Hierarchical organisation in shape-amphiphilic liquid crystals

Article abstract of DOI:10.1039/b819877b

Shape-amphiphilic liquid crystals offer a route towards complex assemblies at the difficult 10 nanometer length scale. We present the preparation and mesomorphic characterisation of three novel shape amphiphiles based on azobenzenes. Their mesophases rang

Full text of DOI:10.1039/b819877b

View Article Online / Journal Homepage / Table of Contents for this issue
www.rsc.org/materials
Volume 19 | Number 11 | 21 March 2009 | Pages 1505–1660
ISSN 0959-9428
PAPER
FEATURE ARTICLE
Paul H. J. Kouwer and Georg H. Mehl Slobodan Ž umer et al.
Hierarchical organisation in shape-
amphiphilic liquid crystals
Liquid crystal elastomer–nanoparticle
systems for actuation
0959-9428(2009)19:11;1-Y

PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Hierarchical organisation in shape-amphiphilic liquid crystals†
*
Paul H. J. Kouwer‡ and Georg H. Mehl
Received 7th November 2008, Accepted 22nd December 2008
First published as an Advance Article on the web 9th February 2009
DOI: 10.1039/b819877b
Shape-amphiphilic liquid crystals offer a route towards complex assemblies at the difficult 10
nanometer length scale. We present the preparation and mesomorphic characterisation of three novel
shape amphiphiles based on azobenzenes. Their mesophases range from a simple nematic to complex
lamellar phases of alternating layers of discotic and calamitic mesogens. In all smectic phases, the
discotic and calamitic moieties are nanophase separated. As the different sub-layers start to tilt and/or
display in-layer order independently, the mesophase assignment becomes more complex. Using
temperature dependent X-ray diffraction we studied their characteristics.
a variety of mesophases, induced by the combination of disc-
Introduction
shaped and rod-shaped liquid crystals.6a,d In this contribution, we
will discuss the organisation of a series of novel of shape-
amphiphilic mesogens based on azobenzenes (see Fig. 1). Azo-
benzenes are well known to display cis–trans isomerisation upon
irradiation with UV and visible light. In the field of liquid crys-
tals, this property has been successfully applied to prepare pho-
toalignment layers, but the versatility and easy accessibility of
azobenzenes allow their application in many other fields
requiring optical switching, for instance displays and optical
In liquid crystals order is combined with fluidity. The organisa-
tion in liquid crystals can be tuned at the molecular scale, but also
much beyond that, on the macroscopic level, by using both
alignment layers as well as external stimuli, such as magnetic,
electric and optical fields. The control at multiple length scales
makes liquid crystals ideal candidates for their use as active
components or scaffolds in self-assembly processes.
Control over the organisation at the nanometer level has been
targeted by many different routes. A classical approach is the
combination of two incompatible units, inducing phase separa-
tion at the sub-molecular level. Incompatibility of ionic head
groups and one or multiple aliphatic tails1,2 forms the basis of self-
organisation schemes of surfactants, which find wide application
in industry and daily life, but also form the building blocks of
complex systems in Nature. Over time, other schemes for nano-
phase segregation have also been developed, such as rigid/flexible
systems (among which there are many thousands of liquid crys-
tals), polar/apolar3 and hydrocarbon/fluorocarbon4 materials, as
well as systems based on the differences of the shapes of the
(sub)molecular units. Shape-amphiphilicity5 is currently receiving
increased attention. After it was shown that an appropriate choice
of the molecular disc-rod geometry yielded rich phase behaviour,6
an increasing number of reports were published.7
A particularly attractive aspect of disc-rod geometries is that
order at the 10 nm scale can be established. Although important
for many photophysical processes, this length scale remains hard
to target by either bottom-up or top-down approaches. The
smectic disc-rod systems introduce two chemically different
moieties that can carry individual functions. In addition, their
order can be tuned independently. This forms an interesting basis
for the design of complex self-assembly in bulk materials.
Previously, we have shown a number of examples that show
Department of Chemistry, University of Hull, Cottingham Road, Hull, UK
HU6 7RX. E-mail: p.kouwer@science.ru.nl; g.h.mehl@hull.ac.uk
† Electronic supplementary information (ESI) available: Synthesis details
and characterisation. See DOI: 10.1039/b819877b
‡ Current address: Institute for Molecules and Materials, Radboud
University Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The
Netherlands. Fax: +31 24 3652929; Tel: +31 24 365 2464.
Fig. 1 Investigated disc-rod mesogens 1–3.
1564 | J. Mater. Chem., 2009, 19, 1564–1575
This journal is ª The Royal Society of Chemistry 2009

