Examination of the interlayer strength of smectic liquid crystals through the study of partially fluorinated and branched fluorinated end-groups

Article abstract of DOI:10.1039/b517913k

Through the incorporation of partially fluorinated and branched fluorinated end-groups of ferroelectric liquid crystals we aimed to examine the effects on physico-chemical structures and properties, and electrical switching behaviour for a family of compo

Full text of DOI:10.1039/b517913k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Examination of the interlayer strength of smectic liquid crystals through the
study of partially fluorinated and branched fluorinated end-groups{
Stephen J. Cowling,*a Alan W. Hall,a John W. Goodby,a Ying Wangb and Helen F. Gleesonb
Received 19th December 2005, Accepted 22nd March 2006
First published as an Advance Article on the web 11th April 2006
DOI: 10.1039/b517913k
Through the incorporation of partially fluorinated and branched fluorinated end-groups of
ferroelectric liquid crystals we aimed to examine the effects on physico-chemical structures and
properties, and electrical switching behaviour for a family of compounds based upon the
MHPOBC motif. We show that the nature of the terminal group strongly affects the liquid crystal
behaviour and the thresholds associated with electro-optic switching. In this article we present the
synthesis, transition temperatures, electro-optical measurements and X-ray diffraction studies for
a series of compounds.
branched chains with methoxy or ethoxy groups located at the
terminal position.9
1. Introduction
Ferroelectricity in smectic liquid crystals was first predicted
and investigated by Meyer et al.1,2 in 1975, and the practical
One of the ways in which mesophase behaviour and physical
properties can be controlled is through interactions occurring
use of ferroelectric liquid crystals was demonstrated by Clark
at the interfaces of the layers, but to date there has been little
and Lagerwall through the invention of the surface stabilized
systematic work on the study of the interactions being made
ferroelectric display (SSFLCD) device.3 The antiferroelectric
at the interfacial regions through structural changes to the
phase was later discovered in the now classical compound
terminal chain. By varying the polarity and the bulky nature of
MHPOBC4 by Chandani et al. Ever since these discoveries and
the terminal group it is possible to increase or decrease the
inventions there has been considerable research into materials
interlayer interactions, and to effectively strengthen or weaken
exhibiting these phases. The majority of materials showing
the interlayer interface and affect the mesophase behaviour
and physical properties.10 Furthermore, we recently showed
ferroelectric and antiferroelectric phases consist of an aromatic
core which can be composed of purely aromatic rings or may
that through the use of large steric groups, such as alicyclic
also include heteroatoms, with terminally appended aliphatic
rings and caged systems located at the terminal position, it was
chains which, when the molecules are arranged into layered
possible to weaken the interlayer regions to such an extent that
structures, partially/weakly interdigitate and form diffuse
frustrated twist grain boundary (TGB) phases were exhibited
in preference to typical smectic phases.11 In this study we show
interlayer regions.
The majority of work in this area has concentrated on the
that via terminal fluorination it is possible to affect properties
study of variations of the aromatic core which may also
such as spontaneous polarization and tilt angle, and hence
include lateral substitution.5–7 Thus, it is believed that the core
show how critically important the interface is to mesophase
structure is the most important component affecting the type
formation and physical properties.
of mesophase exhibited and the temperature ranges of the
In addition, the motivation for this work was due in part to
mesophases formed. Other factors which need to be considered
the proposal by Lagerwall et al. of a device based upon anti-
for potential device applications, however, are physical
switching).12 There have been few examples reported of
properties such as the tilt angle, spontaneous polarization
ferroelectric liquid crystals with tilt angles of 45u (orthoconic
and rotational viscosity.
materials which have tilt angles as high as 45u, however,
Dabrowski and co-workers have shown, through the incor-
There have been many studies on the branched aliphatic
chain systems, which have shown that branching might
poration of perfluorinated end-chains, it is possible to produce
antiferroelectric materials with tilt angles of 45u.13–15 These
affect the type and temperature ranges of mesophases formed.
Examples of this type of study are shown by the formation of
types of materials have a large polar group at the end of
anticlinic phases by swallow tailed branched terminal chains,8
the molecules, and it is generally accepted that perfluoro
and also through the discovery of V-shaped switching via
chains like to microphase segregate and hence may form well
organized interlayer regions. A graphical representation of a
aLiquid Crystal Research Group, Department of Chemistry, University
normal smectic mesophase and a smectic mesophase where the
of York, Heslington, York, UK YO10 5DD. E-mail: sc520@york.ac.uk;
molecules have terminal groups which microphase segregate,
jwg500@york.ac.uk
bDepartment of Physics and Astronomy, Schuster Building, University of
Manchester, Manchester, UK M13 9PL.
E-mail: helen.gleeson@manchester.ac.uk
{ Electronic supplementary information (ESI) available: Synthetic
procedures and characterization details for 2, 3, 8, 15, 22, 25 and 26.
See DOI: 10.1039/b517913k
e.g., via the incorporation of polar, non-polar, bulky, siloxane
units etc., is shown in Fig. 1.
