Structure and ionic conductivity of liquid crystals having propylene carbonate units

Article abstract of DOI:10.1039/c4ta05401f

Liquid-crystalline compounds with a perfluorinated aromatic ring as the mesogenic core and a propylene carbonate unit were prepared and mixed with lithium bis(trifluoro-methanesulfonyl)-imide (LiTFSI). The self-assembly was driven by the interaction of th

Full text of DOI:10.1039/c4ta05401f

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Structure and ionic conductivity of liquid crystals
having propylene carbonate units†
Cite this: J. Mater. Chem. A, 2015, 3,
Andreas Eisele,a Konstantinos Kyriakos,b Rajesh Bhandary,cd Monika Schonhoff,c
2942
¨
b
a
*
Christine M. Papadakis and Bernhard Rieger
Liquid-crystalline compounds with a perfluorinated aromatic ring as the mesogenic core and a propylene
carbonate unit were prepared and mixed with lithium bis(trifluoro-methanesulfonyl)-imide (LiTFSI). The
self-assembly was driven by the interaction of the polar propylene carbonate unit. Thus it is one of the
few examples of liquid crystals with only one phenyl group that is able to self-assemble. Small- and Wide
angle X-ray scattering (SAXS/WAXS) measurements indicate that the molecules spontaneously formed a
smectic phase. Calculation of the Li-salt dissociation from the data obtained by a combination of
impedance spectroscopy measurements and diffusion experiments by 7Li and 19F pulsed field gradient
(PFG) NMR revealed good dissociation properties of the compounds. The complex of (2-Oxo-1,3-
dioxolan-4-yl)methyl 4-(decyloxy)-2,3,5,6-tetrafluorobenzoate (4b) with LiTFSI exhibited anisotropic
conductivity. The conductivity parallel to the smectic orientation was 125 times higher than
perpendicular and is one of the best reported for liquid crystals with propylene carbonate fragments.
Received 9th October 2014
Accepted 9th December 2014
DOI: 10.1039/c4ta05401f
www.rsc.org/MaterialsA
as segregation leads to the formation of nanoscale ion path-
ways.11–14 Until now, the focus has been mainly on liquid
crystals with poly(ethylene oxide) (PEO) moieties.15–21 One of
Introduction
Liquid crystals have gained remarkable attraction as an
alternative to common electrolytes like cyclic carbonates and
poly(ethylene oxide) (PEO).1–3 For the further development of
lithium-ion batteries, it is essential to replace the liquid elec-
trolytes by solid ones because they are less ammable and are
easily processable.4–8 Currently, the most used electrolytes are
organic liquids based on linear and cyclic carbonates as their
high dielectric constant results in good lithium-ion conduc-
tivity.9 To overcome the safety problems of organic liquids,
Wnek and co-workers synthesized polysiloxanes with cyclic
carbonate side chains.10 Although their polymers have a
dielectric constant ranging from 22 to 44, the polymer Li-salt
complex has only a conductivity of 5 ꢀ 10ꢁ7 S cmꢁ1 at room
temperature. An alternative could be liquid crystals with cyclic
carbonates as the moiety enhancing salt dissociation. Liquid
crystals are less ammable and can transport ions efficiently
the few examples of a liquid crystal with a carbonate fragment
was synthesized by Kato and co-workers and it exhibits a
hexagonal columnar structure.22 ꢂHowever, the ionic conduc-
tivity of 2.2 ꢀ 10ꢁ8 S cmꢁ1 at 22 C is relatively low. A reason
for the low ionic conductivity could be the high dipole–dipole
interactions between the polar cyclic carbonate units.10 In
contrast to molecules with PEO side chains, where the ions are
transported through segmental motion of the PEO frag-
ments,23 carbonates on a polymer backbone show a different
conductivity mechanism. The ions are transported via diffu-
sion or an ion-hopping mechanism.24 A different molecular
structure with a lower molecular weight and a smectic liquid
crystalline structure instead of a columnar one can result in an
improved ionic conductivity. It is difficult to align columnar
structures macroscopically; therefore the ion diffusion is
encumbered. Alignment in a smectic structure may enhance
ion diffusion through the material.11,25 Until now the
conductivity mechanism of liquid crystals in the mesogenic
phase is not fully understood.
a
¨
¨
¨
WACKER-Lehrstuhl fur Makromolekulare Chemie, Technische Universitat Munchen,
¨
Lichtenbergstr. 4, 85747 Garching bei Munchen, Germany. E-mail: rieger@tum.de
bFachgebiet Physik weicher Materie, Physik Department, Technische Universitat
¨
¨
¨
Munchen, James-Franck-Str. 1, 85747 Garching bei Munchen, Germany
The goal of this work is to synthesize a liquid crystalline
molecule with a cyclic carbonate unit which has a good thermal
stability and enhanced ionic conductivity when mixed with
lithium salts. On the basis of Small- and Wide-Angle X-ray
Scattering (SAXS/WAXS), conductivity measurements and
diffusion experiments by the PFG-NMR technique, we expect to
get a deeper insight into the lithium coordination and the
conductivity mechanism of aligned liquid crystals.
c
¨
Institut fur Physikalische Chemie, University of Muenster, Corrensstr. 8/30, 48149
¨
Munster, Germany
dNRW Graduate School of Chemistry, University of Muenster, Wilhelm-Klemm-Str. 10,
¨
D-48149 Munster, Germany
† Electronic supplementary information (ESI) available: DSC diagrams for the
mixtures of 4a, 4b and 4c with various amounts of LiTFSI; WAXS/SAXS of 4a
and 4b mixed with 10 mol% LiTFSI, ionic conductivity measurements for 4b, 6
and
8 with various amounts of LiTFSI; NMR spectra of all compounds
discussed in this paper. See DOI: 10.1039/c4ta05401f
2942 | J. Mater. Chem. A, 2015, 3, 2942–2953
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
We have chosen a peruorinated aromatic ring as the
Journal of Materials Chemistry A
mesogenic core, because uorinated compounds have been
found to improve the physical properties in contrast to their
hydrocarbon homologues.3,26–29 The Li-salt mixtures of the per-
uorinated substances may exhibit a higher ionic conductivity
compared to their non-uorinated analogues. In combination
with the polar carbonate function, it is possible to synthesize
liquid crystalline molecules with a lower molar mass. The
higher order in the liquid crystalline state and the cyclic
carbonate moiety inducing a high degree of dissociation of the
lithium salt may result in enhanced ionic conductivity.30,31 The
results for the peruorinated compounds were compared with
those for a non-uorinated molecule. To show the effect of
cyclic carbonates on the formation of a liquid crystalline state
and the ionic conductivity, also molecules with tetra(ethylene
oxide) (TEO) having the same uorinated mesogenic core were
synthesized.
Scheme 2 Synthesis of the investigated non-fluorinated mesogenic
molecule.
The liquid crystalline compounds 4 and 8 were synthesized
by an esterication starting from molecule 3 and 7, respectively,
with glycerine carbonate, EDC and DMAP as coupling agents.22
As the tetrauorobenzoic acid (3) is more reactive than the non-
uorinated benzoic acid used by Kato and co-workers, the
amount of the coupling reagents EDC and DMAP could be
drastically reduced.
Reduction of 3c with BH3–THF solution (1 M) gave the
mesogenic core 5. BH3 was chosen as the reduction agent as
other reduction agents either led to a partial deuorination of
the aromatic ring, mainly on the meta position (LiAlH4) or were
too weak (NaBH4). Compound 6 was synthesized by a standard
Williamson ether synthesis in acetonitrile starting from
substance 5.
Results and discussion
Synthesis of the uorinated mesogens
The synthesis routes of the uorinated and non-uorinated
mesogenic molecules which were analyzed in this study are
shown in Schemes 1 and 2. Substances 4a, 4b and 4c have a
peruorinated aromatic ring as the mesogenic core and differ in
the length of the alkyl chain, whereas 8 has a non-uorinated
mesogenic core.
The syntheses of the mesogenic cores 3 and 7 were carried
out according to a modied literature procedure in quantitative
yields.26,32
Thermal properties
The thermal properties of the synthesized compounds were
analyzed by differential scanning calorimetry (DSC). Addition-
ally, optical textures of the substances which exhibit a liquid
crystalline phase were obtained by polarized optical microscopy
(POM).
Fig. 1 presents the DSC traces from the 1st heating scan, the
subsequent 2nd heating scans and the subsequent 1st cooling
scans obtained for 8, 4a, 4b and 4c. All compounds, except 6
showed a liquid crystalline phase (see ESI, Fig. S1†). Substance 6
is a liquid at room temperature and crystallizes without the
formation of a liquid crystalline state as the intermolecular
interaction of the TEO group is too weak. Cyclic carbonates have
a higher dipole moment than TEO (propylene carbonate m ¼
5.36 D, TEO m ¼ 3.25 D at 40 C).33,34 Hence the dipole–dipole
ꢂ
interaction of the carbonate groups is higher which is expected
to increase the interchain interactions.10,19,35
In POM, the formation of a fan-shaped texture was observed
for the substances 4b and 4c when cooled from the isotropic
ˆ
melt. Substance 4a formed batonnets and 8 formed a focal conic
as well as a fan-shaped texture. These textures are typical for a
smectic A (SA) phase (Fig. 3 and ESI Fig. S15†).36 Aer further
cooling, the compounds started to crystallize, except substance
4a and 8, where no crystallization could be observed within the
measured temperature range. The longer the alkyl chains of the
uorinated liquid crystalline molecules, the higher are the
isotropization temperatures as well as the crystallization
temperatures (Fig. 1, Table 1). Compared to 8, substance 4b has
Scheme
molecules.
