T-shaped ionic liquid crystals based on the imidazolium motif: Exploring substitution of the C-2 imidazolium carbon atom

Article abstract of DOI:10.1002/chem.201001921

In this contribution the first examples of so-called rigid-core, T-shaped imidazolium ionic liquid crystals, in which the C-2 atom of the imidazolium ring is substituted with an aryl moiety decorated with one or two alkoxy chains, are described. The length of the alkoxy chain(s) was varied from six to eighteen carbon atoms (n=6, 10, 14-18). Whereas the compounds with one long alkoxy chain display only smectic A phases, the salts containing two alkoxy chains exhibit smectic A, multicontinuous cubic, as well as hexagonal columnar phases, as evidenced by polarising optical microscopy, differential scanning calorimetry, and powder X-ray diffraction. Structural models are proposed for the self-assembly of the molecules within the mesophases. The imidazolium head groups and the iodide counterions were found to adopt a peculiar orientation in the central part of the columns of the hexagonal columnar phases. The enantiotropic cubic phase shown by the 1,3-dimethyl-2-[3,4-bis(pentadecyloxy) phenyl]imidazolium iodide salt has a multicontinuous Pmβar 3μ structure. To the best of our knowledge, this is the first example of a thermotropic cubic mesophase of this symmetry. Copyright

Full text of DOI:10.1002/chem.201001921

FULL PAPER
DOI: 10.1002/chem.201001921
T-Shaped Ionic Liquid Crystals Based on the Imidazolium Motif: Exploring
Substitution of the C-2 Imidazolium Carbon Atom
Karel Goossens,*[a] Sil Wellens,[a] Kristof Van Hecke,[a] Luc Van Meervelt,[a]
Thomas Cardinaels,[b] and Koen Binnemans[a]
Abstract: In this contribution the first
examples of so-called rigid-core, T-
shaped imidazolium ionic liquid crys-
tals, in which the C-2 atom of the imi-
dazolium ring is substituted with an
aryl moiety decorated with one or two
alkoxy chains, are described. The
length of the alkoxy chain(s) was
varied from six to eighteen carbon
atoms (n=6, 10, 14–18). Whereas the
compounds with one long alkoxy chain
display only smectic A phases, the salts
containing two alkoxy chains exhibit
smectic A, multicontinuous cubic, as
well as hexagonal columnar phases, as
evidenced by polarising optical micros-
copy, differential scanning calorimetry,
and powder X-ray diffraction. Structur-
al models are proposed for the self-as-
sembly of the molecules within the
mesophases. The imidazolium head
groups and the iodide counterions were
found to adopt a peculiar orientation
in the central part of the columns of
the hexagonal columnar phases. The
enantiotropic cubic phase shown by the
1,3-dimethyl-2-[3,4-bis(pentadecyloxy)-
phenyl]imidazolium iodide salt has a
¯
multicontinuous Pm3m structure. To
the best of our knowledge, this is the
first example of a thermotropic cubic
mesophase of this symmetry.
Keywords: cubic phases · imidazoli-
um cations · ionic liquids · liquid
crystals · self-assembly
Introduction
Kouwer and Swager recognised that the imidazolium frag-
ment is a polarisable rigid group that is suitable as part of a
mesogenic core rather than being placed at one end of a
flexible alkyl chain.[26] However, the mesomorphism of ionic
mesogens, wherein the imidazolium group is an integral part
of the rigid core, remained virtually unexplored until very
recently when reports appeared in the literature about 1-
aryl-3-alkylimidazolium salts[26–29] and 1,3-diarylimidazolium
salts.[30,31] However, no examples exist of 2-arylsubstituted
imidazolium ionic liquid crystals (i.e., with a substituted aryl
group directly connected to the carbon atom in the 2-posi-
tion of the imidazolium ring). Kouwer and Swager,[26] and
Kato et al.[21,32] did investigate the influence of a methyl,
ethyl, pentyl or phenyl group in the 2-position on the phase
behaviour of imidazolium salts, but these substituents only
lowered the transition temperatures in some cases, led to a
disappearance of the liquid-crystalline properties or had
only a minor influence. Mukai et al. also reported on a pro-
tonated 2-heptadecylimidazolium tetrafluoroborate salt
showing a smectic A (SmA) phase over a rather narrow
temperature range.[33] In a report by Aida et al., a methyl
group in the 2-position of the six imidazolium groups situat-
ed at the periphery of triphenylene discotic liquid crystals
had an influence on the phase behaviour.[34]
Ionic liquid crystals are a fascinating class of molecular ma-
terials. They combine the characteristics of liquid crystals
(anisotropy of physical properties) with those of ionic liq-
uids (ionic conductivity, thermal stability, tuning possibilities,
etc.).[1] One of the most frequently used cations to obtain
ionic mesogens is the 1,3-disubstituted imidazolium core. By
attaching one or two relatively long alkyl chains to the het-
erocyclic cation, mesomorphism is induced.[2–9] These meso-
gens can be considered as flexible amphiphiles. The alkyl
chains might additionally be decorated with mesogenic
groups to influence the phase behaviour.[10–16] In this way,
rare nematic ionic liquid crystals could be obtained.[16,17] Al-
ternatively, a benzene ring was incorporated into the ali-
phatic chain(s) (benzyl substituent(s)).[18] By attachment of
multiple alkyl chains to this benzene ring, taper-shaped cat-
ions were obtained.[19–25]
[a] Dr. K. Goossens, S. Wellens, Dr. K. Van Hecke,
Prof. Dr. L. Van Meervelt, Prof. Dr. K. Binnemans
Katholieke Universiteit Leuven, Department of Chemistry
Celestijnenlaan 200F (PO Box 2404), 3001 Heverlee (Belgium)
Fax : (+32)16-327992
It is remarkable that no attempts have been made to pre-
pare 2-arylsubstituted imidazolium mesogens, since these
structures might show different intermolecular interactions
(e.g., hydrogen bonding) and thus a different phase behav-
iour than the imidazolium ionic liquid crystals previously re-
ported. Most likely, this is due to the fact that the required
precursors are not commercially available. To be able to
E-mail: karel.goossens@chem.kuleuven.be
[b] Dr. T. Cardinaels
Studiecentrum voor KernACTHNUTRGNENGUenerACHUTNGTRENgNUGN ie
Centre dꢀꢁtude de lꢀꢁnergie Nuclꢂaire (SCK·CEN)
Boeretang 200, 2400 Mol (Belgium)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001921.
Chem. Eur. J. 2011, 17, 4291 – 4306
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4291

attach a substituted aryl moiety to the imidazolium cation in
its 2-position, one has to ꢄbuild upꢀ the cationic core oneself.
alkylating agent instead of CH3Cl or CH3Br, because the
last two compounds are gases at room temperature.
This procedure appeared not to be suitable to obtain com-
pounds 4 (with a 4-(OCnH2n+1) phenyl moiety, n=18) and
Results and Discussion
7–11 (with a 3,4-bisACHTGNURETNN(UG OCnH2n+1) phenyl moiety, n=14–18):
the intermediate imidazoline compounds did not sufficiently
dissolve in either CH2Cl2 or DMSO to transform them into
the corresponding imidazole compounds by using literature
procedures.[38,40] Dehydrogenation of the imidazolines in tol-
uene at reflux in the presence of Pd/C (5 wt%)[41] gave the
imidazole compounds only in very low yields. Therefore, an
alternative synthetic route was followed (Scheme 2). Precur-
Synthesis: Imidazolium salts 1–3 (with a 4-(OCnH2n+1
phenyl moiety, n=6, 10, 14) and 5–6 (with a 3,4-bis-
(OCnH2n+1) phenyl moiety, n=6, 10) were prepared by re-
acting the appropriate alkoxy-substituted benzaldehyde with
ethylenediamine to obtain an imidazoline compound,[35,36]
and by oxidising (Swern oxidation[37]) the latter to the corre-
sponding imidazole compound,[38] which was finally alkylat-
ed and quaternised (Menschutkin reaction[39]) by using NaH
and an excess of CH3I (Scheme 1). CH3I was chosen as the
)
ACHTUNGTRENNUNG
Scheme 1. Synthesis of 2-arylsubstituted imidazolium salts 1–3 and 5–6.
i) CnH2n+1Br, K2CO3, KI, 2-butanone, reflux, Ar; ii) ethylenediamine, N-
bromosuccinimide (NBS), dry CH2Cl2, 08C!RT, Ar; iii) 1) oxalyl chlo-
ride, DMSO, dry CH2Cl2, ꢀ788C, Ar; 2) Et3N, ꢀ788C!RT, Ar; iv) NaH,
CH3I (excess), dry THF, RT!reflux, Ar.
Scheme 2. Synthesis of 2-arylsubstituted imidazolium salts 4 and 7–11.
i) ethylenediamine, NBS, dry CH2Cl2, 08C!RT, Ar; ii) diacetoxyiodoben-
zene (DIB), K2CO3, DMSO, RT, Ar; iii) NaH, CH3I, dry THF, RT, Ar;
iv) pyridinium chloride, 2008C; v) CnH2n+1Br, K2CO3, KI, 2-butanone,
reflux, Ar; vi) CH3I (excess), dry THF, RT!reflux, Ar.
sors 4a and 7a were oxidised to the corresponding imida-
zole compounds 4b and 7b in DMSO by means of DIB.[40]
Compounds 4b and 7b were then monoalkylated with CH3I
to afford 4c and 7c, which were then deprotected in pyridi-
nium chloride at 2008C.[42] Finally, the hydroxyl groups of
4d and 7d were alkylated, followed by quaternisation with
CH3I.
Abstract in Dutch: In deze bijdrage beschrijven we de eerste
voorbeelden van T-vormige ionische vloeibare kristallen ge-
baseerd op een kationische imidazoliumgroep, die deel uit-
maakt van een ruimere rigide kern doordat het koolstof-
atoom in de 2-positie van de imidazoliumring gesubstitueerd
is met een arylgroep die ꢀꢀn of twee alkoxyketens bevat. De
lengte van de alkoxyketen(s) werd gevarieerd van zes tot
achttien koolstofatomen (n=6, 10, 14–18). Terwijl de verbin-
dingen met slechts ꢀꢀn—voldoende lange—alkylketen enkel
smectische A fasen vertonen, werden zowel smectische A, ku-
bische als hexagonaal-kolomvormige mesofasen gevonden
voor de zouten die twee alkylketens bevatten. Voor de studie
van de vloeibaar-kristallijne eigenschappen werd gebruik ge-
maakt van gepolariseerde optische microscopie, differentiꢁle-
scanning calorimetrie, en X-stralenpoederdiffractie. We geven
een nauwkeurige beschrijving van de spontane ordening van
de moleculen in de mesofasen. De imidazoliumkopgroepen
en de jodide-anionen blijken een bijzondere oriꢁntatie te ver-
tonen in het centrale gedeelte van de kolommen in de hexago-
naal-kolomvormige fasen. De enantiotrope kubische fase ver-
toond door het 1,3-dimethyl-2-[3,4-bis(pentadecyloxy)fenyl]i-
Single-crystal X-ray diffraction: Crystal structure of 2c:
Crystals of the neutral precursor 2c suitable for single-crys-
tal X-ray structure determination could be obtained by
slowly evaporating a solution of the compound in chloro-
form. The asymmetric unit contains two 2-[4-(decyloxy)-
phenyl]-1H-imidazole molecules. The imidazole rings are at
an angle to the benzene rings (34.8(1)8 and 33.5(1)8 between
planes through the respective ring atoms). The alkyl chains
of both molecules in the asymmetric unit are completely
staggered and almost planar with respect to the benzene
rings.
Intermolecular hydrogen bonds exist between the proton-
ated imidazole nitrogen atom of one molecule and the non-
protonated imidazole nitrogen atom of another molecule
(N2(H)···N3 and N4(H)···N1 distances of 2.888(2) and
2.922(2) ꢅ, respectively). These interactions link the mole-
cules together in the [001] direction and result in layers par-
¯
midazoliumjodidezout heeft een multicontinue Pm3m-struc-
tuur. Voor zover wij weten, is dit het eerste voorbeeld van een
thermotrope kubische mesofase met deze symmetrie.
4292
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 4291 – 4306