In this report our systematic studies demonstrate effects and
influences of fluorinated terminal polar groups in ferroelectric
and antiferroelectric materials based upon the core structure of
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 2181–2191 | 2181

View Article Online
4.16 The hydroxy group was first protected as the carbonate to
give 5, this intermediate was esterified with 2-octanol in the
presence of diisopropyl azodicarboxylate (DIAD) in the
presence of triphenylphosphine and THF to give 6. In
the esterification using DIAD there is inversion of chirality
i.e., R-2-octanol becomes the S-derivative in the ester. In this
process there may be a slight reduction in the enantiopurity of
the resulting phenols compared to the alcohols. Ester 6 was
deprotected using a mixture of ammonia and ethanol to
generate the phenol 7.
Compound 3 and compound 7 were esterified in the
presence of 1-(3-dimethylaminopropyl)-3-ethyl carbodiimide?
HCl (EDAC) and 4-(dimethylamino)pyridine to give 8. Under
these conditions 8 was formed without any self-esterification of
3. The last step in the pathway was the esterification, under
similar conditions, of 8 with one of the selected acids (RCO2H)
to give products 9 to 20, and where R is either HCF2(CF2)m or
(CF3)2CF(CH2)2. See Table 1 for elemental and mass analyses.
Scheme 2 represents the synthesis of compounds 26–29
which has a biphenyl benzoate arrangement of the aromatic
rings (i.e., essentially reversed to the first set of materials) for
the central core unit, see Table 1.
Fig. 1 A graphical representation of a) a normal smectic mesophase
where the end chains may partially interdigitate within the interlayer
region and b) a smectic mesophase where there is a terminal group
which may enhance or weaken the interface by disrupting or
strengthening the interlayer interactions.
Compound 22, 4-(3-hydroxypropyloxy)benzoic acid, was
prepared by alkylation of compound 21 with 3-bromopropan-
1-ol in the presence of potassium carbonate in butanone.
MHPOBC. The study is broken into two parts, firstly to
examine linear, polar, partially fluorinated groups, and secondly
the examination of bulkier branched fluorinated groups. The
general structures of the materials are shown in Fig. 2.
Compound
24,
1-methylheptyl
49-hydroxybiphenyl-4-
carboxylate, was prepared by esterification of 49-hydroxy-
biphenyl-4-carboxylic acid with (S)-2-octanol in the presence
of triphenylphosphine and DIAD in dry THF. Compound 25
was prepared by esterification of 22 and 24 under similar
conditions as for the preparation of compound 8. Esterifica-
tion of compound 25 with the selected acids (RCO2H) gave the
target materials 26 to 29 where R is either HCF2(CF2)m or
(CF3)2CF(CH2)2. See Table 1 for elemental and mass analyses.
2. Experimental
In the following section we describe the synthesis of the
(R)-4-(1-methylheptyloxycarbonylphenyl) 49-alkoxybipheny-4-
carboxylate target compounds where the alkoxy group has
a v-fluorinated or v-branched fluorinated terminal group
connected to the end of the methyleneoxy chain via an ester
linkage.
2.2 Evaluation of structural and physical properties
2.1 Synthesis of materials
Phase identification by optical and thermal methods. Phase
identifications and determination of phase transition tempera-
tures were carried out by thermal polarized light microscopy
using a Zeiss Universal polarizing transmitted light microscope
equipped with a Mettler FP82HT microfurnace in conjunction
with a FP90 Central Processor. Differential scanning calori-
metry was used to determine enthalpies of transition and to
confirm the phase transition temperatures determined by
optical microscopy. Differential scanning thermograms (scan
rate 10u min21) were obtained using a Perkin Elmer DSC 7 PC
system operating using Pyris software or on a Mettler
DSC822e operating on the Stare software. The results obtained
were standardized to indium (measured onset 156.68 uC, DH
28.47 J g21, lit. value 156.60 uC, DH 28.45 J g21).
General synthetic pathways. The following synthetic path-
way, see Scheme 1, represents a typical synthetic route for the
target liquid crystals 9–20.
Compound 3, 49-(11-hydroxyundecyloxy)biphenyl-4-car-
boxylic acid, was prepared starting from 4-cyano-
49-hydroxybiphenyl, 1, which was first alkylated with
11-bromoundecan-1-ol in the presence of potassium carbonate
in butanone, followed by acid hydrolysis of the nitrile. Com-
pound 7, (R)-1-methylheptyl 4-hydroxybenzoate, was prepared
by standard methods starting from 4-hydroxy-benzoic acid,
X-Ray diffraction studies. The X-ray diffraction experiments
were carried out on powder samples at Daresbury laboratories
using station 2.1 at the synchrotron radiation source. The
diffraction patterns show the intensity as a function of the
modulus of the scattering wave vector (q);
Fig. 2 General structure of the materials prepared in this study.
2182 | J. Mater. Chem., 2006, 16, 2181–2191
q = |q|4psinh/l = 2pn/d
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Scheme 1 The synthesis of compounds 9–20 incorporating either a linear or branched terminal fluorinated group.