1 Synthesis of the investigated fluorinated mesogenic
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 2942–2953 | 2943

View Article Online
Journal of Materials Chemistry A
Paper
Table
1
Thermal transitions of the investigated mesogenic
compounds (enthalpy in J gꢁ1 in parentheses) detected by DSC from
the second heating scan and the first cooling scan
Phase transition (ꢂC) and corresponding enthalpy changesa
(J gꢁ1
)
2nd heating
4a
4b
4c
8
G
ꢁ38
A
I
Cr
A
ꢁ2 (34.5)
Cr
58 (65.2)
I
Cr
Cr
G
71 (88.8)
18 (10.8)
ꢁ21
41 (28.1)
16 (61.4)
Cr
Cr
67 (70.9)
64 (89.4)
I
I
Cooling
4a
4b
4c
8
I
20 (2.5)
37 (2.6)
49 (2.1)
33 (3.0)
SA
SA
SA
SA
ꢁ47
G
I
I
I
13 (53.9)
16 (63.1)
ꢁ23
Cr
Cr
G
a
Cr, crystalline; SA, smectic A; Iso, isotropic; G, glassy; A, amorphous.
ˆ
a batonnet-like texture followed by the formation of a fan-sha-
ped texture.36
Increasing amounts of LiTFSI broadened the temperature
range in the mesogenic phase, see Fig. 2. The highest temper-
ature range was achieved for complex 4b and 4c with 20 mol%
LiTFSI. Compared to the substances without the Li salt, the
temperature range of the mesophase for 4c increases from
33.3 ꢂC up to 82.8 ꢂC and for 4b from 24.6 ꢂC up to 97.4 ꢂC,
respectively (see the ESI, Fig. S3 and S4, and Tables S2 and S3†).
Up to 10 mol% LiTFSI, the isotropization temperature increased
while the crystalline melting transitions were relatively
constant. This is an indication that the Li salt was incorporated
into the polar part of the substance. The incorporation of the
salt enhances the separation of polar and lipophilic groups, as it
makes the polar fragments more polar.37,38 Further increase of
Li salt concentration leads to the formation of a glassy state, and
Fig. 1 DSC traces (10 K minꢁ1) of 8: (a) first heating scan, (b) second
and subsequent heating scans, (c) first and subsequent cooling scans;
4a: (d) first heating scan, (e) second and subsequent heating scans, (f)
first and subsequent cooling scans; 4b: (g) first heating scan, (h)
second and subsequent heating scans, (i) first and subsequent cooling additionally to a decrease of the crystallization temperature and
scans; 4c: (j) first heating scan, (k) second and subsequent heating
scans, (l) first and subsequent cooling scans. Cr, crystalline; SA, smectic
A; Iso, isotropic; G, glassy, A, amorphous.
enthalpy because additional lithium salt disturbs the formation
of the crystalline phase. This suggests that higher amounts of
a higher isotropization temperature and crystallizes on cooling.
This observation suggests that uorination increases the iso-
tropization temperature and leads to the formation of a crys-
talline structure. The formation of a liquid crystalline phase for
4a, 4b, 4c and 8 was only observed on the cooling scans, hence
the substances have a monotropic mesophase.
Substances 4a, 4b, and 4c were mixed with different amounts
of LiTFSI, a common lithium salt, to analyze the inuence of the
Li salt on the mesomorphic behavior. LiTFSI was chosen as the
Li salt because of its good thermal stability and dissociation
behavior.7 The lithium salt mixtures were obtained by slow
evaporation of the THF solution of the substances and LiTFSI.
All compounds were solids at room temperature. On cooling
from the isotropic melt, the mixtures of 4c rst exhibited a
Fig. 2 Phase transition behavior for the mixtures of compound 4c and
transition into a smectic mesophase, followed by crystallization.
LiTFSI, data recorded from the first cooling scans. Iso, isotropic; SA,
smectic A; Cr, crystalline.
This observation was veried by POM. The pictures rst showed
2944 | J. Mater. Chem. A, 2015, 3, 2942–2953
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
SAXS and WAXS measurements
SAXS and WAXS experiments were carried out to examine the
mesomorphic phase structures of 4a, 4b, 4c and their mixtures
with LiTFSI (Fig. 5 and ESI, Fig. S6 and S7†). Substances 4a, 4b
and 4c and their mixtures with 5 mol% and 10 mol% LiTFSI
exhibited a spontaneously macroscopic smectic alignment
when the samples were placed between two mica (silica) plates
in the measurement cell, heated above the isotropization
temperature and subsequently cooled below the isotropic
temperature. The concentration of the scattering intensity
along the y-axes in the 2D SAXS pattern indicates the aligned
smectic layers along the surface of the mica plates in the
measurement cell (Fig. 4b). For the non-aligned samples, the
diffraction rings are homogeneous (Fig. 4a).35
To show the inuence of the lithium salt concentration on
the molecular packing arrangement of 4a, 4b and 4c, WAXS/
SAXS measurements with different amounts of LiTFSI were
performed. At room temperature, all lithium salt mixtures were
crystalline, which was conrmed by WAXS measurements. In
the WAXS curves the complexes gave sharp reections (see the
ESI, Fig. S8†). From the layer spacing and the length of the
molecule the d/l ratio can be calculated, see Table 2. The d/l
ratio gives an indication about the packing structure of the
molecules. A d/l ratio <1 indicates a strongly interdigitated
monolayer, whereas ratios of 1–2 indicate the formation of an
interdigitated bilayer. The 2D SAXS proles for 4c pure and 4c
with 5 mol% LiTFSI exhibit one strong reection with layer
Fig. 3 POM pictures of the synthesized compounds on cooling from
the isotropic melt: (a) compound 4b at 30 ꢂC; (b) substance 4a at
20 ꢂC; (c) substance 4c at 45 ꢂC; (d) substance 4c mixed with 10 mol%
LiTFSI at 70 ꢂC.
lithium salt are not only incorporated into the ionic fragments
but in all parts of the molecule. Compound 4a and the salt
mixtures exhibited a different thermal behavior. On the cooling
scans no crystallization was observed and crystallization occurs
on the subsequent 2nd heating scan. The range of the mesogenic
phase increased up to 5 mol% LiTFSI. Higher amounts of Li-salt
decrease the isotropization temperature and so the range of the
liquid crystalline phase. At concentrations higher than 20 mol%
LiTFSI no liquid crystalline phase was observed any more. The
lithium salts also emphasize the formation of an enantiotropic
mesophase. Whereas compound 4c without and with 5 mol%
LiTFSI only has a monotropic mesophase, the compounds with
higher amounts of the Li salt exhibit the formation of an
enantiotropic mesophase (see the ESI, Fig. S5†).2,15 This
behavior was not observed for substances 4a and 4b.
˚
˚
spacing of 21.4 A and 21.6 A, respectively, thus the molecules
form an interdigitated monolayer (Table 2). Up to 5 mol%, the
Li salt has nearly no inuence on the layer distances. The
compounds with a higher amount of Li salt exhibit two reec-
tions in the SAXS patterns and have a ratio of 1.5 : 1 in the q
range which is typical for a bilayered lamellar structure. The
formation of a bilayer may be explained by an increased polarity
of the ionic group as the Li salt intercalates preferentially into
the polar parts of the molecule. This leads to micro-phase
segregation between the long alkyl chain and the polar
carbonate group. The mixture of 4c with 10 mol% LiTFSI
showed a different SAXS curve (Fig. 5). The rst reection is
broader and has a lower intensity than the second reection.
Probably no clear bilayer is formed. Compound 4b pure and the
mixtures with 5 and 10 mol% LiTFSI formed an interdigitated
monolayer in a smectic phase. Similar to 4c small amounts of
the lithium salt have nearly no inuence on the packing
arrangement. The same behavior is also observed for substance
4a. In 4a the layers are less interdigitated than in 4b and 4c.
The SAXS/WAXS experiments demonstrate that the
substances 4a, 4b and 4c and their mixtures with LiTFSI spon-
taneously align in the smectic phase when they are cooled from
the isotropic melt. To show the inuence of a magnetic eld on
the orientation, the compounds 4a and 4b mixed with 10 mol%
LiTFSI were placed into a magnetic chamber and heated up to
the isotropic state. Then a magnetic eld with eld strength of
9.4 T was applied and the sample was cooled to 20 ꢂC at a rate of
0.5 K minꢁ1. The magnetic eld was applied perpendicular to
the lm surface. Aer the treatment, the SAXS pattern of 4a
Fig. 4 2D SAXS images of compound 4b mixed with 10 mol% LiTFSI.