T-Shaped Ionic Liquid Crystals
FULL PAPER
Crystal structure of 6: Single
crystals of imidazolium salt 6
could be obtained by slowly
evaporating a solution of the
compound in ethanol. The
asymmetric unit contains two
1,3-dimethyl-2-[3,4-bis-
AHCTUNGTRENGN(UN decyloxy)phenyl]imidazolium
iodide molecules. As in the
crystal structure of 2, the imida-
zolium rings are at an angle to
the benzene rings (88.8(2)8 and
51.5(2)8
between
planes
through the respective ring
atoms). The alkyl chains of
both molecules in the asymmet-
ric unit are completely stag-
gered, but they are not lying in
the same plane as the benzene
rings (37.58 and 71.58 between
planes through the respective
atoms for the first molecule,
and 52.78 and 44.98 between
Figure 1. Packing in the crystal structure of compound 2c, showing the linkage of the molecules in the [001] di-
rection through intermolecular hydrogen bonds.
allel with the (010) plane (Figure 1). In the crystal structures
planes through the respective atoms for the second mole-
cule).
of 2-[4-(methoxy)phenyl]-4,5-dihydro-1H-imidazole[43] and
2-[4-(methoxy)phenyl]-1H-benzimidazole,[44]
similar
As in the case of 2, the iodide anions link the imidazolium
cations together through short contacts in the [100] direction
(Figure 2). Iodide anion I1 shows short contacts with hydro-
gen atoms of two methyl groups attached to two different
imidazolium rings, and with the benzene ring belonging to a
N(H)···N hydrogen bonds that link the molecules into one-
dimensional chains (extending in the [100] and [001] direc-
ꢀ
tions, respectively) were found. In the packing diagram, C
H···p ring interactions were observed between H
N
ꢀ
ꢀ
ꢀ
an imidazole ring C20 C22,N3,N4 (C H···ring centroid dis-
tance of 2.88 ꢅ), and between H(C5) and H(C27) and ben-
third molecule (C H···I distances ranging from 3.00 to
3.17 ꢅ). Iodide anion I2 shows short contacts with an imida-
N
ꢀ
ꢀ
zene rings C23–C28 and C4–C9, respectively (C H···ring
centroid distances of 2.84 and 2.93 ꢅ, respectively).
zolium ring hydrogen atom belonging to a fourth cation (C
H···I distance of 3.03 ꢅ), and with an imidazolium ring hy-
drogen atom and a hydrogen atom of a methyl group that
both belong to a cation that is already involved in the inter-
Crystal structure of 2: Single crystals of imidazolium salt 2
were obtained by slowly evaporating a solution of the com-
pound in ethanol. Although the obtained structural model is
very clear, the anisotropical refinement suffered from severe
absorption effects and the attained R and wR2 values are
currently not sufficient for deposition of the crystal structure
in the Cambridge Structural Database. Nevertheless, the
structure is presented in the Supporting Information (Fig-
ures S1–S3 in the Supporting Information). The asymmetric
unit contains two 1,3-dimethyl-2-[4-(decyloxy)phenyl]imida-
zolium iodide molecules. The imidazolium rings are at an
angle to the benzene rings. The alkyl chains of both mole-
cules in the asymmetric unit are completely staggered. The
alkyl chain of one of the two molecules is almost planar
with respect to the benzene ring, whereas the alkyl chain of
the other molecule is non-planar with respect to this aro-
matic ring. The iodide anions link the imidazolium cations
together through short contacts in the [010] direction (Fig-
ures S1 and S2 in the Supporting Information). In this way,
layers parallel with the (100) plane are formed (Figure S3 in
the Supporting Information).
ꢀ
actions with iodide anion I1 (C H···I distances of 3.13 and
ꢀ
3.15 ꢅ, respectively). In the packing diagram, C H···p ring
interactions were observed between H
ring C6–C11 (C H···ring centroid distance of 2.84 ꢅ).
(C43) and benzene
ꢀ
Thermal behaviour: The thermal properties of all com-
pounds were examined by polarising optical microscopy
(POM) and differential scanning calorimetry (DSC), and
most mesophases were additionally investigated by X-ray
diffraction on powder samples (powder XRD) at various
temperatures (see below). Table 1 summarises the transition
temperatures and thermal data. Representative DSC traces
and thermogravimetric analysis (TGA) thermograms can be
found in the Supporting Information.
All of the reported imidazolium salts were liquid crystal-
line, except for compound 1, which has only a single hexyl
chain. The SmA phases could be easily identified by their
defect textures: focal conic fan and oily streak textures were
observed, as well as large homeotropic areas and the forma-
tion of bꢆtonnets on cooling from the isotropic liquid (Fig-
ure 3a). The rather lowly viscous SmX phase shown by 5
Chem. Eur. J. 2011, 17, 4291 – 4306
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4293

K. Goossens et al.
The type of mesophase is
highly dependent on the
number of alkyl chains (and the
length). For the homologous
series 1–4, containing one alkyl
chain, only SmA phases were
found. No smectic phases of
higher order—as observed, for
example, for N-alkyl-N-methyl-
pyrrolidinium ionic liquid crys-
tals[45]—could be detected. A
longer chain results in a higher
clearing point due to a more ef-
ficient microsegregation (see
also below).
For compounds 5, 6 and 7,
with alkyl chains up to 14
carbon atoms, lamellar SmA
phases were also observed. For
5, an underlying highly ordered
smectic phase was found by
POM and DSC in the second
and subsequent heating runs.
On heating a fresh sample of 5,
however, the compound rapidly
recrystallised at 1068C to melt
only at 1248C (see the Support-
ing Information). Salts 8, 9, 10
and 11, with fairly long alkyl
chains (C15–C18), exhibited Colh
phases as a result of the in-
creased curvature at the aro-
matic/ionic–aliphatic interface
(see also below).[46] Compounds
8 and 9 also exhibited a lower-
temperature
Cub
phase.
Figure 2. Top: intermolecular short contacts observed in the crystal structure of compound 6. Iodide anions
are labelled. Bottom: packing in the crystal structure of compound 6, showing the linkage of the molecules in
the [100] direction through intermolecular short contacts.
Powder XRD revealed that the
enantiotropic Cub phase of salt
8 most probably has a multicon-
¯
tinuous Pm3m symmetry (see
below). To the best of our
could not be identified by its optical texture (see the Sup-
porting Information). The hexagonal columnar (Colh)
phases were identified by their pseudo-focal conic fan-
shaped texture and the appearance of homeotropic areas
(Figure 3b–d). The cubic (Cub) phases were identified by
their optical isotropy, in combination with the slow forma-
tion of the texture through the typical appearance of square
edges and black polygonal areas growing across the preced-
ing texture (Figure 3e and f).
knowledge, this is the first example of a thermotropic meso-
gen showing a Cub Pm3m phase.
The appearance of smectic, cubic as well as hexagonal col-
umnar phases for the same homologous series of bis-
AHCTUNGTREG(NNUN alkoxy)-substituted thermotropic ionic mesogens is quite
exceptional. The only other examples of such behaviour for
imidazolium ionic liquid crystals have only very recently
been reported.[23,29] A transition from lamellar mesophases
(with flat interfaces) to columnar mesophases (with curved
interfaces), via an intermediate cubic structure, is typically
found for tetracatenar liquid crystals.[46–51] This is explained
by the increase in the aliphatic volume fraction with temper-
ature and/or chain length. To compensate for this increase,
the rigid cores will tilt to the extent that tilting is no longer
sufficient, and undulations start to occur. These undulations
result in transitions to multicontinuous cubic and columnar
¯
The alkyl chain length has only a very minor influence on
the melting point. The clearing point, however, is largely de-
termined by this parameter, especially for the series 1–4.
This has also been found for simple 1-alkyl-3-methylimid-
ACHTUNGTRENNUNG
azolium halide salts.[6] It is not clear why 6 displays a slightly
higher clearing point than 7.
4294
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 4291 – 4306

T-Shaped Ionic Liquid Crystals
FULL PAPER
pendant mesogenic group,[16] and linkage of imidazolium
Table 1. Transition temperatures and thermal data for the 2-arylsubstitut-
ed imidazolium salts.
ꢀ
cations in the solid state through non-classical C H···X hy-
Subst.
Transition[b]
T [8C][c]
DH [kJmolꢀ1
(DS [JKꢀ1 molꢀ1])
]
drogen bonds is well known.[52,53] This would effectively en-
large the central rigid core from two rings to four rings.
pattern[a]
ACHTUNGTRENNUNG
1
2
4-C6
4-C10
Cr!I
123
122
152
120[d]
220
122
243
33[d]
99[e]
106
124
147
66[d]
75
113
164
91
115
159
103[d]
118
139
172
94
24.3 (61)
40.4 (102)
0.7 (2)
41.6 (106)
1.2 (2)
72.7 (184)
1.5 (3)
0.8 (3)
4.7 (13)
19.6 (52)
0.5 (1.4)
1.7 (4)
0.5 (1.4)
0.5 (1.3)
38.0 (98)
0.7 (1.7)
7.8 (21)
46.9 (121)
0.2 (0.4)
3.8 (10)
47.8 (122)
0.5 (1.2)
0.2 (0.5)
6.9 (19)
0.3 (0.9)[g]
55.0 (141)
0.5 (1.0)
11.6 (30)
58.4 (148)
0.7 (1.5)
0.1 (0.4)
9.7 (26)
63.9 (162)
0.7 (1.5)
ꢀ
Nevertheless, these C H···X hydrogen bonds are rather
Cr!SmA
SmA!I
weak, a fortiori when Xꢀ =Iꢀ instead of Clꢀ or Brꢀ.[54] In the
crystal structures of compounds 2 and 6, intermolecular
3
4
5
4-C14
4-C18
3,4-C6
Cr!SmA
SmA!I
ꢀ
short C H···I contacts were indeed found to link the imida-
Cr!SmA
SmA!I
zolium moieties together, but it is difficult to distinguish any
dimer formation, especially in the case of 6 (Figure 2 and
Figures S1–S3 in the Supporting Information). Therefore,
the concept of a supramolecular polycatenar mesogen being
formed seems to be false (it should also be noted that for
tetracatenar mesogens with shorter chain lengths, SmC (and
nematic) phases are generally expected instead of SmA
phases). Indeed, as will be shown below, the mesophase
structures are governed by microsegregation, space filling
and, in particular, by strong electrostatic interactions. It is
striking that none of the neutral imidazole precursors (1c–
3c, 5c–6c, 4e, 7e–11e) are mesomorphic (despite the phase-
segregated lamellar structure already present in the crystal-
line solid state, see Figure 1), which emphasises the impor-
tance of electrostatic interactions. Hydrogen bonding
Cr1!Cr2
Cr2!Cr3
Cr3!SmX[f]
SmX!SmA
SmA!I
6
3,4-C10
Cr1!Cr2
Cr2!Cr3
Cr3!SmA
SmA!I
7
8
3,4-C14
3,4-C15
Cr1!Cr2
Cr2!SmA
SmA!I
Cr1!Cr2
Cr2!Cub
Cub!Colh
Colh!I
9
3,4-C16
Cr1!Cr2
ꢀ
(through the N H moiety) cannot induce mesomorphism in
A
96[d,g]
118
190
108
120
204
98
103
121
208
precursors 1b–3b, 5b–6b, 1c–3c and 5c–6c, due to too low
a degree of amphiphilicity in these compounds (in contrast
to the compounds reported for example by Seo et al.[55]).
Unlike the models developed for polycatenar mesogens, the
well-established Israelachvili theory concerning the self-as-
sembly of hydrocarbon amphiphiles can be used to explain
the phase behaviour and the observed molecular arrange-
ments in the mesophases (see below).[56] In the past, a
SmA–Cub–Colh phase sequence in a homologous series has
often been observed for non-ionic, amphiphilic compounds,
and microsegregation has been shown to play a crucial role
in the mesomorphism.[57–60] Analogies to lyotropic systems
have already been drawn in these earlier investigations.[61,62]
Cr2!Colh
Colh!I
10
11
3,4-C17
3,4-C18
Cr1!Cr2
Cr2!Colh
Colh!I
Cr1!Cr2
Cr2!Cr3
Cr3!Colh
Colh!I
[a] Substitution pattern of the aryl substituent: 4-(alkoxy) or 3,4-bis-
(alkoxy). [b] Abbreviations: Cr=crystalline phase; SmX=unidentified
smectic phase; SmA = smectic A phase; Cub=cubic phase, most proba-
¯
bly of Pm3m symmetry (see text); Colh = hexagonal columnar phase;
I=isotropic liquid. Transitions between parentheses indicate transitions
to a monotropic mesophase. [c] Onset temperatures obtained by DSC at
heating/cooling rates of 108C minꢀ1 (He atmosphere). Values were taken
from the first heating run for 1, 2, 3, 4, 7, 9 and 10, and from the second
heating run for 5, 6, 8 and 11 (first heating run up to 1808C). [d] Peak
temperature. [e] Prior to this endothermic transition, an exothermic re-
crystallisation process took place at 858C (peak temperature). [f] This
phase does not appear during the first heating run, in contrast to all sub-
sequent heating runs. [g] From the first cooling run.
Powder XRD: The enantiotropic mesophases were investi-
gated by powder XRD to obtain more information about
the molecular packing in these phases. Table 2 summarises
the Bragg reflections collected from the X-ray diffracto-
grams.
Before discussing the powder XRD results, we will give a
brief overview of the imidazolium-based ionic liquid crystals
exhibiting thermotropic cubic mesophases that have been
reported up to now. To the best of our knowledge, only six
examples exist. 1) Bielawski et al. reported one benzobis-
(imidazolium) salt that showed a Cub phase between 188
and 2388C.[63] 2) Kato et al. synthesised an imidazolium bro-
mide salt containing an l-glutamic acid moiety that exhibit-
phases (with the cubic phases having rather slow formation
kinetics due to the large structural rearrangements that
occur). However, the compounds presented herein do not
possess a typical polycatenar structure in the sense that they
only contain flexible alkyl chains on one side of the rigid
core, which is moreover very short (only two ring struc-
tures). One can imagine the formation of some supramolec-
ular polycatenar mesogens by head-to-head coupling of the
imidazolium groups through hydrogen-bonding interactions
between the imidazolium H-4 and H-5 atoms and the halide
anions. Such an arrangement has already been found in the
crystal structure of an imidazolium bromide salt with a
¯
ed a micellar Cub Pm3n phase above 1368C (a lower-tem-
perature Colh phase was found as well) with thermal degra-
dation occurring at about 1808C.[64] [The authors showed
that the anisotropic ionic conductivity measured parallel to
the aligned columnar axes in the low-temperature Colh
phase decreased by one order of magnitude when heating
Chem. Eur. J. 2011, 17, 4291 – 4306
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4295