˚
where h is the diffraction angle, l is the wavelength (1.54 A), n is
Table 1 Elemental and mass analyses for compounds 9–30
an integer and d is the lattice distance. The error in the experiments
˚
is ¡0.1 A.
C (%)
H (%)
Compound
No.
Found Expected Found Expected MS (m/z)
755 (M+ + Na)
855 (M+ + Na)
955 (M+ + Na)
797 (M+ + Na)
897 (M+ + Na)
997 (M+ + Na)
867 (M+ + Na)
967 (M+ + Na)
1067 (M+ + Na)
751 (M+ + Na)
793 (M+ + Na)
863 (M+ + Na)
755 (M+ + Na)
855 (M+ + Na)
955 (M+ + Na)
751 (M+ + Na)
1013 (M+ + Na)
Molecular simulations. Molecular lengths were determined
either by using Dreiding molecular models or via computer
simulations using a PC and ChemDraw3D2 as part of the
ChemDraw Ultra 8.0 program. Space filling models of mole-
cular structures minimized using MM2 computations were
obtained using ChemDraw Ultra assuming that the molecule
was in the gas phase at absolute zero.
9
58.93 59.01
54.60 54.81
51.60 51.51
60.62 60.46
56.06 56.30
53.02 52.98
62.37 62.55
58.49 58.47
55.53 55.18
60.78 60.98
62.12 62.33
63.71 64.27
58.81 59.01
54.51 54.81
51.32 51.51
60.98 60.98
58.06 58.18
5.19
4.23
3.95
5.70
4.99
4.35
6.34
5.65
5.02
5.58
6.17
6.73
4.99
4.26
3.80
5.46
5.68
4.95
4.42
3.89
5.46
4.84
4.34
6.20
5.55
5.01
5.39
5.88
6.59
4.95
4.42
3.89
5.39
5.59
10
11
12
13
14
15
16
17
18
19
20
26
27
28
29
30
Electro-optic measurements. The cells for electrical field tests
were positioned within a Mettler FP82HT hot-stage oven used
in conjunction with a Mettler FP90 temperature controller.
The hot-stage was mounted on a Zeiss Universal microscope
fitted with crossed-polarizers and an 8/0.2 objective. The
applied electric waveform was supplied via a Hewlett Packard
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 2181–2191 | 2183

View Article Online
Scheme 2 The synthesis of compounds 26–29.
33120A Arbitrary Waveform Generator coupled to a linear
620 amplifier (custom-built by QinetiQ). The electrical
response from the cell was fed into a nano-current detector
(20 kV or 100 kV impedance) and to a Hewlett Packard
54600B oscilloscope, which was connected to a PC.
Measurement of tilt angles. The determination of tilt angle
was achieved by measuring the optical transmission of the
switched states using a photodiode (RS303-674; 1 cm2 active
area, high speed >50 ns) in an apparatus custom-built by
QinetiQ. The equipment was fitted with a green-eye response
filter (Coherent-Ealing, 26-7617-000, 1 inch diameter, trans-
mittance 400–700 nm, maximum transmittance 539.5 nm). The
transmission values were determined using a Thurlby Thandar
Instruments 1604 Digital Multimeter.
The type of cells used in the electrical field studies were
constructed from ITO coated glass which had internal separa-
tions 5 ¡ 0.1 mm (Linkam), with the inner surfaces coated
with a polyimide aligning agent buffed in an antiparallel
configuration. The cells were filled by capillary action.
Alignment was achieved by slowly cooling into the smectic
A* mesophase from the isotropic state, and applying a high
electric field in the form of a square wave (typically 16–
20 V mm21) at a frequency of 50 Hz to 1 kHz depending on the
material. Once the material was aligned in the cell, the voltage
was removed and the cell cooled down into the smectic C* or
CA* mesophases. For comparison of results, measurements
were determined as a function of the reduced temperature,
Tc 2 T, where Tc is the Curie point.
3. Results
3.1 Transition temperatures
The transition temperatures, enthalpies of transition, tilt
angles and values of the maximum spontaneous polarization
for materials 9 to 17, which have a linear terminal chain, are
given in Table 2. Similarly, in Table 3 analogous results are
given for the branched perfluoro-terminated compounds 18–20
and Table 4 gives the results for the compounds with
‘‘reversed-core’’ configuration 26–29.
Measurement of spontaneous polarization. Upon alignment
of the liquid crystal, the spontaneous polarization in the tilted
phases was determined, on cooling, using a triangular wave-
form current reversal technique. An AC field in a triangular
waveform (typically 10 V mm21 at 30 Hz) was applied to the
cell. As the polarization is inverted by the field, a current pulse
was observed. The magnitude of the spontaneous polarization
was determined from the area under the current-reversal peak.
In addition, Table 5 lists the transition temperatures of a
similar set of known materials containing either a perfluoro
end-group or a hydrocarbon end-group which give reasonable
comparisons with the new materials reported. Compound 30
was prepared for comparison with the branched compounds
18–20 to examine the effects of a branched end-group relative
to a straight chain end-group.
