From salts to ionic liquids by systematic structural modifications: A rational approach towards the efficient modular synthesis of enantiopure imidazolium salts

Article abstract of DOI:10.1002/chem.200901623

This paper reports a simple and robust modular synthetic strategy that leads to a large variety of configu- rationally and structurally diverse imi- dazole-based chiral ionic liquids (CILs) by lipase-catalyzed resolution. The intimate microscopic interactions of the supramolecular ionic network of these imidazolium chiral salts at the molecular level are investigated both spectro- scopically (NMR, FT-IR-ATR) and theoretically, and a topological analysis of the experimental electron densities obtained by X-ray diffraction of single crystals is performed. Our results support the key role played by the relative configuration of the -OR group on the hydrogen-bonding pattern and its strong influence on the final physical properties of the imidazolium salt. We also obtained a reasonable correlation between the observed melting point and the non-covalent interactions. The spectroscopic data and the topological analysis reflect the key role played by hydrogen bonds between the OH and imidazolium C2H groups in both cation-anion and cation-cation interactions, with the presence of an OH group leading to an additional inter-cation interaction. This interaction sig- nificantly affects the properties of ste- reoisomeric salts. Even more interestingly, we also studied the effect of the chirality by comparing enantiopure CILs with their racemic mixtures and found that, with the exception of trans- Cy6-OH-Im-Bn-Br, the melting points of the racemic mixtures are higher than those of the corresponding enantiomer- ically pure forms. For stereoisomeric examples, we have successfully explained the differences in melting temperatures in light of the corresponding structural data. Chirality should therefore be taken into account as a highly attractive design vector in the preparation of ILs with specifically desired properties.

Full text of DOI:10.1002/chem.200901623

DOI: 10.1002/chem.200901623
From Salts to Ionic Liquids by Systematic Structural Modifications:
A Rational Approach Towards the Efficient Modular Synthesis of
Enantiopure Imidazolium Salts
Nicolꢀs Rꢁos-Lombardꢁa,[a] Eduardo Busto,[a] Vicente Gotor-Fernꢀndez,[a]
Vicente Gotor,*[a] Raffll Porcar,[b] Eduardo Garcꢁa-Verdugo,[b, f] Santiago V. Luis,*[b]
Ignacio Alfonso,[c] Santiago Garcꢁa-Granda,[d] and Amador MenØndez-Velꢀzquez[e]
Abstract: This paper reports a simple
and robust modular synthetic strategy
that leads to a large variety of configu-
rationally and structurally diverse imi-
dazole-based chiral ionic liquids (CILs)
by lipase-catalyzed resolution. The inti-
mate microscopic interactions of the
supramolecular ionic network of these
imidazolium chiral salts at the molecu-
lar level are investigated both spectro-
scopically (NMR, FT-IR-ATR) and
theoretically, and a topological analysis
of the experimental electron densities
obtained by X-ray diffraction of single
crystals is performed. Our results sup-
port the key role played by the relative
configuration of the -OR group on the
hydrogen-bonding pattern and its
strong influence on the final physical
properties of the imidazolium salt. We
also obtained a reasonable correlation
between the observed melting point
and the non-covalent interactions. The
spectroscopic data and the topological
analysis reflect the key role played by
hydrogen bonds between the OH and
imidazolium C2H groups in both
cation–anion and cation–cation interac-
tions, with the presence of an OH
group leading to an additional inter-
cation interaction. This interaction sig-
nificantly affects the properties of ste-
reoisomeric salts. Even more interest-
ingly, we also studied the effect of the
chirality by comparing enantiopure
CILs with their racemic mixtures and
found that, with the exception of trans-
Cy6-OH-Im-Bn-Br, the melting points
of the racemic mixtures are higher than
those of the corresponding enantiomer-
ically pure forms. For stereoisomeric
examples, we have successfully ex-
plained the differences in melting tem-
peratures in light of the corresponding
structural data. Chirality should there-
fore be taken into account as a highly
attractive design vector in the prepara-
tion of ILs with specifically desired
properties.
Keywords: chirality · green chemis-
try
· imidazolium salts · ionic
liquids · lipases
[a] N. Rꢀos-Lombardꢀa, Dr. E. Busto, Dr. V. Gotor-Fernꢁndez,
Prof. V. Gotor
[d] Prof. S. Garcꢀa-Granda
Departamento de Quꢀmica Fꢀsica y Analꢀtica
Universidad de Oviedo, 33071 (Spain)
Departamento de Quꢀmica Orgꢁnica e Inorgꢁnica
Instituto Universitario de Biotecnologꢀa de Asturias
Universidad de Oviedo, 33071 Oviedo (Spain)
Fax : (+34)985103448
[e] Dr. A. Menꢃndez-Velꢁzquez
Centro de Investigaciꢂn en Nanomateriales y Nanotecnologꢀa (CINN)
Instituto Tecnolꢂgico de Materiales de Asturias (ITMA)
Parque Tecnolꢂgico de Asturias, 33428 Llanera (Spain)
E-mail: vgs@fq.uniovi.es
[b] R. Porcar, Dr. E. Garcꢀa-Verdugo, Prof. S. V. Luis
Departamento de Quꢀmica Inorgꢁnica y Orgꢁnica
Universitat Jaume I, 12071 Castellꢂn de la Plana (Spain)
Fax : (+34)964-728-214
[f] Dr. E. Garcꢀa-Verdugo
Instituto de Catꢁlisis y Petroleoquꢀmica, CSIC, Campus de la UAM
C/ Marie Curie, 2; Cantoblanco; Madrid; 28049 (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901623 and includes general
experimental procedures, a full list of the ionic liquids prepared, with
their isolated yields and melting points, DFT calculations, the thermal
stability study by TGA, and the topological analysis of the experimen-
tal electron densities..
E-mail: luiss@qio.uji.es
[c] Dr. I. Alfonso
Departamento de Quꢀmica Biolꢂgica y Modelizaciꢂn Molecular
Instituto de Quꢀmica Avanzada de CataluÇa (IQAC)
Consejo Superior de Investigaciones Cientꢀficas (CSIC)
Jordi Girona 18-26, 08034 Barcelona (Spain).
836
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 836 – 847

FULL PAPER
Introduction
Results and Discussion
Inorganic salts, such as NaCl, usually melt at very high tem-
peratures. However, many salts with organic cations and/or
anions can be found in their liquid state at moderate tem-
peratures. This fact has led to their routine use as solvents
for chemical processing, whereupon they are referred to as
“ionic liquids” (ILs). This term describes organic salts that
are liquid at or near room temperature, with 1008C as an ar-
bitrary upper limit.[1] ILs seem to be ideal replacement sol-
vents since they exhibit low vapor pressures and high boiling
points due to their very strong ion–ion interactions. ILs are
currently attracting considerable attention as reaction sol-
vents,[2,3] extraction liquids,[4] and electrolyte materials[5] as a
result of their remarkable properties.
Molecular design and chemoenzymatic synthesis: A simple
synthetic strategy has been designed for the preparation of
chiral imidazolium salts. Figure 1 summarizes the five tun-
The physicochemical characteristics of ILs depend on the
interaction established between both the cation and the
anion.[6] Thus, the large potential structural diversity of ILs
plays an important role in our ability to define their final
properties.[7] Indeed, ILs are now expected to be tailor-made
with controllable physical and chemical properties or even
specific functions (task-specific ionic liquids, TSILs).[8]
Among the many different design vectors, chirality can be
considered by far the most interesting for organic chemists.
Indeed, a great deal of effort has been expended in the last
few years to design and synthesize chiral ionic liquids
(CILs).[9] However, most of the currently known CILs have
a major structural limitation: they are relatively restricted in
terms of chiral structural diversity as they are derived from
the chiral pool.[10,11] Different studies have established the
influence of side chains as well as the presence of additional
functional groups on the intrinsic physical properties of
ILs.[1,12] However, although some studies have attempted to
correlate the configurational properties of chiral compounds
with their corresponding macroscopic behavior,[13] as far as
we are aware there is no systematic study within the field of
CILs. Such a study would require an in-depth knowledge of
the intimate microscopic interactions at the molecular level.
Herein we report a simple and robust modular synthetic
strategy that leads to a large variety of configurationally and
structurally diverse CILs. This approach allows us to reach a
bottom-up understanding of the structure–property relation-
ship based on the study of these CILs. NMR and ATR-FT-
IR analyses, theoretical calculations, and topological analysis
of experimental electron densities reveal the significant in-
fluence of chirality on the final structure and properties of
ILs. In light of these data, the relationships between the rel-
ative and absolute configurations of the chiral stereocenters
of different chiral imidazolium salts and one of their key
physical macroscopic properties, namely the melting point
(Tm), have been established for the first time. Our results
highlight the importance of the configuration of chiral cen-
ters as new design vectors when preparing ionic liquid with
desired macroscopic properties.
Figure 1. Design vectors employed in the synthesis of chiral imidazolium
salts.
AHCTUNGTREGaNNNU ble molecular engineering vectors selected as the design
variables: 1) the ring size (five- or six-membered ring); 2)
oxygen substitution (R’=H or Ac); 3) the nature of the
counterion (Xꢀ =Clꢀ, Brꢀ, BF4 , or NTf2 ); 4) the imidazole
alkylating side-chain (R=Bn, nBu, or nOct); 5) the configu-
ration at the two stereogenic centers and, in particular, the
relative disposition (cis or trans) of the imidazole and OR
groups on the cycloalkane ring. These modifications lead to
different stereoisomeric chiral imidazolium salts. The three
first variables have been used previously to prepare room-
temperature ionic liquids (RTILs), therefore their influence
on the macroscopic properties of RTILs is beginning to be
understood.[1,12,14] However, geometric modifications such as
ring size or the configuration around the chiral centers have
not been studied previously in a systematic manner.
ꢀ
ꢀ
Scheme 1 summarizes the synthetic strategy followed to
prepare enantiopure imidazoles derived from (ꢁ)-trans-2-
(1H-imidazol-1-yl)cyclopentanol [(ꢁ)-trans-3a] and (ꢁ)-
trans-2-(1H-imidazol-1-yl)cyclohexanol [(ꢁ)-trans-3b]. The
racemic alcohols (3a–b) were prepared in good yields by
treating the corresponding cyclic epoxides 1a–b with imida-
zole (2), following the procedure previously described by
Yus and co-workers.[15] They were then enzymatically re-
solved to yield the enantiopure acetates (R,R)-trans-4a and
-4b and the alcohols (S,S)-trans-3a and -3b under mild reac-
tion conditions by using Candida antarctica lipase type B
(CAL-B) or different Pseudomonas cepacia lipase prepara-
tions immobilized on ceramics (PSL-C I or PSL-C II). The
optimization process is summarized in Table 1. All biocata-
lysts tested showed an excellent stereoselectivity towards
the formation of (R,R)-trans acetates 4a–b and (S,S)-trans
alcohols 3a–b, which were isolated in enantiopure form with
very high yields (entries 2 and 7, Table 1).[16] The absolute
stereochemistries of alcohols (S,S)-trans-3a–b and acetylated
derivatives (R,R)-trans-4a–b were assigned after X-ray anal-
ysis of the alcohol recovered from the enzymatic kinetic res-
olution of racemic (ꢁ)-trans-3a, which showed a clear pref-
erence of the enzyme to acetylate the 1-(R) position.[17]
Chem. Eur. J. 2010, 16, 836 – 847
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
837