Left: not aligned at 40 ꢂC on heating; right: aligned in the SA phase at
40 ꢂC on cooling from the isotropic temperature. The horizontal
stripes of lower intensity are due to the blind regions of the detector.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 2942–2953 | 2945

View Article Online
Journal of Materials Chemistry A
Paper
Table 2 X-ray diffraction analysis and ionic conductivity of compounds 4c, 4b and 4a mixed with various amounts of LiTFSI
Mol%
LiTFSI
T
(ꢂC)
Layer spacing
(A)
Molecular length
(A)
Ratioa d/l
sik at 55 ꢂ
C
Phase
˚
˚
˚
˚
d001 (A)
d002 (A)
4c
0
5
10
20
40
40
45
50
40
40
21.4
21.6
35.9
37.6
39.0
21.4
21.6
35.9
37.6
39.0
25.9
25.9
25.9
25.9
25.9
0.82
0.83
1.38
1.45
1.50
Monomolecular SA
Monomolecular SA
Bimolecular SA
Bimolecular SA
Bimolecular SA
1.9 ꢀ 10ꢁ5
2.0 ꢀ 10ꢁ5
2.1 ꢀ 10ꢁ5
2.0 ꢀ 10ꢁ5
22.1
18.9
19.4
4b
0
5
25
40
40
19.6
19.9
20.1
19.6
19.9
20.1
24.7
24.7
24.7
0.79
0.81
0.81
Monomolecular SA
Monomolecular SA
Monomolecular SA
1.8 ꢀ 10ꢁ5
2.9 ꢀ 10ꢁ5
10
4a
0
10
20
20
18.5
18.4
18.5
18.4
21.0
21.0
0.88
0.88
Monomolecular SA
Monomolecular SA
2.1 ꢀ 10ꢁ6b
a
Ratio of layer spacing, d, to the molecular length, l. b sik at 20 ꢂC.
shows a large number of diffraction spots, which is an indica- temperature was above the isotropic point. Cell B consists of a
tion that the smectic layer plane is now oriented perpendicular pair of indium tin oxide (ITO) electrodes. The thickness
to the surface of the mica plates (see the ESI, Fig. S13†). The between the electrodes was 250 mm, and they were xed with a
smectic layer plane is oriented along the direction of the Teon spacer. The LiTFSI mixtures were placed on the elec-
magnetic eld, which shows that the alignment is inuenced by trodes and heated above the isotropic point and the ionic
external elds. The same behavior was observed for 4b.
conductivity was then measured on cooling to room
temperature.39,40
As shown above (Fig. 4) the substances 4a, 4b and 4c and
their mixtures with LiTFSI align in the smectic phase when they
are cooled from the isotropic melt. This behavior leads to
anisotropic conductivity. In cell A where the current is applied
parallel to the surface, compound 4b mixed with 10 mol%
LiTFSI showed an alignment of the smectic layers parallel to the
surface, when cooled from the isotropic melt. The conductivity
was measured parallel to the lm surface (x- and y-direction)
(sik) (Fig. 6). In cell B the conductivity was measured perpen-
dicular to the lm surface (z-direction, sit). POM pictures of 4b
in cell B showed the formation of a fan-shaped texture (see the
ESI, Fig. S14†) which indicates the formation of polydomains
and thus no macroscopic orientation. The ionic conductivity
(sit) decreased to 1.2 ꢀ 10ꢁ7 S cmꢁ1 at 45 ꢂC on cooling (Fig. 7)
as migration of the cations was hindered on the built
grain boundaries and was 125 times lower than in cell A
(sik ¼ 1.5 ꢀ 10ꢁ5 S cmꢁ1 at 45 ꢂC). The highest conductivity was
achieved for 4a mixed with 10 mol% LiTFSI in cell A. As the
Conductivity measurements
Temperature dependent ionic conductivities were measured for
the LiTFSI mixtures of 4a, 4b, 4c, 6 and 8 by an alternating
current impedance method on cooling with two different cells.
Cell A is an IDA electrode (Interdigitated Array) purchased at
ALS, Japan. With risograph technology, a micro-pattern is
formed on an isolated glass plate to allow the electric eld to be
applied parallel to the substrate surface. To enhance the
alignment of the smectic layer planes in the liquid crystalline
state, the sample was covered with a glass plate when the
ꢂ
smectic transition occurs only at 18 C, the molecules do not
align in the measurement cell till 20 ꢂC and the conductivity in
cell A and cell B is almost the same. Below 20 ꢂC a small decline
of the conductivity is observed in cell B, which indicates an
alignment of the molecules parallel to the surface of the cell,
hence perpendicular to the measured conductivity (Fig. 7).41 The
ionic conductivity for 4a and 4b in cell A is signicantly higher
than the conductivity observed by Kato and co-workers for a
similar molecule with a cyclic carbonate unit which aligned in a
columnar phase.22 This observation suggests that molecules
with a similar structure exhibit in a layered phase a higher
Fig. 5 SAXS curves of compound 4c with various amounts of LiTFSI.
The curves were shifted vertically for clarity.
2946 | J. Mater. Chem. A, 2015, 3, 2942–2953
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
compounds with 10 mol% LiTFSI per carbonate unit. Higher
amounts of Li salt had nearly no inuence on the ionic
conductivity, although they had an inuence on the intermo-
lecular structure, as shown above (Table 2). According to eqn (1)
ionic conductivity depends on the concentration, n+, of the
cationic charge carriers as well as on their mobility, m+.
s ¼ Fn+m+
(1)
where F is the Faraday constant. At higher concentrations the
number of cationic charge carriers increases, but their
mobility is hindered due to the formation of aggregates.42–44
As in solid electrolytes with carbonate units there is no
segmental motion like in electrolytes with PEG chains; the
mobility of the ions only depends on the potential barrier,
Ehop, that the cation must overcome to move from one site to
another. Increasing ion concentration leads to the formation
of ion aggregates which increases this energy barrier.41 The
reduced intensity of the 7Li signal in the NMR spectra also
indicates the formation of molecular aggregates (see ESI,
Fig. S17†). A different behavior is observed for compound 6,
which behaves more like a liquid electrolyte where diffusion
in a liquid is expected to be isotropic. Here, increasing
amounts of LiTFSI increase the ionic conductivity (see ESI,
Fig. S10†).
Fig. 6 Schematic illustration of anisotropic Li+ ion conductivity along
the layer plane in x- and y-direction when the molecules align in an
oriented interdigitated monomolecular SA phase.
As expected the conductivity in the isotropic phase mainly
depends on the length of the alkyl chain. The shorter the alkyl
chain, the lower the molecular mass and the higher is the ionic
conductivity (see ESI, Fig. S9†). Furthermore, uorination of the
aromatic core has a positive effect on the ionic conductivity.
Hence, substance 4a has with 2.7 ꢀ 10ꢁ4 S cmꢁ1 at 100 C the
ꢂ
highest conductivity of all measured compounds.
Diffusion behavior and salt dissociation
The self-diffusion coefficients of both the cation and the
anion species in the LiTFSI mixtures for 4a, 4b and 6 were
measured by the PFG-NMR technique. Aer heating up to
80 ꢂC the diffusion was measured during cooling. Density,
molar mass and molar salt concentration C, which were used
for the calculations, are compiled in Table S4 (see the ESI†).
Because of the lower molecular weight of 4a, the substance
contains a higher molar salt concentration. As described
above, the salt concentration has only a minor effect on the
conductivity.
Table 3 shows the diffusion coefficients, conductivity and the
calculated Haven ratios for the substances 4a, 4b and 6 with 10
mol% LiTFSI. When both the conductivity and the diffusion
data are known, the calculation of the Haven ratio, Limp/LNMR is
a good approach to estimate the dissociation behavior of the Li
salt in the liquid crystalline matrix. The calculation of the Haven
ratio was done according to the approach by Kii and co-
workers.45,46 The molar conductivity Limp was calculated from
the ionic conductivity si obtained from impedance measure-
ments by the following equation
Fig. 7 Arrhenius plots of ionic conductivity for 4a and 4b with 10 mol%
LiTFSI; measured parallel (cell A, sik) and perpendicular (cell B, sit) to
the smectic layer plane of the aligned liquid crystals, respectively.
conductivity than in the columnar phase. Probably, the ion
transport mechanism is hindered in a 1D structure if there is no
perfect macroscopic alignment such that the ion channels are
interrupted and agglomeration between the Li-salt and the
polar coordination sites may be favored. The LiTFSI mixture of
substance 8, the non-uorinated analogue of 4b exhibited a
signicantly lower conductivity. Additionally no alignment
could be observed in the measurement cell; hence the ionic
conductivity is almost the same in both measurement cells (see
the ESI, Fig. S12†). These results indicate that uorination has a
positive effect on both the alignment in the measurement cell
and the ionic conductivity.