K. Goossens et al.
Figure 3. Defect textures observed by polarising optical microscopy (POM): a) formation of bꢆtonnets and fan-shaped texture of the SmA phase of 5 at
1438C on cooling (100ꢇ); b) pseudo-focal conic fan-shaped texture of the Colh phase of 11 at 2058C on cooling (200ꢇ); c) homeotropic domains and
fan-shaped texture of the Colh phase of 9 at 1828C on cooling (200ꢇ); d) fan-shaped texture of the Colh phase of 9 at 1508C on cooling (200ꢇ); e) devel-
opment of the Cub phase of 8 at 1328C on cooling from the Colh phase (200ꢇ); f) development of the monotropic Cub phase of 9 at 968C after fast cool-
ing from the Colh phase (200ꢇ).
the sample into the micellar Cub phase: the ionic moieties
became confined within the micelles and insulated by the
lipophilic parts and this prevented fast, long-range ion con-
duction. A similar decrease in proton conductivity was
found for the micellar Cub phases shown by amphiphilic N-
hexaACTHGNUTERNNUG
fluorophosphate at their apex, respectively.[66] These
¯
dendrons self-assembled into micellar Cub Pm3n phases.
5) Aida et al. synthesised triphenylene discotic liquid crys-
tals with 3-methylimidazolium groups at the ends of the six
peripheral chains.[34] Some of the compounds showed a bi-
(2,3-dihydroxypropyl)benzamides.[65]
pared 1-[3,4-bis(dodecyloxy)benzyl]-3-methylimidazolium
]
3) Douce et al. pre-
continuous Cub Ia3d phase. One compound (namely, that
¯
with peripheral tetradecyl chains and tetrafluoroborate
chloride and bromide salts.[23] These two compounds showed
counterions) additionally showed a bicontinuous Cub Pn3m
¯
¯
a bicontinuous Cub Ia3d phase. 4) Percec et al. reported two
phase. We believe this represents the first example of a ther-
motropic ionic liquid crystal showing this type of Cub phase.
dendrons containing 3-methylimidazolium chloride and
4296
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 4291 – 4306

T-Shaped Ionic Liquid Crystals
FULL PAPER
Table 2. Bragg reflections collected from the X-ray diffractograms of the different enantiotropic mesophases.
dmeas
I[b]
hkl[c]
dcalcd
Parameters of
dmeas
I[b]
hkl[c]
dcalcd
Parameters of
[ꢅ][a]
[ꢅ][a]
the mesophase[d]
[ꢅ][a]
[ꢅ][a]
the mesophase[d]
2
31.11
M
001
31.11
SmA: T=1408C
8[f]
42.89
35.32
30.02
29.15
27.54
26.57
25.23
23.46
22.58
21.45
21.00
20.15
19.62
19.12
18.77
18.31
17.56
34.31
6.61
W
VS
VW
VW
W
200
211
220
221
310
311
222
320
321
400
322
411
331
420
421
332
422
10
43.05
35.15
30.44
28.70
27.23
25.96
24.85
23.88
23.01
21.52
20.88
20.29
19.75
19.25
18.79
18.36
17.57
34.31
Cub: T=1248C
V
A
M =823 ꢅ3
V
M =1343 ꢅ3
M =52.9 ꢅ2
a
V
N
Cub =86.1 ꢅ
Cub =638253 ꢅ3
Cub =475
d/L=1.42
3
34.51
S
001
34.51
SmA: T=1408C
W
W
V
A
M =925 ꢅ3
M =53.6 ꢅ2
VW
VW
VW
VW
VW
VW
VW
VW
VW
W
d/L=1.36
4
38.49
19.50
VS
W
001
002
38.74
19.37
SmA: T=1408C
V
A
M =1026 ꢅ3
M =53.0 ꢅ2
d/L=1.34
5[e]
26.81
M
001
26.81
SmA: T=1298C
V
A
M =896 ꢅ3
M =66.8 ꢅ2
S
Colh: T=1558C
M =1371 ꢅ3
Col =39.6 ꢅ
S=1359 ꢅ2
slice =7.2
d/L=1.58
VW (br)
h1
V
a
6
30.64
15.36
VS
VW
001
002
30.68
15.34
SmA: T=1408C
V
A
M =1104 ꢅ3
N
M =72.0 ꢅ2
d/L=1.39
9
34.91
6.61
VS
W (br)
10
h1
34.91
Colh: T=1558C
M =1423 ꢅ3
Col =40.3 ꢅ
S=1407 ꢅ2
slice =7.2
V
a
7
34.91
S
001
34.91
SmA: T=1408C
V
A
M =1308 ꢅ3
M =74.9 ꢅ2
d/L=1.39
N
10
11
36.17
20.71
6.61
S
W
VW (br)
10
11
h1
36.17
20.88
Colh: T=1468C
M =1465 ꢅ3
Col =41.8 ꢅ
S=1511 ꢅ2
slice =7.4
V
a
N
36.39
21.92
18.65
6.61
S
W
VW
W (br)
10
11
20
h1
36.39
21.51
18.20
Colh: T=1558C
M =1525 ꢅ3
Col =42.0 ꢅ
S=1529 ꢅ2
slice =7.2
V
a
N
[a] dmeas and dcalcd are the measured and calculated diffraction spacings, respectively. [b] I is the intensity of the reflections: VS, very strong; S, strong; M,
medium; W, weak; VW, very weak; br, broad reflection. [c] hkl are the Miller indices of the reflections. h1 denotes the medium-angle reflection that
probably corresponds to the stacking periodicity along the columnar axes (see text). [d] T is the temperature at which the X-ray diffractogram was re-
corded. VM is the molecular volume (estimated to be VM =(M/0.6022)f, in which M is the molecular mass [gmolꢀ1] and f is a temperature-correcting
factor (f=0.9813+7.474ꢇ10ꢀ4T with T in 8C)[16]); AM is the molecular area (calculated from AM =2VM/d). aCub is the lattice parameter of the Cub phase
1
2
2
2
=
2
(=[Shkl
d
G
p
1
2
2
=
2
cules within the cubic cell (estimated to be VCub/VM). aCol is the lattice parameter of the Colh phase (=[(2/ 3)ACTHUNGTRENN[NUG ShkdhkACHTUGNTNERN(UNG h +k +hk) ]]/Nhk, in which Nhk =
the number of (hk) reflections); S is the cross-sectional area of one column in the Colh phase (=( 3/2)aCol2); Nslice is the number of molecules within one
slice of a column in the Colh phase, with a height h1 of about 6.6 ꢅ (simplified model, see text). Nslice was estimated by using the following equation and
ACTHNUTRGNEUNG
assuming a density of 1=1.0 gcmꢀ3: Nslice =h1S(NA/M)1, in which NA =Avogadro constant and M=molecular mass [gmolꢀ1]. L is the calculated length
of the 2-arylsubstituted imidazolium cation in its most extended conformation (estimated with the Chem3D program; the structure of the imidazolium
cation was energy-minimised by an MM2 calculation within Chem3D). [e] No diffraction data could be obtained for the SmX phase on cooling from the
¯
isotropic liquid. [f] The (110) and (111) reflections, which are also allowed in the Pm3m space group, could not be detected with the setup that was at
our disposal. The (210) reflection overlaps with the very intense (211) reflection.
6) Finally, Tschierske et al. reported four rod-like imidazoli-
Bragg reflections at both small and wide angles. The equi-
distant small-angle reflections indicated that the molecules
are arranged in a lamellar fashion (see also the section
about single-crystal X-ray diffraction).
The diffractograms of the SmA phases showed one sharp
reflection, indexed as the (001) reflection, in the small-angle
region. In some cases the much weaker (002) reflection
um bromide compounds containing three tetradecyl or hexa-
[29]
¯
decyl chains that exhibited micellar Cub Pm3n phases.
The following sections describe the X-ray diffraction data
that were obtained for our 2-arylsubstituted imidazolium
salts. The crystalline nature of the solid phases at room tem-
perature was confirmed by the presence of several sharp
Chem. Eur. J. 2011, 17, 4291 – 4306
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4297