V. Gotor, S. V. Luis et al.
trans alcohols, especially for
the cyclohexanol derivatives.[17]
Enzymatic kinetic resolution
was then carried out to prepare
the corresponding enantiopure
cis compounds. Both (ꢁ)-cis-
3a and 3b were treated with
three equivalents of vinyl ace-
tate as acyl donor and different
enzymes as catalysts (see
Table 2). As expected, the cis
compounds were less reactive
than the trans-derivatives in
lipase-catalyzed reactions, es-
pecially in the case of the six-
membered cycles.[18] Cyclopen-
tanol (ꢁ)-cis-3a showed an ex-
cellent enantioselectivity to-
wards formation of the acetate
(R,S)-4a, although a lower re-
action rate was observed in
comparison with the corre-
sponding trans derivative. This
meant that higher tempera-
tures were needed to reach
50% conversion. Both the ace-
tate and the alcohol were iso-
lated with 94% ee (entry 2,
Table 2). Alternatively, PSL-C
II catalyzed the acetylation of
(ꢁ)-cis-3a to give the enan-
tioenriched products in almost
enantiopure form and with
very high isolated yields
(entry 3, Table 2). The absolute
stereochemistries of enantio-
pure alcohol cis-(S,R)-3a and
acetate cis-(R,S)-4a were as-
signed by X-ray analysis of the
crystallized compounds after
Scheme 1. Chemoenzymatic stereoselective synthesis of enantiopure cis- and trans-2-(1H-imidazol-1-yl)cyclo-
pentanol and -cyclohexanol derivatives: i) ref. 14; ii) kinetic resolution, enzyme, vinyl acetate, solvent (see
Table 1); iii) RCOOH (R=Me for 3a and R=p-NO2-C6H4 for 3b), DEAD, PPh3, THF; iv) K2CO3, MeOH,
H2O; v) kinetic resolution, enzyme, vinyl acetate, solvent (see Table 2).
Table 1. Enzymatic kinetic resolution of alcohols (ꢁ)-trans-3a–b using vinyl acetate as acyl donor.
Entry
Substrate
Enzyme
T [8C]
eeS [%][a]
eeP [%][a]
c [%][b]
E[c]
1
2
3
4
5
6
7
(ꢁ)-trans-3a
(ꢁ)-trans-3a
(ꢁ)-trans-3a
(ꢁ)-trans-3b
(ꢁ)-trans-3b
(ꢁ)-trans-3b
(ꢁ)-trans-3b
CAL-B[d]
PSL-C I[d]
PSL-C II[d]
CAL-B[d]
30
30
30
30
30
30
45
99 (92)
>99 (92)
97 (88)
93 (87)
97 (90)
>99 (91)
>99 (94)
>99 (91)
98 (85)
98 (88)
>99 (90)
>99 (91)
50
50
49
49
50
25
50
>200
>200
>200
>200
>200
>200
>200
PSL-C II[d]
PSL-C II[e]
PSL-C II[e]
33 (96)
>99 (93)
[a] Determined by HPLC; yields in brackets refer to the corresponding conversion value. [b] c=ee /
in which the subscripts S and P stand for substrate and product, respectively. [c] E=ln
[(1ꢀc)ꢅ(1ꢀeeS)]/ln-
[(1ꢀc)ꢅ(1+eeS)].[18] [d] THF as solvent. [e] tert-Butyl methyl ether (TBME) as solvent.
S ACHTUNGRTNE(NUNG eeS+eeP)
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
Synthesis of the cis derivatives was also attempted in
the enzymatic process.[17]
order to get cis/trans structural modularity. Thus, the (ꢁ)-
cis-alcohols were easily prepared from the corresponding
trans derivatives by means of a Mitsunobu inversion. Acetic
acid was used for (ꢁ)-trans-3a
The enzymatic kinetic resolution of (ꢁ)-cis-3b was also
conducted with CAL-B and PSL-C II (entries 4 and 5,
Table 2), although the results were rather disappointing as
whereas p-nitrobenzoic acid
Table 2. Enzymatic kinetic resolution of alcohols (ꢁ)-cis-3a–b using three equivalents of vinyl acetate in dry
was used for (ꢁ)-trans-3b due
to the poor reactivity observed
with AcOH. The (ꢁ)-cis esters
THF at 250 rpm.
Entry
Substrate
Enzyme
T [8C]
t [h]
eeS [%][a]
eeP [%][a]
c [%][b]
E[c]
1
2
3
4
5
6
7
8
cis-3a
cis-3a
cis-3a
cis-3b
cis-3b
cis-3b
cis-3b
cis-3b
CAL-B
CAL-B
PSL-C II
CAL-B
PSL-C II
CRL
30
60
60
60
60
60
60
60
161
63
23
53.5
25
72
44 (81)
94 (87)
96 (96)
>99 (83)
94 (76)
99 (93)
31
50
49
50
50
50
31
15
>200
115
>200
4a and
5 were deprotected
with potassium carbonate to
give the racemic alcohols (ꢁ)-
cis-3a and -3b in high overall
yields. The inversion of config-
uration was clearly observed
upon comparing the 1H NMR
spectra for (ꢁ)-cis and (ꢁ)
–
54
–
4
10
–
54
–
9
58
–
6
–
1
4
PPL
CAL-A
72
72
[a] Determined by HPLC and isolated yields in brackets referred to the corresponding conversion value).
[b] c=eeS/A(eeS+eeP). [c] E=ln
[(1-ꢀc)ꢅ(1ꢀeeS)]/ln
C
R
ACHTUNGTRENNUNG
838
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 836 – 847