Compounds 4b and 4c were mixed with different amounts of
LiTFSI. The best ionic conductivity was achieved for the
si
Limp
¼
(2)
C
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 2942–2953 | 2947

View Article Online
Journal of Materials Chemistry A
Paper
Table 3 Diffusion coefficients of the cation and anion at different temperatures. Ionic conductivity and calculated Haven ratio Limp/LNMR for 4a,
4b and 6 with 10 mol% LiTFSI in the isotropic phase
Compound
T (ꢂC)
DLi (cm2 sꢁ1
)
DF (cm2 sꢁ1
)
si (S cmꢁ1
)
Limp/LNMR
4a/10 mol% LiTFSI
80
55
40
80
55
55
30
1.3 ꢀ 10ꢁ7
2.3 ꢀ 10ꢁ8
7.8 ꢀ 10ꢁ9
1.7 ꢀ 10ꢁ7
1.8 ꢀ 10ꢁ8
4.5 ꢀ 10ꢁ7
7.9 ꢀ 10ꢁ8
1.8 ꢀ 10ꢁ7
4.2 ꢀ 10ꢁ8
9.7 ꢀ 10ꢁ9
2.4 ꢀ 10ꢁ7
3.1 ꢀ 10ꢁ8
7.9 ꢀ 10ꢁ7
1.5 ꢀ 10ꢁ7
1.5 ꢀ 10ꢁ4
0.52
4b/10 mol% LiTFSI
6/10 mol% LiTFSI
1.2 ꢀ 10ꢁ4
1.1 ꢀ 10ꢁ5
0.33
0.01
Using the diffusion coefficients DLi and DF from the NMR diffuse in a correlated fashion, as was shown for Li salt-in-
measurements the molar conductivity LNMR can be calculated polymer electrolytes.49 As the anion is not affected by coordi-
by applying the Nernst–Einstein relationship for univalent nation it exhibited a higher diffusion coefficient at lower
electrolytes.
temperature. This is in accordance with the results for polymer
gel electrolytes based on poly(ethylene glycol).35,41–44
F2
LNMR
¼
ðDLi þ DFÞ
(3)
Most interesting, however, are the Haven ratios, as they
reach quite high values. Cyclic carbonates have a high dielectric
constant; therefore the polar sites on the liquid crystalline
molecules 4a and 4b promoted salt dissociation and the Haven
ratio is signicantly higher than for liquid crystalline molecules
with other polar groups and is higher than for molecule 6 with a
TEO unit (Table 3).50 Though the correlation of Li+ and anion
motion is certainly still present, it is less pronounced as
compared to these materials. Since the Haven ratio depends on
the orientation of the smectic layers, it was only determined in
the isotropic state.
RT
where F is the Faraday constant.47 LNMR calculated from the
diffusion measurements is usually higher than Limp as the value
includes the diffusivity of the dissociated ions as well as that of
the ion pairs.48
As expected, the diffusion coefficients decrease with
decreasing temperature and show for all substances strong
temperature dependence. While this trend is a general conse-
quence of diffusion resulting from thermally activated ion
jumps, in the present case the phase transitions as well as the
alignment of the smectic layers might play a role. The magnetic
eld orientation was parallel to the direction of the magnetic
eld gradients; thus, in the oriented smectic phases, the diffu-
sion is measured parallel to the smectic layer plane. Between 80
ꢂC and 55 ꢂC, the diffusion coefficient of 4b decreases stronger
Conclusions
Liquid crystalline substances with a propylene carbonate unit as
the ionically conductive group were prepared. It was shown that
carbonate units enhance the formation of a liquid crystalline
state. SAXS/WAXS measurements demonstrated that the mole-
cules align spontaneously in the smectic phase when they are
cooled from the isotropic melt. The LiTFSI mixtures of 4a and
4b aligned in the measurement cells and anisotropic conduc-
tivity could be observed. Fluorination of the aromatic ring
signicantly increases the ionic conductivity of the Li-salt
mixtures. Determination of the diffusion coefficients and
calculation of the Haven ratio revealed that the cyclic carbonate
moiety on the molecules enhanced Li-salt dissociation
compared to the molecules containing a TEO unit and the Li
cation was probably transported by a hopping mechanism
through the liquid crystalline matrix. The formation of a
smectic layer enhances ion mobility parallel to the layer plane
which enhances the ionic conductivity. Therefore one of the
best ionic conductivities for liquid crystalline molecules with
cyclic carbonate units could be achieved.
(factors of 9 and 8 for Li and 19F, respectively) than that of 4a
7
(factors of 6 and 4, respectively). Knowing that 4b is in the
smectic phase and 4a is in the isotropic phase at 55 ꢂC (see also
Fig. 7), and assuming an at least partial orientation of the
smectic layers of 4b in the magnetic eld, it would imply that
the net effect of the formation of smectic layers and their
orientation leads to a lower diffusivity of the ions. This seems
surprising, but might be an effect of a strong binding of Li (and
thus also anions by contact ion pairs) to the high concentration
of carbonates in the conductive, polar region. Another reason
might be a not ideal orientation in the eld, such that poly-
domains are formed and migration is hindered by the formed
grain boundaries.
The diffusion coefficient of Li+ is generally lower than that of
the anions. This reveals that Li+ is more affected by the polar
sites of the molecules. Li+ interacts with the oxygen of the
carbonate units via coordination to free electron pairs of oxygen
and tends to hop from site to site through the liquid crystalline
matrix. The ratio of Li and F diffusion coefficients is in the range
between 0.5 and 0.8 without any systematic dependencies. It
reects the fact that the anion is generally faster, but its diffu-
sivity is correlated to that of the cation. Most probably there are
contact ion pairs formed, such that Li+ and the anion partially
Experimental section
Materials
Unless otherwise noted the reagents and solvents were
purchased from Aldrich, ABCR, Alfa Aesar and Gruessing. The
2948 | J. Mater. Chem. A, 2015, 3, 2942–2953
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
dry solvents were taken from a solvent-purication-system MB Lorentzian t from the rst-order Bragg reection with the
SPS-800 from MBraun.
Bragg relationship d ¼ 2p/q0.
Pulsed eld gradient NMR studies
Characterization
1H NMR spectra were measured on an AVIII-300 from Bruker at
300 K and at 300 MHz. Chemical shis were quoted relative to
(CH3)4Si (d ¼ 0.00) as an internal standard or to the residual
protons of the deuterated solvents. The multiplicity was char-
acterized by the following abbreviations: s – singlet, d – doublet,
t – triplet, q – quartet, p – quintet, m – multiplet. 19F NMR
spectra were recorded on a Bruker AV-250 at 300 K and
250 MHz.
Diffusion coefficients of the ionic species in the electrolytes
were measured using the pulsed gradient stimulated echo (PFG)
NMR technique on a Bruker 400 MHz FT-NMR spectrometer
(Bruker, Avance) employing a probe head (Bruker, DIFF 30) with
RF-selective inserts for the respective nuclei and providing
magnetic eld gradient strength of up to 12 T mꢁ1. For the
measurements, the electrolytes were lled in 5 mm NMR tubes,
which were sealed. A stimulated echo pulse sequence was used
for the measurements. Gradient pulses of a duration of d ¼ 1.9–
2.5 ms and an observation time of D ¼ 200 ms for 19F and D ¼
150 ms, d ¼ 2.5 ms for 7Li experiments, respectively, were
applied. The evaluation of the self-diffusion coefficients of
TFSIꢁ and Li+ was carried out by analyzing the area of the peak
IR measurements were conducted using a Bruker Vertex
70 FT/IR with an ATR unit. Electrospray ionization mass spec-
trometry (ESI-MS) was performed on a Varian 500 MS.
Differential scanning calorimetry (DSC)
7
of 19F and Li, respectively. The decay of the signal in depen-
DSC measurements were performed on a Q 2000 from TA
Instruments. Heating and cooling rates were 10 K minꢁ1. The
transition temperatures were taken at the maximum of
exothermic and minimum point of endothermic peaks,
respectively.
dence of the gradient strength g resulted in the diffusion coef-
cient by tting the Stejskal–Tanner equation.51–54 The gradient
pulses were applied in the direction of the magnetic eld, B0.
Synthesis
General procedure for the synthesis of 4-(alkoxy)-2,3,5,6-
tetrauorobenzene (2a, 2b, 2c). In a round bottom ask tted
with a reux condenser, a nitrogen inlet and a magnetic stirrer
were placed 2,3,5,6-tetrauorophenol (50.0 mmol, 1.0 eq.),
anhydrous potassium carbonate (75.0 mmol, 1.5 eq.) and
acetonitrile (80 mL). Under stirring, the reaction mixture was
heated to reux (oil bath, T ¼ 85 ꢂC). Once reux was reached,
bromoalkane (52.5 mmol, 1.1 eq.) was slowly added with a
syringe over 10 min. Then, the reaction mixture was stirred
under reux for 16 h. Aer cooling the mixture to room
temperature, water (200 mL) was added and the aqueous phase
was extracted with petroleum ether (4 ꢀ 100 mL). The organic
layer was washed with NaOH (10%, 200 mL), neutralized with
water and dried over Na2SO4. Evaporation of the solvent with a
rotary vapor led to a clear, pale yellow oil in high yields. The
substance was used without further purication.
Polarized optical microscopy (POM)
POM was done on a Leica MZ8 optical microscope equipped
with a Mettler FP52 hot stage.