K. Goossens et al.
could be observed as well. These reflections are related to
the consecutive smectic layers of a specific layer thickness d
(Table 2). In the wide-angle region, a diffuse signal centred
at about 4.7 ꢅ was observed, corresponding to the lateral
short-range order of the disordered (molten) aliphatic
chains.
sponding to the liquid-like order of the molten aliphatic
chains. The small-angle reflections could be indexed as the
(200), (211), (220), (221), (310), (311), (222), (320), (321),
(400), (322), (411), (331), (420), (421), (332) and (422) re-
1
2
2
2
=
2
flections of a cubic lattice (for which dhkl =aCub/ACTHNGUTRENNU(G h +k +l ) ,
with aCub =the lattice parameter of the cubic lattice). Al-
though an unequivocal space group determination of a ther-
motropic cubic mesophase is possible only by X-ray studies
on aligned monodomain samples, the initial number of theo-
retical possibilities (36 cubic space groups) can be reduced
by logical analysis of the data.[23,29,51] First of all, because the
presence or absence of a centre of symmetry cannot be de-
tected on the basis of standard powder diffraction measure-
ments (Friedelꢀs law), all non-centrosymmetric groups
(groups with Laue classes 23, 432, or 4m) can be disregard-
ed.[67] This leaves only 17 cubic space groups to consider.
The observed reflections and the ratios of the corresponding
reciprocal spacings are not compatible with a face-centred
cubic network (F), nor with a body-centred cubic network
(I) (for F: h+k=2n, h+l=2n, k+l=2n; for I: h+k+l=
2n). This leaves 7 cubic space groups with a primitive Brav-
ais lattice (P). However, the reflection conditions that apply
The diffractograms of the Colh phases showed one sharp
reflection in the small-angle region, indexed as the (10) re-
flection of the two-dimensional hexagonal lattice (for which
1
2
2
=
2
dhk =aCol/ACHTUNGTRENNUNG[(4/3)ACHTUNGTRENNUNG(h +k +hk)] , in which aCol is the lattice pa-
rameter of the hexagonal lattice). For compound 11 both
the (11) and (20) reflections (reciprocal spacings of the (10),
(11) and (20) reflections in the characteristic ratio of
p
1: 3:2) could be observed. For compound 10, the character-
istic (11) reflection was found at 1468C, but not the (20) re-
flection. For 8 and 9, the Colh phase type was assigned
based on the textural similarities with 10 and 11. At 1558C,
the lattice parameter aCol of the Colh phase was found to be
39.6, 40.3, 41.3 and 42.0 ꢅ for compounds 8, 9, 10 and 11, re-
spectively. In addition, all diffractograms showed a diffuse
wide-angle scattering centred at about 4.6 ꢅ, indicating the
liquid-like nature of the mesophases (molten aliphatic
chains). Along with this broad scattering, another weak halo
was seen at about 6.6 ꢅ for 8, 9, 10 and 11. This signal prob-
ably corresponds to the stacking periodicity along the col-
umnar axes (see below). The presence of this signal for all
of the compounds in the series 8–11 is additional support of
the Colh phase type assignment.
223
205
¯
¯
for the space groups Pm3n (hhl: l=2n; 00l: l=2n), Pa3
222
¯
(0kl: k=2n; h0l: l=2n; hk0: h=2n; 00l: l=2n) and Pn3n
(0kl: k+l=2n; hhl: l=2n; 00l: l=2n) are not compatible
with the observed reflections. Then, the space groups
200
221
201
224
¯
¯
¯
¯
Pm3 , Pm3m , Pn3 and Pn3m are theoretically still
¯
possible. But the (320) reflection is not allowed in Pn3 or
Pn3m symmetry. This leaves only Pm3 and Pm3m. Aggrega-
¯
¯
¯
The diffractogram of the Cub phase exhibited by com-
pound 8 showed one very intense reflection, as well as mul-
tiple weak reflections in the small- and medium-angle
region (Figure 4). The reciprocal spacings of the reflections
within the 2–58 2q angular region were in the ratios of
tion into the highest symmetry is generally assumed, and
[68]
¯
therefore, the Pm3m space group is finally retained. We
are not aware of any reports on thermotropic Cub meso-
¯
phases with Pm3m symmetry. There have been reports on
p
p
p
p
p
p
p
p
p
p
p
p
p
p
¯
4: 6: 8: 9: 10: 11: 12: 13: 14: 16: 17: 18: 19:
thermotropic bicontinuous Cub Pn3m phases shown by dis-
p
p
p
cotic liquid crystals,[34,69] and bicontinuous Cub Pn3m phases
¯
20: 21: 22: 24. In the wide-angle region, a weak and
diffuse scattering centred at about 4.7 ꢅ was found, corre-
are frequently found for lyotropic systems.[70,71] Usually, rod-
like and polycatenar mesogens tend to exhibit bicontinuous
(or tricontinuous[72]) thermotropic Cub phases of Im3m or
¯
[51,70,73–78]
¯
Ia3d symmetry.
Therefore, the observation of a Cub
¯
phase of Pm3m symmetry for compound 8 merits further in-
vestigation. Direct Fourier reconstruction of the electron
density profile on the basis of the observed small-angle X-
ray scattering (SAXS) intensities might be difficult because
the (211) reflection is too dominating. Synchrotron measure-
ments, however, can provide an unambiguous confirmation
¯
of Pm3m as a new Cub phase structure. These experiments
will be carried out in due course. It should be noted that
judging from the position of the Cub phase in the phase se-
quence Cub!Colh it is expected to have a multicontinuous
structure, consisting of multiple interwoven, infinite three-
dimensional (columnar) networks (probably formed by the
cationic rigid cores and the anions) separated from each
other by lipophilic regions (formed by the molten aliphatic
chains; the surfaces obtained by connecting all the midway
points between the networks are referred to as infinite peri-
odic minimum surfaces).[79] The lattice parameter aCub was
found to be about 86.1 ꢅ for the Cub phase shown by 8. As
Figure 4. X-ray diffractogram of 8 in the Cub phase at 1248C, represent-
ed as the scattering intensity (in counts) versus the scattering angle, 2q.
The broad diffraction signal between 2qꢁ5.68 and 7.68 is due to the cov-
ering foil used in the experimental setup.
4298
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 4291 – 4306

T-Shaped Ionic Liquid Crystals
FULL PAPER
deduced from the estimated molecular volume VM (VM
ꢁ1343 ꢅ3, 1ꢁ1.0 gcmꢀ3 (Table 2)), the elementary cell of
the Cub phase (volume VCub =aCub3) contains about 475 mol-
ecules.
ð1Þ
ð2Þ
dSmA ðin ꢅÞ ¼ 21:4ðꢂ0:9Þ þ 0:95ðꢂ0:06Þn
dSmA ðin ꢅÞ ¼ 20:7ðꢂ0:3Þ þ 1:01ðꢂ0:03Þn
We assume that the cationic rigid cores and the iodide
anions separate into a distinct sublayer (microphase) inside
the smectic layers. From Equations (1) and (2), the thickness
of the ionic sublayers d0 can be estimated to be approxi-
mately 21.4 ꢅ for the series 2–4, and 20.7 ꢅ for the series 5–
7. These values correspond well to twice the length of a 1,3-
dimethyl-2-(4-hydroxyphenyl)imidazolium cation (estimated
from the crystal structure of 6 to be about 8.2 ꢅ; a similar
value is found by using Chem3D) when taking into account
that, in addition, the iodide anions have an ionic radius of
2.15 ꢅ. This supports our assumption that a sublayer is
formed in which the cationic rigid cores and the iodide
anions of different molecules interact in a head-to-head ar-
rangement. The alkyl chains are homogeneously distributed
on either side of this sublayer. Driving forces for this ar-
rangement are microsegregation, electrostatic interactions
and, to a lesser degree, intermolecular short contacts be-
tween the imidazolium cations and iodide anions (see also
Figure 2).
Mesophase structure of the SmA phases: As commonly ob-
served, the layer thickness d of the SmA phases decreased
with increasing temperature (Figure 5). This is due to the
Interestingly, an almost identical layer thickness d is
found at 1408C for both compounds 2 and 6 (dꢁ30.9 ꢅ),
which carry decyl chains, and for both compounds 3 and 7
(dꢁ34.7 ꢅ), which carry tetradecyl chains (Table 2 and
Figure 6). Thus, while one might expect—intuitively and on
the basis of the crystal structures of 2 and 6[80]—an arrange-
ment with interdigitated alkyl chains in the case of the
single-chain compounds 2 and 3, in contrast to an arrange-
ment with non-interdigitated chains in the case of the
double-chain compounds 6 and 7, all of the alkyl chains of
the two molecules belonging to different ionic sublayers are
always interdigitated in both series. When one subtracts the
thickness of the ionic sublayers d0 from the observed d
values, one can see that the thickness of the aliphatic sublay-
ers is even smaller than the length of one decyl or tetradecyl
chain (Table 3). This means that the alkyl chains are not
only interdigitated, but also highly folded.
Figure 5. Evolution of the layer thickness d of the SmA phases of 3 (&)
and 7 (&) as a function of temperature.
higher thermal mobility of the alkyl chains at higher temper-
atures, resulting in a compression of the smectic layers. The
layer thickness also increases linearly with the number of
carbon atoms n in the alkyl chain(s) [Eqs. (1) for the series
2–4 (at 1408C) and (2) for the series 5–7 (at 1408C), and
Figure 6]:
One can estimate the cross-sectional area, schain, of one
fully stretched aliphatic chain as schain ꢁ(VCH /1.27)ꢁ
2
Table 3. Analysis of the layer thicknesses found for compounds 2, 6, 3
and 7.
Subst.
d [ꢅ]
d0
(dꢀd0)
Lchain
pattern[a]
(at 1408C)[b]
[ꢅ][c]
[ꢅ][d]
[ꢅ][e]
2
6
3
7
4-C10
3,4-C10
4-C14
31.11
30.68
34.51
34.91
ꢁ21.4
ꢁ20.7
ꢁ21.4
ꢁ20.7
ꢁ9.7
ꢁ10.0
ꢁ13.1
ꢁ14.2
ꢁ12.4
ꢁ12.4
ꢁ17.5
ꢁ17.5
3,4-C14
[a] Substitution pattern of the aryl substituent: 4-(alkoxy) or 3,4-bis-
(alkoxy). [b] d is the observed layer thickness of the SmA phase
(Table 2). [c] d0 is the estimated thickness of the ionic sublayer (see text).
[d] (dꢀd0) corresponds to the thickness of the aliphatic sublayer (see
text). [e] Lchain is the calculated length of one decyl or tetradecyl chain,
respectively, in the most extended conformation (estimated with the
Chem3D program, after energy minimisation by an MM2 calculation).
Figure 6. Evolution of the layer thickness d of the SmA phase at 1408C
as a function of the number of carbon atoms n in the alkyl chain(s) for
the series 2–4 (&) and for the series 5–7 (&).
Chem. Eur. J. 2011, 17, 4291 – 4306
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4299

K. Goossens et al.
[(26.5616+0.02023T)/1.27] (in which VCH is the volume of
2
one methylene group and T is the temperature in 8C).[81] At
1408C, schain equals about 23.1 ꢅ2. Furthermore, the molecu-
lar volume, VM, can be estimated by using the relationship
VM =(M/0.6022)f, in which M is the molecular mass (in
gmolꢀ1) and
f is a temperature-correcting factor (f=
0.9813+7.474ꢇ10ꢀ4T with T in 8C).[16] This allows us to cal-
culate the molecular area, AM, as 2VM/d, when considering a
supramolecular assembly of two molecules with a head-to-
head arrangement of ionic head groups. The following
values were found at 1408C: AM ꢁ52.9 ꢅ2 for 2, AM
ꢁ53.6 ꢅ2 for 3, AM ꢁ72.0 ꢅ2 for 6, and AM ꢁ74.9 ꢅ2 for 7
(Table 2). This means that the fully interdigitated alkyl
chains are folded to a larger extent in the case of 2 and 3
than in the case of 6 and 7, in which two and three alkyl
chains, respectively, counterbalance the area occupied by
the cationic head group and the iodide anion (see below).
Series 2–4: An AM value of about 53 ꢅ2 is acceptable when
one considers literature values for the cross-sectional areas
of rod-like mesogenic groups (for example, 22–24 ꢅ2 for cy-
anobiphenyl groups) and for the area occupied by iodide
anions (about 14.5 ꢅ2, since these anions have an ionic
radius of 2.15 ꢅ). Although the exact cross-sectional area of
a 1,3-dimethyl-2-phenylimidazolium core is presently not
known, it is clearly larger than that of a cyanobiphenyl
group because of the protruding methyl groups. For com-
pound 2, the values found for d, d0 and AM suggest that two
orthogonally oriented ionic cores are arranged in a head-to-
head manner in such a way that a strict alternation of posi-
tive and negative charges is maintained. However, a slight
tilt of the rigid cores with respect to the layer normal cannot
be excluded. Within the polar sublayers, the imidazolium
cations are probably additionally linked by intermolecular
Figure 7. Structural models (idealised) for the SmA phases exhibited by
a) 2 and b) 6. Iodide anions are represented by grey spheres (not in
space-filling mode). The smectic layer thickness is indicated by d.
structural model. The SmA phases shown by compounds 5
and 7 have similar structures (slightly larger tilt for longer
chain lengths). Of course, the SmA phase is an optically uni-
axial mesophase, and therefore, there is no correlation in
the direction of the tilt of the rigid cores. When compared
with the crystal structure of compound 6 (Figure 2), the
melting process involves softening of the alkyl chains to
result in interdigitation, whereas the clearing process in-
volves the breakdown of the polar sublayers. For longer
rigid cores, SmC phases might appear instead of SmA
phases (it should be noted that, surprisingly, the SmC phase
is rarely found for ionic liquid crystals[1,82]).
ꢀ
short C H···I contacts (see above). The alkyl chains at-
tached to rigid cores that belong to adjacent polar sublayers
need to interdigitate and fold to compensate for the cross-
sectional area of the ionic head groups (see above). Fig-
ure 7a shows a structural model of this system. When com-
pared with the packing in the crystal structure of compound
2 (Figures S2 and S3 in the Supporting Information), this
structure is reminiscent of the crystalline solid state. The
SmA phases shown by compounds 3 and 4 have similar
structures.
Series 5–7: In the case of compound 6, the calculated AM
value (ꢁ72.0 ꢅ2 at 1408C) is smaller than 4ꢇschain, but
larger than 2ꢇschain. In fact, it is only slightly larger than 3ꢇ
schain (ꢁ69.3 ꢅ2 at 1408C). Consequently, the cationic head
group and the iodide anion are now counterbalanced by
three alkyl chains (instead of two). Since 6 contains only
two aliphatic chains, the third one originates from an imida-
zolium cation in the adjacent polar sublayer. The alkyl
chains are folded to only a minor degree and the cationic
head group is necessarily tilted (increase of AM from 52.9 ꢅ2
for 2 to 72.0 ꢅ2 for 6). This also explains the slightly lower
d0 value found for 6 in comparison with 2. Figure 7b shows a
The proposed models are supported by an analysis based
on Israelachviliꢀs theory concerning the self-assembly of hy-
drocarbon amphiphiles.[56] In this concept, packing con-
straints are considered by the critical parameter P=
Valkyl/ACTHGNUETRUNN(G AMLalkyl) (in which Valkyl is the volume of the alkyl
chain(s) and Lalkyl is the length of the alkyl chains in the
most extended conformation). For a decyl chain, Lalkyl was
estimated with Chem3D to be 12.37 ꢅ (Table 3). At 1408C,
for the single-chain compound 2, P=[1ꢇ(9VCH
+
2
VCH )]/
G
ACHTUNGTRENNUNG
3
AHCTUNGTRENNUNG
2
3
AHCTUNGTRENNUNG
3
4300
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 4291 – 4306

