The effect of carborane, bicyclo[2.2.2]octane and benzene on mesogenic and dielectric properties of laterally fluorinated three-ring mesogens

Article abstract of DOI:10.1039/b600068a

Six series of structurally similar compounds containing 12- and 10-vertex p-carborane (A and B), bicyclo[2.2.2]octane (C), and benzene (D) were prepared and their mesogenic and dielectric properties investigated. Comparative analysis showed that all carborane derivatives form significantly less stable mesophases than their carbocyclic analogs, however they exhibit a relatively high shielding ability for lateral fluorination. Depression of the clearing temperature upon fluorination of series 1, 3, and 5 is approximately constant for each series A-D and correlates with the diameter of the ring (the slope = 14.8 °C A^-1 and R² = 0.997). Compounds in series 2 (X = F) were used as low concentration additives to a nematic host, 6-CHBT. Dielectric parameters were extrapolated to pure additives and analyzed using the Maier-Meier equation. The Kirkwood factors g and apparent order parameters S_app that are required to reproduce the extrapolated dielectric values follow the trend in the size of ring . The smallest g (0.47) and the largest S_app (6.3) are obtained for carborane 2A, and the largest g (0.69) and the smallest S_app (0.7) are obtained for the terphenyl derivative 2D. The increase of S_app in the series D→A corresponds to the increasing disorder of the nematic solution with increasing size of ring script A sign. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b600068a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
The effect of carborane, bicyclo[2.2.2]octane and benzene on mesogenic and
dielectric properties of laterally fluorinated three-ring mesogens{
a
a
a
a
b
Adam Januszko, Kristin L. Glab, Piotr Kaszynski,* Kaushik Patel, Robert A. Lewis, Georg H. Mehl
b
c
and Michael D. Wand
Received 4th January 2006, Accepted 18th May 2006
First published as an Advance Article on the web 20th June 2006
DOI: 10.1039/b600068a
Six series of structurally similar compounds containing 12- and 10-vertex p-carborane (A and B),
bicyclo[2.2.2]octane (C), and benzene (D) were prepared and their mesogenic and dielectric
properties investigated. Comparative analysis showed that all carborane derivatives form
significantly less stable mesophases than their carbocyclic analogs, however they exhibit a
relatively high shielding ability for lateral fluorination. Depression of the clearing temperature
upon fluorination of series 1, 3, and 5 is approximately constant for each series A–D and
2
1
2
correlates with the diameter of the ring (the slope = 14.8 uC A˚ and R = 0.997). Compounds
in series 2 (X = F) were used as low concentration additives to a nematic host, 6-CHBT. Dielectric
parameters were extrapolated to pure additives and analyzed using the Maier–Meier equation.
The Kirkwood factors g and apparent order parameters Sapp that are required to reproduce the
extrapolated dielectric values follow the trend in the size of ring . The smallest g (0.47) and the
largest Sapp (6.3) are obtained for carborane 2A, and the largest g (0.69) and the smallest Sapp (0.7)
are obtained for the terphenyl derivative 2D. The increase of Sapp in the series DAA corresponds
to the increasing disorder of the nematic solution with increasing size of ring
.
1
3,15,16
Introduction
in large conformational freedom of their derivatives.
Recently, we demonstrated that with its large size, 12-vertex
Liquid crystals with negative dielectric anisotropy (De , 0) are
carborane A provides a particularly effective ‘‘shield’’ for
17
lateral substitution as compared to the carbocyclic analogs.
important components of materials for flat panel display
1
,2
applications. One of the most effective ways to introduce a
transverse molecular dipole moment and to form materials
This ability of carboranes to sterically shield and promote
nematogenic properties is of interest for fundamental
structure–property relationship studies and in the context of
developing nematic materials for electro-optical applications.
In continuation of our comparative studies of carboranes
and carbocycles as structural elements for liquid crystals,
we investigate here the effect of the ring structure and
fluorination on the mesogenic behavior of a series of
compounds 1–6 and also on the dielectric parameters of
series 2 in a nematic host. The dielectric results for the binary
mixtures and observed trends for extrapolated values are
analysed by, and rationalized with, the aid of the Maier–Meier
relationship.
3
–8
with De , 0 is by using lateral fluorination of benzene rings.
Such materials typically exhibit low viscosity and high
resistivity desired for nematic and ferroelectric applica-
2
,7,9
tions.
Accumulated results also show that lateral substitu-
tion with fluorine depresses the clearing temperature of the
1
0
liquid crystal compounds by broadening the molecular
width and consequently lowering the molecular aspect
ratio. The extent to which the TNI is affected by such
substitution depends on the size of the ring adjacent to the
1
1
substitution site.
Derivatives of carboranes A and B (Fig. 1) exhibit
nematogenic properties and a lower tendency for the forma-
tion of smectic phases than their isostructural analogs con-
1
2–15
taining bicyclo[2.2.2]octane (C) and benzene (D) rings.
This behavior has been ascribed to the relatively large size
of the carboranes and high order rotational axes which results
a
Organic Materials Research Group, Department of Chemistry,
Vanderbilt University, Nashville, TN, 37235.
E-mail: piotr.kaszynski@vanderbilt.edu; Fax: +1-615-343-1234;
Tel: +1-615-322-3458
b
The Department of Chemistry, Hull University, Hull, HU6 7RX, UK
LC Vision LLC, Boulder CO, 80305, USA
c
Fig. 1 Four
ring
systems:
1,12-dicarba-closo-dodecaborane
{
Electronic supplementary information (ESI) available: analytical
(
12-vertex p-carborane, A), 1,10-dicarba-closo-decaborane (10-vertex
data for synthesized mesogens and their intermediates, DSC for 1D,
texture photomicrographs for 3D, complete dielectric data, computa-
tional method selection, and quantum-mechanical and Maier–Meier
computational details. See DOI: 10.1039/b600068a
p-carborane, B), bicyclo[2.2.2]octane (C), and benzene (D). In A
and B each vertex represents a BH fragment and each sphere is a
carbon atom.
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3183–3192 | 3183

View Article Online
Scheme 2
Scheme 1
Scheme 3
alkylation of bromobenzene with 1-bromo-4-pentylbicyclo-
Results
2
2.2.2]octane following a literature procedure. The prepara-
2
[
2
tion of bromides 7A and 7B was reported before.
3
24
Synthesis
2
5
4
26
4
27
9
Boronic acids 8a, 8b, 8c, 8d, 8e, and 8f were pre-
pared as described in the literature. 4-Bromoheptylbenzene (9),
required for the preparation of 8c, was conveniently obtained
by Pd-catalyzed coupling of heptylmagnesium bromide and
Mesogens 1–6 were prepared by the coupling of aryl halide 7
and boronic acids 8 (or their esters) according to either the
1
original or modified Suzuki procedures (Scheme 1). The use
8
of a mixture of polar N-methylpyrrolidone (NMP) and conc.