Ionic conductivity measurements
Electronic impedance spectroscopy (EIS) was carried out on a
BioLogic Science Instruments VMP3 potentiostat (frequency
range 20 Hz to 500 kHz) and a custom setup temperature
controller. Analysis of the data was done with a BioLogic EC-
Lab. Sample preparation and the measurements were per-
formed in a glove box under an argon atmosphere. For the
conductivity measurements of the liquid crystals, two different
electrodes were used. An IDA electrode (Interdigitated Array)
purchased at ALS, Japan (cell constant: 0.07 cmꢁ1) and a home-
made ITO electrode (cell constant: 0.19 cmꢁ1).
4-(Octyloxy)-2,3,5,6-tetrauorobenzene (2a). Yield 98%. 1H-
3
NMR (300 MHz, CDCl3, 300 K) d (ppm) ¼ 0.89 (t, J ¼ 6.71 Hz,
3H), 1.30 (m, 8H), 1.46 (p, 3J ¼ 6.96 Hz, 2H), 1.78 (p, 3J ¼ 6.65 Hz,
Small- and wide-angle X-ray scattering (SAXS/WAXS)
SAXS/WAXS measurements were carried out using a Ganesha 2H), 4.22 (t, 3J ¼ 6.57 Hz, 2H), 6.75 (m, 1H). 13C-NMR (63 MHz,
300XL SAXS–WAXS system (SAXSLAB ApS, Copenhagen/Den- CDCl3, 300 K) d (ppm) ¼ 13.77 (1C), 22.39 (1C), 25.28 (1C), 26.67
mark). The X-ray source was a GENIX 3D microfocus X-ray (1C), 28.64 (1C), 29.64 (1C), 31.53 (1C), 75.11 (1C), 98.88 (1C),
source which was operated at 50 kV/0.6 mA with a Cu anode (Ka, 138.34–138.71 (1C), 139.59–139.88 (1C), 142.85–143.09 (1C),
l ¼ 0.1542 nm). A movable two-dimensional (2D) Pilatus 300K 144.85 (1C), 148.17 (1C). 19F-NMR (235 MHz, CDCl3, 300 K) d
detector was used. All measurements were performed in a fully (ppm) ¼ ꢁ157.30 (m, 1F), ꢁ157.24 (m, 1F), ꢁ140.46 (m, 1F),
evacuated chamber. To measure the structural changes, the ꢁ140.41 (m, 1F).
1
samples were placed between two mica sheets (5–7 mm thick-
4-(Decyloxy)-2,3,5,6-tetrauorobenzene (2b). Yield 97%. H-
3
ness). They were heated above the isotropization temperatures NMR (300 MHz, CDCl3, 300 K) d (ppm) ¼ 0.88 (t, J ¼ 6.71 Hz,
and then cooled down with a rate of 10 K minꢁ1. Sample- 3H), 1.27 (m, 12H), 1.46 (p, 3J ¼ 6.96 Hz, 2H), 1.79 (p, 3J ¼ 6.65
detector-distances (SDD) of 101 mm (WAXS) and 401ꢁm1 m (SAXS) Hz, 2H), 4.22 (t, J ¼ 6.57 Hz, 2H), 6.75 (m, 1H). 13C-NMR (63
3
˚
were used, resulting in a q-range of 0.01 to 2.5 A (q is the MHz, CDCl3, 300 K) d (ppm) ¼ 14.23 (1C), 22.87 (1C), 25.71 (1C),
momentum transfer). The 2D images obtained were azimuth- 29.44 (1C), 29.51 (1C), 29.72 (2C), 30.06 (1C), 32.09 (1C), 75.41–
ally averaged. The layer spacing was determined from a 75.52 (1C), 99.39 (1C), 138.27–138.70 (1C), 139.52–139.82 (1C),
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 2942–2953 | 2949

View Article Online
Journal of Materials Chemistry A
Paper
142.79–143.16 (1C), 144.84 (1C), 148.12 (1C). 19F-NMR (235 148.51–148.75 (1C), 165.60 (1C). 19F-NMR (235 MHz, CDCl3, 300
MHz, CDCl3, 300 K) ¼ ꢁ157.33 (m, 1F), ꢁ157.20 (m, 1F), K) d (ppm) ¼ ꢁ156.68 (m, 1F), ꢁ156.53 (m, 1F), ꢁ138.30 (m, 1F),
ꢁ140.50 (m, 1F), ꢁ140.37 (m, 1F).
ꢁ138.13 (m, 1F).
4-(Dodecyloxy)-2,3,5,6-tetrauorobenzene (2c). Yield 98%.
General procedure for the synthesis of (2-oxo-1,3-dioxolan-4-
1H-NMR (300 MHz, CDCl3, 300 K) d (ppm) ¼ 0.88 (t, 3J ¼ 6.71 Hz, yl)methyl 4-(alkoxy)-2,3,5,6-tetrauoro-benzoate (4a, 4b, 4c). To
3H), 1.26 (m, 16H), 1.44 (p, 3J ¼ 6.96 Hz, 2H), 1.77 (p, 3J ¼ 6.65 a solution of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide
3
Hz, 2H), 4.21 (t, J ¼ 6.57 Hz, 2H), 6.75 (m, 1H). 13C-NMR (63 hydrochloride (EDC) (24.2 mmol, 1.1 eq.) and 4-dimethylami-
MHz, CDCl3, 300 K) d (ppm) ¼ 14.27 (1C), 22.85 (1C), 25.69 (1C), nopyridine (DMAP) (24.2 mmol, 1.1 eq.) in dry DCM (140 mL) 4-
29.40 (1C), 29.51 (1C), 29.68 (1C), 29.72 (1C), 29.79 (2C), 30.04 (alkyloxy)-2,3,5,6-tetrauorobenzoic acid (3) (22.0 mmol, 1.0
(1C), 32.08 (1C), 75.49 (1C), 99.32 (1C), 138.34–138.71 (1C), eq.) was added under cooling with ice/water and stirred for one
139.60–139.81 (1C), 142.86–143.06 (1C), 144.81 (1C), 148.08 hour. Then, glycerol 1,2-carbonate (44.0 mmol, 2.0 eq.) was
(1C). 19F-NMR (235 MHz, CDCl3, 300 K) d (ppm) ¼ ꢁ157.16 (m, added, and the mixture was stirred for 2 days at room temper-
1F), ꢁ157.11 (m, 1F), ꢁ140.27 (m, 1F), ꢁ140.32 (m, 1F).
ature. The reaction mixture was then poured into a saturated
General procedure for the synthesis of 4-(alkoxy)-2,3,5,6- NH4Cl aqueous solution, the organic phase was separated and
tetrauorobenzoic acid (3a, 3b, 3c). Compound 2 (67.4 mmol, the aqueous phase was extracted with CH3Cl. The organic phase
1.0 eq.) was dissolved in dry diethyl ether (200 mL) in a dried was pre-dried with brine and dried over Na2SO4. The solvent was
and argon lled three necked ask and cooled to ꢁ70 ꢂC in an evaporated in vacuo and the crude product puried by column
acetone/dry ice bath. Then, n-butyl lithium (2.5 M in hexane) chromatography (eluent: DCM–MeOH ¼ 95/5) to obtain the
(67.4 mmol, 1.0 eq.) was added, and the reaction mixture was product as a viscous liquid which started to crystallize aer
stirred for 3 h at this temperature. Dried CO2 was then bubbled several minutes to a white solid.
through the mixture for about 30 min at ꢁ70 ꢂC. Under
(2-Oxo-1,3-dioxolan-4-yl)methyl 4-(octyloxy)-2,3,5,6-tetra-u-
continuous CO2 bubbling, the mixture was allowed to warm to orobenzoate (4a). Yield 55%. 1H-NMR (300 MHz, CDCl3, 300 K)
room temperature. The solvent was removed; the crude product d (ppm) ¼ 0.88 (t, 3J ¼ 6.69 Hz, 3H), 1.30 (m, 8H), 1.45 (p, J ¼
3
was dissolved in water and acidied with HCl (1 N) to bring the 6.86 Hz, 2H), 1.79 (p, 3J ¼ 6.59 Hz, 2H), 4.35 (t, 3J ¼ 6.54 Hz, 2H),
solution to pH z 2. The mixture was extracted with diethyl 4.40–4.51 (m, 2H), 4.60–4.71 (m, 2H), 5.01 (m, 1H). 13C-NMR (75
ether, pre-dried with brine and dried over Na2SO4. Evaporation MHz, CDCl3, 295 K) d (ppm) ¼ 14.14 (1C), 22.70 (1C), 25.50 (1C),
of the solvent with a rotary vapor led to a white-yellow solid. The 29.21 (2C), 29.92 (1C), 31.81 (1C), 62.23 (1C), 65.91 (1C), 73.43
crude product was puried by recrystallization in cyclohexane (1C), 75.50 (1C), 103.59 (1C), 139.69–139.88 (1C), 141.49–141.86
yielding the nal product as a white solid.
