.6 ppm (s); 13C NMR (300 MHz, CD3CN): d=22.2, 23.6, 52.0,
53.3 ppm; MS (FAB+): m/z: 197 [MꢀBF4ꢀ]+, 112 [Mꢀ
species considered were fully optimised and the nature of every station-
ary point was characterised by means of frequency calculations and anal-
ysis of the vibrational modes. Additional geometry optimisations starting
from the transition states were performed to establish which reactants
and products were linked by a specific transition state. Zero-point vibra-
tional energy (ZPE) corrections to the total energies were obtained from
frequency calculations.
(BF4ꢀ)ꢀ
A
G
(C6H13N)]+; FTIR: n˜ =3137 (m),
ACHTUNGTRENNUNG
2932 (s), 2858 (m), 2784 (w), 1458 (m), 1284 (w), 1121 (s), 1063 cmꢀ1 (vs).
[N-Methyl piperidinium]BF4:
A 50 mL round-bottomed flask was
charged with N-methyl piperidine (1 g, 50.41 mmol) and diethyl ether
(5 mL). Then equimolar amounts of tetrafluoroboric acid/diethyl ether
were added dropwise into an ice bath and the mixture was stirred at
room temperature for 1 h. Active carbon was added and the reaction
mixture was filtered off. Solvent was evaporated under vacuum and the
viscous liquid was dried under vacuum to give the protonated compound
(3.39 g, 90%). 1H NMR (300 MHz, CDCl3): d=1.5 (s), 1.8 (s), 2.8 (s),
3.05 ppm (m); 13C NMR (300 MHz, CDCl3): d=55.6, 44.2, 23.4,
21.0 ppm; MS (FAB+): m/z: 100 [MꢀBF4ꢀ]+, 84 [MꢀCH3ꢀBF4ꢀ]+;
FTIR: n˜ =3575 (m), 3170 (s), 2960 (s), 2863 (s), 2576 (w), 2530, 1839,
1634, 1465 (vs), 1409 (vs), 1276 (m), 1122 (br), 846, 768, 518 (s), 456,
Acknowledgements
The authors thank CICYT (MAT 2006-14274-C02-01) for financial sup-
port.
[1] a) T. Welton, Chem. Rev. 1999, 99, 2071–2083; b) M. J. Earle, K. R.
Seddon, Pure Appl. Chem. 2000, 72, 1391–1398; c) P. Wasserscheid,
410 cmꢀ1
.
1230
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 1221 – 1231

Ionic Liquid Organocatalysts
FULL PAPER
W. Keim, Angew. Chem. 2000, 112, 3926–3945; Angew. Chem. Int.
Ed. 2000, 39, 3772–3789; d) J. Dupont, R. Souza, P. Suarez, Chem.
Rev. 2002, 102, 3667–3692.
[26] A. Noda, M. Susan, K. Kudo, S. Mitsushima, K. Hayamizu, M. Wa-
tanabe, J. Phys. Chem. B 2003, 107, 4024–4033.
[27] a) H. A. Staab, T. Saupe, Angew. Chem. 1988, 100, 895–909; Angew.
Chem. Int. Ed. Engl. 1988, 27, 865–909; b) A. L. Llamas-Saiz, C.
Foces-Foces, J. Elguero, J. Mol. Struct. 1994, 328, 297–323.
[28] I. Rodriguez, G. Sastre, A. Corma, S. Iborra, J. Catal. 1999, 183, 14–
23.
[2] a) D. Zhao, M. Wu, Y. Kou, E. Min, Catal. Today 2002, 74, 157–189;
b) Z. C. Zhang, Adv. Catal. 2006, 49, 153–273.
[3] a) K. Qiao, Y. Deng, J. Mol. Catal. A: Chem. 2001, 171, 81–84;
b) G. Wang, W. Qiao, L. Cheng, Appl. Catal. A 2004, 262, 69–73.
[4] Y. Deng, F. Dhi, J. Beng, K. Quiao, J. Mol. Catal. A: Chem. 2001,
165, 33–36.
[29] M. J. Climent, A. Corma, I. Dominguez, S. Iborra, M. J. Sabater, J.
Catal. 2007, 246, 136–146.
[5] M. H. Valkenberg, C. de Castro, W. F. Holderich, Appl. Catal. A
2001, 215, 185–190.
[30] A. Corma, S. Iborra, I. Rodriguez, F. Sanchez, J. Catal. 2002, 211,
208–215.
[6] A. Kumar, S. S. Pawar, J. Mol. Catal. A: Chem. 2004, 208, 33–37.