1,4-dibromobenzene (Scheme 3), according to an analogous
28
literature procedure.
solution of K PO4 in the latter reaction was found to be
3
generally more efficient. A coupling reaction of triflate 7D
(
Hal = OTf) with an ester of 8b gave a low yield of the expected
D, presumably due to hydrolytic instability of the triflate
Thermal properties
2
in the homogenous aqueous reaction mixture. Only alkoxy
Transition temperatures and enthalpies of the mesogens were
determined by differential scanning calorimetry (DSC) and the
results are shown in Table 1. Phases were identified by com-
derivatives in series B containing the 10-vertex p-carborane
(1B, 2B, 5B, and 6B) were prepared; the synthesis of 3B and 4B
was not attempted due to expected significant difficulties with
parison of the microscopic textures observed under polarized
29–31
1
purification. Transition temperatures for three members of
9
light with those published for reference compounds
based on the established order of phase stability.
and
2
series 5 and 6D were taken from the literature.
0
4
Iodides 7C and 7D (Hal = I) were prepared from the
Carborane derivatives A and B exhibit only a nematic phase,
2
0
corresponding hydrocarbons by iodination with I
2
–HIO
3
with the exception of 5B and the reported 5A which also
show a monotropic SmA. In contrast, all bicyclo[2.2.2]octane
and benzene derivatives, C and D, show smectic polymorphism
2
according to a general literature procedure (Scheme 2). The
1
bromide 7C (Hal = Br) was prepared by a Friedel–Crafts
2
1
a
Table 1 Transition temperatures (uC) and enthalpies (kJ mol ) for carborane-containing liquid crystals
A
B
C
D
b
1
2
3
4
5
6
a
, X = H Cr 103 N 160 I
Cr 63 N 153 I
(17.7) (0.9)
Cr 66 N 121 I
(28.6) (0.7)
Na
Cr 100 SmB 195 SmA 221 N 227 I
(18.9) (3.1) (3.8) (1.2)
Cr 95 E 207 SmB 218 SmA 230 I
(6.1) (5.1) (4.8) (8.7)
Cr 99 SmC 145 N 165 I
(
27.0) (1.4)
, X = F Cr 98 Cr
6.9) (21.0) (0.9)
, X = H Cr 62 Cr 71 N 119 I
0.4) (25.6) (0.8)
, X = F Cr 90.0 N 91.5 I
33.5) (0.9)
c
1
2
99 N 132 I
Cr 64 SmA 157 N 188 I
(18.7) (1.1) (1.0)
Cr 46 SmB 196 SmA 198 I
(
(13.2) (1.0) (1.3)
Cr 135 E 180 SmB 203.5 SmA 204.5 I
e
1
2
d
d
(
(15.2) (12.7)
(0) (5.0) (19.5)
Cr 56 SmC 103 SmA 132 N 137 I
f
Na
Cr 83 SmB 101 SmA 137 N 160 I
(21.9) (1.0) (0.8) (0.7)
73 N 142 I Cr 89 SmB 190 SmA 216.4 N 217 I Cr 97 E 194 SmB 210.5 SmA 221.5 I
d
(
(21.8) (0.05) (5.2)
i
g
k
g
g
, X = H Cr 71 (SmA 45) N 149 I Cr 46 S
A
d
(
20.1) (1.3)
, X = F Cr 75 N 120 I
30.9) (0.8)
(14.1) (0.1) (1.4)
Cr 64 N 111 I
(33.4) (0.6)
(16.3) (2.7) (6.7)
Cr 55 SmC 77 SmA 155 N 177 I
(12.6) (5.0) (6.7) (11.3)
Cr 93.5 SmC 144 SmA 159 I
j
e
h
j
( ) (0.0) (0.8) (1.0)
(
b
Cr = crystal, Sm = smectic, N = nematic, I = isotropic. Lit.: Cr 205 SmB 216 SmA 228.5 I; ref. 32. Lit.: Cr 97.5 SmC 145.5 N 166 I;
c
d
ref. 4. Sum of two phase transition enthalpies. From microscopic observation—see text for details. Lit.: Cr 56 SmC 105.5 SmA 131 N
e
f
g
h
i
21
j
21
136 I; ref. 4. Ref. 20. Ref. 4. Cr–Cr transition at 64 uC (6.1 kJ mol
)
Cr–Cr transition at 33 C (6.3 kJ mol )
Cr–Cr transition at 48 uC (6.1 kJ mol ) and 52 uC. Combined
21
k
21
enthalpy for the latter and melting is 13.1 kJ mol
.
3
184 | J. Mater. Chem., 2006, 16, 3183–3192
This journal is ß The Royal Society of Chemistry 2006

View Article Online
either in combination with a nematic phase or exclusive (3C,
2
D, 3D, 5D, and 6D ). These results are consistent with
0
4
1
3
2
4
4
literature reports for 1D, 2D, and 4D. We found, however,
3
a broad range E phase in addition to the reported SmB
2
and SmA for 1D. This is evident from a broad peak at 95 uC
corresponding to the enthalpy of 6.1 kJ mol , which is
2
1
2
1
consistent with a typical enthalpy of 10 kJ mol for melting to
1
an E phase. Moreover, the enthalpy of 5.1 kJ mol for the
0
21
phase transition at 207 uC does not correspond to a usual value
2
1 10
.
for Cr–SmB transition, which typically is >30 kJ mol
The
E phase was also found in two other terphenyl derivatives 3D
and 5D. DSC analysis of 3D revealed no melting peak, and
only Sm–Sm and Sm–I transitions were observed. Microscopic
studies of 3D showed softening of the virgin solid crystal at
about 135 uC at which temperature the sample became a
viscous liquid susceptible to distortion upon mechanical stress.
This fluid phase was designated as an E phase based on the
characteristic texture formed upon cooling the SmB. The DSC
trace for 1D and photomicrographs of textures for 3D are
provided in the Electronic supplementary information (ESI{).
A comparative analysis of the series of compounds offered
insight into the effect of structural changes on mesogenic
behavior. For better visualization, the series was grouped
according to structural features, and clearing temperatures for
derivatives A, B and C relative to those of the benzene analogs
D are plotted in Fig. 2. Analysis showed that for the non-
fluorinated derivatives (1, 3, 5) the order of the mesophase
stability follows D > C > A > B, while in the fluorinated series
Fig. 3 Average change in clearing temperatures (DTMI
) upon
fluorination in series 1, 3, and 5 plotted as a function of effective
VDW diameter of the ring . Best fit line DTNI = 14.8 6 d 2 138.4
e
2
(R = 0.997).