(2C), 145.21–145.40 (1C), 147.26–147.42 (1C), 154.44 (1C),
4-(Octyloxy)-2,3,5,6-tetrauorobenzoic acid (3a). Yield 57%. 159.41 (1C). 19F-NMR (235 MHz, CDCl3, 300 K) d (ppm) ¼
1H-NMR (300 MHz, CDCl3, 300 K) d (ppm) ¼ 0.89 (t, 3J ¼ 6.74 Hz, ꢁ156.22 (m, 1F), ꢁ156.16 (m, 1F), ꢁ139.23 (m, 1F), ꢁ139.17 (m,
3H), 1.31 (m, 8H), 1.43 (p, 3J ¼ 6.84 Hz, 2H), 1.80 (p, 3J ¼ 6.61 Hz, 1F). IR (ATR): n ¼ 2919, 2853, 1800, 1780, 1731, 1645, 1505, 1486,
2H), 4.37 (t, 3J ¼ 6.54 Hz, 2H), 10.61 (s, 1H). 13C-NMR (63 MHz, 1405, 1384, 1314, 1211, 1183, 1091, 1051, 1006, 850, 776, 713.
CDCl3, 300 K) d (ppm) ¼ 14.24 (1C), 22.79 (1C), 25.59 (1C), 29.30 ESI-MS (m/z): [M+] calculated for C19H22F4O6, 422.4; found,
(2C), 30.01 (1C), 31.90 (1C), 75.52 (1C), 103.61 (1C), 139.07– 445.4 ([M + Na+]). EA (%) calculated: C 54.03, H 5.25, F 17.99, O
139.33 (1C), 141.91 (1C), 142.31–142.62 (1C), 144.95–145.22 22.73. Found: C 54.05, H 5.52, F 17.70.
(1C), 148.43–148.84 (1C), 165.29 (1C). 19F-NMR (235 MHz,
(2-Oxo-1,3-dioxolan-4-yl)methyl 4-(decyloxy)-2,3,5,6-tetra-u-
CDCl3, 300 K) d (ppm) ¼ ꢁ156.67 (m, 1F), ꢁ156.62 (m, 1F), orobenzoate (4b). Yield 58%. 1H-NMR (300 MHz, CDCl3, 300 K)
ꢁ138.23 (m, 1F), ꢁ138.17 (m, 1F).
d (ppm) ¼ 0.86 (t, 3J ¼ 6.67 Hz, 3H), 1.26 (m, 12H), 1.44 (p, 3J ¼
4-(Decyloxy)-2,3,5,6-tetrauorobenzoic acid (3b). Yield 88%. 6.86 Hz, 2H), 1.78 (p, 3J ¼ 6.60 Hz, 2H), 4.34 (t, 3J ¼ 6.54 Hz, 2H),
1H-NMR (300 MHz, CDCl3, 300 K) d (ppm) ¼ 0.88 (t, 3J ¼ 6.74 Hz, 4.40–4.47 (m, 2H), 4.59–4.70 (m, 2H), 5.03 (m, 1H). 13C-NMR (75
3H), 1.27 (m, 12H), 1.46 (p, 3J ¼ 6.84 Hz), 1.80 (p, 3J ¼ 6.65 Hz), MHz, CDCl3, 295 K) d (ppm) ¼ 14.25 (1C), 22.81 (1C), 25.56 (1C),
4.37 (t, 3J ¼ 6.44 Hz, 2H), 9.56 (s, 1H). 13C-NMR (63 MHz, CDCl3, 29.30 (1C), 29.42 (1C), 29.62 (2C), 29.88 (1C), 32.00 (1C), 64.21
300 K) d (ppm) ¼ 14.12 (1C), 22.70 (1C), 25.45 (1C), 29.20 (1C), (1C), 65.91 (1C), 73.31 (1C), 75.52 (1C), 103.59 (1C), 139.71–
29.32 (1C), 29.57 (2C), 29.87 (1C), 31.90 (1C), 75.38 (1C), 103.49 139.86 (1C), 141.52–141.87 (2C), 145.23–145.42 (1C), 147.28–
(1C), 138.88–139.14 (1C), 141.77 (1C), 142.22–142.48 (1C), 147.47 (1C), 154.41 (1C), 159.43 (1C). 19F-NMR (235 MHz, CDCl3,
144.86–145.17 (1C), 148.30–148.86 (1C), 165.80 (1C). 19F-NMR 300 K) d (ppm) ¼ ꢁ156.17 (m, 1F), ꢁ156.12 (m, 1F), ꢁ139.19 (m,
(235 MHz, CDCl3, 300 K) d (ppm) ¼ ꢁ156.69 (m, 1F), ꢁ156.64 1F), ꢁ139.13 (m, 1F). IR (ATR): n ¼ 2919, 2852, 1800, 1780, 1731,
(m, 1F), ꢁ138.24 (m, 1F), ꢁ138.19 (m, 1F).
1644, 1504, 1486, 1405, 1384, 1314, 1211, 1183, 1091, 1050,
4-(Dodecyloxy)-2,3,5,6-tetrauorobenzoic acid (3c). Yield 1004, 850, 776, 715. ESI-MS (m/z): [M+] calculated for
1
3
95%. H-NMR (300 MHz, CDCl3, 300 K) d (ppm) ¼ 0.88 (t, J ¼
C
21H26F4O6, 450.4; found, 473.4 ([M + Na+]). EA (%) calculated:
6.74 Hz, 3H), 1.26 (m, 16H), 1.46 (p, 3J ¼ 6.84 Hz, 2H), 1.80 (p, 3J C 56.00, H 5.82, F 16.87, O 21.31. Found: C 56.06, H 5.97, F
¼ 6.64 Hz, 2H), 4.37 (t, 3J ¼ 6.54 Hz, 2H), 11.42 (s, 1H). 13C-NMR 16.80.
(63 MHz, CDCl3, 300 K) d (ppm) ¼ 14.26 (1C), 22.85 (1C), 25.61
(2-Oxo-1,3-dioxolan-4-yl)methyl 4-(dodecyloxy)-2,3,5,6-tetra-
(1C), 29.34 (1C), 29.50 (1C), 29.64 (1C), 29.70 (1C), 29.78 (2C), uorobenzoate (4c). Yield 63%. 1H-NMR (300 MHz, CDCl3, 300
30.03 (1C), 32.07 (1C), 75.55 (1C), 103.64 (1C), 139.05–139.36 K) d (ppm) ¼ 0.87 (t, 3J ¼ 6.69 Hz, 3H), 1.25 (m, 16H), 1.44 (p, 3J
(1C), 141.90 (1C), 142.39–142.60 (1C), 145.04–145.30 (1C), ¼ 6.86 Hz, 2H), 1.78 (p, 3J ¼ 6.59 Hz, 2H), 4.35 (t, 3J ¼ 6.54 Hz,
2950 | J. Mater. Chem. A, 2015, 3, 2942–2953
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
2H), 4.41–4.70 (m, 4H), 5.05 (m, 1H). 13C-NMR (75 MHz, CDCl3,
295 K) d (ppm) ¼ 14.23 (1C), 22.81 (1C), 25.56 (1C), 29.30 (1C),
29.46 (1C), 29.60 (1C), 29.66 (1C), 29.74 (2C), 29.97 (1C), 32.03
(1C), 64.22 (1C), 65.92 (1C), 73.35 (1C), 75.54 (1C), 103.60 (1C),
139.71–139.86 (1C), 141.52–141.87 (2C), 145.25–145.44 (1C),
147.30–147.49 (1C), 154.41 (1C), 159.46 (1C). 19F-NMR (235
MHz, CDCl3, 300 K) d (ppm) ¼ ꢁ156.22 (m, 1F), ꢁ156.17 (m, 1F),
ꢁ139.23 (m, 1F), ꢁ139.18 (m, 1F). IR (ATR): n ¼ 2919, 2853,
1799, 1780, 1731, 1644, 1504, 1485, 1405, 1384, 1314, 1211,
1182, 1091, 1050, 1005, 852, 758, 714, 611. ESI-MS (m/z): [M+]
calculated for C23H30F4O6, 478.4; found, 501.4 ([M + Na+]). EA
(%) calculated: C 57.70, H 6.30, F 15.88, O 20.06. Found: C 58.17,
H 6.34, F 15.86.
Synthesis of 2-{2-[2-(2-Methoxyethoxy)ethoxy]ethoxy}ethyl-4-
(dodecyloxy)-2,3,5,6-tetrauorophenyl (6).
A
solution of
compound 5 (4.4 g, 12.2 mmol, 1.0 eq.) and Cs2CO3 (12.0 g, 26.7
mmol, 3.0 eq.) in dry THF (70 mL) was stirred at RT for 1 h
before the addition of 2-{2-[2-(2-methoxyethoxy)ethoxy]ethoxy}
ethyl p-tosylate (4.7 g, 12.8 mmol, 1.1 eq.). Aer 16 h stirring at
75 ꢂC, the reaction mixture was cooled to room temperature,
and the white precipitate was ltered off. The precipitate was
washed several times with THF and EtOAc. The organic extracts
were combined, washed with brine and dried over Na2SO4. The
solvent was evaporated and the crude product puried by
column chromatography (eluent: hexane–ethyl acetate ¼ 55/45)
to give compound 6 in 4.3 g (65%) yield as a pale yellow liquid.