[7] C. Lee, Tetrahedron Lett. 1999, 40, 2461–2464.
[31] G. Jones in Organic Reactions, Vol. 15 (Ed.: A. C. Cope), Wiley,
New York, 1967, p. 204.
[8] A. C. Cole, J. L. Jensen, I. Ntai, K. L. T. Tran, K. J. Weave, D. C.
Forbes, J. H. Davis, J. Am. Chem. Soc. 2002, 124, 5962–5963.
[9] Y. Gu, F. Shi, Y. Deng, Catal. Commun. 2004, 5, 597–601.
[10] F. Gu, Y. Shi, J. Deng, J. Mol. Catal. A: Chem. 2004, 212, 71–75.
[11] K. Qiao, C. Yokoyama, Chem. Lett. 2004, 33, 472–473.
[12] a) M. Yoshizawa, W. Xu, C. A. Angell, J. Am. Chem. Soc. 2003, 125,
15411–15419; b) H. Ohno, M. Yoshizawa, W. Ogihara, Electrochim.
Acta 2004, 50, 255–261.
[13] Z. Du, Z. Li, S. Guo, J. Zhang, L. Zhu, Y. Deng, J. Phys. Chem. B
2005, 109, 19542–19546.
[14] Z. Duan, Y. Gu, J. Zhang, L. Zhu, Y. Deng, J. Mol. Catal. A: Chem.
2006, 250, 163–168.
[15] H. P. Zhu, F. Yang, J. Tang, M. Y. He, Green Chem. 2003, 5, 38–39.
[16] H. H. Wu, F. Yang, P. Cui, J. Tang, M. Y. He, Tetrahedron Lett. 2004,
45, 4963–4965.
[17] G. Zhao, T. Jiang, H. Gao, B. Han, J. Huang, D. Sun, Green Chem.
2004, 6, 75–77.
[18] T. L. Greaves, C. J. Drummond, Chem. Rev. 2008, 108, 206–237.
[19] a) A. R. Gholap, K. Venkatesan, T. Daniel, R. J. Lahoti, K. V. Srini-
vasan, Green Chem. 2003, 5, 693–696; b) T. Welton, Coord. Chem.
Rev. 2004, 248, 2459–2477.
[20] a) M. J. Earle, P. B. MacCormac, K. R. Seddon, Green Chem. 1999,
1, 23–25; b) A. Zhu, T. Jiang, B. Han, J. Huang, J. Zhang, X. Ma,
New J. Chem. 2006, 30, 736–740.
[32] O. Attanasi, P. Filippone, A. Mei, Synth. Commun. 1983, 13, 1203–
1208.
[33] R. P. Shanthan Rao, R. V. Venkataratnam, Tetrahedron Lett. 1991,
32, 5821–5822.
[34] W. Bao, Y. Zhang, J. Wang, Synth. Commun. 1996, 26, 3025–3028.
[35] a) M. J. Climent, A. Corma, V. Fornes, R. Guil-Lopez, S. Iborra,
Adv. Synth. Catal. 2002, 344, 1090–1096; b) M. J. Climent, A.
Corma, S. Iborra, A. Velty, J. Mol. Catal. A: Chem. 2002, 182–183,
327–342.
[36] Gaussian 03, Revision B.04, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[21] N. Bicak, J. Mol. Liq. 2005, 116, 15–18.
[22] a) B. C. Ranu, R. Jana, Eur. J. Org. Chem. 2006, 3767–3770; b) J. M.
Xu, C. Qian, B. K. Liu, Q. Wu, X. F. Lin, Tetrahedron 2007, 63, 986–
990; c) L. Yang, L. W. Xu, W. Zhou, L. Li, C. G. Xia, Tetrahedron
Lett. 2006, 47, 7723–7726.
[37] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) J. P. Perdew,
Y. Wang, Phys. Rev. B 1992, 45, 13244–13249.
[23] D. R. MacFarlane, S. A. Forsyth, J. Golding, G. B. Deacon, Green
Chem. 2002, 4, 444–448.
[38] P. C. Hariharan, J. A. Pople, Theor. Chim. Acta 1973, 28, 213–222.
[24] M. Yoshizawa-Fujita, K. Johansson, P. Newman, D. R. MacFarlane,
Tetrahedron Lett. 2006, 47, 2755–2758.
[25] M. J. Earle, K. R. Seddon, S. Forsyth, U. Frohlich, N. Gunaratne, S.
Katdare, WO 2006/072775, 2006.
Received: June 4, 2009
Published online: December 9, 2009
Chem. Eur. J. 2010, 16, 1221 – 1231
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1231