Fluorination generally decreases the clearing temperatures
for all series and the trend in the DTMI follows the order in
the series AAD (Fig. 3). The smallest destabilization of the
mesophase by about 28 uC is observed for the carborane pairs
1
A–2A, 3A–4A, and 5A–6A, while for the benzene analogues
the average destabilization of TMI is 64 uC. The DTMI values
for the three pairs are within 1 uC in series A, B, and C, while
in series D the values differ by as much as 7 uC. The observed
(
2, 4, 6) the BCO derivatives exhibit the highest clearing
trend in the DT parallels the decrease of effective van der
MI
33
temperatures. In the first group of compounds, clearing
temperatures of BCO derivatives are close to those of benzene
analogues D with an average DTMI = 25 uC, while the largest
difference in TMI of about 279 uC is observed for the 10-vertex
carborane derivatives B. The difference between TMI for
Waals ring size
d
e
of the ring
from the largest for
carborane ring (A) to the smallest for benzene (D).
Other phase transitions in series 2 are also affected by
introduction of the fluorine atoms. A comparison of data for
the Sm–N or SmA–I transitions shows that fluorination
depresses smectic phases in the hydrocarbons by about 62 uC
for C and by an average of 73 uC for benzene derivatives D.
It also induces a SmC phase in benzene derivatives D and
also in 5C. Melting points of the alkoxy derivatives A and D
are little affected by fluorination. In contrast, introduction
of fluorine to derivatives 1C and 5C decreases their melting
points by about 35 uC while 5B melts 18 uC lower than 6B.
Heptyl derivatives 3 are more strongly affected by fluorination.
The melting points of the 12-vertex carborane 3A and BCO 3C
are increased by fluorination, while the melting point of 3D is
depressed by almost 80 uC in the fluoro derivative 4D.
1
2-vertex and benzene derivatives is almost constant for the
alkoxy derivatives, but it increases significantly for the alkyl
derivatives 3 and 4. Thus, the graph reflects differences in the
2
effect of fluorination and OACH substitution on mesogenic
behaviour between each series A–D.
Replacement of the oxygen linking the alkyl chain and the
2
core in 1 and 2 with a CH group in 3 and 4 results in
significant depression of the clearing temperatures (Fig. 4).
For the hydrocarbon series C and D this destabilization
is about 30 uC, while for the carborane A the TNI is lower by
41 uC. This is consistent with our other observations of
generally larger stabilization of the nematic phase upon
CH AO replacement in carborane mesogens than in the
2
3
4
benzene analogs.
Fig. 2 Clearing temperatures for 12-vertex p-carborane (A, r),
Dielectric measurements
10-vertex p-carborane (B, m), and bicyclo[2.2.2]octane (C, $)
derivatives relative to the TMI for the benzene derivatives D. The lines
To assess the effect of ring
on electro-optical properties,
are guides for the eye.
compounds in series 2 were investigated as low concentration
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3183–3192 | 3185

View Article Online
Table 2 Dielectric parameters for 2 extrapolated from 6-CHBT
solutions at 25 uC
a
A
B
C
D
e
e
||
1.1 ¡ 0.1
6.4 ¡ 0.2
2.5 ¡ 0.5
6.8 ¡ 0.3
24.3 ¡ 0.7
4.2 ¡ 0.25
6.5 ¡ 0.3
22.3 ¡ 0.2
4.7 ¡ 0.8
6.8 ¡ 0.3
22.1 ¡ 0.55
H
De 25.3 ¡ 0.2
a
Full experimental details in ESI.
2
Fig. 4 Effect of O ACH change on the clearing temperature TMI for
fluorinated (m) and non-fluorinated ($) series. The lines are guides
for the eye.
1
0 mol%. The observed deviation from linear behaviour and
additives to a nematic host with positive dielectric anisotropy
3
De > 0). On the basis of our previous experience we selected
5
consequently the higher error for extrapolated values of 2B in
(
6
-CHBT parallels other observations for binary mixtures
3,36
6
-CHBT as a suitable host, and dielectric constants were
1
involving 10-vertex mesogens.
Each individual datapoint
measured for solutions of each compound 2 in three
concentrations ranging from about 4 mol% to 17 mol%.
Dielectric constants for each compound 2 were obtained by
linear extrapolation of the data for the solutions and the pure
host. To lower the error for the extrapolated values, the
intercept in the fitting function was set at the appropriate value
for the pure host. A sample of data analysis is presented for 2A
in Fig. 5, and all results are collected in Table 2. Results show
for binary mixtures used for the extrapolation represents an
average of 10 repetitive measurements in a single cell and has
an associated error of ,0.02 units. Since cell parameters vary
slightly, reproducibility of the measurement from cell to cell is
lower and a typical error for a multi-cell analysis is about
0
.08 units. Thus, dielectric parameters for the host have an
error of 0.08 established by averaging results for three cells.
This relatively high uncertainty for the measured values is
partially eliminated in the fitting process of three datapoints
from three different cells.
that the extrapolated transverse dielectric permittivity e is
H
similar for all compounds in series 2 and falls in a narrow
range of 6.4–6.8. In contrast, the longitudinal values vary
substantially from e|| = 4.7 for 2D to e|| = 1.1 for 2A, which
results in marked differences in De in the series.
Computation of molecular parameters
The quality of the data fit to linear functions was excellent
for 2A and 2C and the resulting error on the extrapolated
values was ,0.3 units. For 2D a noticeable non-linear
behaviour was observed for a concentration of 15 mol%, and
the highest concentration used for the extrapolation was about
To rationalize the observed trends in the extrapolated
dielectric properties, dipole moments m and electronic polar-
izabilities a were calculated for 2 at standard orientation using
3
7
the HF the B3LYP methods respectively. Reorientation of
the molecules from Gaussian standard orientation (nuclear
charge based) to the principal moment of inertia coordinates
(
mass based) had a negligible effect on the angle b (+0.3u)
between m and m . Results in Table 3 show that the transverse
||
dipole moment m
H
is the dominant (about 3.8 D) and virtually
invariant for all compounds in the series. It originates from the
two polar C–F bonds reinforced by the C–O bond of the
coplanar alkoxy substituent whose alkyl chain is oriented anti
to the fluorine atoms. In contrast, the longitudinal dipole
moment m|| varies significantly in the series. It reflects the
electron withdrawing ability of the ring
and correlates
39
3
parameters: A (0.14) B (0.05)
8
39
well with the ring r
p
C
4
0
20.13), and D (4-EtPh,20.02). The two dipole moment
41
(
components m and m|| define the orientation of the net
H
molecular dipole moment vector which is approximately
orthogonal to the long molecular axis. The smallest angle b
of 73u is calculated for the 12-vertex carborane derivative with
the largest longitudinal dipole moment component, and the
vector closest to orthogonality is found for the BCO derivative
Fig. 5 Plot of e|| (r), e
H
(m), and De ($) vs. concentration x of 2A in
-CHBT. Datapoints were fitted to functions y = 11.0 + xe||, y = 7.1 +
with the smallest m||
.