recording.8 Key in many applications, however, is the control of
the organisation in these materials on the nanometer level. Our
results show that highly organised systems can be achieved, while
a high molecular mobility is maintained. The order in these
systems can be fine-tuned by changing the discotic group and its
functionality.
Results and discussion
Synthesis
The synthesis of the linked disc rod mesogens proceeded in
a modular manner. Separately, the functionalised disc-shaped
and rod-shaped mesogens were prepared. The calamitic moieties
were linked using a gallic acid group and, in the final step of the
synthesis, this trimer was coupled to the discotic mesogens. This
modular approach was designed to generate a substantial
number of shape-amphiphilic liquid crystal dimers from a limited
number of starting materials, allowing a systematic study of the
properties of this relatively new class of materials. Synthesis
procedures and characterisation are provided in the ESI.†
The azobenzene mesogen was obtained in high yields by
reacting the diazonium salt of 4-pentylaniline with phenol, see
Scheme 1. The product was alkylated with an excess of dibro-
mohexane under straightforward Williamson etherification
conditions to yield 6. Subsequent etherification under similar
conditions with ethyl gallate yielded the ‘‘rod-trimer’’ mesogen 7,
which was fully characterised. Saponification of the carboxylic
ester group under basic conditions yielded the free carboxylic
acid 8 quantitatively. The poor solubility of 8 in a wide variety of
organic solvents limited our ability to characterise the product
and, hence, 8 was used in subsequent steps without further
purification.
Scheme 2 Synthesis of the bifunctional triphenylene 13. Key: (i)
Br(CH2)11OH, K2CO3, KI, butanone, heat to reflux 40 hrs. (ii) AcCl,
C5H5N, CH2Cl2, room temperature for 70 hrs, 67% over 2 steps. (iii) 1.
FeCl3, CH2Cl2, room temperature for 20 hrs; 2. dry MeOH, 62%. (iv)
Ethanol, para-toluenesulfonic acid, heat to reflux for 2 hrs, 98%.
alcohol functionalities were protected by acetyl groups from the
oxidising conditions of the triphenylene reaction. Tetrahexyloxy-
biphenyl was prepared according to literature procedures.6d
Oxidative coupling of 11 and 12 and subsequent removal of the
acetyl protecting groups under acidic conditions yielded the
bifunctional triphenylene 13.
The synthesis of discotic mesogens 14 and 15 was described
previously.6d,10 Coupling of 13–15 with the linked rod-shaped
mesogens 8 gave the target compounds 1–3 in moderate yields,
see Scheme 3. We found that the linked disc-rods are readily
soluble and straightforward to purify using standard column
chromatography techniques
Thermal behaviour
The bifunctional triphenylene was prepared via the classical
biphenyl route,9 see Scheme 2. Catechol was equipped with two
undecyl spacers via a Williamson etherification and the aliphatic
The thermal behaviour of the novel linked disc-rod mesogens
and their precursors was studied by differential scanning calo-
rimetry (DSC, Fig. 2) and optical polarising microscopy (OPM,
Fig. 3). The results are summarised in Table 1. Note that the
mesomorphic properties of the mono-functional disc-shaped
mesogens 146d and 1510 were published previously, but have been
included here for comparison. Monofunctional triphenylene 14
exhibits a monotropic columnar phase, while 15 displays
a broad, enantiotropic nematic phase. The bifunctional discotic
mesogen 13 is not liquid crystalline and shows a melting
Scheme 1 Synthesis of the linked-rod mesogens 7 and 8. Key: (i) 1.
NaNO2, aq. HCl, 0 ꢀC 1 hr, room temperature 1 hr; 2. PhOH, aq. NaOH,
addition at room temperature, then heat to reflux 4 hrs, 86%. (ii)
Br(CH2)6Br (excess), K2CO3, KI, butanone, heat reflux 16 hrs, 74%. (iii)
K2CO3, KI, butanone, heat to reflux 16 hrs, 62%. (iv) 1. aq. KOH MeOH/
THF (1:1), heat to reflux 4 hrs; 2. aq. HCl, quantitative.
Scheme 3 Schematic representation of the synthesis of the target
compounds 1–3. Key: (i) DCC, DMAP, pTSA, CH2Cl2, room tempera-
ture, 3–5 days, 42–65%.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1564–1575 | 1565

Fig. 2 Normalised DSC traces of precursors 6 (blue) and 7 (green); and
final products 1 (red, magnified 2 ꢂ); 2 (orange) and 3 (purple). The
(broad) glass transitions of 7 and 3 are visualised by magnifications of the
DSC signal. The arrows indicate (small) transitions. Phases are indicated
in the figure (cooling traces), monotropic phases between square
brackets.
Fig. 3 Optical polarising microscopy photographs of (a) 1 at T ¼ 80 ꢀC
(N phase; marbled texture); (b) 2 at T ¼ 75 ꢀC (SmA phase; a grainy
texture right and a homeotropically aligned texture left); (c) 2 at T ¼ 65
ꢀC (SmX phase; light birefringent grainy texture); and (d) 3 at T ¼ 92 ꢀC
(SmA phase; focal conic texture).
temperature of 82 ꢀC. Both azobenzene precursors 6 (monomer)
and 7 (trimer) show, on cooling from the isotropic phase,
a narrow nematic range, followed by a SmA phase. The nematic
phase of both materials is characterised by an optical schlieren
texture and a small latent heat value at the nematic–isotropic
mesogens that we prepared so far.6a,d,e At lower temperatures
a highly organised lamellar structure is obtained, most likely
a soft crystalline phase. Introduction of the triphenylene-based
mesogen in the disc-rod mesogen changes the phase behaviour
dramatically. The triphenylene moiety, usually organised in
two-dimensional positionally ordered columnar mesophases,
excludes the formation of the nematic phase entirely. Instead,
a SmA phase is observed for both 2 and 3. Microscopy experi-
ments show a grainy texture, even after annealing (Fig. 3b and c),
in combination with the formation of homeotropic areas. The
large latent heat values at the smectic to isotropic transition
(DHSmA–I ¼ 24.3 and 45.8 kJ molꢁ1 for 2 and 3, respectively) are
indicators of the exceptionally high degree of order established in
these optically conventional SmA phases. X-Ray diffraction
measurements, discussed later in the paper, confirm this.
phase transition: for 6 DHN–I ¼ 0.9 kJ molꢁ1 and for 7 DHN–I
¼
2.1 kJ molꢁ1 for three mesogens, which corresponds to 0.7 kJ per
mol mesogen. Also, the latent heat at the SmA–N transition is
roughly three times larger for trimer 7 than for 6. At lower
temperatures 7 shows a small transition to a more ordered
smectic phase and below room temperature a second order
transition to a glassy phase, which effectively freezes in the order
of the SmX phase. Monomer 6 crystallises at low temperatures
and exhibits on cooling a similar, but monotropic, SmX phase
due to slow crystallisation kinetics. Complete characterisation of
the organisation in the higher ordered smectic phases of 6 and 7
has not been pursued.
Mixing studies of the nematic mesophases
The linked disc-rod mesogen 1, with the strongly nematogenic
discotic mesogen attached, shows a narrow nematic phase at
elevated temperatures. OPM experiments of the nematic phase
initially show a grainy texture; however, after annealing over-
night, just below the clearing temperature, the domains grow and
a marbled texture can be observed (Fig. 3a). DSC experiments
show the characteristically small latent heat at the clearing
temperature, which was observed for all nematic linked disc-rod
Optical microscopy studies of contact samples of 1, 7 and 15
provided insight into the miscibility of the disc and rod-shaped
species. Contact samples of 7 (calamitic) and 15 (discotic)
showed homogeneous mixing in the isotropic phase at elevated
temperatures. Upon cooling a broad region remained isotropic
between the bulk mesophases of the two pure species. The
isotropic range persisted at room temperature. This result shows
1566 | J. Mater. Chem., 2009, 19, 1564–1575
This journal is ª The Royal Society of Chemistry 2009