Modular Synthesis of Enantiopure Imidazolium Salts
FULL PAPER
no or only poor enantioselectivity for acylation of the race-
mic alcohol was achieved. Similarly low activities and enan-
tioselectivities were observed with other biocatalysts [Can-
dida rugosa lipase (CRL), Porcine pancreas lipase (PPL) or
Candida antarctica lipase type A (CAL-A)] (entries 6–8,
Table 2).
In light of these results, a different synthetic strategy start-
ing from the enantiopure alcohols (S,S)-trans-3a and -3b,
previously prepared by enzymatic kinetic resolution, and in-
volving Mitsunobu inversion and deprotection steps was
evaluated. This synthetic methodology led to enantiopure al-
cohols (R,S)-cis-3a and (R,S)-cis-3b with 78% and 85% iso-
lated yields, respectively.
4008C.[20] The Tm is an important parameter because it can
be related to other properties of ionic liquids, such as viscos-
ity or conductivity.[1]
In general, cyclopentanol-derived imidazolium salts were
found to have lower melting points than those derived from
cyclohexanol. As expected, the introduction of nonbasic,
ꢀ
noncoordinating anions such as BF4 , or organic amides
ꢀ
such as NTf2 , into the chiral salts led to lower melting
points than with halide anions (Brꢀ or Clꢀ) as counterions.[1]
The tendency towards the formation of solid ionic salts for
ꢀ
different anions decreases in the order Clꢀ >Brꢀ >BF4
>
ꢀ
ꢀ
ꢀ
NTf2 . Indeed, organic salts containing BF4 or NTf2 can
be considered to be chiral RTILs for all the cases studied
(Tm <1008C). The size and nature of the alkylating agent
(Step i in Scheme 2) also play a key role in determining the
melting point of the different enantiopure chiral imidazoli-
um salts, with the general trend observed for the melting
Finally, the preparation of optically active imidazolium
salts was performed from either the cis/trans acetates or al-
cohols by treatment with the corresponding alkyl or aryl
halide (Scheme 2). Exchange of the halide present in the ini-
temperatures being as follows: Bn>nBu>nOct. As
a
matter of fact, the chiral salts obtained upon alkylation with
benzyl bromide and benzyl chloride showed high melting
points and were easily obtained as crystalline materials, in
some cases forming crystals appropriate for subsequent crys-
tallographic analysis (see below).
More interesting, however, were the results observed for
the series of compounds shown in Figure 2. For this family
Scheme 2. Schematic representation of the synthesis of chiral 2-(1H-imi-
dazol-1-yl)cyclopentanol and -cyclohexanol salts from enantiopure com-
pounds prepared according to Scheme 1: i) R’X, 70 or 1008C; ii) NaBF4
or LiNTf2, MeOH/H2O, room temperature.
ꢀ
ꢀ
tial imidazolium salts with different anions (BF4 or NTf2 )
occurred with very high yields to give a family of more than
30 enantiopure chiral imidazolium salts with significant
structural diversity.[16,19] Since some of the obtained salts are
hygroscopic, they were carefully dried in an oven at high
vacuum before characterization.
Phase transition (melting points) for the chiral imidazolium
salts: The phase transitions (melting points, Tm) of the newly
synthesized CILs were measured in order to compare their
physical properties. Thus, a dried sample of the correspond-
ing salt was placed in an open DSC pan and heated at
1308C for 120 min. Three complete cycles of heating and
cooling at a rate of 58C/min were then performed to clearly
identify the corresponding phase transitions. All these oper-
ations were carried out under nitrogen. Table 3 illustrates
some of the solid–liquid phase transitions for the newly pre-
pared imidazolium salts, as measured by differential scan-
ning calorimetry (DSC) in the temperature range ꢀ50 to
Table 3. Effect of the alkyl chain (Bn, Bu, or Oct) and counteranion
(Brꢀ, Clꢀ, BF4ꢀ, or NTf2ꢀ) on the melting points (Tm) for some trans
chiral imidazolium salts.
A
E
ꢀ
ꢀ
ꢀ
ꢀ
Brꢀ
Clꢀ
Brꢀ
Clꢀ
-Bn-
-Bu-
-Oct-
112
–
–
–
ꢀ21
ꢀ30
ꢀ9
ꢀ47
ꢀ48
ꢀ42
<ꢀ50
<ꢀ50
188
–
–
219
7
2
6
ꢀ22
ꢀ35
ꢀ29
ꢀ50
ꢀ51
Figure 2. DSC plots for a family of enantiopure chiral imidazolium salts
with controlled structural modifications [Tm (8C) calculated at onset;
heating rate: 58Cminꢀ1]. The curves correspond to the third DSC cycle
of melting.
[a] Melting temperatures [8C] for a heating rate of 108Cminꢀ1 calculated
at onset.
Chem. Eur. J. 2010, 16, 836 – 847
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
839

V. Gotor, S. V. Luis et al.
Table 4. Interaction energies for the computed (B3LYP/6-31+G*)
model ionic liquids.
of compounds, controlled structural modifications in ring
size, oxygen substitution, and cis/trans geometry were sys-
tematically introduced whilst keeping the bromide anion
and the benzyl group as common structural elements. The
corresponding melting temperatures and DSC traces are
shown in Figure 2.
The first conclusion that can be derived from Figure 2 is
that salts containing five-membered rings have lower melt-
ing points than those bearing six-membered rings (left side
in Figure 2), as previously observed for other series of salts.
Furthermore, the structural modifications have a more sig-
nificant impact in the C5 derivatives than in the C6 ones.
The nature of the OR groups also seems to play an impor-
tant role in determining the final macroscopic properties of
the chiral salt. Indeed, the melting point varies significantly
from the alcohols to the corresponding acetylated deriva-
tives. Finally, it can also be seen that for the isomeric molec-
ular chiral salts, the cis derivatives lead to lower melting
points than the corresponding trans isomers.
ꢀEint (syn)[a]
ꢀEint (anti)[a]
Entry
[kcalmolꢀ1
]
[kcalmolꢀ1
]
1
2
3
4
5
6
7
8
9
U
88.23
87.32
83.39
87.86
88.40
83.39
93.78
85.01
82.89
88.05
88.85
84.55
86.09
89.06
79.81
93.87
84.80
–
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
[a] The syn/anti conformations refer to the relative disposition between X
and the imidazolium C2-H group (see the Supporting Information for de-
tails).
The thermal stability of the ionic liquids was also assessed
by thermogravimetric analysis. In general, the TGA plots
for the different NTf2 derivatives show the high stability of
the compounds up to 2008C (see Supporting Information).
For example, when (S,S)-trans-Cy6-OH-Im-Bn-NTf2 was
heated at 1508C for a period of three hours, the compound
Our theoretical calculations also reflected a stabilization
effect of the OH group. Thus, elimination of the OH group
in the cyclohexane derivative weakens the anion–cation in-
teraction with bromide by around 6.2 kcalmolꢀ1 (Table 4,
entries 5 and 9). Moreover, acetylation of the OH group
also decreases the interaction energy by 5.7 kcalmolꢀ1
(Table 4, entries 5 and 6). Both experimental data and calcu-
lations strongly support the stabilization effect of the
OH···anion hydrogen-bonding interactions, although secon-
dary inter-cationic interactions (such as additional H bond-
ing, p–p or C-H···p bonding involving benzyl aromatic
rings), which were not considered in the calculations but
were observed in the crystal structures (see X-ray section
below), could also contribute significantly to the global sta-
bility of the salts. The presence of additional functionalities
in the ILs, which are able to form hydrogen-bond interac-
tions with the counterion, has a significant influence on
their final macroscopic properties[12,22]
As regards the conformation of the chiral cations, the imi-
dazolium group is located in one equatorial position of the
cycle in all cases. The cyclohexane derivatives have perfect
chair-like conformations, thus placing the OH group equato-
rial for the trans isomers and axial for cis isomers. The cyclo-
pentane rings display an envelope conformation. The imida-
zolium ring can have the C2-H moiety in two different posi-
tions: syn [for (R,R)-trans-Cy6-OAc-Im-Me-Br and for
(R,S)-cis-Cy6-OH-Im-Me-Br] or anti [for (S,S)-trans-Cy6-
OH-Im-Me-Cl, (S,S)-trans-Cy5-OH-Im-Bn-Br, and (R,S)-
cis-Cy5-OH-Im-B-Br] to the neighboring OR group. In
some cases, the favored dispositions obtained by theoretical
calculations (see entries 2 and 4 in Table 4) are in good
agreement with those observed experimentally in the solid
state (see X-ray section below). In contrast, both syn (for
bromide) and anti (for chloride) dispositions were found in
the solid state for the trans six-membered ring (see
below).[16] Interestingly, our theoretical model predicts a
1
did not show any signals of decomposition upon H NMR
spectroscopic analysis after cooling. We can therefore con-
clude that the disposition and geometry of both OH/OR
groups and the imidazolium ring are very important for de-
termining the macroscopic properties of the imidazolium
salts. The former data suggest that interactions between OH
and C2-H are likely to play a key role in determining the
structural properties of the CILs. This is also highlighted by
the fact that the bromide salts of cyclopentylbenzyl imidazo-
lium (Cy5-Im-Bn-Br) and cyclohexylbenzyl imidazolium
(Cy6-Im-Bn-Br) are liquids at room temperature.
Theoretical and spectroscopic studies: We performed theo-
retical quantum mechanics calculations in order to try to un-
derstand the differences in physical properties observed for
the CILs. DFT calculations at the B3LYP/6-31+G* level of
theory were performed on isolated ionic pairs of model sys-
tems bearing a methyl group at N3 of the imidazolium
ring.[21] Since we suspected that the physical properties
could be related to the inter-ion contacts, the anion–cation
interaction energy (Eint) was computed using the following
[Eqn (1)]:
Eint ¼ E ðisolated ionic pairÞꢀ½E ðisolated cationÞ
ð1Þ
þE ðisolated anionÞꢂ
Some results from the theoretical analysis are shown in
Table 4. As regards the effect of the anion, the interaction
ꢀ
energy decreases in the series Clꢀ >Brꢀ >BF4 (Table 4, en-
tries 5, 7 and 8), as expected on the basis of previously re-
ported ILs and in accordance with the melting temperatures
measured for our compounds.[6]
840
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 836 – 847