6
2
. Correlation parameters R > 0.99 for e|| and De,
xDe , y = 3.9 + xe
2
and R = 0.98 for e
H
The calculated average polarizability aavrg decreases from
H
.
the most polarizable 12-vertex carborane to the terphenyl
3
186 | J. Mater. Chem., 2006, 16, 3183–3192
This journal is ß The Royal Society of Chemistry 2006

View Article Online
a
Table 3 Calculated molecular parameters for 2
assumption, apparent order parameters, Sapp, were obtained
from the extrapolated dielectric anisotropy De by system-
4
atically varying the Kirkwood factor g in such a way that the
6
reverse calculations reproduce the extrapolated dielectric
parameters e and e for each compound. The factor g relates
2
||
H
and meff and is the only adjustable quantity in eqn 1.
2
m
A
B
C
D
Varying the angle b had little effect on e||, which appears to
be the most critical of the two dielectric components. The
resulting g and Sapp values are listed in Table 4, and details of
calculations are described in the ESI{.
b
m
m
|| /D
/D
1.14
3.63
3.80
0.90
3.76
3.86
77
52.0
55.1
0.51
3.84
3.88
82
43.9
49.4
0.71
3.78
3.85
79
53.1
48.7
Analysis of results in Table 4 shows that both experimental
H
H
dielectric components, e|| and e , are very closely reproduced
m /D
c
by calculations. This is remarkable considering that dielectric
anisotropy De was the only experimental input value for the
dopants. The Kirkwood factor g necessary to reproduce the
experimental permittivities follows the trend in the ring size
and varies from the smallest, 0.47, for the carborane 2A, to the
highest, 0.69, for 2D. The observed range of g is typical for
b (u) 73
3
Da/A 50.4
˚
3
˚
avrg/A 57.1
a
a
Dipole moments obtained with the HF/6-31G(d) method and
b
electronic polarizabilities at the B3LYP/3-21G level of theory. The
dipole moment vector is orientated from (negative) to OC
positive). Angle between the net dipole vector and long
molecular axes calculated from the vector components. For details
see ESI.
6
H
13
c
(
m
4
7
nematogens such as 6-CHBT and 5CB. The most surprising
result of this analysis, however, is that the apparent order
parameter Sapp is required to be greater than unity(!).
To verify the correctness of the calculations, dielectric
properties of 2 in 6-CHBT were estimated using arbitrarily
chosen S and g values. Thus, for assumed parameters S = 0.7
and g = 1.0, values typically used in analogous calcula-
derivative (2A > 2B > 2C > 2D). In contrast, the terphenyl
derivative 2D has the largest anisotropy of polarizability Da,
and the carboranes exhibit intermediate values of Da in the
series (2D > 2B > 2A > 2C). The observed trends are consistent
4
8,49
3
5
tions,
narrow ranges, and De increases in the series from 2A to 2C
Table 4). This order of De values is in agreement with the
all calculated dielectric parameters fall in relatively
with our findings for another series of mesogens and are
related to the electronic structure of the two carborane cages.
Both of the carboranes A and B have highly polarizable sigma-
delocalized electrons, which results in high average electronic
polarizability. Due to their ellipsoidal distribution, however,
(
expectations and results from the decreasing magnitude of the
longitudinal dipole moment in the series. Also the estimated De
for 2C is consistent with extrapolated values for similar
4
2
anisotropy of polarizability is small.
2
,3,5
compounds.
These estimated dielectric values for 2 are,
however, significantly different from those extrapolated
from 6-CHBT solutions. Thus, not only is the calculated
range too narrow, but also the trend is opposite relative to
the experimentally observed De values (Table 2), if normal
behaviour of all additives 2 in solution is assumed. Similar
dielectric values and trends were predicted for pure mesogens 2
Dielectric data analysis. The experimental values for the
extrapolated transverse dielectric permittivity e are approxi-
H
mately constant, which is consistent with a narrow range of
m
H
values calculated for series 2 (Table 3). In contrast, the
longitudinal component of the dipole moment varies signifi-
cantly in the series from the lowest for the BCO derivative 2C
(
Table 2), which are consistent with other computational
4
results.
(m|| = 0.51 D) to the highest for the carborane 2A (m|| = 1.14 D).
8
A comparison of m|| values with the experimental dielectric
results for 2 (Table 2) shows that the increasing dipole
corresponds to decreasing e|| values. This is contrary to the
expectations based on the general relationship between these
Analysis of the dielectric results clearly shows that the
extrapolated highly negative De requires a small e which can
||
be calculated only by allowing physically unrealistic Sapp
values. So what does the apparent order parameter Sapp mean,
and what is the physical significance of Sapp > 1? The Maier–
Meier equation (eqn 1) can be presented in a simplified form in
which De is the product of the molecular parameter C and
order parameter S. Consequently, ideal additivity of De for a
binary mixture can be expressed as eqn 2, in which x represents
2
two parameters (e|| y (m||) ). To shed more light on this
unexpected result, dielectric data were analyzed quantitatively
4
using the Maier–Meier equation, which relates the dielectric
3
anisotropy of the nematic phase and molecular parameters
4
4
(
eqn 1).
Using this equation, dielectric behaviour of
compounds 2 was evaluated in solution with 6-CHBT and
i
mole fraction and S is the individual order parameter. Since
also in the pure state.
the original extrapolation ignored the effect of the additive 2
on properties of the host (including the order parameter), Sapp
ꢀ
ꢁ
2
eff
NFh
Fm
À
Á
2
1{3cos b
De~
Da{
S
(1)
represent
a factor which corrects for this assumption.
e0
2kBT
Consequently, a large value for Sapp indicates a large change
in the mixture’s order parameter S. In reality both individual
order parameters Si have close values in a homogenous
mixture, and the assumption of their equality leads to eqn 3.