Table 1 Mesomorphic properties of the final products and the precursors
Mesophase behavioura b
,
Mesogen
1
2
3
6
CrX
Cr
GSmX
Cr
GSmX
Cr
Cr
72 (12.5)
43 (24.1)
5
65 (29.2)
5
82 (57)
55 (49)
132 (37)
N
81 (0.3)
I
I
I
I
I
I
I
I
[SmY 43 (<0.1)]c
[Colh 51 (3.7)]c
SmX
SmX
69 (1.7)
70 (4.4)
SmA
SmA
SmA
SmA
81 (24.3)
101 (45.8)
79 (3.4)
N
N
84 (0.9)
111 (2.1)
7
SmX
72 (2.3)
104 (8.1)
13
14
15
Cr
N
246 (0.2)
a
b
Transition temperatures in ꢀC and corresponding latent heat values (between brackets) in kJ molꢁ1
.
Abbreviations: Cr ¼ crystalline; GSmX ¼ glassy
phase with SmX texture frozen in; SmX/SmY ¼ unknown smectic phase with in-layer positional order (see discussion); SmA ¼ smectic A; N ¼ nematic; I
¼ isotropic. c Monotropic phase transition.
that, although 7 and 15 are miscible over the entire phase
diagram, the formation of a nematic phase, or any other ordered
phase for that matter, is strongly destabilised.
The monotropic phase behaviour of these mixtures hampered
further investigation of the phase diagram in greater detail, but
from the information available from the contact studies, a qual-
itative phase diagram can be sketched, see Fig. 5.
Contact samples of 1 and 7 showed (nearly) linear mixing
behaviour, similar to what was observed for cyanobiphenyl
systems.6b,c Also similar to these systems are the results of mixing
studies between the linked disc-rod mesogen 1 with the discotic
nematogen 15. In the centre of the phase diagram a minimum in
the clearing temperature is found. By cooling the sample from the
isotropic phase, initially a wide isotropic gap is observed, which
slowly grows smaller by movement of both nematic–isotropic
phase boundaries towards each other, see Fig. 4a–d. At T ¼ 61
ꢀC, both nematic phases meet and a homogeneously mixed
nematic phase is formed,6b,c see Fig. 4e. For 1/15, however, the
nematic phase at this concentration is metastable and crystallises
upon annealing, clearly seen in the reheated sample in Fig. 4f.
The crystals appear exactly in the area of the minimum in the
clearing temperature and slowly grow outward. After annealing,
the crystalline phase melts in the reverse order; slowly from the
sides to the centre area. A maximum clearing temperature of
91 ꢀC is detected.
X-Ray diffraction
To further investigate the organisation of the new class of disc-
rod liquid crystals, the materials were subjected to detailed X-ray
diffraction (XRD) studies. The integrated diffractograms are
shown as a function of the length of the scattering vector q:
4p sin q
q h jqj ¼
(1)
nl
Here d is the spacing, q is half the diffraction angle, n is an integer
˚
and l is the wavelength (l ¼ 1.54 A). Spacings were calculated
from eq. (1) after fitting the intensities and determining the local
maxima. Sharp peaks (Cr, SmX and small angle reflections in the
SmA phase) were satisfactorily fitted with Gaussian distribution
functions and diffuse peaks were fitted with Lorentzian func-
tions. From the fits the half-width at half maximum (HWHM) of
Fig. 4 Optical polarising microscopy photographs of contact samples of 1 (left) and 15 (right). Upon cooling from the isotropic phase with (rate: 1 ꢀC
minꢁ1): (a) T ¼ 78 ꢀC; (b) T ¼ 75 ꢀC; (c) T ¼ 70 ꢀC; (d) T ¼ 65 ꢀC; and (e) T ¼ 61 ꢀC. All pictures taken at the same spot in the sample. Note that at lower
temperatures the bulk of 15 and later also of 1 shows dark crystalline patches in the microscope. (f) The sample was annealed (2 hrs at 60 ꢀC) and
reheated to 78 ꢀC, showing the monotropic crystals formed in the mixture sided by the isotropic regions of the mixture at different concentrations.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1564–1575 | 1567