Modular Synthesis of Enantiopure Imidazolium Salts
FULL PAPER
[24,25]
ꢀ
small interaction energy difference (see entries 5 and 7 in
Table 4) between the syn and anti conformers. This agree-
ment represents a good validation of our theoretical model.
The theoretical calculations also reflect the differences be-
tween cis and trans isomers obtained experimentally from
the melting data. Thus, these calculations consistently pro-
vide larger interaction energies for the trans isomers (com-
pare entries 1 vs. 2 and 4 vs. 5 in Table 4). A closer inspec-
tion of the absolute values of the energies for the corre-
sponding optimized structures (Supporting Information)
shows that these differences are due to the higher stability
of the isolated cations for the cis compounds. This extra sta-
bilization could be due to a C2-H···OH hydrogen-bonding
interaction, which seems to be more favorable for the cis
isomers. However, the differences in Tm between cyclohex-
ane and cyclopentane derivatives are not reflected in the
computed interaction energy. This observation implies that
these differences are more likely due to lattice packing ef-
fects rather than an effect of the intrinsic cation–anion inter-
action. The volume of the cation with a C6 cycle is slightly
bigger than that with a C5 cycle. However, the latter has a
larger conformational freedom, which may well contribute
to increasing the number of possible conformations with an
effect on the melting point.
nostic of the aromatic C H bonds of the cations.
Vibra-
tional bands at higher wavenumbers in this region corre-
spond to C4-H and C5-H stretching modes whereas those at
lower wavenumbers can be assigned to C2-H stretching
modes.[26] The spectra of Cy6-Im-Bn-Br, a reference com-
pound with no OR groups (Figure 3a), shows broad bands
in this region. In the deconvoluted spectrum, the bands for
C4-H, C5-H, and C2-H stretching modes are found at 3123,
3059, and 3028 cmꢀ1, respectively, although these bands are
clearly differentiated for the cis and trans imidazolium salts
(Figure 3b and 3c). It is noteworthy that the bands at
around 3170 cmꢀ1, which are characteristic for the symmet-
ric stretching vibration modes of the HC(4)-C(5)H moiety
and are usually of weak intensity, are present in these spec-
tra, thus indicating a stronger interaction between the anion
and the cation, particularly for the trans diastereoisomer.
The spectra also show a band corresponding to the C2-H
stretching vibration with a lower vibrational frequency and
higher intensity [peaks at 3090 cmꢀ1 for the C(2)-H moiety
of (R,S)-cis-Cy6-OH-Im-Bn-Br and 3061 cmꢀ1 for (S,S)-
trans-Cy6-OH-Im-Bn-Br]. This also suggests a stronger C2-
H···X hydrogen bond and is in good agreement with the
data obtained by using other experimental techniques.[27]
ꢀ
The C2 H bond is weakened upon formation of a C2-H···X
The spectroscopic data also confirmed the differences ob-
served between the cis and trans stereoisomers. Hydrogen-
bonding effects in ionic liquids can be studied experimental-
ly by techniques such as IR[23] or NMR[24] spectroscopy. In
this case, the NMR spectra showed that the C2-H proton in
trans derivatives resonates at
hydrogen bond therefore the frequency of its stretching vi-
bration decreases. Such a decrease is usually accompanied
by an increase in intensity and broadening of the band.[28]
The same tendency is also observed when liquid cis/trans
diastereoisomeric imidazolium salt derivatives are compared
lower field than that in the cis
isomers [d=10.13 vs. 9.88 ppm
for (S,S)-trans-Cy5-OH-Im-Bn-
Br and (R,S)-cis-Cy5-OH-Im-
Bn-Br, respectively, and d=
9.93 vs. 9.75 ppm for (S,S)-
trans-Cy6-OH-Im-Me-Br and
(R,S)-cis-Cy6-OH-Im-Me-Br,
respectively], thereby suggest-
ing a weaker interaction be-
tween this proton and the
anion in the cis diastereo-
mers.[25] ATR-FTIR spectros-
copy can also be used to study
the interactions present in the
cis/trans stereoisomers. NMR
spectroscopy provides an aver-
age chemical shift for C2-H,
C4-H, and C5-H interacting
with the anion, whereas ATR-
FTIR gives separate vibration-
al bands for distinct species,
such as ion pairs, due to the
different time scales of these
techniques. In this case, the
Figure 3. ATR-IR spectra for the CH stretching region. Insets show the deconvoluted infrared spectra: a) Cy6-
Im-Bn-Br; b) (R,S)-cis-Cy6-OH-Im-Bn-Br; c) (S,S)-trans-Cy6-OH-Im-Bn-Br; d) (R,S)-cis-Cy6-OH-Im-Bu-Cl
(line) and (S,S)-trans-Cy6-OH-Im-Bu-Cl (dashed line).
bands between 3300 and
3000 cmꢀ1 are considered diag-
Chem. Eur. J. 2010, 16, 836 – 847
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
841

V. Gotor, S. V. Luis et al.
(see Figure 3d). For example, the spectrum of (S,S)-trans-
Cy6-OH-Im-Bu-Cl shows stronger peaks corresponding to
the C–H vibrational modes with an overall red shift when
compared to those of the cis diastereoisomer. All these data
suggest a more efficient interaction between the anion and
the cation for the trans isomer.[29]
2.715 ꢆ) for the cis isomer.[16,31] In principle, this would
mean a less compact lattice for the cis than for the trans dia-
stereomer. Such differences can be correlated with the de-
crease in Tm found for cis derivatives [Figure 2; Tm =188 vs.
1568C for (S,S)-trans-Cy6-OH-Im-Bn-Br and (R,S)-cis-Cy6-
OH-Im-Bn-Br, respectively].
We also analyzed the differences between the cis/trans rel-
ative configurations by means of a topological analysis of
the Laplacian of the electron density as obtained from Four-
ier maps.[32] This analysis provided us with more unambigu-
ous and incisive information regarding bonding interactions
than that obtainable from comparing interatomic distances
and allowed us to observe bonding beyond mere geometry.
This analysis also provided valuable information about the
topological magnitudes, which gives a rather accurate pic-
ture of the bonding. In (R,S)-cis-Cy6-OH-Bn-Br, for exam-
Solid-state structures—X-ray diffraction analysis: In order
to better understand the differences observed for these
chiral imidazolium salts, their intimate microscopic interac-
tions should be studied at the molecular level. In this
regard, we were able to grow crystals of several imidazolium
salts suitable for X-ray diffraction analysis (Figure 4). Fortu-
ple, the O atom in the OH group exhibits two large non-
2
bonding
charge
concentrations
(5 1=ꢀ9.8
and
ꢀ10.0 eꢆꢀ5), which correspond to lone pairs, and two small-
2
er bonded charge concentrations (5 1=ꢀ5.1 and
ꢀ4.8 eꢆꢀ5), which correspond to the C1–O1 and O1–H11
interactions, whereas the O atoms in the OH group of (S,S)-
2
trans-Cy6-OH-Bn-Br exhibit three such maxima (5 1=
ꢀ4.8, ꢀ4.7 and ꢀ10.1 eꢆꢀ5) for the C1–O1 and O1–H11 in-
teractions and one lone pair, respectively. However, an addi-
2
tional maximum of smaller magnitude (5 1=ꢀ7.3 eꢆꢀ5),
which indicates an additional noncovalently bonded charge
concentration between the electron pairs of O11 and H22 of
adjacent molecules, is also observed. The trans isomer there-
fore shows an additional interaction compared to the cis
compound involving OH and C2-H (C-22 in the crystal
structure) at the imidazolium ring.
Figure 4. X-ray diffraction studies for a) (S,S)-trans-Cy6-OH-Im-Bn-Br;
b) (R,S)-cis-Cy6-OH-Im-Bn-Br; c) crystal packing of (S,S)-trans-Cy6-OH-
Im-Bn-Br; d) crystal packing of (R,S)-cis-Cy6-OH-Im-Bn-Br.
nately, the effect of the cis/trans relative configurations of
the chiral centers could be systematically studied as we ob-
tained crystals for both cis- and trans-Cy6-OH-Im-Bn-Br
(Figure 4).[30] All the crystal structures showed short contacts
between the anion and different hydrogen atoms of the
chiral cation (See X-ray figure in the Supporting Informa-
tion).[16] Both the hydroxyl and imidazolium moieties are in-
volved in hydrogen bonding with the anion (either chloride
or bromide), with those hydrogen bonds involving the imi-
dazolium C2-H (C22-H in the crystal structure, Figure 4)
moiety being especially noteworthy.[27] If we evaluate the hy-
drogen-bonding ability on the basis of the donor–acceptor
distances a general trend can be seen for different X-H
donors. Thus, the hydrogen-bonding distances increase in
the sequence O-H<C22-H<C24-H/C25-H<C2-H<C26-H.
However, it should be noted that the OH and imidazolium
C2-H groups from a given cation are bound to different
anions within the crystal lattice (see Figure 4c and 4d).
A more careful study of the crystal packing in the cis and
trans forms of the cyclohexane derivatives [(S,S)-trans-Cy6-
OH-Im-Bn-Br; and (R,S)-cis-Cy6-OH-Im-Bn-Br] showed a
lower crystal density (1.3369 vs. 1.4128 gcmꢀ3), larger cell
volume (1675.3 vs. 1585.3 ꢆ3), a lower number of intermo-
lecular interactions per molecule (13 vs. 14), and a longer
mean distance for the intermolecular contacts (2.826 vs.
Overall, the structural data obtained by X-ray diffraction,
NMR and ATR-FTIR spectroscopy, and theoretical calcula-
tions suggest stronger cation–anion interactions for the trans
isomers than for the cis ones, which nicely correlates with
the observed physical (Tm) properties of the ionic liquids.
Moreover, our structural study located these distinguishing
interactions at the OH group and the C2-H moiety of the
imidazolium ring. Apart from anion–cation interactions, the
topological analysis of experimental electron densities, as re-
constructed from X-ray diffraction data for diastereomeric
crystals, has allowed us to locate a noncovalent cation–
cation bonding which must also have an important influence
on the macroscopic properties.
Racemic versus enantiopure chiral imidazolium salts: We
have demonstrated the key role played by the relative con-
figuration of the chiral centers in the non-covalent interac-
tions at the molecular level and, ultimately, in the physical
properties of the CILs. In order to further confirm the im-
portance of chirality, we also measured the melting tempera-
tures of the corresponding racemic mixtures of the salts and
found significant differences between the racemic and enan-
tiopure compounds (see Table 5). Thus, with the exception
of trans-Cy6-OH-Im-Bn-Br, the melting points of the race-
mic mixtures were found to be higher than those of the
842
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 836 – 847