For calculations involving binary mixtures, the medium
was assumed to be the pure host, and the effect of the additive
was ignored. Therefore, the reaction field factor F and
the cavity factor h in eqn 1 were calculated for pure 6-CHBT
Thus, knowing molecular parameters C , the mixture’s order
i
4
5
using experimental optical and dielectric data. With this
parameter S can be calculated. The relationship between S,
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3183–3192 | 3187

View Article Online
a
Table 4 Bulk parameters for 2 calculated using the Maier–Meier equation
A
B
C
D
Computed values
In 6-CHBT solution
g = 0.47 ¡ 0.04
1.1 ¡ 0.2
6.4 ¡ 0.2
25.4
g = 0.59 ¡ 0.09
2.5 ¡ 0.5
6.8 ¡ 0.4
24.3
2.2 (1.7–3.0)
For Sapp = 0.7, g = 1.0
5.3
8.3
23.1
For S = 0.7, g = 1.0
g = 0.67 ¡ 0.05
4.2 ¡ 0.25
6.5 ¡ 0.25
22.3
g = 0.69 ¡ 0.08
4.7 ¡ 0.5
6.8 ¡ 0.5
22.1
b
e
e
||
H
De
S
app
In 6-CHBT solution
b
b
b
6.3 (5.1–8.2)
0.70 (0.63–0.78)
0.69 (0.57–0.87)
e
e
||
5.4
7.7
22.4
4.7
8.6
23.9
5.3
8.9
23.6
H
De
Pure compound
e
e
||
5.2
7.4
5.0
7.7
4.6
8.2
5.1
8.5
H
De
22.2
6.43
22.7
7.03
23.6
6.65
23.4
7.61
e
s
a
Computed using molecular parameters listed in Table 3 and experimental De in Table 2. The e
s
is an isotropic static permittivity obtained by
numerical solving the Debye equation using theoretical a and m. It is used to calculate parameters F and h. For details see text and ESI.
b
Range of values which correspond to the uncertainty of the g parameter.
S
app, and order parameter of the pure host Shost is expressed
The data in Fig. 6 clearly show a large decrease of the
mixture’s order parameter S for solutions of carborane
derivatives 2A and 2B. Thus, addition of about 15 mol% of
either compound results in a decrease of S by nearly 10%! In
contrast, the BCO and terphenyl derivatives 2C and 2D have
practically no effect on the S parameter within the assumed
modest error limits of ¡2%. It is interesting to note that the
computed S values change monotonically with concentration
for 2A and 2C for which the highest correlation of the
dielectric data was observed (Table 2). Conversely, the
observed scattering of datapoints for 2B and 2D reflects lower
quality of dielectric data correlation for these compounds
by eqn 4.
De = C
1
S
1
x + C
2
S
2
(1 2 x)
(2)
(3)
De = S[C x + C (1 2 x)]
2
1
1
À
Á
S~Shostz
Sapp{Shost
1
za
(
4)
Chost.ð1{xÞ
C.x
where a~
The molecular parameter Chost for the host was calculated
5
0
from De and literature order parameter Shost = 0.67 at 25 uC.
Parameters C for the additives 2 were calculated according to
the Maier–Meier equation and using the estimated Kirkwood
parameter g (Table 4). It should be noted that the resulting
parameters listed in Table 5 depend on the medium (F and h),
g, and T, and also the Kirkwood parameter g depends on T
and the medium (host). For simplicity, factors F and h were
assumed to be the same in the range of the concentrations as
for the pure host. Estimates show that these values are less
than 1% different than those for the actual mixtures.
(vide supra).
i
From parameters C for the mixture’s components and
experimental De for the solutions, order parameter S was
calculated for each concentration according to eqn 4, and the
results are shown in Fig. 6.
a
Table 5 Molecular components of dielectric anisotropy
6-CHBT
2A
2B
2C
2D
Fig. 6 Plot of order parameter S for 6-CHBT solutions vs. mole
fraction of 2A (r), 2B (m), 2C ($), and 2D (&). Parameter S was
calculated using eqn 4 and data in Table 5. The lines are guides for the
eye. The error bar is set at 2%.
C
a
10.60
20.84
21.97
23.30
22.99
Calculated from experimental data for 6-CHBT and using eqn 1
for 2. For details see text and ESI.
3
188 | J. Mater. Chem., 2006, 16, 3183–3192
This journal is ß The Royal Society of Chemistry 2006

View Article Online
1
7
Results presented here and those recently published for
series 10 suggest that shielding of lateral substitution can be
numerically correlated with the ring size. Previous studies
of the ring’s shielding effect were rare and qualitative, and
involved only three rings: benzene (D), bicyclo[2.2.2]octane (C)
1
1
and cyclohexane. The inclusion of the two carboranes in
structure–property relationship investigations has broadened
the range of the ring sizes and permitted the formulation of
quantitative correlations. The presented dependence of DTMI
on the ring size resembles other, more numerous studies which
demonstrated a linear correlation of the size of the lateral
5
2
Fig. 7 A plot of e|| (r), De (m), and order parameter Sapp ($) for 2 as
a function of the diameter d of the cylinder of rotation for ring . Best
fit lines: e|| = 24.45 6 d + 34.2 (R = 0.993); De = 24.22 6 d
R = 0.999). Data and error bars taken from Tables 2 and 4.
substituent and the clearing temperature. More studies
with substituents other than F and more variety of molecular
systems are necessary to establish the generality of the
presented relationship.
c
2
c
c
+ 26.0
2
(
The size of ring appears to be an important factor also in
the modulation of the order parameter and consequently
dielectric parameters for solutions of 2 in 6-CHBT. A measure
of the ring’s impact on the properties of a mixture is the
apparent order parameter Sapp. The value of Sapp is required
by theory to reproduce the extrapolated dielectric values when
the order parameter of the host is assumed to be unchanged. In
other words, Sapp represents a correction for this assumption
or an ‘‘activity coefficient’’ of the dopant. It can be inferred
from eqn 4 that for Sapp = Shost, the host’s order parameter
remains unchanged, for Sapp > Shost it decreases, and for
Results suggest that the disruption of the host’s
nematic order and the values of De and Sapp are related to
the size of ring . Indeed, a plot of e|| and De in Fig. 7
c
shows a linear dependence on the ring size d . Thus, with
increasing steric demands of the molecule, the apparent
order parameter Sapp increases, which implies a decrease
in order parameter S of the solution, and consequent
decrease of dielectric anisotropy. The latter is affected mainly
by a decrease in the longitudinal component of dielectric
permittivity e .
||
Sapp , Shost the order parameter increases, when a dopant
with De , 0 is added to a host with De > 0 and |De(host)| >
53
Discussion
|
De(dopant)|. The Sapp values obtained for 2C and 2D,
0
6
.70 and 0.69 respectively, are not much higher than that of
A comparison of three series of non-fluorinated compounds 1,
, and 5 shows that the clearing temperatures increase in the
order B , A , C , D. This observed trend appears to be
-CHBT (Shost = 0.67), which indicates that both compounds
3
are largely compatible with the host. In contrast, derivatives
of the more bulky carboranes, 2A and 2B, have Sapp values
significantly larger than Shost, which suggests their much lower
compatibility with the 6-CHBT host.