1H-NMR (300 MHz, CDCl3, 300 K) d (ppm) ¼ 0.86 (t, 3J ¼ 6.5 Hz,
3H), 1.24 (m, 16H), 1.43 (m, 2H), 1.74 (p, 3J ¼ 6.7 Hz, 2H), 3.36
(s, 3H), 3.54 (m, 2H), 3.64 (m, 14H), 4.19 (t, 3J ¼ 6.8 Hz, 2H), 4.61
(t, 4J ¼ 1.7 Hz, 2H). 13C-NMR (75 MHz, CDCl3, 295 K) d (ppm) ¼
14.19 (1C), 22.78 (1C), 25.62 (1C), 29.33 (1C), 29.44 (1C), 29.61
(1C), 29.65 (1C), 29.73 (2C), 29.97 (1C), 32.01 (1C), 59.09 (1C),
60.17 (1C), 70.07 (1C), 70.54 (1C), 70.61 (1C), 70.69 (2C), 70.71
(1C), 70.77 (1C), 72.04 (1C), 75.43 (1C), 109.43 (1C), 138.06 (1C),
139.62–139.88 (1C), 142.35–142.56 (1C), 144.57 (1C), 147.30
(1C). 19F-NMR (235 MHz, CDCl3, 300 K) d (ppm) ¼ ꢁ144.91 (1F),
ꢁ144.96 (1F), ꢁ157.50 (1F), ꢁ157.55 (1F). IR (ATR): n ¼ 2923,
2854, 1653, 1493, 1468, 1386, 1353, 1294, 1248, 1199, 1130,
1103, 1048, 998, 935, 881, 852, 761, 722, 633. ESI-MS (m/z): [M+]
calculated for C28H46F4O6, 554.7; found, 577.7 ([M + Na+]). EA
(%) calculated: C 60.60, H 8.40. Found: C 60.24, H 8.56.
Synthesis of 3-(dodecyloxy)-2,3,5,6-tetrauoromethanol (5).
To a solution of 4-(dodecyloxy)-2,3,5,6-tetrauorobenzoic acid
(3c) (15.0 g, 39.6 mmol, 1.0 eq.) in dry THF (100 mL), BH3–THF
solution (1 M) (139 mL, 139.0 mmol, 3.5 eq.) was slowly added
ꢂ
with a dropping funnel at ꢁ15 C. Then, the reaction mixture
was allowed to warm to room temperature and stirred for 1
hour, followed by 2 hours at 50 ꢂC. To complete the reaction, the
mixture was cooled to 0 ꢂC and diluted HCl was added, and the
mixture was stirred for 1 hour. The organic phase was sepa-
rated, and the aqueous phase was extracted three times with
diethyl ether. The combined organic phases were pre-dried with
brine and dried over Na2SO4. The solvent was evaporated and
the crude product puried by column chromatography (eluent:
hexane–EtOAc ¼ 95/5) to obtain the product as a viscous liquid
in 13.1 g (91%) yield. 1H-NMR (300 MHz, CDCl3, 300 K) d (ppm)
¼ 0.88 (t, 3J ¼ 7.00 Hz, 3H), 1.26 (m, 16H), 1.45 (m, 2H), 1.76 (p,
3J ¼ 6.7 Hz, 2H), 4.21 (t, 3J ¼ 6.7 Hz, 2H), 4.78 (t, 3J ¼ 4.8 Hz, 2H).
13C-NMR (75 MHz, CDCl3, 295 K) d (ppm) ¼ 14.15 (1C), 22.71
(1C), 25.52 (1C), 29.25 (1C), 29.37 (1C), 29.53 (1C), 29.58 (1C),
29.65 (2C), 29.86 (1C), 31.93 (1C), 55.76 (1C), 75.34 (1C), 111.92
(1C), 137.79 (1C), 139.59 (1C), 142.81 (1C), 143.93 (1C), 147.15
(1C). 19F-NMR (235 MHz, CDCl3, 300 K) d (ppm) ¼ ꢁ146.49 (m,
1F), ꢁ146.55 (m, 1F), ꢁ157.21 (m, 1F), ꢁ157.58 (m, 1F). EA (%)
calculated: C 62.62, H 7.74, F 20.85, O 8.78. Found: C 62.73, H
7.70, F 20.70.
Synthesis of 4-decyloxybenzoic acid (7). 4-Hydroxybenzoic
acid (5.5 g, 40.0 mmol, 1.0 eq.) was dissolved in ethanol solution
(100 mL) of KOH (6.7 g, 120 mmol, 3.0 eq.). To this solution 1-
bromodecane (10.0 g, 40.0 mmol, 1.0 eq.) was added drop by
drop with a syringe under constant stirring. The reaction
mixture was reuxed over 16 h. Then conc. HCl was added to
hydrolyze the obtained potassium salt (pH ¼ 2). A white
precipitate was obtained. The precipitate was dissolved in hot
toluene. The insoluble fraction was ltered off and the solvent
was evaporated under reduced pressure from the remaining
product solution yielding the crude product as a white solid.
The crude product was recrystallized to give compound 7 in 8.5
g (69%) yield as a white crystalline solid.
Synthesis of 2-{2-[2-(2-methoxyethoxy)ethoxy] ethoxy}ethyl p-
tosylate. An aqueous solution of NaOH (w ¼ 0.5, 1.5 mL) was
slowly added to tetra(ethylene glycol) monomethyl ether (4.2 g,
20.0 mmol, 1.0 eq.) with vigorous stirring at 0 ꢂC. Aer stirring
for 30 min, a solution of p-toluenesulfonyl chloride (TsCl) (4.6 g,
24.0 mmol, 1.2 eq.) dissolved in THF (20 mL) was slowly added
1H-NMR (300 MHz, CDCl3, 300 K) d (ppm) ¼ 0.87 (t, 3J ¼ 6.73
3
3
Hz, 3H), 1.26 (s, 12H), 1.45 (p, J ¼ 6.82 Hz), 1.79 (p, J ¼ 6.60
Hz), 4.00 (t, 3J ¼ 6.55 Hz, 2H), 6.94 (d, 3J ¼ 8.89 Hz, 2H), 8.07 (d,
3J ¼ 8.85, 2H), 9.56 (s, 1H). 13C-NMR (63 MHz, CDCl3, 300 K) d
(ppm) ¼ 13.86 (1C), 25.69 (1C), 28.80 (1C), 29.04 (2C), 29.08 (1C),
29.27 (2C), 31.62 (1C), 68.10 (1C), 113.89 (2C), 121.05 (1C),
132.04 (2C), 163.38 (1C), 171.58 (1C). EA (%) calculated: C 73.35,
H 9.41. Found: C 73.16, H 9.46.
ꢂ
to the reaction mixture and the solution was stirred at 0 C for
another 3 h. Then the solution was poured into water and
extracted with CHCl3. The combined organic solution was
washed with water and brine. Aer drying with MgSO4, the
solvent was removed in vacuo. The crude product was puried
by column chromatography (eluent: ethyl acetate) to obtain 2-
{2-[2-(2-methoxyethoxy) ethoxy]ethoxy}ethyl p-tosylate in 4.9 g
(68%) yield as a colorless oily liquid. 1H-NMR (CDCl3, 300 MHz)
Synthesis of (2-Oxo-1,3-dioxolan-4-yl)methyl 4-decylox-
ybenzoate (8). To a solution of 1-ethyl-3-(3-dimethylamino-
propyl) carbodiimide hydrochloride (EDC) (30.0 mmol, 6.0 eq.)
and 4-dimethylaminopyridine (DMAP) (5.5 mmol, 1.1 eq.) in dry
DCM (120 mL) 4-decyloxybenzoic acid (7) (5.0 mmol, 1.0 eq.)
was added under cooling with ice/water and stirred for one
hour. Then, glycerol 1,2-carbonate (30.0 mmol, 6.0 eq.) was
3
d (ppm) ¼ 2.43 (s, 3H), 3.36 (s, 3H), 3.60 (m, 14H), 4.14 (t, J ¼
3
4.9 Hz, 2H), 7.33 (d, J ¼ 7.5 Hz, 2H), 7.78 (d, 3J ¼ 7.5 Hz, 2H).
13C-NMR (75 MHz, CDCl3, 300 K) d (ppm) ¼ 21.75 (1C), 59.13
(1C), 68.74 (1C), 69.35 (1C), 70.60 (2C), 70.67 (2C), 70.81 (1C),
71.99 (1C), 128.07 (2C), 129.91 (2C), 133.01 (1C), 144.89 (1C).
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 2942–2953 | 2951

View Article Online
Journal of Materials Chemistry A
Paper
added, and the mixture was stirred for 2 days at room temper- 17 K. Kishimoto, M. Yoshio, T. Mukai, M. Yoshizawa, H. Ohno
ature. The reaction mixture was then poured into a saturated and T. Kato, J. Am. Chem. Soc., 2003, 125, 3196.