Fig. 5 Sketched phase diagrams of (left) 1/7 and (right) 1/15. Note that the relative fraction x was only estimated from contact sample experiments. 1/7
shows linear mixing between the two entities. The phase diagram of 1 and 15 shows a minimum in the clearing temperature at intermediate ratios and the
induction of a crystalline phase with a maximum at a similar concentration. The latter can be supercooled giving rise to a narrow monotropic nematic
phase (shaded area).
every reflection was determined, which is related to the extent of
order in the mesophase, expressed by the correlation length x that
can be calculated using the Scherrer equation:
from calculated molecular lengths is an indication of the validity
of this analytical approach.
In the higher ordered phases of 2 and 3, the observed layer
spacing showed a linear dependence on the temperature (equa-
tion 6), yielding values for both a and b:
x ¼ 2p/Dq
(2)
where Dq is the half width of the reflection in reciprocal space.
Temperature dependent XRD measurements allowed us to
analyse the evolution of the layer spacing in the different liquid
crystal phases. The SmA phases of 2 and 3, as well as the SmX
phase of 1, could be fitted satisfactorily with the empirically
derived equation (3), commonly used to describe the temperature
dependence of the tilt angle qt of SmC phases:11
d(T) ¼ a + bT
(6)
The results of the X-ray diffraction measurements and the
analysis will be discussed for all three final products individually.
The powder patterns of 1 in the isotropic, nematic and soft
crystal phase are shown in Fig. 6. Fig. 7a shows the evolution of
the radially integrated diffractograms with temperature. Quan-
titative results are summarised in Table 2. The diffraction
pattern of 1 in the nematic phase shows weak alignment to the
applied magnetic field. The pattern shows three diffuse halos at
ꢀ
ꢁ
z
TSm-Iso ꢁ T
qtðTÞ ¼ k
(3)
TSm-Iso
Here k, an empirical material parameter, is a fitting constant,
TSm/Iso is the transition temperature of the tilted smectic phase
(commonly the SmC phase) and z is the numerically determined
critical exponent z ¼ 0.35.11 Equation (4) defines the tilting angle
as the ratio between the experimentally observed layer spacing
d(T) and the layer spacing at zero tilt angle dq ¼ 0, i.e. the layer
spacing that corresponds to the SmA phase and, in the case of
t
limited interdigitation, to the molecular length.
dðTÞ
cos½qtðTÞꢃ ¼
(4)
d
qt ¼ 0
By combining equations (3) and (4) an expression for the layer
spacing as a function temperature is obtained:
"
#
ꢀ
ꢁ
z
TSm-Iso ꢁ T
dðTÞ ¼ dq ¼ 0 cos k
(5)
t
TSm-Iso
Equation (5) can be fitted to the experimental results from
temperature dependent X-ray diffraction experiments using the
free fitting parameters k and dq ¼ 0. Although during the fitting
t
procedure of the layer spacing at zero tilt (dq ¼ 0) was set as a free
t
fitting parameter, the resulting values for the molecular length
(or twice the molecular length) have to match those obtained
from molecular modelling studies. This correlation between layer
spacings obtained from fits of the XRD data and those gained
Fig. 6 Optical diffraction patterns of 1 at (a) 70 ꢀC (CrX phase); (b) 80
ꢀC (N phase) and (c) 100 ꢀC (I phase).
1568 | J. Mater. Chem., 2009, 19, 1564–1575
This journal is ª The Royal Society of Chemistry 2009