T-Shaped Ionic Liquid Crystals
FULL PAPER
estimated to be 27.14+0.01713T+0.0004181T2, with T in 8C
[83]). Since for a lamellar arrangement P should be approxi-
mately 1, the aliphatic chains of opposing imidazolium head
groups completely interdigitate and the theoretical number
of chains per head group is two for 2 (2ꢇ0.46ꢁ1), and three
for 6 (1.5ꢇ0.68ꢁ1), in agreement with the models in
Figure 7.
Finally, it should be noted that AM increases linearly as a
function of temperature (Figure 8). This corresponds to an
increased disordering of the molten aliphatic chains. Since
compound 7 contains one alkyl chain more than compound
3, the slope of the AM(T) curve is steeper in the case of 7.
At the clearing point, the chain fluctuations have become so
strong that the ionic layers collapse.
in the Colh phases do consist of stacked slices that contain
multiple taper-shaped molecules.
From the lattice parameter aCol determined for 11
(ꢁ42.0 ꢅ at 1558C), the cross-sectional area, S, of one
p
column can be estimated as ( 3/2)aCol2 =1529 ꢅ2 (Table 2).
Consequently, the column radius, r, is about 22.1 ꢅ at
1558C; the calculated length L of the 1,3-dimethyl-2-[3,4-bi-
s(octadecyloxy)phenyl]imidazolium cation in its most ex-
tended conformation (estimated by using the Chem3D pro-
gram; the structure of the imidazolium cation was energy-
minimised by an MM2 calculation within Chem3D) is
32.1 ꢅ. When comparing the values for r and L, it appears
that the aliphatic chains of adjacent columns need to interdi-
gitate. This is also required for an efficient filling of the
space remaining at the periphery of a column slice (see
below).[29] The columnar cross-section, S, decreases with in-
creasing temperature (Figure 9), indicating a shrinkage of
Figure 8. Evolution of AM in the SmA phases of 3 (&) and 7 (&) as a
function of temperature.
Figure 9. Evolution of the columnar cross-section, S. in the Colh phases of
8 (&), 9 (&), 10 (*) and 11 (*) as a function of temperature.
Mesophase structure of the Colh phases: For longer chain
lengths (n>14), the double-chain compounds exhibit Colh
phases due to the increased curvature at the aromatic/ionic–
aliphatic interface. Indeed, the increased chain volume
cannot be counterbalanced anymore by further tilting of the
cationic rigid cores within a layer structure, and the layers
collapse to form cylindrical superstructures. Conceptually,
one can regard these cylinders as stacked slices that contain
a certain number of molecules, which self-aggregate in such
a way that the centre of one slice contains the polar moieties
(microsegregation).[19,23,29,84] In reality, for these taper-
shaped amphiphilic molecules there are no discrete slices
(as known from the Col phases formed in lyotropic systems,
and previously reported thermotropic systems[61]), and the
plane of the aromatic core can nearly freely rotate (no dis-
crete p–p stacking distance has been observed by powder
XRD in any of the Col phases formed by simple taper-
shaped amphiphilic compounds, and this is also the case for
the compounds presented herein). However, in the follow-
ing sections we will analyse the powder XRD data using a
simplified/idealised model wherein the columnar structures
the column slices due to the higher thermal mobility of the
alkyl chains and increasing disorder at higher temperatures.
The slope of the S(T) curves is comparable for all com-
pounds in the series 8–11.
The area S also increases linearly with the number of
carbon atoms n in the alkyl chains [Eq. (3) for the series 8–
11 (at 1558C) and Figure 10]:
S ðin ꢅ2Þ ¼ 490ðꢂ42Þ þ 58ðꢂ3Þn
ð3Þ
From Equation (3), the cross-sectional area S0 of the non-
aliphatic centre of the columns can be estimated to be ap-
proximately 490 ꢅ2 for the series 8–11 at 1558C. Thus, the
radius r0 of the central part of the columns is about 12 ꢅ.
With the data currently at our disposal, we cannot deduce
the precise arrangement of the cationic rigid cores and the
iodide anions within the central part of the columns. It is im-
portant to note, however, that our compounds must show a
different arrangement in the Colh phase than the N-alkyl-
Chem. Eur. J. 2011, 17, 4291 – 4306
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4301

K. Goossens et al.
Figure 11. Structural model (idealised) for one column in the Colh phase
Figure 10. Evolution of the columnar cross-section, S, in the Colh phase
at 1558C as a function of the number of carbon atoms n in the alkyl
chains for the series 8–11.
exhibited by compound 11 (at 1558C). The columns are arranged in a
two-dimensional hexagonal lattice with the lattice parameter aCol
=
42.0 ꢅ. The outer, solid circle indicates the imaginary edge of the column
with a diameter of rꢁ22.1 ꢅ. The inner, dashed circle indicates the imag-
inary edge of the non-aliphatic column centre with a diameter of r0
ꢁ12 ꢅ. The imidazolium planes, on average perpendicular to the plane
of the column slice, are represented by grey rectangles. The methyl
groups and hydrogen atoms attached to the imidazolium rings are repre-
sented by small grey (C atoms) and white (H atoms) circles, respectively.
Iodide anions are represented by grey spheres (not in space-filling
mode).
substituted imidazolium-based mesogens reported by the
groups of Kato,[19,21] Douce,[23] Percec,[66] and Tschierske.[29]
When one positions the imidazolium planes perpendicular
to the columnar axes and the aliphatic chains in a slice radi-
ally pointing away from the column centre, the central part
of the columns cannot be filled by the alkyl groups that are
attached to the nitrogen atoms. Indeed, the H-4 and H-5
atoms of the imidazolium rings would point towards the
centre in such an arrangement. At the same time, different
molecules in a slice, with more or less co-planar imidazolium
rings, cannot approach each other very closely because of
steric hindrance between the protruding methyl groups of
adjacent molecules. It is highly improbable that the resulting
central void is completely filled by the iodide anions due to
mutual repulsions. A model in which the long axes of the
molecules in a slice (these axes coincide with the bond be-
tween the imidazolium ring and the aryl substituent) nearly
coincide with tangent lines to the column centre, so that one
or both methyl groups point towards the centre, also seems
improbable. Therefore, we propose a model in which the
imidazolium planes of adjacent ionic head groups are on
average more or less perpendicular to the plane of the
column slice (i.e., on average parallel to the columnar long
axis). This allows a closer packing of the imidazolium
groups because of a reduced steric hindrance between the
protruding methyl groups, as well as an efficient charge al-
ternation between cations and anions. Recently, a Colh mes-
ophase in which rod-like mesogenic units with hydrogen-
bonding polar groups lie parallel to the columnar long axes
was reported, which supports the structure proposed in
Figure 11.[85] The aromatic rings connected to the imidazoli-
um groups are assumed to lay more or less in the plane of
the column slice (but let us recall that the concept of dis-
crete slices is a simplification for taper-shaped amphiphilic
molecules). As seen in the crystal structures of compounds 2
and 6, the imidazolium rings are at an angle to the benzene
rings. The observation of a halo centred at about 6.6 ꢅ in
the diffractograms of the Colh phases (denoted as h1 in
Figure 12) completely supports the proposed model. Indeed,
the width of a 1,3-dimethylimidazolium cation (with the
methyl groups lying in the plane of the imidazolium ring) is
about 6.2 ꢅ (estimated from the crystal structure of 6, and
with Chem3D). Thus, the height h1 of one column slice is
about 6.6 ꢅ. The influence of the replacement of one of the
methyl groups by a longer alkyl chain is currently being in-
vestigated. This newly found cation–anion arrangement is
Figure 12. X-ray diffractogram of 11 in the Colh phase at 1558C, repre-
sented as the scattering intensity (in counts) versus the scattering angle,
2q. The broad diffraction signal between 2qꢁ5.68 and 7.68 is due to the
covering foil used in the experimental setup.
4302
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 4291 – 4306

T-Shaped Ionic Liquid Crystals
FULL PAPER
expected to have an influence on the dynamic ionic conduc-
tivity measured parallel to the columnar axes of an aligned
sample (experiments by Kato et al.[19–21]).
compound 8 is the first example of a thermotropic Cub mes-
ophase of Pm3m symmetry. Structural models were pro-
¯
posed to describe the molecular organisation within the
SmA and Colh phases. It turned out that the imidazolium
head groups and the iodide counterions adopt a peculiar ori-
entation in the central part of the columns of the columnar
phases.
The number of molecules that self-aggregate within one
slice of height h1 ꢁ6.6 ꢅ can be estimated as Nslice =h1S
ACHTUNGRTEN(NUNG NA/
M)1 (in which NA is the Avogadro constant and M is the
molecular mass (in gmolꢀ1); Table 2). By assuming a density
of 1=1.0 gcmꢀ3, this gives a value of 7.2 for 11 at 1558C.
The model that we propose for the Colh phase shown by
compound 11, by considering the values for aCol, S, S0, h1
and Nslice, and the conclusions drawn above, is displayed in
Figure 11. For the most efficient packing, the molecules in
adjacent column slices are rotated by about 268 (=[3608/(7
molecules per slice)]/2). The Colh phases shown by com-
pounds 8, 9 and 10 have a similar structure. The model pre-
sented in Figure 11 is different from the structure of the
Colh phases shown by polycatenar mesogens, which arrange
into columns in a totally different way.[49]
The new rigid-core structure offers opportunities as a ver-
satile scaffold to synthesise more elaborate molecular archi-
tectures by using appropriately substituted benzaldehyde
compounds. Substitution of the nitrogen atoms of the imida-
zolium ring and the choice of the anion will allow the transi-
tion temperatures and phase behaviour to be fine-tuned.
Current research efforts focus on the replacement of the
iodide anions by more asymmetric and/or functional anions.
The mesomorphic behaviour of 2-arylsubstituted imidazoli-
um salts with these anions and that of the corresponding 2-
arylsubstituted benzimidazolium salts will be published else-
where. The thermal behaviour of 2-arylsubstituted imidazoli-
um salts with an elongated aryl substituent (i.e., containing
more than one aromatic ring, resulting in increasing shape
anisotropy) is being investigated as well. The latter com-
pounds might be interesting as materials displaying non-
linear optical properties because of the inherent donor–ac-
ceptor structure (provided that no centrosymmetric struc-
ture is formed in the mesophase).[86]
Again, a packing constraint analysis similar to that de-
scribed above for 2 and 6 can be applied to the Colh phase
of compound 11. For an octadecyl chain, Lalkyl was estimated
with the Chem3D program to be 22.57 ꢅ (after energy mini-
misation by an MM2 calculation). At 1558C, the critical pa-
rameter P=V
[(54.5 ꢅ2)
alkyl/ACHTUNGTRENNUNG(AMLalkyl)=[2ꢇ(17VCH +VCH )]/ACHTNUGTRENNNUG
2
3
ACHTUNGTRENNUNG
tained for compound 4 in the SmA phase at 1558C; in
Figure 11 the cross-sectional area of a cationic rigid core
and an iodide anion in a tangent plane to the column
formed by the polar moieties is comparable to that of the
untilted head groups in the SmA phases of compounds 2–4].
Because of interdigitation (see above), the theoretical
number of chains per head group is actually four, Pꢁ1.78.
A value for P that exceeds one indeed corresponds to an in-
verted micellar structure of amphiphiles (with the micellar
structures having a cylindrical shape in the case of a colum-
nar mesophase).
As noted by Douce et al., it is reasonable to assume that
the columnar structure described above is retained during
the Colh-to-Cub phase transition on cooling the Colh phase
of compound 8 (i.e., the minicolumns forming the rod net-
works in the multicontinuous Cub phase reflect the structure
of the columns in the higher-temperature columnar
phase).[23] They point out that molecular flexibility is impor-
tant in this aspect and that this transformation may occur
through regular column undulations and interconnections,
which increase substantially with decreasing temperature.
Experimental Section
General: NMR spectra were recorded on a Bruker Avance 300 spectrom-
eter (operating at 300 MHz for 1H), a Bruker AMX-400 spectrometer
1
(operating at 400 MHz for H) or a Bruker Avance II+ 600 spectrometer
1
(operating at 600 MHz for H). Abbreviations used for the description of
the spectra are s=singlet, d=doublet, dd=doublet of doublets, t=triplet
and m=multiplet. Chemical shifts are given in ppm relative to tetrame-
thylsilane (TMS) (when CDCl3 or CD3OD was used as a solvent) or to
the solvent peak (when CD2Cl2 was used as a solvent). Elemental analy-
ses (carbon, hydrogen and nitrogen) were obtained from a CE Instru-
ments EA-1110 elemental analyser. The results in percentages were inter-
preted allowing a deviation of ꢂ0.4%. ESI mass spectra were recorded
on a Thermo Finnigan LCQ Advantage mass spectrometer. Optical tex-
tures of the mesophases were observed with an Olympus BX60 polarising
optical microscope equipped with a LINKAM THMS600 heating stage
and a LINKAM TMS93 programmable temperature controller. DSC
traces were recorded under helium with a Mettler-Toledo DSC822e
module. Heating/cooling rates are specified in the captions of the ther-
mograms. Indium was used as a standard for temperature and enthalpy
calibrations. Thermal stability was assessed by TGA under nitrogen with
a TA Instruments SDT Q600 thermal analyser. Heating rates are speci-
fied in the captions of the thermograms. Powder XRD patterns were re-
corded with a Bruker AXS D8 Discover diffractometer mounted with a
copper X-ray ceramic tube, working at 1.6 kW. The emitted CuKa radia-
tion (l=1.5418 ꢅ) was focused on the sample by a Gçbel mirror. All of
the samples were prepared by spreading the powders on a thin cleaned
silicon wafer. Diffraction patterns were collected by using the Bragg–
Brentano reflection geometry (q/2q setup) at an angular resolution (in
2q) of 0.038 per step. The scattered signal was recorded by a one-dimen-
sional detector (LynxEye detector). Indexation of the powder X-ray dif-
fractograms was performed with the WinXPOW program package, with the
Index & Refine program by using Wernerꢀs TREOR algorithm pro-
Conclusion
Synthetic strategies were developed to prepare a homolo-
gous series of 1,3-dimethyl-2-[4-(alkoxy)phenyl]imidazolium
and 1,3-dimethyl-2-[3,4-bisACHTNUTRGNE(NUG alkoxy)phenyl]imidazolium salts.
Whereas only SmA phases were found for the former com-
pounds, the latter salts showed SmA, multicontinuous Cub
as well as Colh phases, depending on the chain length. To
the best of our knowledge, the enantiotropic Cub phase of
Chem. Eur. J. 2011, 17, 4291 – 4306
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4303