NH4Cl aqueous solution, the organic phase was separated and 18 Y. Akachi, K. Kawamura and N. Koide, Polym. Adv. Technol.,
the aqueous phase was extracted with CH3Cl. The organic phase 2001, 12, 422.
was pre-dried with brine and dried over Na2SO4. The solvent was 19 T. Kato, Chem. Mater., 2000, 12, 782.
evaporated in vacuo and the crude product puried by column 20 K. Kishimoto, T. Suzawa, T. Yokota, T. Mukai and T. Kato, J.
chromatography (eluent: DCM–MeOH ¼ 95/5) to obtain the
Am. Chem. Soc., 2005, 127, 15618.
product as a white solid in 0.8 g (42%) yield. 1H-NMR (300 MHz, 21 K. Hoshina, K. Kanie, T. Ohtake, T. Mukai, M. Yoshizawa,
CDCl3, 300 K) d (ppm) ¼ 0.87 (t, 3J ¼ 6.95 Hz, 3H), 1.26 (m, 12H),
S. Ujiie, H. Ohno and T. Kato, Macromol. Chem. Phys.,
2002, 203, 1547.
1.45 (p, 3J ¼ 7.46 Hz, 2H), 1.79 (p, 3J ¼ 6.62 Hz, 2H), 4.00 (t, 3J ¼
6.57 Hz, 2H), 4.20 (dd, 2J ¼ 8.74 Hz, 3J ¼ 5.62 Hz, 1H), 4.49 (dd, 2J 22 H. Shimura, M. Yoshio, A. Hamasaki, T. Mukai, H. Ohno and
¼ 12.60 Hz, 3J ¼ 3.83 Hz, 1H), 4.53 (dd, 2J ¼ 12.60 Hz, 3J ¼ 3.10
T. Kato, Adv. Mater., 2009, 21, 1591.
3
3
Hz, 1H), 4.63 (t, J ¼ 8.59 Hz, 1H), 5.04 (m, 1H), 6.91 (d, J ¼ 23 T. Ohtake, K. Ito, N. Nishina, H. Kihara, H. Ohno and
8.93, 2H), 7.95 (d, 3J ¼ 8.91, 2H). 13C-NMR (75 MHz, CDCl3, 295
T. Kato, Polym. J., 1999, 11–2, 1155.
K) d (ppm) ¼ 14.25 (1C), 22.79 (1C), 26.06 (1C), 29.16 (1C), 29.42 24 N. A. A. Rossi and R. West, Polym. Int., 2009, 58, 267.
(1C), 29.46 (1C), 29.66 (2C), 32.00 (1C), 63.45 (1C), 66.25 (1C), 25 I. Ichikawa, M. Yoshio, A. Hamasaki, T. Mukai, H. Ohno and
68.41 (1C), 74.16 (1C), 114.44 (2C), 120.80 (1C), 131.99 (2C),
T. Kato, J. Am. Chem. Soc., 2007, 129, 10662.
¨
154.69 (1C), 163.69 (1C), 165.82 (1C). IR (ATR): n ¼ 2953, 2918, 26 D. W. Bruce, P. Metrangolo, F. Meyer, T. Pilati, C. Prasang,
2851, 1781, 1714, 1605, 1511, 1399, 1282, 1254, 1162, 1047, 780.
EA (%) calculated: C 66.65, H 8.00. Found: C 65.84, H 7.98.
G. Rednati, G. Terraneo, S. G. Wainwright and
A. C. Whitwood, Chem.–Eur. J., 2010, 16, 9511.
27 G. Sandford, Tetrahedron, 2003, 59, 437.
28 Z. Luo, Q. Zhang, Y. Oderaotoshi and D. P. Curran, Science,
2001, 291, 1766.
29 Y. Sasada, H. Monobe, Y. Ueda and Y. Shimizu, Mol. Cryst.
Liq. Cryst., 2009, 509, 186.
30 M. P. Kra and J. G. Riess, Chem. Rev., 2009, 109, 1714.
31 X. Cheng, M. Kumar Das, S. Diele and C. Tschierske,
Langmuir, 2002, 18, 6521.
Acknowledgements
¨
Funding from Consortium fur elektrochemische Industrie der
Wacker Chemie AG is gratefully acknowledged. We thank Prof.
Gasteiger for the opportunity to measure conductivity at his
chair.
32 R. Nandi, H. K. Singh, S. K. Singh, B. Singh and R. K. Singh,
Spectrochim. Acta, Part A, 2014, 128, 248.
33 http://www.dow.com/ethyleneglycol/about/properties.htm,
accessed Nov 12, 2014.
Notes and references
1 R. L. Kerr, S. A. Miller, R. K. Shoemaker, B. J. Elliott and
D. L. Gin, J. Am. Chem. Soc., 2009, 131, 15972.
34 Y. Chernyak, J. Chem. Eng. Data, 2006, 51, 416.
2 T. Ichikawa, M. Yoshio, A. Hamasaki, J. Kagimoto, H. Ohno 35 L. M. Blinov, in Structure and Properties of Liquid Crystals,
and T. Kato, J. Am. Chem. Soc., 2011, 133, 2163.
Springer, 2011.
3 K. Hoshino, K. Kanie, T. Ohtake, T. Mukai, M. Yoshizawa, 36 I. Dierking, in Texture of Liquid Crystals, Wiley-VCH,
S. Ujiie, H. Ohno and T. Kato, Macromol. Chem. Phys.,
2002, 203, 1547.
4 B. Scrosati, Electrochim. Acta, 2000, 45, 2461.
5 M. Armand and J.-M. Tarascon, Nature, 2008, 451, 652.
6 J. B. Goodenough and Y. Kim, Chem. Mater., 2010, 22, 587.
7 K. Xu, Chem. Rev., 2004, 104, 4303.
8 M. Wakihara, Mater. Sci. Eng., 2001, 109.
9 J.-P. Gabano, in Lithium Batteries, Academic Press, New York,
1983, p. 17.
10 Z. Zhu, A.-G. Einset, C.-Y. Yang, W.-X. Chen and G. E. Wnek,
Macromolecules, 1994, 27, 4076.
Weinheim, 2003.
37 M. Yoshio, T. Mukai, H. Ohno and T. Kato, J. Am. Chem. Soc.,
2004, 126, 994.
38 M. Lee and B.-K. Cho, Chem. Mater., 1998, 10, 1894.
39 http://www.als-japan.com, accessed June 6, 2014.
40 K. Ito-Akita, N. Nishina, Y. Asai, H. Ohno, T. Ohtake,
Y. Takamitsu and T. Kato, Polym. Adv. Technol., 2000, 11,
529.
41 S. Liang, U. H. Choi, W. Liu, J. Runt and R. H. Colby, Chem.
Mater., 2012, 24, 2316.
42 M. Duclot, F. Alloin, O. Brylev, J. Y. Sanchez and
J. L. Souquet, Solid State Ionics, 2000, 136–137, 1153.
43 D. Fragiadakis, S. Dou, R. H. Colby and J. Runt, J. Chem.
Phys., 2009, 130, 064907.
44 C. T. Imrie, M. D. Ingram and G. S. McHattie, J. Phys. Chem.
B, 1999, 103, 4132.
45 H. Kataoka, Y. Saito, Y. Uetani, S. Murata and K. Kii, J. Phys.
Chem. B, 2002, 106, 12084.
11 T. Kato, Science, 2002, 295, 2414.
12 T. Kato, Angew. Chem., Int. Ed., 2010, 49, 7847.
13 T. Kato, N. Mizoshita and K. Kishimoto, Angew. Chem., 2006,
118, 44.
14 V. Percec and G. Ungar, Adv. Mater. Opt. Electron., 1994, 4,
303.
15 T. Ohtake, M. Ogasawara, K. Ito-Akita, N. Nishina, S. Ujiie,
H. Ohno and T. Kato, Chem. Mater., 2000, 12, 782.
46 G. E. Murch, Solid State Ionics, 1982, 7, 177.
16 Y. Zheng, J. Lui, G. Ungar and P. V. Wright, Chem. Rec., 2004, 47 I. M. Ward, N. Boden, J. Cruickshank and S. A. Leng,
176.
Electrochim. Acta, 1995, 40, 2071.
2952 | J. Mater. Chem. A, 2015, 3, 2942–2953
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
48 I. Nicotera, C. Oliviero, W. A. Henderson, G. B. Appetecchi 51 J. E. Tanner, J. Chem. Phys., 1970, 52, 2523.
and S. Passerini, J. Phys. Chem. B, 2005, 109, 22814. 52 P. Stilbs, Prog. Nucl. Magn. Reson. Spectrosc., 1987, 19, 1.
¨
¨
¨
49 M. Kunze, Y. Karatas, H.-D. Wiemhofer, H. Eckert and 53 J. Kosters, M. Schonhoff and A. N. Stolwijk, J. Phys. Chem. B,
¨
M. Schonhoff, Phys. Chem. Chem. Phys., 2010, 12, 6844.
2013, 117, 2527.
¨
50 Y. Saito, K. Hirai, S. Murata, Y. Kishii, K. Kii, M. Yoshio and 54 M. Joost, M. Kunze, S. Jeong, M. Schonhoff, M. Winter and
T. Kato, J. Phys. Chem. B, 2005, 109, 11563.
S. Passerini, Electrochim. Acta, 2012, 86, 330.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 2942–2953 | 2953