Modular Synthesis of Enantiopure Imidazolium Salts
FULL PAPER
Table 5. Effect of the enantiopurity on the melting points (Tm) for some
cis- and trans-imidazolium salts.
significant differences. Thus, the oxygen atom in the OH
moiety of enantiopure (S,S)-trans-Cy5-OH-Im-Bn-Br exhib-
Tm [8C][a]
Tm [8C][a]
2
Imidazolium salt
its two large nonbonded charge concentrations (5 1=ꢀ10.1
and ꢀ10.3 eꢆꢀ5), which correspond to lone pairs, and two
(enantiopure)
(racemic mixture)
2
trans-Cy6-OH-Im-Bn-Br
cis-Cy6-OH-Im-Bn-Br
trans-Cy5-OH-Im-Bn-Br
cis-Cy5-OH-Im-Bn-Br
188 (S,S)
156 (R,S)
112 (S,S)
19 (R,S)
181 (ꢁ)
166 (ꢁ)
168 (ꢁ)
300 (ꢁ)
smaller bonded charge concentrations (5 1=ꢀ5.3 and
ꢀ5.1 eꢆꢀ5), which correspond to C1–O1 and O1–H11 inter-
actions. In the racemic compound [(ꢁ)-trans-Cy5-OH-Im-
Bn-Br], in contrast, the O atoms in the OH moiety exhibit
[a] Melting temperatures for a heating rate of 58Cminꢀ1 calculated at
2
three such maxima (5 1=ꢀ5.2, ꢀ5.1, and=ꢀ10.3 eꢆꢀ5) for
onset.
the C1–O1 and O1–H11 interactions and one lone pair, re-
2
spectively. However, a maxima of lower magnitude (5 1=
enantiomerically pure forms. This observation suggested
that CILs tend to interact to form stable racemic conglomer-
ates.
Fortunately, we also obtained experimentally conclusive
data to support this assumption at the molecular level as we
were able to grow crystals of a chiral imidazolium salt in
both its enantiopure [(S,S)-trans-Cy5-OH-Im-Bn-Br, Fig-
ꢀ7.1 eꢆꢀ5), which suggests an additional noncovalently
bonded charge concentration between the electron pairs of
O11 and H22 of adjacent molecules within the racemic crys-
tal, is also observed. We therefore expanded the racemic
crystal structure through the contacts corresponding to this
additional interaction and obtained a heterochiral (racemic)
non-covalent dimer connected through O1 and H22 of the
imidazolium moiety and interrelated by a center of symme-
try (one molecule with an R,R and the other with an S,S rel-
ative configuration; see Figure 5e). These data strongly sup-
port our initial hypothesis and correlate nicely with the
melting values, thus highlighting the strong impact of chirali-
ty on the physical properties of imidazolium salts.
ure 5a] and racemic [(ꢁ)-trans-Cy5-OH-Im-Bn-Br, Fig-
Both the racemic and enantiopure salts were also studied
by ATR-FTIR spectroscopy. In general, the absorptions for
the vibrational modes of OH and imidazolium C–H aromat-
ic groups in the racemic salts appear at lower frequency, but
are of higher intensity than those for the enantiopure form
(Figure 6). It should be noted that the different trend in
melting points observed for (S,S)-trans-Cy6-OH-Im-Bn-Br
and the racemic mixture is also reflected in the C–H vibra-
tion modes of their imidazoliun rings, with the peaks corre-
sponding to these stretching modes being slightly red-shifted
compared to those in the spectra of the racemic mixture.
Thus, there is once again a perfect correlation between the
binding interactions obtained from the topological analysis
and the “classical” hydrogen-bonding interaction character-
ized by FT-IR spectroscopy, which ultimately allows us to
explain the observed differences in physical macroscopic
properties (Tm).
Figure 5. X-ray analysis and corresponding crystal packing of enantiopure
(S,S)-trans-Cy5-OH-Im-Bn-Br (a and c) and racemic (ꢁ)-trans-Cy5-OH-
Im-Bn-Br (b and d). An expansion of the heterochiral dimer observed in
the crystal of the racemic form is shown in e).
Conclusions
We have reported the chemoenzymatic modular synthesis
and physical/structural study of a new family of more than
30 enantiopure imidazolium salts. This has allowed us to sys-
tematically map all the possible structural variables within
the chiral molecules. The melting temperatures of these
compounds have been found to differ significantly. Some of
these variables (alkylating group or nature of the anion)
were found to have similar effects to those reported previ-
ously for imidazolium-based ILs. Other structural vectors,
such as cycloalkane ring size, the presence of polar groups
in the cation structure and, especially, the configurations of
the chiral centers have also been systematically studied. In
ACHTUNGTRENNUNG
,
respectively). Furthermore, the disposition of the imidazoli-
um ring with respect to the OH group is syn in the racemic
compound (Figure 5b), which contrasts with the configura-
tion obtained for the enantiopure crystal (Figure 5a) and
that obtained from theoretical calculations on isolated ion
pairs (Table 4). This conformational change must be pro-
duced by additional cation–cation interactions within the
racemic compound. The topological analysis also produced
Chem. Eur. J. 2010, 16, 836 – 847
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
843