Table 2 X-Ray diffraction data of 1 in the N and the CrX phase
T/ꢀC
Phase
N
Reflectiona
Peak shapeb
q/nmꢁ1
˚
d/A
80
A
B
d
d
d
s
s
s
s
s
d
s
s
2.2
4.3
29.0
14.6
4.5
55.0
37.0
27.8
22.2
15.9
4.4
Cd
002
003
004
005
007
Cd
Cs
14.0
1.14
1.70
2.26
2.82
3.96
14.4
14.7
15.7
70
CrX
4.3
4.0
D
a
Reflections: Miller indices are used for layer reflections; diffuse
reflections A–D are assigned: A, B ¼ small angle reflection in the N
phase (see inset in Fig. 7b); Cs/Cd ¼ sharp/diffuse wide angle reflection
in the SmX phase, always diffuse in the N and I phase; D ¼ small wide
b
angle reflection attributed to the face-to-face disc–disc distance. d ¼
diffuse, s ¼ sharp.
˚
than the total size of the dyad (ꢄ55 A as obtained from molecular
modelling experiments) indicates that the discotic and calamitic
components are homogeneously mixed in the nematic phase
and no nanophase segregation between the two components
occurs.
Upon cooling the sample, a highly ordered lamellar phase
develops that shows multiple harmonics in the small angle area,
see Fig. 7b. These (00l) reflections and the absence of cross
reflections highlight the strong one-dimensionality of the orga-
nisation in this phase. The non-uniform azimuthal distribution of
the small angle reflections (Fig. 6a) shows that the alignment
obtained in the high temperature nematic phase is retained. At
wide angles the diffractogram is dominated by a single sharp
reflection. Fitting of the experimental data reveals also a diffuse
component under this peak. In addition, a shoulder is observed
˚
at 4 A that was assigned to the face-to-face disc–disc distance
(reflection D). Despite appearing rather sharp, its spacing and
corresponding correlation length have characteristic values of
a disordered (nematic-like) organisation.12 The evolution of the
layer spacing with temperature (Fig. 7c) shows a small but
significant change. After fitting the experimental data to equa-
tion (5) an overall 10ꢀ tilt angle (averaged over the entire mole-
cule) could be determined.
Fig. 7 (a) Integrated diffractograms of 1 as a function of temperature.
Colour coding: red: isotropic; blue: nematic; orange: CrX. (b) Integrated
diffraction patterns at T ¼ 80 ꢀC (blue) and T ¼ 70 ꢀC (orange). The inset
shows the magnification of the two small angle reflections in the nematic
phase, fitted by two Lorentzian distribution functions. (c) Temperature
dependence of the layer spacing (open circles, left y-axis) and the corre-
lation length (filled triangles, right y-axis). The solid line though the
experimental data points represent a fit to equation (5).
In the CrX phase, the system is clearly highly organised, but
has no three-dimensional positional order, required for a formal
crystal phase. The correlation length of the layer spacing is
constant, but not very high (x002/d002 ¼ 6.9), which indicates that
the organisation in the CrX phase is more or less frozen in and
that the molecules have a limited mobility. In this sense, the CrX
phase may very well be a one-dimensional equivalent to the two-
dimensional columnar plastic phase, which is frequently
observed for discotic liquid crystals. Fig. 8 summarises the
structures of the two phases observed for 1.
˚
small and wide angles (Fig. 6b). The wide angle reflection at 4.5 A
is attributed to the average distance between alkyl tails, convo-
luted with the distances along the short axes of the mesogenic
˚
components in the dyad (typically 4.5 A for calamitics and
˚
4.0–4.5 A for discotics in the nematic phase). Two small angle
The magnetic field in our XRD setup was too weak to align the
SmA phase displayed by 2 and only azimuthally isotropic
diffraction rings were obtained. The radially integrated spectra
as a function of temperature are shown in Fig. 9; a summary of
the reflections in the different phases is given in Table 3. In the
isotropic phase, 2 shows only two diffuse halos.
halos are observed (see the inset in Fig. 7b) with spacings of
˚
29.0 and 14.6 A, respectively. Note that their different alignment
˚
to the magnetic field shows that B (14.6 A) is not a higher
harmonic of A, which also would not agree with a nematic
organisation. The fact that both reflections are much smaller
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1564–1575 | 1569

Table 3 X-Ray diffraction data of 2 in both SmA phases and the higher
ordered SmX and SmY phases
T/ꢀC Phase
Reflectiona Peak shapeb q/nmꢁ1 d/A
˚
81
79
interdigitated SmAd 001
s
s
s
s
d
d
s
s
s
s
d
s
d
s
s
s
s
d
s
d
s
0.95
1.87
2.79
3.71
3.89
66.1
33.6
22.5
16.9
16.2
4.3
58.4
39.8
29.7
23.8
17.1
17.0
4.3
58.2
39.8
29.4
23.8
17.4
16.9
4.4
002
003
004
B
C
14.5
bilayer SmA2
002c
003
004
005
B
1.08
1.58
2.12
2.65
3.66
3.71
007
C
14.5
51
SmX
002
003
004
005
B
1.08
1.58
2.14
2.64
3.62
3.72
14.4
14.7
1.09
1.58
2.14
2.65
3.63
3.72
007
Cd
Cs
˚
Fig. 8 Sketch of the order in the mesophases of 1 (l ¼ 55 A). (a) Mixed
nematic. Although biaxiality is likely at small length scales as a result of
the large aspect rations of the components, we have no direct evidence for
the formation of a biaxial nematic phase. (b) CrX: bilayered lamellar
phase with in-layer organisation of both the calamitic and the discotic
4.27
40
SmY
002
003
004
005
B
s
s
s
s
d
s
d
s
57.7
39.7
29.3
23.7
17.3
16.9
4.3
˚
species (L ¼ 110.9 A). We have insufficient evidence to reliably discuss the
tilt angles observed in the different layers, their correlation between
subsequent layers or the correlation of the columnar phase normal
(perpendicular to the smectic phase normal) between different layers.
007
Cd
Cs
14.5
14.7
4.27
a
Reflections: Miller indices are used for layer reflections; diffuse
reflections B and C are assigned: B ¼ reflection in layers of discotic
moieties in SmA or SmX phase; C ¼ wide angle reflection with
b
a diffuse (Cd) or sharp (Cs) component. d ¼ diffuse, s ¼ sharp.
c
Additional SAXS measurements elucidated the (small) fundamental
reflection as well: q001 ¼ 0.54 nmꢁ1 and d001 ¼ 116 A (T ¼ 79 ꢀC).
˚
The difference with the SmA phase is striking. Many high
order reflections are observable in the small angle area. The
diffuse reflection in the wide angle area confirms the absence of
long-range in-layer order, confirming the phase structure to be
a SmA phase. A zoom of the small angle area (Fig. 9b) shows
clearly the higher harmonics. Under the sharp reflections in the
small angle area, a diffuse reflection, indexed B, is observed
which is attributed to the packing of the discotic groups within
the layer (best seen in Fig. 10). The diffuse character of this
reflection highlights the short range positional order that is
associated with the organisation of the discotic moieties. In other
words, the two-dimensional organisation of the discotic moieties
inside the layered structure is either isotropic or, more likely due
to their large aspect ratio, nematic-like. Either way, the organi-
sation will always be confined between two layers of calamitic
mesogens and communication (i.e. long-range orientational
order) between two adjacent layers of discotics seems unlikely.
Interestingly, the (006) reflection is entirely missing from the
diffractogram, which we, at this moment, cannot explain by any
means other than a non-systematic absence from the diffraction
experiment. It is, however, noted that the (006) reflection is
missing in the SmX phase of dyad 1 too.13 The fundamental (001)
Fig. 9 XRD results of 2. Integrated diffractograms of 2 as a function of
temperature: (a) full pattern; (b) zoom of the small angle area. Colour
coding: red: isotropic; light blue: SmAd; blue: SmA2; green: SmX; purple
SmY (monotropic); orange: crystal.
˚
reflection corresponds to a spacing >100 A and is positioned
˚
behind the beam stop of our setup. It was observed at 117 A
1570 | J. Mater. Chem., 2009, 19, 1564–1575
This journal is ª The Royal Society of Chemistry 2009