K. Goossens et al.
gram.[87,88] Molecular models were obtained with the Chem3D software
package from CambridgeSoft.
Compound 3: Yield: 92% (1.19 g); 1H NMR (300 MHz, CDCl3, 258C):
d=0.88 (t, J=6.0 Hz, 3H; CH3-CH2), 1.20–1.52 (m, 22H; CH2), 1.82
(quintet, J = 7.0 Hz, 2H; O-CH2-CH2), 3.87 (s, 6H; N-CH3), 4.04 (t, J =
6.5 Hz, 2H; O-CH2), 7.12 (d, Jo =7.9 Hz, 2H; H-aryl), 7.65 (d, Jo =7.9 Hz,
2H; H-aryl), 7.85 ppm (s, 2H; N-CH=CH-N); 13C NMR (75 MHz,
CD2Cl2, 258C): d=14.2, 22.8, 26.1, 29.1, 29.4, 29.6, 29.7, 29.8, 32.0, 37.0,
68.7, 111.7, 115.9, 123.6, 132.5, 145.2, 162.5 ppm; ESI-MS (methanol): m/z
(%): 385.9 (100) [MꢀI]+; elemental analysis calcd (%) for C25H41IN2O
(512.51): C 58.59, H 8.06, N 5.47; found: C 58.37, H 8.36, N 5.55.
For the determination of the crystal structures of compounds 2c, 2 and 6,
X-ray intensity data were collected at 100 K on a SMART 6000 diffrac-
tometer equipped with
a CCD detector, using CuKa radiation (l=
1.5418 ꢅ) and making use of f and w scans. The images were interpreted
and integrated with the program SAINT from Bruker.[89] All structures
were solved by direct methods and refined by full-matrix least-squares
techniques on F2 by using the SHELXTL program package.[90] Non-hy-
drogen atoms were refined anisotropically and hydrogen atoms in riding
mode with isotropic temperature factors fixed at 1.2 U(eq) of the parent
atoms (1.5 U(eq) for methyl groups). CCDC-782251 (2c) and 782252 (6)
contains the supplementary crystallographic data for this paper. These
data can be obtained free of charge from The Cambridge Crystallograph-
ic Data Centre via www.ccdc.cam.ac.uk/data_request/cif. The provisional
crystal data for compound 2 are given in the Supporting Information. 2c:
C19H28N2O; M=300.44 gmolꢀ1; monoclinic; P21/c (no. 14); a=24.064(1),
b=7.2810(4), c=19.9105(9) ꢅ; b=103.898(2)8; V=3386.4(3) ꢅ3; T=
Compound 5: Additional purification on a silica column with chloro-
form/methanol (95:5) as the eluent gave 5 (0.80 g, 55%). 1H NMR
(300 MHz, CD2Cl2, 258C): d=0.84–0.93 (m, 6H; CH3-CH2), 1.17–1.58
(m, 12H; CH2), 1.81 (quintet, J=7.6 Hz, 4H; O-CH2-CH2), 3.81 (s, 6H;
N-CH3), 4.05 (t, J=5.9 Hz, 4H; O-CH2), 7.00–7.40 (m, 3H; H-aryl),
7.74 ppm (s, 2H; N-CH=CH-N); 13C NMR (100 MHz, CDCl3, 258C): d=
14.1, 22.6, 25.8, 29.0, 29.2, 31.56, 31.63, 36.8, 69.3, 70.1, 111.9, 113.4, 115.0,
123.5, 124.0, 145.2, 149.9, 152.7 ppm; ESI-MS (methanol): m/z (%): 373.7
(100) [MꢀI]+, 873.1 (5) [M+MꢀI]+; elemental analysis calcd (%) for
C23H37IN2O2 (500.46): C 55.20, H 7.45, N 5.60; found: C 55.14, H 7.82, N
5.51.
100(2) K; Z=8; 1calcd. =1.179 gcmꢀ3
(CuKa)=0.561 mmꢀ1
; mACHTUNGRTENNUNG ; FACHTUNGTRENGN(UN 000)=
1312; crystal size 0.4ꢇ0.15ꢇ0.1 mm; 5964 independent reflections (Rint
=
Compound 6: Yield: 81% (1.08 g); 1H NMR (300 MHz, CD2Cl2, 258C):
d=0.87 (t, J=5.6 Hz, 6H; CH3-CH2), 1.17–1.60 (m, 28H; CH2), 1.82
(quintet, J=7.5 Hz, 4H; O-CH2-CH2), 3.81 (s, 6H; N-CH3), 4.03 (t, J=
6.2 Hz, 2H; O-CH2), 4.07 (t, J=6.3 Hz, 2H; O-CH2), 7.00–7.40 (m, 3H;
H-aryl), 7.70 ppm (s, 2H; N-CH=CH-N); 13C NMR (150 MHz, CDCl3,
258C): d=14.2, 22.8, 26.08, 26.14, 29.1, 29.3, 29.45, 29.46, 29.5, 29.66,
29.70, 32.0, 36.9, 69.3, 70.2, 111.9, 113.5, 115.2, 123.5, 124.0, 145.4, 150.0,
152.7 ppm; ESI-MS (methanol): m/z (%): 486.0 (100), [MꢀI]+; elemental
analysis calcd (%) for C31H53IN2O2·0.5H2O (621.68): C 59.89, H 8.76, N
4.51; found: C 59.93, H 8.89, N 4.40.
0.0887). Final R=0.0552 for 3850 reflections with I>2s(I) and wR2 =
0.1457 for all data. 6: C31H53IN2O2; M=612.67 gmolꢀ1; triclinic; P1 (no.
¯
2); a=7.3299(4), b=11.2460(5), c=39.019(2) ꢅ; a=88.599(2), b=
87.604(2), g=88.138(3)8; V=3211.0(3) ꢅ3; T=100(2) K; Z=4; 1calcd.
=
1.267 gcmꢀ3; m(CuKa)=8.032 mmꢀ1; F
ACHTUNGTRENNUNG ACHTUNGTREN(NNGU 000)=1288; crystal size 0.4ꢇ0.3ꢇ
0.15 mm; 11130 independent reflections (Rint =0.0481). Final R=0.0453
for 10072 reflections with I>2s(I) and wR2 =0.1159 for all data.
General procedure for the synthesis of 1-3 and 5–6: A solution of the ap-
propriate precursor (1c–3c, 5c–6c) (1 equiv) in dry THF was added
dropwise to an ice-cooled stirred suspension of NaH (1.5 equiv of a 60%
dispersion in mineral oil) in dry THF in a three-necked round-bottomed
flask. The mixture was stirred for 1 h at room temperature and for 1 h at
558C under an argon atmosphere (the formation of hydrogen gas could
be observed by the formation of bubbles in the solution). After cooling
to room temperature, iodomethane (20 equiv) was added dropwise and
the mixture was further stirred at the same temperature for 1 h, protected
from light. Then the reaction mixture was quickly transferred from the
round-bottomed flask to a screw-cap vial. The mixture was further stirred
overnight at 708C, protected from light. After cooling to room tempera-
ture, dichloromethane was added and the reaction mixture was filtered.
The solvent and excess iodomethane were removed under reduced pres-
sure. The residue was dissolved in dichloromethane and washed with
water. After removal of the solvent under reduced pressure, the product
was stirred in n-hexane at 508C for 30 min. The precipitate was filtered
off and washed with n-hexane. The pure compound was obtained as a
pale yellow to yellow powder. It was dried in vacuo at 508C for 24 h, and
stored in a container protected from light.
General procedure for the synthesis of 4, 7–11: Iodomethane (20 equiv)
was added dropwise to a solution of the appropriate precursor (4e, 7e–
11e) (1 equiv) in dry THF. The mixture was stirred for 1 h at room tem-
perature under an argon atmosphere, protected from light. Then, it was
transferred from the round-bottomed flask to a screw-cap vial, and was
further stirred overnight at 708C, protected from light. The solvent and
excess iodomethane were removed under reduced pressure. The pure
product was obtained as a pale yellow to yellow powder and was dried in
vacuo at 508C for 24 h, and stored in a container protected from light.
Compound 4: Yield: 89% (0.60 g); 1H NMR (400 MHz, CDCl3, 258C):
d=0.83 (t, J=6.9 Hz, 3H; CH3-CH2), 1.20–1.54 (m, 30H; CH2), 1.78
(quintet, J=7.1 Hz, 2H; O-CH2-CH2), 3.83 (s, 6H; N-CH3), 4.00 (t, J=
6.5 Hz, 2H; O-CH2), 7.08 (d, Jo =8.7 Hz, 2H; H-aryl), 7.61 (d, Jo =
8.7 Hz, 2H; H-aryl), 7.81 ppm (s, 2H; N-CH=CH-N); 13C NMR
(100 MHz, CDCl3, 258C): d=14.1, 22.7, 26.0, 29.1, 29.35, 29.37, 29.57,
29.61, 29.65, 29.69, 31.9, 36.9, 68.6, 111.7, 115.9, 123.5, 132.5, 145.2,
162.5 ppm; ESI-MS (methanol): m/z (%): 442.0 (100) [MꢀI]+, 1009.3 (2)
[M+MꢀI]+; elemental analysis calcd (%) for C29H49IN2O (568.62): C
61.26, H 8.69, N 4.93; found: C 60.98, H 8.84, N 4.91.
Compound 7: Yield: 96% (0.19 g); 1H NMR (300 MHz, CDCl3, 258C):
d=0.88 (t, J=6.4 Hz, 6H; CH3-CH2), 1.18–1.65 (m, 44H; CH2), 1.77–
1.93 (m, 4H; O-CH2-CH2), 3.85 (s, 6H; N-CH3), 4.07 (t, J=6.4 Hz, 4H;
O-CH2), 7.04 (d, Jo =8.3 Hz, 1H; H-aryl), 7.17 (dd, Jo =8.3 Hz, Jm =
1.7 Hz, 1H; H-aryl), 7.31 (d, Jm =1.7 Hz, 1H; H-aryl), 7.72 ppm (s, 2H;
N-CH=CH-N); 13C NMR (75 MHz, CDCl3, 258C): d=14.3, 22.8, 26.2,
29.2, 29.3, 29.5, 29.6, 29.78, 29.82, 29.9, 32.1, 36.9, 69.4, 70.2, 111.8, 113.4,
115.3, 123.4, 123.9, 145.6, 150.1, 152.8 ppm; ESI-MS (methanol): m/z
(%): 598.1 (100) [MꢀI]+, 1321.5 (5) [M+MꢀI]+; elemental analysis calcd
(%) for C39H69IN2O2 (724.88): C 64.62, H 9.59, N 3.86; found: C 64.41, H
9.90, N 4.01.
Compound 8: Yield: 94% (0.58 g); 1H NMR (300 MHz, CDCl3, 258C):
d=0.88 (t, J=6.4 Hz, 6H; CH3-CH2), 1.17–1.62 (m, 48H; CH2), 1.77–
1.93 (m, 4H; O-CH2-CH2), 3.85 (s, 6H; N-CH3), 4.07 (t, J=6.2 Hz, 4H;
O-CH2), 7.04 (d, Jo =8.3 Hz, 1H; H-aryl), 7.17 (dd, Jo =8.3 Hz, Jm =
1.6 Hz, 1H; H-aryl), 7.29 (d, Jm =1.6 Hz, 1H; H-aryl), 7.75 ppm (s, 2H;
N-CH=CH-N); 13C NMR (75 MHz, CDCl3, 258C): d=14.3, 22.8, 26.1,
26.2, 29.2, 29.3, 29.5, 29.6, 29.78, 29.81, 29.9, 32.1, 36.9, 69.4, 70.2, 111.9,
113.4, 115.2, 123.4, 123.9, 145.5, 150.0, 152.8 ppm; ESI-MS (methanol):
Compound 1: Additional purification on a silica column with chloroform/
methanol (95:5) as the eluent gave 1 (0.68 g, 42%). 1H NMR (300 MHz,
CD2Cl2, 258C): d=0.90 (t, J=6.2 Hz, 3H; CH3-CH2), 1.28–1.46 (m, 6H;
CH2), 1.81 (quintet, J=6.9 Hz, 2H; O-CH2-CH2), 3.81 (s, 6H; N-CH3),
4.05 (t, J=6.5 Hz, 2H; O-CH2), 7.13 (d, Jo =8.6 Hz, 2H; H-aryl), 7.56 (d,
Jo =8.6 Hz, 2H; H-aryl), 7.74 ppm (s, 2H; N-CH=CH-N); 13C NMR
(100 MHz, CDCl3, 258C): d=14.0, 22.6, 25.7, 29.0, 31.5, 36.8, 68.6, 100.0,
111.7, 115.9, 123.6, 132.4, 145.1, 162.5 ppm; ESI-MS (methanol): m/z
(%): 273.5 (100) [MꢀI]+, 673.1 (5) [M+MꢀI]+; elemental analysis calcd
(%) for C17H25IN2O (400.30): C 51.01, H 6.29, N 7.00; found: C 50.90, H
6.61, N 6.91.
Compound 2: Yield: 82% (0.50 g); 1H NMR (300 MHz, CD2Cl2, 258C):
d=0.87 (t, J=5.9 Hz, 3H; CH3-CH2), 1.23–1.51 (m, 14H; CH2), 1.81
(quintet, J=6.8 Hz, 2H; O-CH2-CH2), 3.81 (s, 6H; N-CH3), 4.05 (t, J=
6.4 Hz, 2H; O-CH2), 7.14 (d, Jo =8.3 Hz, 2H; H-aryl), 7.55 (d, Jo =
8.3 Hz, 2H; H-aryl), 7.71 ppm (s, 2H; N-CH=CH-N); 13C NMR
(100 MHz, CDCl3, 258C): d=14.2, 22.8, 26.1, 29.1, 29.38, 29.41, 29.6,
32.0, 37.0, 68.7, 111.7, 115.9, 123.6, 132.5, 145.3, 162.6 ppm; ESI-MS
(methanol): m/z (%): 329.6 (100) [MꢀI]+, 785.1 (5) [M+MꢀI]+; elemen-
tal analysis calcd (%) for C21H33IN2O·0.5H2O (465.41): C 54.19, H 7.36,
N 6.02; found: C 54.05, H 7.58, N 5.97.
4304
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 4291 – 4306