2184 | J. Mater. Chem., 2006, 16, 2181–2191
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Table 2 The phase classification, phase transition temperatures (uC), enthalpies of transition [DH, kcal g21], maximum values of the spontaneous
polarization (Ps, nC cm22) and optical tilt angles (h, u) for compounds 9 to 17
Compound No.
m
n
9
3
3
K
63.6
[28.9]
SmCA*
80.3
[0.1]
SmC*
83.0
[1.3]
SmA*
85.1
[1.9]
Iso Liq
Iso Liq
Ps 170 nC cm22; h 41u; SmCA* threshold 15 V mm21
K
10
11
12
13
14
15
16
17
a
5
7
3
5
7
3
5
7
3
54.9
[22.5]
SmCA*
67.6
[0.03]
SmC*
81.7
[0.9]
SmA*
92.6
[1.9]
Ps 150 nC cm22; h 44u; SmCA* threshold 10 V mm21
K
3
61.5
[24.3]
SmC*
90.9
[0.6]
SmA*
SmC*
110.4
[2.4]
Iso Liq
SmA*
Ps 153 nC cm22; h 42u
K
6
,250
SmCA*
81.5
[1.2]a
82.9
[1.2]a
85.6
[0.8]
Iso Liq
Ps 130 nC cm22; h 42.5u; SmCA* threshold 10 V mm21
K
6
39.2
[27.5]
SmC*
90.4
[1.1]
SmA*
SmA*
SmA*
SmA*
SmA*
98.6
[1.3]
Iso Liq
Iso Liq
Iso Liq
Iso Liq
Iso Liq
Ps 120 nC cm22; h 44.5u
K
6
57.2
[27.8]
SmC*
100.6
[0.8]
114.2
[1.9]
Ps 115 nC cm22; h 45u
K
11
11
11
62.9
[9.2]
SmC*
97.8
[0.8]
110.2
[4.4]
Ps 92 nC cm22; h 33.5u
K
20.8
[15.2]
SmC*
100.0
[0.4]
115.4
[1.9]
Ps 83 nC cm22; h 34u
K
30.3
[31.1]
SmC*
106.9
[0.6]
125.5
[3.7]
Ps 85 nC cm22; h 40.5u
The enthalpy could not be determined due to overlapping peaks.
Table 3 The phase classification, phase transition temperatures (uC), enthalpies of transition [DH, kcal g21], maximum values of the spontaneous
polarization (Ps, nC cm22) and optical tilt angles (h, u) for compounds 18 to 20
Compound No.
n
18
3
K
51.9
[27.5]
SmCA*
98.8
[0.03]
SmC*
116.5
[1.1]
SmA*
SmA*
Iso Liq
128.8
[5.7]
Iso Liq
Iso Liq
Ps 187 nC cm22; h 43u; SmCA* threshold 14 V mm21
K
19
20
6
57.7
[32.2]
SmCA*
116.3
[0.03]
SmC*
123.4
[1.5]
129.3
[4.4]
Ps 180 nC cm22; h 43.5u; SmCA* threshold 10 V mm21
K
11
62.3
[61.6]
SmC*
114.4
[0.9]
SmA*
124.8
[6.6]
Ps 126 nC cm22; h 39u
A close examination of the transition temperatures and the
mesophase types in Table 2 shows that the compounds
predominantly exhibit smectic A* and smectic C* mesophases.
Compounds with short perfluoro chains (9 and 12) also
possess antiferroelectric phases, SmCA*, but as the perfluoro
chain length is increased the antiferroelectric phase is
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 2181–2191 | 2185

View Article Online
Table 4 The phase classification, phase transition temperatures (uC), enthalpies of transition [DH, kcal g21], maximum values of the spontaneous
polarization (Ps, nC cm22) and optical tilt angles (h, u) for compounds with ‘‘reversed-ring’’ arrangement of the core, 26 to 29
Compound No.
R
26
HCF2(CF2)3
K
51.0
[21.2]
SmCA*
76.1
[0.3]
SmC*
SmA*
SmA*
SmC*
80.7
[3.9]
Iso Liq
Iso Liq
Iso Liq
Iso Liq
SmCA* threshold . 20 V mm21
K
27
28
29
HCF2(CF2)5
52.0
[21.7]
SmC*
SmC*
SmCA*
84.7
[0.7]
87.7
[1.5]
Ps 166 nC cm22; h 45u
K
HCF2(CF2)7
65.2
[45.3]
94.5
[0.7]
105.7
[2.4]
Ps 157 nC cm22; h 42.5u
K
(CF3)2CF(CH2)2
99.0
[47.4]
101.2
[0.05]
121.7
[8.8]
SmCA* threshold . 20 V mm21
Table 5 Transition temperatures (uC) for known high tilt antiferro-
electric compounds.15 The values in parenthesis represent monotropic
transitions
This behaviour may also be examined in terms of the total
length of the perfluoro unit and the aliphatic spacer, as shown
in Fig. 3. It can be argued that at short combined chain lengths
there is increased rigidity of the terminal chain and as such the
antiferroelectric phase is encouraged by the interlayer interac-
tions occurring through the interface, but at longer aliphatic
chain lengths there is an increase in flexibility, thus allowing
decoupling of the mesogenic core from the interfaces, and
thereby the interfacial interactions play a much smaller role in
mesophase formation for these materials.