Fig. 11 Temperature dependence of the layer spacing (open circles, left
y-axis) and the correlation length (filled triangles, right y-axis). The
colour coding of Fig. 9 has been used again. The solid lines through the
experimental data points represent linear fits (soft crystal SmY and SmX
phases) and a fit using equation (5) in the SmA2 phase.
Fig. 11 shows how the layer spacing and the correlation length
develop with temperature. Phase transitions are easily recognised
as a discontinuity in the layer spacing and the related values for
the correlation lengths (or the temperature derivatives thereof).
Interestingly, at 81 ꢀC in the SmA phase, the layer spacing shows
a maximum, after which it gradually decreases on cooling; the
common behaviour typically observed for a tilted phase. In fact,
the experimental data of the (narrow) SmA2 phase is best fitted
with equation (5) that is often used to describe SmC phases. In
the lower temperature SmX and monotropic SmY phase, the
layer spacing shows a linear relation with temperature. The
inflection point in d002 and x002 clearly marks the phase transition
that initially was missed in the DSC experiments because of its
small latent heat.
Fig. 10 Integrated diffraction patterns at T ¼ 79 ꢀC (top), T ¼ 80 ꢀC
(middle, both SmA phases coexisting) and T ¼ 81 ꢀC, all fitted by
multiple Gaussian functions as well as one Lorentzian function (at q ꢄ
3.8 nmꢁ1). The pattern at 80 ꢀC is a mixture of the SmA2 phase at 79 ꢀC
and the SmAd at 81 ꢀC. The solid lines represent the fits of the dif-
fractograms, all obtained by a single multi-peak fitting procedure.14
(T ¼ 79 ꢀC) during experiments recorded with a SAXS setup,
albeit low in intensity compared to the (002) reflection.
The soft-crystal CrZ phase at low temperatures is charac-
terised by a number of additional reflections that do not corre-
spond to a simple lamellar structure. In this phase the five (00l)
reflections remain present, albeit with different intensities, indi-
cating that this phase also possesses a strongly layered character.
The limited number of cross reflections, however, do not allow us
to index and characterise this phase further. The disappearance
A further interesting structural change was detected close to
the transition from the isotropic liquid. As the SmA phase is
formed at 81 ꢀC, peak positions of the sꢀmall angle reflections are
remarkably different from those at 79 C and at lower temper-
atures, see Fig. 10. Analysis shows that the structure of the SmA
phase changes rapidly on cooling. The layer spacing at 80 ꢀC was
˚
of the diffuse reflection B at 17 A and the appearance of sharp
˚
found to be 64 A, which corresponds to a lamellar phase, where
reflections in this area point towards a change in the organisation
of the triphenylene moieties at the SmY–CrZ transition, which
we attribute to their organisation in columns. If one assumes that
the direction of column formation is uncorrelated between the
layers, a macroscopically highly disordered three-dimensional
structure is expected. The reduced spacing of these reflections
associated with the discs with lowering the temperature may
indicate some tilting of the discs with respect to the layer normal,
which in turn explains the continuing reduction of the layer
spacing, also in the three-dimensional ordered phase, see Fig. 11.
The phase behaviour of 2 is graphically summarised in Fig. 12.
In the SmAd and SmA2 phases, the system shows a well-defined
periodicity without in-layer positional order. The decrease in the
layer spacing upon cooling is a signature of tilt within one of the
layers (or both). This, in combination with the presence of
homeotropic optical textures, is a signature of De Vries-type SmA
phases,15 as will be discussed later. At the transition to the SmX
both the discotic and calamitic moieties are separated in sub-
layers but where the two classes of mesogenic groups are
completely interdigitated (SmAd). The SmA phase that forms at
ꢀ
˚
79 C and below has a layer spacing of 117 A, which corresponds
to a bilayered structure of stretched mesogens (SmA2). The
correlation lengths x of the lattice spacings of the SmAd and
˚
SmA2 phase are 457 and 359 A, respectively. With the difference
in the one dimensional lattice, this translates into 7.0 layers in the
SmAd phase and only 6.1 layers in the SmA2 phase. Although the
material displays a higher degree of (one-dimensional) ordering
in the SmAd phase, the substantial free volume it costs (based on
a simple molecular model) causes the system to reorganise into
the interdigitated bilayered phase at lower temperatures. This
transition is a clear marker of the subtle forces and interactions
that are involved in the phase formation in these shape-amphi-
philic liquid crystals.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1564–1575 | 1571