1
2,17,51
general
and can be ascribed mainly to the conforma-
tional mobility and the size of ring . The introduction of
lateral fluorine atoms results in a lowering of the clearing
temperatures TMI for all compounds, and the magnitude of the
depression correlates well with the effective size of ring
The excessively low permittivity e|| = 1.1 extrapolated for 2A
3
5
is consistent with our other results for a binary mixture of a
carborane mesogen in 6-CHBT. We found that De extra-
polated from 6-CHBT solutions for a carborane ester was
significantly more negative (21.3) than the recently measured
value of practically 0.0 for the pure mesogen. Moreover, an
estimate of the order parameter from optical data for this
mesogen showed the lowest value in the series of structurally
(
Fig. 3). Since the DTMI for each ring is practically the same for
the OC H (series 1 and 2), C H (series 3 and 4), and
6
13
7
15
8
OC H17 (series 5 and 6) pairs of compounds, the effect appears
to depend only on ring and not on the nature of the terminal
group (chain length and the linking group). This suggests
that conformations of the non-fluorinated and fluorinated
analogues are very similar.
3
5
similar compounds. One possible reason for these extra-
polated low permittivity values in carborane derivatives is the
non-linear behaviour of solutions which has been reported in
A comparison of results for series 1–6 with recently
1
7
reported series 10 shows that the change in clearing tem-
peratures DTMI depends on the core structure and presumably
on the number of fluorine atoms and their spatial relationship
with the ring . This is reflected in the slope of the best-fit line
for each correlation. Thus, the slope obtained for series 10 is
about 1/3 greater than that for series 1–6 calculated per one F
atom. This is rationalized by more effective shielding of the
substitution site closer to ring in series 10 as compared to
mesogens 1–6.
4
5,54
the literature for other binary mixtures.
The dielectric values extrapolated for 2 and listed in Table 2
should be considered as relative to the others in the series
rather than as absolute values. For instance, they strongly
depend on the precise values for the pure host. Since the host’s
values have a relatively large error of approximately 0.1,
precise absolute dielectric values for additives 2 are unattain-
able, however, the resulting trend is sufficient for the
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3183–3192 | 3189

View Article Online
structure–property analysis. Further, dielectric parameters
for the host and mixtures, and consequently the resulting
extrapolated values depend also on the type of cell used, the
orienting layer, the pretilt angle, and the definition of the
threshold voltage VTH (% of raise). Thus, for polyimide-coated
cells used in this work with the pretilt of 2u–3u, the values
obtained for the host are lower than those reported in the
parameter Sapp reflects the effect of the dopant on the order
parameter of the mixture. Thus, a large apparent order
parameter Sapp required by the theory for the extrapolated
dielectric values for 2 implies a large decrease of the actual
order parameter S of the solution in 6-CHBT. In series 2 the
Sapp increases with the increasing size of the ring ; the larger
the ring the higher steric demands of the additive, the higher
apparent order parameter Sapp, and the lower actual order
parameter S of the solution.
2
literature. In contrast, preliminary experiments in SiO -coated
cells with practically 0.0 pretilt angle showed that although the
measured permittivity values were higher, the trend in the
extrapolated values for additives 2 remains the same. Thus,
the observed trends are fully reproducible and suggest a
strong correlation with molecular parameters (such as in
Fig. 7), but the quantitative relationship between the structure
and dielectric behaviour will require more detailed and precise
experiments.
Experiment and theory are consistent and show that an
increase in the size of ring
(and hence decrease in order
of a
parameter) strongly affects the e|| but weakly affects e
H
host with De > 0. In consequence, the addition of a sterically
demanding dopant such as 2A with De , 0 excessively lowers
the e|| component, which results in lower De values as
compared to more sterically compatible dopants such as 2C.
Further studies are needed to provide higher quality dielectric
data and quantitative correlations, and also to understand how
additives 2 affect dielectric properties of a host with De , 0
and whether similar excessive increase of the negative De values
can be observed upon addition of 2A and 2B.
Results presented here contribute to a better understanding
of the impact of carborane derivatives on materials properties
and the formulation of mixtures in general. They also shed
more light on the dielectric values that are typically
extrapolated from 10%–20% solutions of either mesogenic or
2
,7,8,55
The protocol used in this work to analyze dielectric
results for binary mixtures consists of three steps. First, using
extrapolated De, computed molecular values, and field para-
meters for the pure host, Kirkwood parameter g is fitted
non-mesogenic compounds.
For structurally similar
compounds and very low concentrations, the impact of the
additive on a material’s order parameter is minimal. For
higher concentrations and less sterically compatible compo-
nents, such as carborane-containing compounds, the order
parameter of the mixture may be affected strongly by the
dopant. Consequently, dielectric and optical properties of such
mixtures may exhibit strong non-linear behavior.
H
to calculate back the extrapolated values e|| and e . This
procedure also gives the Sapp. Secondly, using the resulting g
value, the molecular constant C is calculated for the additive
from the Maier–Meier equation. Finally, the order parameter
of the mixtures is computed from the additivity of De.
Finally, the use of a differently defined ring size requires a
comment. Thus, correlation in Fig. 3 uses the effective ring
diameter d
e
, while the VDW diameter of the cylinder of ring
Experimental
rotation, d , is used for analysis of data in Fig. 7. It can be
c
Optical microscopy and phase identification were performed
using a PZO ‘‘Biolar’’ polarized microscope equipped with a
HCS250 Instec hot stage. Thermal analysis was obtained using
a TA Instruments 2920 DSC. Transition temperatures (onset)
and enthalpies were obtained using small samples (1–2 mg)
e
argued that effective ring size d is appropriate for correlation
of clearing temperatures (such as in Fig. 3), since it is directly
5
6
related to the packing fraction of the mesogen. Steric
effects exerted by a molecule on the environment and
order parameter, may be correlated with the diameter of
2
1
and a heating rate of 5 uC min under a flow of nitrogen gas.
The clearing transition was typically less than 0.3 uC wide.
E-4-(4-Hexylcyclohexyl)phenyl isothiocyanate (6-CHBT)
was purified by vacuum distillation before use. The measure-
ments for the pure host at 23 uC: VTH = 1.73 ¡ 0.02 V;
the rotating ring d
c e
. For rings A–C the two diameters d
and d are the same, while for benzene the diameter of the
c
˚
cylinder of rotation d is 6.66 A instead of the effective ring size
c
˚
d = 5.03 A.
e
e
values:
|| = 11.0 ¡ 0.1; e
H
= 3.9 ¡ 0.1; De = 7.1 ¡ 0.1; Literature
5
7
Conclusions and summary
V
TH = 1.63; e|| = 12.0; e = 4.0; De = 8.0.
H
Experimental results demonstrate that thermal and dielectric
properties of pure compounds and binary mixtures depend on
the steric factors of the compounds and scale with the ring size.