m
n
2
3
K 83.3 (SmIA* 54.0) SmCA* 121.3 SmC* 123.8 SmA*
128.9 Iso Liq
3
6
3
3
3
5
K 72.0 SmCA* 114.9 SmC* 123.7 SmA* 134.4 Iso Liq
K 68.6 SmCA* 97.5 SmC* 124.3 SmA* 153.5 Iso Liq
K 60.5 SmCA* 117.6 SmC* 123.7 SmA* 136.9 Iso Liq
Furthermore, if we assume that there is an increased overlap
of the perfluoro chains, which in turn are considered to be
rather rigid and polar groups, then at the interface the
overlapping interactions that are perpendicular to the layers
will promote the formation of smectic A* phases. Fig. 4 shows
the range of the smectic A* phase as a function of the total
m
2
3
4
6
n
3
3
3
3
K 66.6 (SmIA* 43.0) SmCA* 92.4 SmA* 117.3 Iso Liq
K 62.2 (SmIA* 31.9) SmCA* 92.8 SmA* 111.7 Iso Liq
K 73.6 (SmIA* 26.3) SmCA* 92.5 SmA* 109.1 Iso Liq
K 49.5 (SmIA* 25.0) SmCA* 89.9 SmC* 91.9 SmA*
105.4 Iso Liq
suppressed (see compounds 11, 13, 14, 16 and 17 in Table 2). It
is also possible to observe that the shorter aliphatic spacer
length enables the antiferroelectric phase to be present with
longer perfluoro chains, as shown by compound 10.
This behaviour may be explained by assuming that the
short perfluoro unit is partially interdigitated at the interface
between the layers and that the interface encourages the
formation of an antiferroelectric mesophase, however, as the
perfluoro unit is increased in length, such that the fluoro units
now resemble another separate core unit due to their rigidity,
the antiferroelectric phase is suppressed. The packing of the
bulky groups at the interface now frustrate the formation of an
antiferroelectric mesophase.
Fig. 3 Transition temperatures as a function of the total chain length
of the spacer and the terminal group for the linear semifluorinated
compounds 9–17.
2186 | J. Mater. Chem., 2006, 16, 2181–2191
This journal is ß The Royal Society of Chemistry 2006

View Article Online
clearing points and transition temperatures when compared to
the analogous compounds 9–11 and, in addition, they tend to
exhibit higher ordered smectic phases. This comparison shows
that the inclusion of a single hydrogen atom at the terminal
point of the perfluoro chain causes a large perturbation in the
interlayer organization and as a result the mesophase stability
of these compounds is significantly lower than those that are
purely alkyl or perfluoro. Examination of the perfluoro
compounds shows that as the perfluoro chain length is
increased, the antiferroelectric mesophase stability is reduced,
whereas for the alkyl compounds there appears to be very little
variation. The lowering of the mesophase stability of the
antiferroelectric phase for compounds 9–11, on increasing
fluoro chain length, shows a much more exaggerated reduction
than for the perfluoro compounds described by Dabrowski
et al.,15 which in the case of the materials we report can only be
due to the interactions associated with the proton at the end of
the fluorinated chain, i.e., they will be weaker than those
associated with the fluoro-substituent it replaces. Moreover, if
one considers that fluorocarbon and hydrocarbon systems do
not like to mix and tend to phase separate; in the Dabrowski
et al. materials there is the potential for the perfluoro end-
groups to microphase segregate to provide a fluorocarbon
environment but in the mixed system considered for com-
pounds 9–11 there is an enforced hydrocarbon element in the
middle of the fluorocarbon environment which would cause
disruptions within the microphase segregating regions. This
disruption of the interlayer region can be used to explain the
large temperature reduction for all of the phase transitions
when comparing compounds 9–11 to either the purely
fluorinated or purely aliphatic systems.
Fig. 4 Temperature range of the SmA* phase at different total chain
lengths. The circles represent a spacer length of n = 3, the triangles
represent a spacer length of n = 6 and the squares represent a spacer
length of n = 11.
chain length for aliphatic chain lengths of n = 3, 6 and 11.
Quite clearly, for all combined chain lengths, it can be seen
that with an increase in perfluoro chain length there is an
increase in the smectic A* phase temperature range. For the
compounds examined, the SmA* to Iso Liq transition appears
to increase linearly with increasing chain length. The largest
increase in the clearing point is observed with the short spacer
showing that when the core is closest to the interface, small
changes will have the largest effect on phase behaviour.
Examination of the branched fluorinated compounds,
18–20, Table 3, shows an increase in the stability of the
antiferroelectric mesophase relative to the linear compounds.