˚
Fig. 12 Sketch of the order in the mesophases of 2 (l ¼ 58 A). (a) SmAd: bilayer SmA phase wherein both layers are completely interdigitated (L ¼ 66
˚
˚
A). (b) SmA2: bilayer SmA phase without interdigitation (L ¼ 118.5 A). (c,d) SmX, SmY: bilayered lamellar phase with one of the layers showing
positional order and further tilting of the mesogens in the layer. (e) Suggested crystal phase. It is important to stress that for all smectic phases shown
here there is no correlation between the layers of the tilt angles of the mesogens or direction of the phase normal in the 2D ordered discotic layers.
phase, the layer of azobenzenes moieties shows in-layer positional
order, while the XRD evidence suggests that the organisation in
the layer of discotic mesogens does not change considerably.
In other words, this points toward an assembly behaviour where
the sub-layer of the calamitic mesogens exhibits a soft crystal
structure whereas the discotic sub-layer is still disordered or, at
most, oriented in nematic like domains. The rapid further decrease
in layer spacing in the monotropic SmY phase indicates an addi-
tional tilting process that could originate from the discotic
moieties. The Cr phase at room temperature shows in-layer
organisation in both the calamitic and the discotic layer.
The XRD pattern of 3 shows fewer features than observed for
the dimers 1 or 2. It was not possible to align 3 in a magnetic field
and as a result only azimuthally isotropic reflections were
observed. Radially integrated spectra at different temperatures
are shown in Fig. 13 and XRD data in different phases are given
in Table 4.
Table 4 X-ray diffraction data of 3 in the SmAd and the SmX phase
T/ꢀC
Phase
SmAd
Reflectiona
Peak shapeb
q/nmꢁ1
˚
d/A
90
002
004
006
B
s
s
s
d
d
s
s
d
s
1.35
2.70
3.9
46.4
23.3
15.9
15.5
4.4
45.6
23.0
15.8
15.4
4.4
4.05
14.4
1.38
2.74
3.97
4.09
14.4
14.82
C
24
SmX
002
004
B
006
Cd
Cs
s
d
Fig. 13 Integrated small angle diffractograms of 3 at 90 ꢀC (top) and 30
ꢀC (bottom). The diffractograms are fitted by multiple Gaussian func-
tions and two Lorentzian functions. The individual peaks fits are shown
in orange lines, the cumulative fit function as a solid red line. The inset
shows the small angle region as a function of temperature. Colour coding:
red: isotropic; blue: SmAd; green: SmX.
4.24
a
Reflections: Miller indices are used for layer reflections; diffuse
reflections B and C are assigned: B ¼ reflection in layers of discotic
moieties in SmA or SmX phase; C ¼ wide angle reflection with
a diffuse (Cd) or sharp (Cs) component. b d ¼ diffuse, s ¼ sharp.
1572 | J. Mater. Chem., 2009, 19, 1564–1575
This journal is ª The Royal Society of Chemistry 2009

In the SmA phase three layer spacings are observed, the first
˚
reflection (d ¼ 49 A) represents a spacing considerably smaller
˚
that the molecular length (l ¼ 58 A). The absence of a strong
fundamental reflection in 1 and 2 encouraged us to interpret the
XRD results of 3 analogously. This interpretation fits duly with
a bilayered structure, but now with only the discotic mesogens
completely interdigitated (SmAd). Considering that in 3 the cross
section of the discotic groups was designed to be approximately
half of that of the calamitic groups, this assembly model is very
plausible. The first strong reflection in the small angle area is then
the (002) reflection and further reflections are the higher
harmonics. It is remarkable that for 3 (and in contrast to other
disc-rod systems, presented here and before) only the even
harmonics are observed.16
Fig. 13 shows the integrated diffractogram at 90 ꢀC and the fits
of the reflections. As described earlier for 2, the small angle area
shows, in addition to the sharp lamellar spacings, a diffuse halo
˚
at around 17 A, assigned to the two-dimensional isotropic or
nematic-like organisation of the discotic moieties. This reflection
remains present (and diffuse) in the lower temperature SmX
phase (Fig. 13). The major difference between the XRD patterns
in the SmAd and the SmX phase in 3 is the sharpening of the wide
angle reflection, which suggests the formation of in-layer orga-
nisation of the azobenzene mesogens. The inset in Fig. 13 shows
a zoom of the small angle area with temperature measured on
a SAXS setup. No small angle reflection could be observed. Even
at low temperatures, the small angle region of the diffractogram
remains unaltered.19
˚
Fig. 15 Sketch of the order in the mesophases of 3 (l ¼ 58 A). (a) SmAd:
bilayer SmA phase wherein the layer with the discotic mesogens is
completely interdigitated. (b) SmX: lamellar phase similar to the SmA
with in-layer positional order in the layer with the calamitic mesogens.
T ¼ 79 ꢀC). This enhanced lamellar organisation may be one of
the reasons why further in-layer organisation is disfavoured, and
hence no higher ordered phases are observed for 3. A graphical
interpretation of the mesophases of 3 is shown in Fig. 15.
The temperature effect on the layer spacing (Fig. 14) shows
a very similar picture to that observed for 2. In the SmAd phase,
the (002) spacings are satisfactorily fitted to equation (5). A high
R2 value of the fit could be obtained because of the large
temperature window of the SmAd phase, when compared to 2.
Incidentally this result strengthened our view that the layer
spacing of the SmA phase of 2 can also be described best using
equation (5). In the SmX phase, the layer spacing of 3 shows
a linear dependence with temperature. At the transition, the
correlation length of the layer spacing does not change consid-
Organisation and tilting in the smectic phases
The presence of the specific XRD reflections from the discotic
moiety (reflections B and D) in the products allows us to evaluate
the organisation of the different layers individually.17 The
pronounced difference in the organisation of the adjacent layers
highlights a nomenclature problem in the identification of mes-
ophases that consist of alternating layers of ordered calamitic
mesogens (e.g. SmI or SmF phases) with disordered (nematic or
isotropic) discotic mesogens.18 Current procedures use miscibility
studies for proper nomenclature, but for this class of complex
materials such methods cannot be applied for mesophase
assignment. In addition, we believe that calamitic or discotic
mesogens with conventional mesophases may readily mix with
these complex assemblies by simply inserting the liquid crystal
into the appropriate layer, thereby expanding the layers without
affecting the symmetry of the system.
erably. It can be observed that the value for x002 in 3 (x002/d002
14.4 at T ¼ 90 ꢀC) is much larger than for 1 or 2 (x002/d002 ¼ 6.1 at
¼
Additionally, our microscopy results indicate the presence of
optically uniaxial phases, whilst the XRD experiments show on
cooling decreasing layer spacing for all lamellar phases. This
behaviour is characteristic of so-called De Vries phases15 in which
the mesogens are tilted with respect to the layer normal, but as
the direction of the tilt in subsequent layers is uncorrelated, the
phases appear uniaxial under crossed polarisers. Cooling in a De
Vries mesophase commonly results in a minor increase in the tilt
angle and therefore a minor decrease in the layer spacing
(as observed for 1–3). In the design of materials exhibiting a De
Vries SmA phase, the correlation between the layers is reduced
by separating the mesogens with small incompatible blocks
Fig. 14 Temperature dependence of the layer spacing (B, left y-axis)
and the correlation length (:, right y-axis). Solid lines in the experi-
mental data points represent data fits (power law in the SmAd phase,
linear in the SmX phase). The same colour coding as in Fig. 9 was used.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1564–1575 | 1573