T-Shaped Ionic Liquid Crystals
FULL PAPER
m/z (%): 626.0 (100) [MꢀI]+; elemental analysis calcd (%) for
C41H73IN2O2 (752.94): C 65.40, H 9.77, N 3.72; found: C 65.25, H 9.62, N
3.75.
[9] X. J. Wang, F. W. Heinemann, M. Yang, B. U. Melcher, M. Fekete,
A. V. Mudring, P. Wasserscheid, K. Meyer, Chem. Commun. 2009,
7405–7407.
Compound 9: Yield: 92% (0.56 g); 1H NMR (300 MHz, CDCl3, 258C):
d=0.88 (t, J=6.4 Hz, 6H; CH3-CH2), 1.17–1.60 (m, 52H; CH2), 1.78–
1.96 (m, 4H; O-CH2-CH2), 3.84 (s, 6H; N-CH3), 4.07 (t, J=6.4 Hz, 4H;
O-CH2), 7.04 (d, Jo =8.3 Hz, 1H; H-aryl), 7.16 (dd, Jo =8.3 Hz, Jm could
not be determined because the doublet of doublets was not fully re-
solved, 1H; H-aryl), 7.31 (d, Jm =1.9 Hz, 1H; H-aryl), 7.69 ppm (s, 2H;
N-CH=CH-N); 13C NMR (75 MHz, CDCl3, 258C): d=14.3, 22.8, 26.15,
26.21, 29.2, 29.3, 29.5, 29.6, 29.8, 29.9, 32.1, 36.9, 70.2, 111.8, 113.4, 115.3,
123.4, 145.5, 150.1, 152.8 ppm; ESI-MS (methanol): m/z (%): 654.0 (100)
[MꢀI]+; elemental analysis calcd (%) for C43H77IN2O2 (780.99): C 66.13,
H 9.94, N 3.59; found: C 66.07, H 10.32, N 3.63.
[10] K. Hoshino, M. Yoshio, T. Mukai, K. Kishimoto, H. Ohno, T. Kato,
J. Polym. Sci., Part A: Polym. Chem. 2003, 41, 3486–3492.
[11] H. Yoshizawa, T. Mihara, N. Koide, Mol. Cryst. Liq. Cryst. 2004,
423, 61–72.
[12] H. Yoshizawa, T. Mihara, N. Koide, Liq. Cryst. 2005, 32, 143–149.
[13] Q. X. Zhang, L. S. Jiao, C. S. Shan, P. Hou, B. Chen, X. Y. Xu, L.
Niu, Liq. Cryst. 2008, 35, 765–772.
[14] Q. X. Zhang, C. S. Shan, X. D. Wang, L. L. Chen, L. Niu, B. Chen,
Liq. Cryst. 2008, 35, 1299–1305.
[15] S. Yazaki, M. Funahashi, T. Kato, J. Am. Chem. Soc. 2008, 130,
13206–13207.
[16] K. Goossens, P. Nockemann, K. Driesen, B. Goderis, C. Gçrller-Wal-
rand, K. Van Hecke, L. Van Meervelt, E. Pouzet, K. Binnemans, T.
Cardinaels, Chem. Mater. 2008, 20, 157–168.
[17] W. Li, J. Zhang, B. Li, M. L. Zhang, L. X. Wu, Chem. Commun.
2009, 5269–5271.
[18] W. Dobbs, L. Douce, L. Allouche, A. Louati, F. Malbosc, R. Welter,
New J. Chem. 2006, 30, 528–532.
[19] M. Yoshio, T. Mukai, H. Ohno, T. Kato, J. Am. Chem. Soc. 2004,
126, 994–995.
[20] M. Yoshio, T. Kagata, K. Hoshino, T. Mukai, H. Ohno, T. Kato, J.
Am. Chem. Soc. 2006, 128, 5570–5577.
[21] M. Yoshio, T. Ichikawa, H. Shimura, T. Kagata, A. Hamasaki, T.
Mukai, H. Ohno, T. Kato, Bull. Chem. Soc. Jpn. 2007, 80, 1836–
1841.
[22] H. Shimura, M. Yoshio, K. Hoshino, T. Mukai, H. Ohno, T. Kato, J.
Am. Chem. Soc. 2008, 130, 1759–1765.
[23] W. Dobbs, B. Heinrich, C. Bourgogne, B. Donnio, E. Terazzi, M. E.
Bonnet, F. Stock, P. Erbacher, A. L. Bolcato-Bellemin, L. Douce, J.
Am. Chem. Soc. 2009, 131, 13338–13346.
Compound 10: Yield: 95% (0.57 g); 1H NMR (300 MHz, CDCl3, 258C):
d=0.88 (t, J=6.5 Hz, 6H; CH3-CH2), 1.15–1.60 (m, 56H; CH2), 1.78–
1.95 (m, 4H; O-CH2-CH2), 3.84 (s, 6H; N-CH3), 4.07 (t, J=6.4 Hz, 4H;
O-CH2), 7.04 (d, Jo =8.3 Hz, 1H; H-aryl), 7.17 (dd, Jo =8.3 Hz, Jm =
1.9 Hz, 1H; H-aryl), 7.32 (d, Jm =1.9 Hz, 1H; H-aryl), 7.70 ppm (s, 2H;
N-CH=CH-N); 13C NMR (75 MHz, CDCl3, 258C): d=14.3, 22.8, 26.2,
29.2, 29.3, 29.5, 29.6, 29.8, 29.9, 32.1, 36.9, 69.4, 70.2, 111.8, 113.4, 115.3,
123.4, 123.9, 145.6, 150.1, 152.8 ppm; ESI-MS (methanol): m/z (%): 682.0
(100) [MꢀI]+; elemental analysis calcd (%) for C45H81IN2O2 (809.04): C
66.81, H 10.09, N 3.46; found: C 67.03, H 10.42, N 3.49.
Compound 11: Yield: 98% (0.33 g); 1H NMR (300 MHz, CDCl3, 258C):
d=0.88 (t, J=6.5 Hz, 6H; CH3-CH2), 1.15–1.62 (m, 60H; CH2), 1.78–
1.95 (m, 4H; O-CH2-CH2), 3.85 (s, 6H; N-CH3), 4.07 (t, J=6.4 Hz, 4H;
O-CH2), 7.04 (d, Jo =8.3 Hz, 1H; H-aryl), 7.17 (dd, Jo =8.3 Hz, Jm could
not be determined because the doublet of doublets was not fully re-
solved, 1H; H-aryl), 7.30 (s, 1H; H-aryl), 7.72 ppm (s, 2H; N-CH=CH-
N); 13C NMR (75 MHz, CDCl3, 258C): d=14.3, 22.8, 26.15, 26.21, 29.2,
29.3, 29.5, 29.6, 29.8, 29.9, 32.1, 36.9, 69.4, 70.2, 111.8, 113.4, 115.3, 123.4,
123.9, 145.6, 150.1, 152.8 ppm; ESI-MS (methanol, m/z) (%): 710.2 (100)
[MꢀI]+, 1545.6 (5) [M+MꢀI]+; elemental analysis calcd (%) for
C47H85IN2O2 (837.09): C 67.44, H 10.23, N 3.35; found: C 67.33, H 10.51,
N 3.41.
[24] J. H. Olivier, F. Camerel, J. Barberꢈ, P. Retailleau, R. Ziessel, Chem.
Eur. J. 2009, 15, 8163–8174.
[25] J. H. Olivier, F. Camerel, G. Ulrich, J. Barberꢈ, R. Ziessel, Chem.
Eur. J. 2010, 16, 7134–7142.
[26] P. H. J. Kouwer, T. M. Swager, J. Am. Chem. Soc. 2007, 129, 14042–
14052.
[27] J. Fouchet, L. Douce, B. Heinrich, R. Welter, A. Louati, Beilstein J.
Org. Chem. 2009, 5, No. 51.
[28] X. J. Li, D. W. Bruce, J. M. Shreeve, J. Mater. Chem. 2009, 19, 8232–
8238.
[29] X. H. Cheng, X. Q. Bai, S. Jing, H. Ebert, M. Prehm, C. Tschierske,
Chem. Eur. J. 2010, 16, 4588–4601.
[30] J.-M. Suisse, S. Bellemin-Laponnaz, L. Douce, A. Maisse-Francois,
R. Welter, Tetrahedron Lett. 2005, 46, 4303–4305.
[31] J.-M. Suisse, L. Douce, S. Bellemin-Laponnaz, A. Maisse-Francois,
R. Welter, Y. Miyake, Y. Shimizu, Eur. J. Inorg. Chem. 2007, 3899–
3905.
Acknowledgements
K.G. is a postdoctoral fellow of the FWO-Flanders (Belgium). S.W.
thanks the IWT-Flanders and Umicore for a doctoral fellowship. Finan-
cial support by the K.U. Leuven (projects GOA 08/03 and IDO/05/005) is
gratefully acknowledged. The authors would like to thank Dirk Henot
for performing the CHN elemental analyses; Dimitri Soccol for recording
the thermogravimetric traces; Dr. Jan DꢀHaen and Bart Ruttens (UHas-
selt, Institute for Materials Research) for recording the powder X-ray dif-
fractograms; and the referees for valuable suggestions.
[32] T. Mukai, M. Yoshio, T. Kato, H. Ohno, Chem. Lett. 2004, 33, 1630–
1631.
[33] T. Mukai, M. Yoshio, T. Kato, M. Yoshizawa-Fujita, H. Ohno, Elec-
trochemistry 2005, 73, 623–626.
[34] M. A. Alam, J. Motoyanagi, Y. Yamamoto, T. Fukushima, J. Kim, K.
Kato, M. Takata, A. Saeki, S. Seki, S. Tagawa, T. Aida, J. Am.
Chem. Soc. 2009, 131, 17722–17723.
[1] K. Binnemans, Chem. Rev. 2005, 105, 4148–4204.
[2] C. J. Bowlas, D. W. Bruce, K. R. Seddon, Chem. Commun. 1996,
1625–1626.
[35] H. Fujioka, K. Murai, Y. Ohba, A. Hiramatsu, Y. Kita, Tetrahedron
Lett. 2005, 46, 2197–2199.
[3] C. M. Gordon, J. D. Holbrey, A. R. Kennedy, K. R. Seddon, J.
Mater. Chem. 1998, 8, 2627–2636.
[36] H. Fujioka, K. Murai, O. Kubo, Y. Ohba, Y. Kita, Tetrahedron 2007,
63, 638–643.
[4] J. D. Holbrey, K. R. Seddon, J. Chem. Soc. Dalton Trans. 1999,
2133–2139.
[37] A. J. Mancuso, D. Swern, Synthesis 1981, 165–185.
[38] D. H. Huh, H. Ryu, Y. G. Kim, Tetrahedron 2004, 60, 9857–9862.
[39] N. Menschutkin, Z. Phys. Chem. 1890, 6, 41–57.
[40] M. Ishihara, H. Togo, Synlett 2006, 227–230.
[41] P. Parik, S. Senauerova, V. Liskova, K. Handlir, M. Ludwig, J. Heter-
ocycl. Chem. 2006, 43, 835–841.
[5] C. K. Lee, H. W. Huang, I. J. B. Lin, Chem. Commun. 2000, 1911–
1912.
[6] A. E. Bradley, C. Hardacre, J. D. Holbrey, S. Johnston, S. E. J.
McMath, M. Nieuwenhuyzen, Chem. Mater. 2002, 14, 629–635.
[7] M. Yoshio, T. Mukai, K. Kanie, M. Yoshizawa, H. Ohno, T. Kato,
Chem. Lett. 2002, 320–321.
[8] M. Tosoni, S. Laschat, A. Baro, Helv. Chim. Acta 2004, 87, 2742–
2749.
[42] V. Prey, Chem. Ber. B 1942, 75, 350–356.
[43] R. Kia, H. K. Fun, H. Kargar, Acta Crystallogr. E 2009, 65, o798.
Chem. Eur. J. 2011, 17, 4291 – 4306
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4305