An increase in the spacer length from three methylene units
to six methylene units actually enhances the antiferroelectric
phase rather than reducing it, as seen for the linear com-
pounds. This result can be considered analogous to the
formation of the alternating smectic C phase shown by
compounds containing a branched or ‘‘swallow-tailed’’ end-
group, as described by Nishiyama and Goodby.8 This shows
that the interface and the arrangement of the molecular
fragments at the interface have a large effect on meso-
morphism. When the spacer group is increased to eleven
methylene units, however, there is sufficient decoupling of the
mesogenic core from the interlayer interface, and hence the
antiferroelectric phase no longer persists.
In order to fully appreciate the behaviour of the end-group
in the formation of smectic mesophases it is possible to
compare another series of materials11 shown in Fig. 5. In this
series of compounds the bulky alicyclic end-groups cause a
disruption to the interlayer interfacial region and as a result
the layers become weakly associated, with the result that
frustrated twist grain boundary phases are observed, replacing
the smectic A* phase. An increase in the size of the terminal
ring reduces the temperature range of the TGB phase until a
direct isotropic liquid to smectic C* transition is observed.
This result is in direct contrast to the behaviour of the
perfluoro end-group materials where there is an increase in the
smectic A* temperature range with an increase in the number
of perfluoromethyl units.
Another related series of compounds reported by Petrenko17
show that the inclusion of a terminal siloxane at the end of the
The compounds with a ‘‘reversed-core’’ system, 26–29,
Table 4, exhibit similar behaviour to the analogous com-
pounds 9–11 and 18, but interestingly there is a much lower
tendency for these compounds to exhibit smectic A* phases.
This suggests that the packing arrangement of the cores in
conjunction with the interface promotes tilted phases over
orthogonal phases.
There are interesting comparisons which need to be drawn
with compounds prepared by Dabrowski et al.15 A selection of
these compounds is shown in Table 5. Interestingly, both
families of compounds (with a perfluoro end-group and the
alkyl end-groups) described by Dabrowski, exhibit higher
Fig. 5 Structure of bulky-end-group compounds exhibiting TGB
phases.
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 2181–2191 | 2187

View Article Online
Fig. 8 Minimised model of compound 19.
Fig. 6 Siloxanes exhibiting smectic C* mesophases.
molecule also weakens the interlayer interactions with these
compounds only showing direct isotropic liquid to smectic C*
transitions. There is no evidence of a smectic A* phase
even upon increasing the length of the siloxane unit or by
shortening the spacer, see Fig. 6. The siloxane unit is con-
sidered to be a microphase segregating unit, but due to its size
and flexibility it disrupts the interlayer interface and weakens
the layers unlike the perfluoro unit which is relatively rigid and
strengthens the interfacial interactions and thereby the layers.
Fig. 9 Optical and X-ray tilt angles as a function of reduced
temperature for compound 19.
3.2 X-Ray diffraction
The minimized model for compound 19 is shown in Fig. 8
˚
and the average molecular length was calculated to be 41.2 A.
Compound 19 was studied by X-ray diffraction in order to
examine the effects of the branched terminal fluorinated chain
on the layer spacing. From the X-ray results the layer spacing
was determined and also the X-ray tilt angle was calculated,
allowing comparison with the optical tilt angle.
Using this result we can assume that the smectic A* layer
˚
spacing determined from the X-ray measurements of 34 A,
either represents an interdigitated smectic structure where the
˚
interlayer region possesses an approximate 7 A overlap, or that
The X-ray layer spacing for compound 19 is shown in
Fig. 7a. From the X-ray measurements we can determine that
˚
the smectic A* layer spacing is approximately 33.75 A and that
this type of compound has a de Vries type SmA* phase where
the molecules have random tilts within the smectic A* layers.
The X-ray tilt compared with the optical tilt angle is shown
in Fig. 9. There is a large difference between these two
measurements. The optical tilt angle is in the region of 43–44u,
however, the X-ray tilt is only in the region of 22–23u, which
suggests that the core is much more tilted within the layers
than the chains since the optical tilt is associated with the
there is a rapid reduction in the layer spacing upon transition
to the smectic C* phase. The layer spacing reduces to a
˚
minimum of 31.5 A in the antiferroelectric phase. From Fig. 7b
we can see that there is only a 4% reduction in the layer spacing
from the smectic A* phase to the smectic C* phase.
Fig. 7 a) X-Ray layer spacing for compound 19. b) X-Ray layer spacing/average SmA X-ray layer spacing (d/dA) for compound 19.
2188 | J. Mater. Chem., 2006, 16, 2181–2191
This journal is ß The Royal Society of Chemistry 2006

View Article Online
transition moments of the cores, whereas the X-ray tilt is
associated with the gross molecular structure. This result
would allow for an extensive interpenetration of the terminal
groups at the interlayer interface, or alternatively the ends of
the molecules are fixed and the core units become decoupled
from one layer to the next resulting in a large tilt angle.
3.3 Electro-optical measurements
3.3.1 Spontaneous polarization. The spontaneous polariza-
tion as a function of the reduced temperature for compounds
9–17 is shown in Fig. 10. For compound 9 there is a high
voltage threshold for switching in the antiferroelectric phase,
and consequently the measurements needed to be carried out
at 16 V mm21. This is another indication that the fluorinated
interlayer region acts as a strong interface, increasing the
stability of the antiferroelectric mesophase and requiring a
high energy to break the antiferroelectric organization during
the switching process.