Thus the depression of the clearing point upon lateral
fluorination can be quantitatively correlated with the effective
ring size. It is suggested that the slope of the linear fit provides
a measure of the effectiveness of shielding of the substituent
Dielectric measurements
Properties of compounds in series 2 were measured by a Liquid
Crystal Analytical System (LCAS - Series I, LC Analytical
Inc.) using GLCAS software version 0.59 which implements
5
8
literature procedures for dielectric constants.
Approximately 5, 10 and 15 mol% solutions of 2 in 6-CHBT
were prepared and conditioned for 24 h at 50 uC. Thermal
analysis of each mixture (DSC) showed linear dependence of
both onset and peak temperature of the transition on mole
fraction x. The onset transition temperatures were fitted to
TNI = 42.0 + xTonset and gave the extrapolated values which
are higher by +12 uC, 22 uC, +28 uC, +6 uC than those for
by ring
and is related to the number and position of the
substituents relative to ring . More experimental work
involving other substituents and structural motifs is needed
to establish the generality of this correlation.
Dielectric results for low concentration nematic solutions
give information about host–additive interactions and
their structural compatibility. The calculated apparent order
2
pure 2A–2D, respectively (r > 994, error , ¡2.5 uC).
3
190 | J. Mater. Chem., 2006, 16, 3183–3192
This journal is ß The Royal Society of Chemistry 2006

View Article Online
The mixtures were loaded into ITO electro-optical cells by
preparation of 6A, Mr Po-Chih Chen for technical assistance,
and Ms Krystyna Kulikiewicz for help with obtaining several
NMR spectra. We are also grateful to Dr Eike Poetsch of
E. Merck for the gift of 4-bromo-49-pentylbiphenyl (7D).
capillary forces at ambient temperature. The cells (about 5 mm
2
thick, electrode area of 0.28 cm and antiparallel rubbed
polyimide layer 2u–3u pretilt) were obtained from LCA Inc.
and their precise thickness (¡0.05 mm) was measured by
optical methods. The filled cells were heated to an isotropic
phase and were left for an hour at room temperature before
measurement.
References
1
P. Kirsch and M. Bremer, Angew. Chem., Int. Ed., 2000, 39,
216–4235.
2 D. Pauluth and K. Tarumi, J. Mater. Chem., 2004, 14, 1219–1227.
4
Default parameters were used for measuring dielectric
constants of the mixtures: triangular shaped voltage bias
ranging from 0.1–20 V at 1 kHz frequency. The threshold
voltage Vth was measured as a 10% of change. For each
mixture the measurement was repeated at least ten times. The
first two results were rejected and the rest of consistent results
were averaged to calculate the mixture’s parameters. Values
for dielectric permittivity e were plotted as a function of
concentration and extrapolated to pure 2. The intercept in the
fitting functions was fixed at the value for the pure host. The
results are presented in Table 2 and details are listed in ESI{.
3
V. Reiffenrath, J. Krause, H. J. Plach and G. Weber, Liq. Cryst.,
989, 5, 159–170.
1
4
G. W. Gray, M. Hird, D. Lacey and K. J. Toyne, J. Chem. Soc.,
Perkin Trans. 2, 1989, 2041–2053.
5 P. Kirsch, V. Reiffenrath and M. Bremer, Synlett, 1999, 389–396.
6
7
M. Bremer and L. Lietzau, New J. Chem., 2005, 29, 72–74.
M. Klasen, M. Bremer and K. Tarumi, Jpn. J. Appl. Phys., 2000,
3
9, L1180–L1182.
8 G.-X. Sun, B. Chen, H. Tang and S.-Y. Xu, J. Mater. Chem., 2003,
3, 742–748.
1
9
C. C. Dong, P. Styring, J. W. Goodby and L. K. M. Chan,
J. Mater. Chem., 1999, 9, 1669–1677.
10 V. Vill, LiqCryst 4.3, ed. N. Harris, H. Sajus and G. Peters,
002, LCI Publisher GmbH, Hamburg, Germany, database and
references therein.
2
Suzuki coupling of halides 7 and boronic acid 8. General
procedure for the preparation of 1–6
1
1
1 K. J. Toyne, in Thermotropic Liquid Crystals, ed. G. W. Gray,
Wiley, New York, 1987, pp. 28–63.
2 A. Januszko, P. Kaszynski, M. D. Wand, K. M. More,
S. Pakhomov and M. O’Neill, J. Mater. Chem., 2004, 14,
1
Method A . A mixture Pd(PPh ) (0.03 mmol), toluene
8
3
4
1
544–1553.
3 W. Piecek, J. M. Kaufman and P. Kaszynski, Liq. Cryst., 2003, 30,
9–48.
4 P. Kaszynski and A. G. Douglass, J. Organomet. Chem., 1999, 581,
28–38.
5 A. Januszko, P. Kaszynski, K. Ohta, T. Nagamine, V. G. Young,
Jr. and Y. Endo, in preparation.
16 A. Januszko, P. Kaszynski and W. Drzewinski, J. Mater. Chem.,
2006, 16, 452–461.
(
5.0 mL), aryl halide 7 (1.0 mmol), and an aqueous solution of
1
1
1
Na CO3 (1.0 mL of a 2 M solution) under nitrogen atmo-
2
3
sphere, and then boronic acid 8 (1.1 mmol) in EtOH (1.0 mL)
was added. The mixture was refluxed at a temperature of
1
10 uC for approximately 24 h under vigorous stirring and the
progress monitored by TLC analysis (hexane: CH Cl , 9 : 1).
2
2
The reaction mixture was poured into water, the product was
extracted with CH Cl , extracts washed with brine, and dried
Na SO ). The solution was passed through a silica gel plug,
1
7 B. Ringstrand, J. Vroman, D. Jensen, A. Januszko, P. Kaszynski,
J. Dziaduszek and W. Drzewinski, Liq. Cryst., 2005, 32,
2
2
(
2
4
1061–1070.
18 N. Miyaura, Y. Yanagi and A. Suzuki, Synth. Commun., 1981, 11,
513–519.
9 Pure carborane derivatives with sharp transition temperatures can
be obtained only after repeated recrystallization, but 10-vertex
derivatives in general do not crystallize well.
solvent was evaporated and the resulting solid was purified
chromatographically on silica gel.
1
Final purification for analysis was performed as follows:
2 2
each compound was dissolved in CH Cl , solution filtered
through cotton to remove particles, evaporated and the
product recrystallized from the indicated solvent until constant
temperature. The resulting crystals were dried in vacuum
overnight at ambient temperature. The purity was confirmed
by combustion analysis. All analytical data (NMR, combus-
tion analysis and MS) are provided in ESI{.
20 K. Czuprynski, A. G. Douglass, P. Kaszynski and W. Drzewinski,
Liq. Cryst., 1999, 26, 261–269.