V. Gotor, S. V. Luis et al.
tions between the OH group and the imidazolium ring.
Thus, in one case we observed the formation of heterochiral
(racemic) conglomerates through non-covalent interactions
between OH and C2-H, which increases the melting temper-
ature of the racemates.
Overall, we have been able to prepare and characterize a
broad family of new enantiopure CILs by systematic struc-
tural modifications. Moreover, the differences in macroscop-
ic properties (Tm) have been successfully correlated with
non-covalent interactions observed by microscopic structural
data from experimental (NMR, FT-IR-ATR, X-ray) and
theoretical studies. We expect that these findings will help
us to better understand the properties of these neoteric sol-
vents, especially the effect of chirality. A full understanding
will allow an improved bottom-up design and lead to the
preparation of new CILs with preselected properties.
Our study demonstrates that chirality has significant
impact on the physical properties of imidazolium salts. The
relative configuration of the chiral centers and the presence
of both enantiomers within the mixture also significantly af-
fects the macroscopic properties. These results show that
compounds with a similar structure do not necessarily have
similar interactions.
Experimental Section
Materials: Candida antarctica lipase type B (CAL-B, Novozyme 435,
7300 PLU/g) was a gift from Novozymes. Pseudomonas cepacia lipase
PSL-C II (1019 U/g) and PSL-C I (1638 U/g) were purchased from
Amano Pharmaceutical Co. and Sigma–Aldrich respectively. Porcine pan-
creas lipase (PPL, 46 U/mg) and Candida cylindracea lipase (CRL, 1141
U/mg) were purchased from Sigma. Candida antarctica lipase type A
[Chirazyme l-5, c-f, lyophilized, (1000 U/g using tributyrin)] was pur-
chased from Roche. All other reagents were purchased from different
commercial sources and used without further purification. Solvents were
distilled over an appropriate desiccant under nitrogen. Flash chromatog-
Figure 6. ATR-FT-IR spectra for enantiopure (red lines) and racemic
mixtures (blue dashed lines) of: a) (S,S)-trans-Cy6-OH-Im-Bn-Br;
b) (R,S)-cis-Cy6-OH-Im-Bn-Br;
d) (R,S)-cis-Cy5-OH-Im-Bn-Br.
c) (S,S)-trans-Cy5-OH-Im-Bn-Br;
general, five-membered rings lead to salts with lower melt-
ing temperatures than six-membered rings, probably due to
a different degree of conformational freedom. The presence
of hydrogen-bonding groups also increases the Tm of the
imidazolium salts. Finally, a trans configuration also increas-
es Tm with respect to the cis isomers. A multidisciplinary ap-
proach involving spectroscopic (NMR, FT-IR-ATR) and
theoretical studies has shown that the observed differences
mainly arise from the anion–cation interactions. Thus, we
obtained a reasonable correlation between the observed
melting points and the theoretical anion–cation interaction
energies for model compounds. Furthermore, the spectro-
scopic data reflect the key role played by hydrogen bonds
between the OH and imidazolium C2H moieties, and an X-
ray diffraction study (including topological analysis of cis/
trans isomers) of several compounds allowed us to locate ad-
ditional inter-cation interactions involving these groups.
These interactions are a major cause of the differences in
properties of stereoisomeric salts.
1
raphy was performed on silica gel 60 (230–240 mesh). H, 13C, DEPT, and
1H-13C heteronuclear NMR experiments were performed on AV-300 (1H:
300.13 MHz; 13C: 75.5 MHz), DPX 300 (1H: 300.13 MHz; 13C: 75.5 MHz),
AV-400 (1H: 400.13 MHz; 13C: 100.6 MHz), or AV-600 (1H; 600.13 MHz)
spectrometers. Chemical shifts are given in delta (d) units and the cou-
pling constants (J) in hertz (Hz). ESI+ or APCI+ mass spectra were re-
corded with an HP1100 chromatograph mass spectrometer. Microanaly-
ses were performed using a Perkin–Elmer model 2400 instrument. Opti-
cal rotations were measured with a Perkin–Elmer 241 polarimeter. High
performance liquid chromatography (HPLC) analyses were completed by
using a Hewlett Packard 1100 chromatograph UV detector at 210 nm
using a Daicel CHIRALCEL OD, CHIRALCEL OB-H, or a CHIRAL-
PAK AS column (25 cm ꢅ 4.6 mm I.D.). Further detailed data are pro-
vided in the Supporting Information.
General procedures for substrate synthesis
(ꢁ)-trans-3a: Cyclopentene oxide (3.18 mL, 36.72 mmol) was added to a
solution of imidazole (2.00 g, 29.38 mmol) in 1,4-dioxane (16 mL) and the
solution stirred for 18 h at 1008C. The solvent was then evaporated
under reduced pressure and the resulting crude mixture purified by flash
chromatography (10% MeOH/CHCl3) to give 3.92 g of (ꢁ)-trans-3a as a
white solid (88% yield).
Even more interestingly, we also studied the effect of chir-
ality by comparing enantiopure CILs with their racemic
mixtures. The surprising differences observed can also be ex-
plained by the corresponding spectroscopic and crystallo-
graphic data and are also intimately related with interac-
(ꢁ)-trans-4a: Et3N (277 mL, 1.97 mmol), DMAP (27 mg, 0.22 mmol), and
Ac2O (123 mL, 1.31 mmol) were added successively to a solution of (ꢁ)-
trans-3a (100 mg, 0.66 mmol) in dry CH2Cl2 (10 mL) under nitrogen. The
mixture was stirred at room temperature for 4 h until complete consump-
844
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 836 – 847

Modular Synthesis of Enantiopure Imidazolium Salts
FULL PAPER
tion of the starting material, then the solvent was evaporated under re-
duced pressure. The crude reaction mixture was then purified by flash
chromatography on silica gel (5% MeOH/CHCl3) to give 125 mg of (ꢁ)-
trans-4a as a pale-yellow oil (97% yield).
90–1008C and 10ꢀ2–10ꢀ3 mbar for 4 h, and was further dried in situ in the
differential scanning calorimeter by holding the sample at 1308C for
120 min as the presence of volatiles, especially water, can affect the glass-
transition and melting temperatures. Melting transition temperatures
were determined by multiple cycles (typically three) involving heating
the sample from ꢀ65 to 2508C, followed by cooling from 250 to ꢀ658C,
both at a rate of 58Cminꢀ1. The melting temperatures were determined
at the onset of the transition.
Enzymatic kinetic resolution of racemic cis- and trans-alcohols 3a–b:
Vinyl acetate (1.64 mL, 17.82 mmol) was added to a suspension of the
racemic alcohol (5.94 mmol) and the corresponding enzyme (1:1 weight
ratio with respect to the alcohol) in dry THF (59 mL) under nitrogen and
the mixture shaken, depending on the substrate, at between 30 and 608C
at 250 rpm. Aliquots were regularly analyzed by HPLC until 50% con-
version, then the reaction was quenched and the enzyme filtered and
washed with CH2Cl2 (3ꢅ50 mL). The solvent was evaporated under re-
duced pressure and the crude reaction mixture purified by flash chroma-
tography on silica gel (5–10% MeOH/CHCl3) to afford the correspond-
ing optically enriched acetate and alcohol (see Tables 1 and 2).
Acknowledgements
Financial support from the Ministerio de Ciencia e Innovaciꢂn (MICINN
CTQ2007-61126, CTQ2008-04415, and CTQ2008-04309), CSIC-I3
(200780I001), and BANCAIXA (P1 1A2007-11) is gratefully acknowl-
edged. V.G.-F and E.G-V. (Ramꢂn y Cajal Program), and R.P. (FPU)
thank the MICINN for personal funding. N.R-L. thanks the FICYT for a
pre-doctoral fellowship.
Mitsunobu reaction: The corresponding acid (4.32 mmol), PPh3 (1.36 g,
5.18 mmol), and DEAD (0.82 mL, 5.18 mmol) were successively added to
a solution of the corresponding trans-alcohol 3a–b (4.32 mmol) in dry
THF (22 mL) under nitrogen. The mixture was stirred for 2 h until start-
ing material could no longer be detected by TLC analysis. The organic
solvent was evaporated under reduced pressure and the crude mixture
used directly for the deprotection step.
[1] a) S. Z. E. Abedin, F. Endres, Acc. Chem. Res. 2007, 40, 1106–1113;
b) R. D. Rogers, G. A. Voth, Acc. Chem. Res. 2007, 40, 1077–1078;
c) R. D. Rogers, K. R. Seddon, Ionic Liquids III: Fundamentals,
Progress, Challenges, and Opportunities, ACS, Washington, 2005;
d) P. Wasserscheid, T. Welton, Ionic Liquids in Synthesis, Wiley-
VCH, Weinheim, 2007.
Deprotection step: K2CO3 (595 mg, 4.32 mmol) and H2O (6.8 mL) were
added to a solution of the crude product from the Mitsunobu reaction in
MeOH (6.8 mL). This mixture was stirred for 1 h and then the MeOH
was evaporated under reduced pressure. The resulting suspension was re-
dissolved in H2O (10 mL), a solution of brine was added (5 mL), and the
mixture extracted with EtOAc (3ꢅ10 mL). The corresponding cis-alcohol
was isolated as a white solid (78–85% overall yield for both steps) by
flash chromatography on silica gel (5–10% MeOH/CH2Cl2).
[2] a) N. V. Plechkova, K. R. Seddon, P. Tundo, A. Perosa, F. Zecchini,
Methods and Reagents for Green Chemistry, Wiley, New York, 2007,
pp. 103–130; b) R. D. Rogers, K. R. Seddon, Ionic Liquids as Green
Solvents, Progress and Prospects, ACS, Washington, 2003; c) R. D.
Rogers, K. R. Seddon, S. Volkov, Green Industrial Applications of
Ionic Liquids, Kluwer Academic, Norwell, 2003; d) R. D. Rogers,
K. R. Seddon, Ionic Liquids: Industrial Applications for Green
Chemistry, ACS, Washington 2002; e) M. J. Earle, K. R. Seddon,
Pure Appl. Chem. 2000, 72, 1391–1398.
Reaction between imidazole derivatives and benzyl bromide
a) trans-derivatives: A mixture of the alcohol (3 mmol) and benzyl bro-
mide (3.62 mmol) in MeCN (2.3 mL) was stirred for 2.5 h at 708C, then
the reaction was cooled to room temperature and the resulting gummy
product washed with Et2O (5ꢅ10 mL) to afford the corresponding bro-
mide salt as a white solid (93–96% yield).
ˇ
ˇ
[3] a) S. Toma, M. Meꢇiarovꢁ, R. Sebesta, Eur. J. Org. Chem. 2009,
321–327; b) N. V. Plechkova, K. R. Seddon, Chem. Soc. Rev. 2008,
37, 123–150; c) S. Chowdhury, R. S. Mohan, J. L. Scott, Tetrahedron
2007, 63, 2363–2389; d) N. Jain, A. Kumar, S. Chauhan, S. M. S.
Chauhan, Tetrahedron 2005, 61, 1015–1060; e) T. Welton, Coord.
Chem. Rev. 2004, 248, 2459–2477; f) P. Wasserscheid, W. Keim,
Angew. Chem. 2000, 112, 3926–3945; Angew. Chem. Int. Ed. 2000,
39, 3772–3789; g) T. Welton, Chem. Rev. 1999, 99, 2071–2083.
[4] a) X. Han, D. W. Armstrong, Acc. Chem. Res. 2007, 40, 1079–1086;
b) S. T. Anjan, Chem. Eng. Prog. 2006, 102, 30–39; c) H. Zhao, S.
Xiao, P. Ma, J. Chem. Technol. Biotechnol. 2005, 80, 1089–1096;
d) A. E. Visser, R. P. Swatloski, W. M. Reichert, R. Mayton, S. Sheff,
A. Wierzbicki, J. H. Davis, R. D. Rogers, Environ. Sci. Technol.
2002, 36, 2523–2529; e) J. G. Huddleston, R. D. Rogers, Chem.
Commun. 1998, 1765–1766.
b) cis-derivatives: A mixture of the alcohol (3 mmol) and benzyl bromide
(60 mmol) was stirred for 2.5 h at 708C, then the reaction was cooled to
room temperature and the resulting gummy product washed with Et2O
(5ꢅ10 mL) to afford the corresponding bromide salt as a white solid (96–
97% yield).
Anion-exchange reaction with LiNTf2: A solution of lithium bis(trifluoro-
methane) sulfonimide (467 mg, 1.63 mmol) in H2O (1.5 mL) was added
to a solution of the corresponding bromide or chloride salt (1.48 mmol)
in MeOH (40 mL), and the resulting mixture stirred for the required
time at room temperature (see Table 3). MeOH was then evaporated at
reduced pressure, further H2O (20 mL) added, and the mixture extracted
with CH2Cl2 (3ꢅ20 mL). The organic phases were combined, dried over
Na2SO4, and the solvent evaporated under reduced pressure to afford the
corresponding imidazole salt as a light yellowish liquid (78–97% yield).
Anion-exchange reaction with NaBF4: A solution of sodium tetrafluoro-
borate (92 mg, 0.84 mmol) in H2O (580 mL) was added to a solution of
the corresponding bromide or chloride salt (0.56 mmol) in MeOH
(14.4 mL) and the resulting mixture stirred for 90 h at room temperature.
The solvent was then evaporated under reduced pressure and the result-
ing solid residue washed with CH2Cl2 (3ꢅ8 mL). The insoluble salts were
filtered off and the solvent evaporated under reduced pressure to afford
the corresponding imidazole salt as a light yellowish liquid (85–94%
yield).
[5] D. MacFarlane, M. Forsyth, The Handbook of Ionic Liquids: Elec-
trochemistry, Wiley-VCH, Weinheim, 2008.
[6] a) H. Weingꢈrtner, Angew. Chem. 2008, 120, 664–682; Angew.
Chem. Int. Ed. 2008, 47, 654–670; b) H. Tokuda, K. Ishii,
M. A. B. H. Susan, S. Tsuzuki, K. Hayamizu, M. Watanabe, J. Phys.
Chem. B 2006, 110, 2833–2839; c) H. Tokuda, K. Hayamizu, K.
Ishii, M. A. B. H. Susan, M. Watanabe, J. Phys. Chem. B 2005, 109,
6103–6110; d) Z.-B. Zhou, H. Matsumoto, K. Tatsumi, Chem. Eur.
J. 2004, 10, 6581–6591; e) H. Tokuda, K. Hayamizu, K. Ishii,
Differential scanning calorimetry: Glass-transition temperatures and
melting points were measured by using a Mettler–Toledo differential
scanning calorimeter (DSC), model DSC822e. The instrument was cali-
brated for temperature and heat flow with zinc and indium reference
samples provided by Mettler-Toledo. Samples were placed in a 40-mL,
hermetically sealed aluminum pan with a pinhole in the top. An empty
aluminum pan was used as the reference. Samples were exposed to a
flowing N2 atmosphere. Before the DSC test, each sample was dried at
M. A. B. H. Susan, M. Watanabe, J. Phys. Chem.
B 2004, 108,
16593–16600; f) J. G. Huddleston, A. E. Visser, W. M. Reichert,
H. D. Willauer, G. A. Broker, R. D. Rogers, Green Chem. 2001, 3,
156–164; g) J. D. Holbrey, K. R. Seddon, J. Chem. Soc. Dalton
Trans. 1999, 2133–2140.
[7] a) W. L. Hough-Troutman, M. Smiglak, S. Griffin, W. M. Reichert, I.
Mirska, J. Jodynis-Liebert, T. Adamska, J. Nawrot, M. Stasiewicz,
R. D. Rogers, J. Pernak, New J. Chem. 2009, 33, 26–33; b) M. Smi-
Chem. Eur. J. 2010, 16, 836 – 847
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
845