Fig. 11 Spontaneous polarization as a function of reduced tempera-
ture for compounds 18–20.
As the length of the terminal perfluoro-group is increased,
the magnitude of the spontaneous polarization decreases. This
behaviour is consistent for all the different spacer lengths
that have been examined, i.e., 9 > 10 > 11, 12 > 13 > 14 and
15 > 16 > 17. Secondly, as the aliphatic spacer length is
increased there is also a reduction in the magnitude of
spontaneous polarization. This decrease in the polarization is
attributed to a volume effect, i.e., an undecyl spacer occupies a
greater volume than a propyl spacer. The reduction in the
polarization with an increase in perfluoro chain length could
also be attributed to a volume effect but it may also be due to a
subtractive effect of the fluoro-substituents to the molecular
dipoles associated with the generation of the polarization in
the smectic C* phase.
phase, but this phase could not be detected by DSC or
polarized optical microscopy (sandwiched between coverslip
and slide or as a free-standing film). The current peaks are
shown as a function of temperature in Fig. 12. This phase has
also been observed in similarly related compounds by Marzec
et al.18 in a situation where the phase could only be detected
during switching studies in an electro-optic cell.
For the reversed-core system, compounds 26–29, we find
that there is an incredibly high threshold to overcome in order
to observe any switching in the antiferroelectric mesophase.
This threshold is so high (>20 V mm21) that in the cases of
compounds 26 and 29 we could not align or switch the
materials in the 5 mm cells. For compounds 27 and 28 we
observe a drop in the magnitude of spontaneous polarization,
which is comparable with that observed for compounds 10
and 11. The spontaneous polarization for compounds 27 and
28 is shown in Fig. 13.
Examination of both the branched system and the
‘‘reversed-core’’ system reveals identical behaviour. In the case
of the branched system, see Fig. 11, we can see that the
spontaneous polarization drops as the length of the aliphatic
spacer is increased. There is some unusual behaviour which
is observed for these compounds upon transition from the
smectic A* phase to the smectic C* phase. On initial examina-
tion, the polarization current peaks resemble a smectic Ca
Comparison of the behaviour of the fluorinated materials
9–20 and 26–29 with the compounds that have a cyclic end-
group11 or the siloxane compounds17 show that the behaviour
of the terminal group at the interlayer interface has a large
effect on the switching behaviour. Compounds with a cyclic
Fig. 10 Spontaneous polarization as a function of reduced tempera-
ture for compounds 9–17.
Fig. 12 Current peaks observed on SmA*–SmC* transition for
compound 19.
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 2181–2191 | 2189

View Article Online
Fig. 13 Spontaneous polarization as a function of reduced tempera-
ture for compounds 27 and 28.
Fig. 15 Tilt angle as
compounds 18–20.
a
function of reduced temperature for
end-group exhibit conventional switching where there is a
single hysteresis loop observed for the ferroelectric switching,
whereas the siloxane materials exhibit V-shaped switching
behaviour where no hysteresis loop is observed. This suggests
that the fluorinated end-groups produce a strong interlayer
interaction, whereas the siloxane materials produce a dis-
organized or weak interlayer interaction.
in the figure because the tilt angle appears to reach a value of
45u but neither compound aligns well enough to be able to give
a definitive reproducible result.
The increase in tilt angle, caused by the increase of the
aliphatic spacer length from three to six methylene units,
reveals that there is a volume effect which affects the behaviour
of the spontaneous polarization. Usually an increase in the
tilt angle would lead to an increase in the spontaneous
polarization, but for these compounds a large reduction in the
spontaneous polarization is observed upon increasing the
aliphatic spacer length.
3.3.2 Tilt angle. The tilt angle measurements for compounds
9 to 17 are shown in Fig. 14. The tilt angle for each of the
different length fluorinated end-groups shows very little
difference although there is a tendency for the tilt to increase
for a mid-length fluorinated chain, and then to drop for long
fluorinated chains. As the length of the spacer is increased
from a short spacer of three methylene units (compounds
9–11) to the intermediate length of six methylene units
(compounds 12–14) an increase in the tilt angle is found.
Compounds 12–14 all exhibit tilt angles in the region of 45u
whereas the tilt angle for the longer chain length of eleven
methylene units (compounds 15–17) are all much lower, i.e.,
less than 40u. Data for compounds 10 and 14 are not included
Materials with the branched fluorinated end-group, 18–20,
show similar tilt angles for both the short and intermediate
spacer lengths, but there is a large reduction in tilt angle for the
long spacer length, see Fig. 15. This behaviour is consistent
with the results found for compounds 9–17.