K. Goossens et al.
[44] H. Moreno-Diaz, G. Navarrete-Vazquez, S. Estrada-Soto, H. Tla-
huext, Acta Crystallogr. E 2006, 62, o2601–o2602.
[45] K. Goossens, K. Lava, P. Nockemann, K. Van Hecke, L. Van Meer-
velt, K. Driesen, C. Gçrller-Walrand, K. Binnemans, T. Cardinaels,
Chem. Eur. J. 2009, 15, 656–674.
[46] K. E. Rowe, D. W. Bruce, J. Mater. Chem. 1998, 8, 331–341.
[47] J. MalthÞte, H. T. Nguyen, C. Destrade, Liq. Cryst. 1993, 13, 171–
187.
[67] C. Hammond, The Basics of Crystallography and Diffraction, 2nd
ed., Oxford University Press, New York, 2001.
[68] E. Terazzi, C. Bourgogne, R. Welter, J. L. Gallani, D. Guillon, G.
Rogez, B. Donnio, Angew. Chem. 2008, 120, 500–505; Angew.
Chem. Int. Ed. 2008, 47, 490–495.
[69] M. Ichihara, A. Suzuki, K. Hatsusaka, K. Ohta, Liq. Cryst. 2007, 34,
555–567.
[70] S. Diele, Curr. Opin. Colloid Interface Sci. 2002, 7, 333–342.
[71] M. Impꢂror-Clerc, Curr. Opin. Colloid Interface Sci. 2005, 9, 370–
376.
[48] H. T. Nguyen, C. Destrade, J. MalthÞte, Adv. Mater. 1997, 9, 375–
388.
[72] X. B. Zeng, G. Ungar, M. Impꢂror-Clerc, Nat. Mater. 2005, 4, 562–
567.
[49] B. Donnio, D. W. Bruce, J. Chem. Soc. Dalton Trans. 1997, 2745–
2755.
[73] D. W. Bruce, B. Donnio, S. A. Hudson, A. M. Levelut, S. Megtert, D.
Petermann, M. Veber, J. Phys. II 1995, 5, 289–302.
[74] A. M. Levelut, M. Clerc, Liq. Cryst. 1998, 24, 105–115.
[75] S. Diele, P. Gçring in Handbook of Liquid Crystals. Volume 2B:
Low Molecular Weight Liquid Crystals II (Eds.: D. Demus, J. W.
Goodby, G. W. Gray, H.-W. Spiess, V. Vill), Wiley-VCH, Weinheim,
1998, pp. 887–900.
[76] S. Kutsumizu, T. Ichikawa, S. Nojima, S. Yano, Chem. Commun.
1999, 1181–1182.
[77] S. Kutsumizu, Curr. Opin. Solid State Mater. Sci. 2002, 6, 537–543.
[78] M. Impꢂror-Clerc, Curr. Opin. Colloid Interface Sci. 2005, 9, 370–
376.
[79] C. Tschierske, Chem. Soc. Rev. 2007, 36, 1930–1970.
[80] It should be noted that the crystal structure for only one of the crys-
tal polymorphs of 6 was obtained.
[81] M. Marcos, R. Gimꢂnez, J. L. Serrano, B. Donnio, B. Heinrich, D.
Guillon, Chem. Eur. J. 2001, 7, 1006–1013.
[82] K. Goossens, K. Lava, P. Nockemann, K. Van Hecke, L. Van Meer-
velt, P. Pattison, K. Binnemans, T. Cardinaels, Langmuir 2009, 25,
5881–5897.
[83] B. Donnio, B. Heinrich, H. Allouchi, J. Kain, S. Diele, D. Guillon,
D. W. Bruce, J. Am. Chem. Soc. 2004, 126, 15258–15268.
[84] T. Kato, N. Mizoshita, K. Kishimoto, Angew. Chem. 2006, 118, 44–
74; Angew. Chem. Int. Ed. 2006, 45, 38–68.
[50] D. Fazio, C. Mongin, B. Donnio, Y. Galerne, D. Guillon, D. W.
Bruce, J. Mater. Chem. 2001, 11, 2852–2863.
[51] T. Cardinaels, J. Ramaekers, P. Nockemann, K. Driesen, K. Van
Hecke, L. Van Meervelt, G. Wang, S. de Feyter, E. Fernandez-Igle-
sias, D. Guillon, B. Donnio, K. Binnemans, D. W. Bruce, Soft Matter
2008, 4, 2172–2185.
[52] A. K. Abdul-Sada, A. M. Greenway, P. B. Hitchcock, T. J. Moham-
med, K. R. Seddon, J. A. Zora, J. Chem. Soc. Chem. Commun. 1986,
1753–1754.
[53] A. Elaiwi, P. B. Hitchcock, K. R. Seddon, N. Srinivasan, Y. M. Tan,
T. Welton, J. A. Zora, J. Chem. Soc. Dalton Trans. 1995, 3467–3472.
[54] T. Steiner, Angew. Chem. 2002, 114, 50–80; Angew. Chem. Int. Ed.
2002, 41, 48–76.
[55] S. H. Seo, J. H. Park, G. N. Tew, J. Y. Chang, Tetrahedron Lett. 2007,
48, 6839–6844.
[56] J. N. Israelachvili, D. J. Mitchell, B. W. Ninham, J. Chem. Soc. Fara-
day Trans. 2 1976, 72, 1525–1568.
[57] K. Borisch, S. Diele, P. Gçring, H. Muller, C. Tschierske, Liq. Cryst.
1997, 22, 427–443.
[58] K. Borisch, S. Diele, P. Gçring, H. Kresse, C. Tschierske, J. Mater.
Chem. 1998, 8, 529–543.
[59] U. Stebani, G. Lattermann, M. Wittenberg, R. Festag, J. H. Wen-
dorff, Adv. Mater. 1994, 6, 572–574.
[60] U. Stebani, G. Lattermann, R. Festag, M. Wittenberg, J. H. Wen-
dorff, J. Mater. Chem. 1995, 5, 2247–2251.
[85] M. Prehm, F. Liu, X. B. Zeng, G. Ungar, C. Tschierske, J. Am.
Chem. Soc. 2008, 130, 14922–14923.
[61] C. Tschierske, J. Mater. Chem. 1998, 8, 1485–1508.
[62] C. Tschierske, J. Mater. Chem. 2001, 11, 2647–2671.
[63] A. J. Boydston, C. S. Pecinovsky, S. T. Chao, C. W. Bielawski, J. Am.
Chem. Soc. 2007, 129, 14550–14551.
[64] S. Yazaki, Y. Kamikawa, M. Yoshio, A. Hamasaki, T. Mukai, H.
Ohno, T. Kato, Chem. Lett. 2008, 37, 538–539.
[65] K. Borisch, S. Diele, P. Gçring, H. Kresse, C. Tschierske, Angew.
Chem. 1997, 109, 2188–2190; Angew. Chem. Int. Ed. Engl. 1997, 36,
2087–2089.
[86] Z. Q. Zhu, S. J. Xiang, Q. Y. Chen, C. S. Chen, Z. Zeng, Y. P. Cui,
J. C. Xiao, Chem. Commun. 2008, 5016–5018.
[87] WinXPOW 1.06, STOE & Cie GmbH, Darmstadt (Germany), 1999.
[88] P. E. Werner, L. Eriksson, M. Westdahl, J. Appl. Crystallogr. 1985,
18, 367–370.
[89] SAINT, Version 5/6.0, Bruker Analytical X-ray Systems Inc., Madi-
son, Wisconsin, 1997.
[90] SHELXTL-PC, Version 5.1, Bruker Analytical X-ray Systems Inc.,
Madison, Wisconsin, 1997.
[66] M. R. Imam, M. Peterca, U. Edlund, V. S. K. Balagurusamy, V.
Percec, J. Polym. Sci. Part A: Polym. Chem. 2009, 47, 4165–4193.
Received: July 7, 2010
Published online: March 7, 2011
4306
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 4291 – 4306