V. Gotor, S. V. Luis et al.
glak, A. Metlen, R. D. Rogers, Acc. Chem. Res. 2007, 40, 1182–
1192.
of the terms Cy5 for cyclopentane rings and Cy6 for cyclohexane
rings [for instance: (S,S)-trans-Cy5]. The fourth descriptor indicates
the presence of OH or OR at C2 of the cycloalkyl ring [for instance:
(S,S)-trans-Cy5-OH-]. The nature of the alkyl chain at N3 of the
imidazolium moiety is indicated by the use of Bn (for benzyl), Bu
(for butyl), or Oct (for octyl) [for instance: (S,S)-trans-Cy5-OH-Bu-
]. Finally, the term Im refers to the central core of imidazolium,
while the last descriptor refers to the specific anion included in the
structure [for instance: (S,S)-trans-Cy5-OH-Bu-Im-Cl].
[8] a) K. E. Gutowski, E. J. Maginn, J. Am. Chem. Soc. 2008, 130,
14690–14704; b) M. D. Soutullo, C. I. Odom, B. F. Wicker, C. N.
Henderson, A. C. Stenson, J. H. Davis, Jr., Chem. Mater. 2007, 19,
3581–3583; c) S. Lee, Chem. Commun. 2006, 1049–1063; d) J. H.
Davis, Jr., Chem. Lett. 2004, 33, 1072–1077.
[9] For general reviews of chiral ionic liquids, see: a) K. Bica, P. Gaert-
ner, Eur. J. Org. Chem. 2008, 3235–3250; b) A. Winkel, P. V. G.
Reddy, R. Wilhelm, Synthesis 2008, 999–1016; c) X. Chen, X. Li, A.
Hu, F. Wang, Tetrahedron: Asymmetry 2008, 19, 1–14; d) A. D.
Headley, B. Ni, Aldrichimica Acta 2007, 40, 107–117; e) C. Baude-
quin, D. Brjgeon, J. Levillain, F. Guillen, J.-C. Plaquevent, A.-C.
Gaumont, Tetrahedron: Asymmetry 2005, 16, 3921–3945; f) J. Ding,
D. W. Armstrong, Chirality 2005, 17, 281–292; g) C. Baudequin, J.
Baudoux, J. Levillain, D. Cahard, A.-C. Gaumont, J.-C. Plaquevent,
Tetrahedron: Asymmetry 2003, 14, 3081–3093.
[10] a) K. Schneiders, A. Bçsmann, P. S. Schulz, P. Wasserscheid, Adv.
Synth. Catal. 2009, 351, 432–440; b) D. Chen, M. Schmitkamp, G.
Franciꢉ, J. Klankermayer, W. Leitner, Angew. Chem. 2008, 120,
7449–7451; Angew. Chem. Int. Ed. 2008, 47, 7339–7341; c) T. Yu, T.
Yamada, G. C. Gaviola, R. G. Weiss, Chem. Mater. 2008, 20, 5337–
5344; d) T. Yamada, P. J. Lukac, T. Tao Yu, R. G. Weiss, Chem.
Mater. 2007, 19, 4761–4768; e) B. Ni, A. D. Headley, Tetrahedron
Lett. 2006, 47, 7331–7334; f) S. V. Malhotra, Y. Wang, Tetrahedron:
[20] See the Supporting Information for a complete list of yields and
Tmꢊs for all the ionic liquids prepared in this work.
[21] a) R. Ludwig, Phys. Chem. Chem. Phys. 2008, 10, 4333–4339;
b) B. L. Bhargava, S. Balasubramanian, M. L. Klein, Chem.
Commun. 2008, 3339–3351; c) S. A. Katsyuba, E. E. Zvereva, A.
ˇ
Vidis, P. J. Dyson, J. Phys. Chem. A 2007, 111, 352–370; d) H. Mar-
kusson, J. P. Belires, P. Johansson, C. A. Angell, P. Jacobsson, J.
Phys. Chem. A 2007, 111, 8717–8723; e) P. A. Hunt, B. Kirchner, T.
Welton, Chem. Eur. J. 2006, 12, 6762–6775; f) C. J. Margulis, Mol.
Phys. 2004, 102, 829–838; g) A. T. Elizabeth, C. C. Pye, R. D.
Singer, J. Phys. Chem. A 2003, 107, 2277–2288.
[22] a) G. D. Smith, O. Borodin, L. Li, H. Kim, Q. Liu, J. E. Bara, D. L.
Ginc, R. Nobel, Phys. Chem. Chem. Phys. 2008, 10, 6301–6312;
b) Q. Zhang, Z. Li, J. Zhang, S. Zhang, L. Zhu, J. Yang, X. Zhang,
Y. Deng, J. Phys. Chem. B 2007, 111, 2864–2872; c) Z. Fei, W. H.
Ang, D. Zhao, R. Scopelliti, E. E. Zvereva, S. A. Katsyuba, P. J.
Dyson, J. Phys. Chem. B 2007, 111, 10095–10108; d) J. Y. Z. Chiou,
J. N. Chen, J. S. Lei, I. J. B. Lin, J. Mater. Chem. 2006, 16, 2972–
2977.
[23] a) O. Hçfft, S. Bahr, V. Kempter, Langmuir 2008, 24, 11562–11566;
b) K. Fumino, A. Wulf, R. Ludwig, Angew. Chem. 2008, 120, 3890–
3894; Angew. Chem. Int. Ed. 2008, 47, 3830–3834; c) A. Yokozeki,
D. J. Kasprzak, M. B. Shiflett, Phys. Chem. Chem. Phys. 2007, 9,
5018–5026; d) N. E. Heimer, R. E. Del Sesto, Z. Meng, J. S. Wilkes,
W. R. Carper, J. Mol. Liq. 2006, 124, 84–95; e) S. A. Katsyuba, P. J.
´
Asymmetry 2006, 17, 1032–1035; g) C. Luys, L. C. Branco, P. M. P.
Gois, N. M. T. LourenÅo, V. B. Kurteva, C. A. M. Afonso, Chem.
Commun. 2006, 2371–2372; h) F. Guillen, D. Bregeond, J.-C. Pla-
quevent, Tetrahedron Lett. 2006, 47, 1245–1248; i) R. Gausepohl, P.
Buskens, J. Kleinen, A. Bruckmann, C. W. Lehmann, J. Klankermay-
er, W. Leitner, Angew. Chem. 2006, 118, 3772–3775; Angew. Chem.
Int. Ed. 2006, 45, 3689–3692.
[11] a) J.-C. Plaquevent, J. Levillain, F. Guillen, C. Malhiac, A.-C. Gau-
mont, Chem. Rev. 2008, 108, 5035–5060; b) H. Ohno, K. Fukumoto,
Acc. Chem. Res. 2007, 40, 1122–1129.
ˇ
Dyson, E. E. Vandyukova, A. V. Chernova, A. Vidis, Helv. Chim.
[12] a) F. Neve, A. Crispini, CrystEngComm 2007, 9, 698–703; b) D.
Zhao, Z. Fei, W. H. Ang, R. Scopelliti, P. J. Dyson, Eur. J. Inorg.
Chem. 2007, 279–284; c) N. K. Sharma, M. D. Tickell, J. L. Ander-
son, J. Kaar, V. Pino, B. F. Wicker, D. W. Armstrong, J. H. Davis, Jr.,
A. J. Russell, Chem. Commun. 2006, 646–648; d) Z. Fei, T. J. Geld-
bach, D. Zhao, P. J. Dyson, Chem. Eur. J. 2006, 12, 2122–2130;
e) J. Y. Z. Chiou, J. N. Chen, J. S. Lei, I. J. B. Lin, J. Mater. Chem.
2006, 16, 2972–2977; f) D. Zhao, Z. Fei, R. Scopelliti, P. J. Dyson,
Inorg. Chem. 2004, 43, 2197–2205; g) L. C. Branco, J. N. Rosa, J. J.
Moura Ramos, C. A. M. Afonso, Chem. Eur. J. 2002, 8, 3671–3677.
[13] A. J. Bailey, C. Lee, R. K. Feller, J. B. Orton, C. Mellot-Draznieks,
B. Slater, W. T. A. Harrison, P. Simoncic, A. Navrotsky, M. C. Gros-
sel, A. K. Cheetham, Angew. Chem. 2008, 120, 8762–8765; Angew.
Chem. Int. Ed. 2008, 47, 8634–8637.
[14] I. Krossing, J. M. Slattery, C. Daguenet, P. J. Dyson, A. Oleinikova,
H. Weingꢈrtner, J. Am. Chem. Soc. 2006, 128, 13427–13434.
[15] R. Torregrosa, I. M. Pastor, M. Yus, Tetrahedron 2007, 63, 469–473.
[16] Some preliminary synthetic results regarding the preparation of
chiral imidazolium salts have been reported previously: E. Busto, N.
Rꢀos-Lombardꢀa, V. Gotor-Fernꢁndez, E. Garcꢀa-Verdugo, I. Alfon-
so, A. Menꢃndez-Velꢁzquez, S. Garcꢀa-Granda, M. I. Burguete, S. V.
Luis, V. Gotor, Tetrahedron Lett. 2007, 48, 5251–5254.
Acta 2004, 87, 2556–2565.
[24] a) R. C. Remsing, J. L. Wildin, A. L. Rapp, G. Moyna, J. Phys.
Chem. B 2007, 111, 11619–11621; b) J. Palomar, V. R. Ferro, M. A.
Gilarranz, J. J. Rodriguez, J. Phys. Chem. B 2007, 111, 168–180;
c) S. H. Chung, R. Lopato, S. G. Greenbaum, H. Shirota, E. W. Cast-
ner, Jr., J. F. Wishart, J. Phys. Chem. B 2007, 111, 4885–4893; d) D.
Bankmann, R. Giernoth, Prog. Nucl. Magn. Reson. Spectrosc. 2007,
51, 63–90; e) J.-F. Huang, P.-Y. Chen, I.-W. Sun, S. P. Wang, Inorg.
Chim. Acta 2001, 320, 7–11.
[25] T. Kçddermann, C. Wertz, A. Heintz, R. Ludwig, ChemPhysChem
2006, 7, 1944–1949.
[26] K. Fumino, A. Wulf, R. Ludwig, Angew. Chem. 2008, 120, 8859–
8862; Angew. Chem. Int. Ed. 2008, 47, 8731–8734.
[27] a) S. Zahn, F. Uhlig, J. Thar, C. Spickermann, B. Kirchner, Angew.
Chem. 2008, 120, 3695–3697; Angew. Chem. Int. Ed. 2008, 47, 3639–
3641; b) S. Tsuzuki, H. Hiroyuki Tokuda, M. Mikami, Phys. Chem.
Chem. Phys. 2007, 9, 4780–4784; c) K. Dong, S. Zhang, D. Wang, X.
Yao, J. Phys. Chem. A 2006, 110, 9775–9782.
[28] A. V. Iogansen, Spectrochim. Acta Part A 1999, 55, 1585–1612.
[29] The computed vibrational frequencies of the C2 H bond for the ob-
ꢀ
tained minima showed similar trends to the experimental ones (see
the Supporting Information). Thus, the trans isomers showed ab-
sorptions at lower frequencies than the cis ones, in good accordance
with the experimental data and reflecting the somehow stronger
anion–cation interaction in the trans diastereoisomer. Although the
computed (uncorrected) quantitative values of vibrational frequen-
cies are far from the experimental ones, we believe that the correla-
tion is qualitatively acceptable considering the number of approxi-
mations made.
[17] See the Supporting Information for additional data.
[18] L. M. Levy, I. Lavandera, V. Gotor, Org. Biomol. Chem. 2004, 2,
2572–2577.
[19] The nomenclature of imidazolium salts is defined by seven descrip-
tors: The first one includes the absolute configuration of C1 and C2
at the cycloalkyl ring [for in-
stance: (S,S)-]. The relative con-
figuration at these carbon atoms is
reported by the second descriptor
trans or cis [for instance: (S,S)-
trans]. The size of the cycloalkyl
ring is described through the use
[30] We also attempted to record circular dichroism spectra to further
support these observations. Unfortunately, the very low UV absorp-
tion of the imidazolium group precluded suitable spectra from being
obtained.
846
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 836 – 847