The ‘‘reversed-core’’ materials 27 and 28 exhibit high tilt
angles of >42u (see Fig. 16) and show a reduction in tilt angle
with increasing fluorinated chain length, which is consistent
Fig. 14 Tilt angle as
compounds 9–17.
a function of reduced temperature for
Fig. 16 Tilt angle as
compounds 27 and 28.
a function of reduced temperature for
2190 | J. Mater. Chem., 2006, 16, 2181–2191
This journal is ß The Royal Society of Chemistry 2006

View Article Online
with the results obtained for compounds 9–11. The tilt angles
for these compounds are essentially identical to the analogous
compounds 10 and 11, however, there is a higher voltage
threshold for switching for compounds 27 and 28.
for the X-ray measurements. They would also like to thank
Kingston Chemicals for financial support.
References
1 R. B. Meyer, L. Liebert, L. Strzele and P. Keller, J. Phys. Lett.,
1975, 36, L69.
4. Conclusion
2 R. B. Meyer, Mol. Cryst. Liq. Cryst., 1975, 40, 33.
3 N. A. Clark and S. T. Lagerwall, Appl. Phys. Lett., 1980, 36, 899.
4 N. Hiji, A. D. L. Chandani, S. Nishiyama, Y. Ouchi, H. Takezoe
and A. Fukuda, Ferroelectrics, 1988, 85, 99.
5 S. J. Cowling, J. W. Goodby and K. J. Toyne, J. Mater. Chem.,
2001, 11, 1590.
6 M. Hird, J. W. Goodby, N. Gough and K. J. Toyne, J. Mater.
Chem., 2001, 11, 2732.
7 D. J. Byron, L. Komitov, A. Matharu, I. McSherry and
R. C. Wilson, J. Mater. Chem., 1996, 6, 1871.
8 I. Nishiyama and J. W. Goodby, J. Mater. Chem., 1992, 2, 1015.
9 A. Fukuda, Proceeding of the 15th International Display Research
Conference, 1995, 61, 177; S. Inui, N. Imura, T. Suzuki, H. Iwane,
K. Miyachi, Y. Takanishi and A. Fukuda, J. Mater. Chem., 1996,
6, 671.
We have prepared three series of partially fluorinated and
branched fluorinated materials in order to investigate the
effects of the end-group on the liquid crystal properties and the
physical properties of the materials. From this work we can
make the following conclusions:
Antiferroelectric mesophases are only evident when there is
a short fluorinated chain in the terminal position.
Long aliphatic spacer groups between the mesogenic core
and the fluorinated end-group do not promote antiferroelectric
mesophases due to decoupling of the core from the interfacial
interactions.
10 G. Y. Cosquer, S. J. Cowling, J. W. Goodby, K. J. Toyne,
N. Gough and M. Hird, Patent, WO03024903, 2003; G.Y. Cosquer,
S. J. Cowling, J. W. Goodby, K. J. Toyne, N. Gough and M. Hird,
Patent, WO03040812, 2003; S. J. Cowling, A. W. Hall and
J. W. Goodby Patent P103296WO, 2004.
An increase in the fluorinated chain length enhances the
smectic A* temperature range. The largest effect is observed
for short methylene spacer groups.
Comparison of the partially fluorinated end-groups to fully-
fluorinated and fully-hydrocarbon analogues shows that
the mixed fluorocarbon/hydrocarbon system has caused a
perturbation in the interlayer organization which results in
reduced mesophase stability.
11 S. J. Cowling, A. W. Hall and J. W. Goodby, Liq. Cryst., 2005, 32,
11–12, 1483.
12 K. D’Have, P. Rudquist, S. T. Lagerwall, M. Matuszczyk,
H. Pauwels and R. Dabrowski, Proc. SPIE-Int. Soc. Opt. Eng.,
2000, 3955, 33; S. T. Lagerwall, A. Dahlgren, P. Jagemalm,
P. Rudquist, K. D’Have, H. Pauwels, R. Dabrowski and
W. Drzewinski, Adv. Funct. Mater., 2001, 11, 2, 87.
13 A. Fafara, B. Gestblom, S. Wro´bel, R. Dabrowski, W. Drzewin´ski,
D. Kilian and W. Haase, Ferroelectrics, 1998, 212, 79.
The relative strength of the interlayer interactions of the
fluorinated end-groups is shown by comparison to cyclic and
siloxane end-groups. The fluorinated system promotes smectic
A* character whereas the cyclic groups show TGBA/C phases
and the siloxanes show no smectic A* character at all.
´
´
14 W. Drzewinski, K. Czuprynski, R. Dabrowski and M. Neubert,
Mol. Cryst. Liq. Cryst., 1999, 328, 401.
´
´
15 W. Drzewinski, R. Dabrowski and K. Czuprynski, Polish J. Chem.,
2002, 76, 273.
16 M. Negishi, K. Kanie, T. Ikeda and T. Hiyama, Chem. Lett., 1996,
8, 583.
Acknowledgements
17 A. S. Petrenko, PhD Thesis, University of Hull, 2003.
18 M. Marzec, A. Mikulko, S. Wro´bel, R. Da˛browski, M. Darius and
W. Haase, Liq. Cryst., 2004, 31, 2, 153.
The authors would like to thank EPSRC for funding this work
and Daresbury Laboratories, UK for the use of station 2.1
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 2181–2191 | 2191