2
1 H. O. Wirth, O. K o¨ nigstein and W. Kern, Liebigs Ann. Chem.,
960, 634, 84–104.
22 G. W. Gray and S. M. Kelly, J. Chem. Soc., Perkin Trans. 2, 1981,
6–31.
1
2
2
3 A. G. Douglass, S. Pakhomov, B. Reeves, Z. Janousek and
P. Kaszynski, J. Org. Chem., 2000, 65, 1434–1441.
24 Z. Janousek and P. Kaszynski, Polyhedron, 1999, 18, 3517–3526.
25 S. Sharma, D. Lacey and P. Wilson, Liq. Cryst., 2003, 30, 451–461.
Method B. Boronic acid 8 (0.17 mmol), aryl halide
2
2
2
2
6 M. R. Friedman, K. J. Toyne, J. W. Goodby and M. Hird, Liq.
Cryst., 2001, 28, 901–912.
7
2
(0.14 mmol), and Pd(AcO) (1.0 mg) were dissolved in
degassed N-methylpyrrolidinone (NMP, 1.5 mL). The flask was
flushed with argon and heated to 50 uC. Tricyclohexylphosphine
7 M. C. Artal, K. J. Toyne, J. W. Goodby, J. Barber a´ and
D. J. Photinos, J. Mater. Chem., 2001, 11, 2801–2807.
8 H. Dorn, J. M. Rodenzo, B. Brunnh o¨ fer, E. Rivard, J. A. Massey
and I. Manners, Macromolecules, 2003, 36, 291–297.
9 I. Dierking, Textures of Liquid Crystals, Wiley-VCH, Weinheim,
2003.
(
2.0 mg) was added, and the mixture was stirred for 10 min.
A 2.5 M solution of K PO (1.0 mL) was added and the reaction
3
4
was stirred at 90 uC for 8 h under Ar. The reaction mixture
was poured onto water, and the product was extracted into
EtOAc. The organic extract was washed with brine and dried
30 G. W. Gray and J. W. G. Goodby, Smectic Liquid Crystals-
Textures and Structures, Leonard Hill, Philadelphia, 1984.
3
3
3
1 D. Demus and L. Richter, Textures of Liquid Crystals, 2nd edition,
VEB, Leipzig, 1980.
2 G. W. Gray, M. Hird and K. J. Toyne, Mol. Cryst. Liq. Cryst.,
991, 195, 221–237.
3 The geometrical parameters for each ring were calculated (HF/6-
1G(d)) and corrected for H and C (benzene) van der Waals radii
(
Na SO ). Further purification as described in Method A.
2 4
1
Acknowledgements
3
This project was supported by the NSF grant (DMR-0111657).
We thank Dr Serhii Pakhomov for his help with the
˚ ˚
(1.2 A and 1.7 A respectively). Benzene was treated as an ellipsoid
and the diameter d represents an average of the two axes a and b:
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3183–3192 | 3191

View Article Online
˚
A: d = 7.43 A; B: d = 7.17 A; C: d = 6.72 A; D: d = 5.03 A
˚
˚
˚
47 P. Kedziora and J. Jadzyn, Acta Phys. Pol., 1990, A77, 605–610.
48 G. Saitoh, M. Satoh and E. Hasegawa, Mol. Cryst. Liq. Cryst. Sci.
Technol., Sect. A, 1998, 301, 13–18.
49 M. Klasen, M. Bremer, A. G o¨ tz, A. Manabe, S. Naemura and
K. Tarumi, Jpn. J. Appl. Phys., 1998, 37, L945–L948.
50 S. Urban, J. Kedzierski and R. Dabrowski, Z. Naturforsch., 2000,
55A, 449–456.
51 K. Ohta, A. Januszko, P. Kaszynski, T. Nagamine, G. Sasnouski
and Y. Endo, Liq. Cryst., 2004, 31, 671–682.
52 For example see:D. Demus, in Hadbook of Liquid Crystals, ed.
D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, and V. Vill,
Willey-VCH, 1998, vol. 1, pp. 151–153. Also ref. 10 and citation
therein.
53 Based on eqn 4, a reverse relationship between S and Sapp is
expected when both dopant and the host have the same sign of
dielectric anisotropy or |De(host)| , |De(dopant)|.
54 Z. Raszewski, Liq. Cryst., 1988, 3, 307–322.
55 S. Naemura, in Physical Properties of Liquid Crystals: Nematics,
ed. D. Dunmur, A. Fukuda, and G. Luckhurst, IEE, London,
2001, pp. 523–581.
˚
a = 6.66 A, b = 3.40 A).
˚
(
3
3
4 T. Nagamine, A. Januszko, K. Ohta, P. Kaszynski and Y. Endo,
Liq. Cryst., 2005, 32, 985–995.
5 A. G. Douglass, K. Czuprynski, M. Mierzwa and P. Kaszynski,
J. Mater. Chem., 1998, 8, 2391–2398.
6 K. Czuprynski and P. Kaszynski, Liq. Cryst., 1999, 26, 775–778.
7 Details including method selection are given in the ESI{.
8 C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91, 165–195
and references therein.
9 L. I. Zakharkin, V. N. Kalinin and E. G. Rys, Bull. Acad. Sci.
USSR, Div. Chem. Sci., 1974, 2543–2545.
3
3
3
3
4
0 The s
E. C. H. Keung and R. A. Partis, J. Org. Chem., 1971, 36,
352–1355.
p
value for adamantane was used instead: H. Alper,
1
4
4
1 M. Charton, Prog. Phys. Org. Chem., 1981, 13, 119–251.
2 P. Kaszynski, S. Pakhomov, K. F. Tesh and V. G. Young, Jr.,
Inorg. Chem., 2001, 40, 6622–6631.
4
3 W. Maier and G. Meier, Z. Naturforsch., 1961, 16A, 262–267 and
470–477.
4
4 S. Urban, in Physical Properties of Liquid Crystals, ed.
D. Nematics, A. Dunmur, Fukuda, and G. Luckhurst, IEE,
London, 2001, pp. 267–276.
56 D. Demus and A. Hauser, in Selected Topics In Liquid Crystal
Research, ed. H. D. Koswig, Akademie–Verlag, Berlin, 1990,
pp. 19–44.
45 Z. Raszewski, R. Dabrowski, Z. Stolarzowa and J. Zmija, Cryst.
Res. Technol., 1987, 22, 835–844.
46 P. Bordewijk, Physica, 1974, 75, 146–156.
57 R. Dabrowski, Mol. Cryst. Liq. Cryst., 1990, 191, 17–27.
58 S.-T. Wu, D. Coates and E. Bartmann, Liq. Cryst., 1991, 10,
635–646.
3
192 | J. Mater. Chem., 2006, 16, 3183–3192
This journal is ß The Royal Society of Chemistry 2006