Modular Synthesis of Enantiopure Imidazolium Salts
FULL PAPER
[31] All these parameters were calculated with the Mercury 1.3.2 soft-
ware, developed by The Cambridge Crystallographic Data Centre
and freely downloadable at: www.ccdc.cam.ac.uk/products/mercury/.
[32] a) R. F. W. Bader, J. Phys. Chem. A 2007, 111, 7966–7972; b) “The
Lagrangian Approach to Chemistry”: R. F. W. Bader, C. F. Matta, R.
Boyd, in Quantum Theory of Atoms in Molecules: From Solid State
to DNA and Drug Design, Wiley-VCH, Weinheim, 2007, pp. 37–59;
c) A. Menꢃndez-Velꢁzquez, S. Garcꢀa Granda, J. Appl. Crystallogr.
2003, 36, 193–205; d) R. F. W. Bader, Chem. Rev. 1991, 91, 893–928;
e) R. F. W. Bader, Atoms in Molecules—A Quantum Theory, Oxford
University Press, Oxford, 1990; f) R. F. W. Bader, R. J. Gillespie,
P. J. MacDougall, J. Am. Chem. Soc. 1988, 110, 7329–7336.
Received: June 14, 2009
Published online: November 27, 2009
Chem. Eur. J. 2010, 16, 836 – 847
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
847