Table 5 Fitting results of the temperature dependence of the layer spacing of 1–3
fitting parameters eq. (6)
fitting parameters eq. (5)
˚
a
˚
l /A
Mesogen
Phase
d002(qt ¼ 0)/A
k
R2/%
qtb/ꢀ at (T/ꢀC)
ꢁ1
˚
a/A
˚
b/mA K
1
2
55.0
58.0
CrXc
—
—
55.0 ꢅ 0.1
58.4 ꢅ 0.1
—
0.43 ꢅ 0.01
0.16 ꢅ 0.01
—
99.5
71/76
99.7
10.2(30)
2.7(70)
6.8(48)
9.9(36)
6.8(62)
10.3(30)
c d
,
SmA2
SmXd
SmYd
SmAd
SmXd
58.02 ꢅ 0.07
57.63 ꢅ 0.02
55.45 ꢅ 0.06
—
4.7 ꢅ 0.9
12 ꢅ 1
57 ꢅ 1
—
46.4 ꢅ 0.1
—
99.0
99.4
99.9
c
3
48.9
—
15 ꢅ 2
0.28 ꢅ 0.02
41.24 ꢅ 0.05
—
—
a
˚
Molecular length l is calculated after minimisation of the (stretched) structure in Chem3D; for 3, the full molecular length is 58.0 A (same as 2), but
after considering full interdigitation of the discotic moieties a smaller value is calculated. b Tilt angle qt determined from equation (5) or as the cosine of
c
d
d002(T)/d002(qt ¼ 0), the latter a result of the fitting procedure of a higher temperature phase. Fit using equation (5). Fit using equation (6).
(e.g. siloxanes and perfluorinated materials) at the terminal end
of the tails.20 The present case with a discotic mesogen positioned
at the end of the tail could, in principle, be considered an extreme
example of the above approach.
The shape-amphiphiles with triphenylene groups show
a variety of smectic phases. The highest temperature smectic
phase is characterised by a relatively low viscosity, a homeo-
tropic texture, sharp layer reflections and the absence of sharp
wide angle reflections. All features are characteristic of a SmA
phase. The ratio in cross sections of the rods and the disc
determines the details of packing inside the layers: partly inter-
digitated for 3 and bilayer for 2. Although the organisation of the
azobenzene rods fits very well with the classical picture of a SmA
phase, the discs inside the layer are forced into a two-dimensional
structure, resulting in a SmAbiax type organisation, dependent on
the correlation length of the orientation order. Another example
of such a structure has been recently published.21
Table 5 summarises the observed tilting behaviour of 1–3 as
determined by fitting the XRD analysis. Experimental layer
spacings were either fitted linearly or to equation (5).11 It is
interesting to see that the different lamellar phases show rather
different temperature dependences for the constants b and k
associated with the layer spacing, which in the absence of a wider
set of materials and considering that the measured values are
averages over both (different) moieties in the molecule is not
surprising.
One of the fitting parameters of equation (5) is dq ¼ 0, which
The smectic phases at lower temperature are all characterised
by some extent of in-layer organisation observed from the XRD
experiments. However for 2 and 3 the limited number of reflec-
tions (only one) does not allow at present t the organisation in the
azobenzene layer to be established (scattering from the discotic
mesogens remains diffuse).
t
corresponds to the stretched molecular length of the mesogen l
(taking into account interdigitation for 3). For all materials, the
fits give values close to those gained from calculating the fully
extended molecular conformations. From these values and the
experimental layer spacing, it is possible to deduce a tilt angle qt,
which, for all materials, is rather low. However, one has to take
into account that this calculated tilt angle is an average value for
the entire molecule. In the case that only part of the mesogen, for
instance the azobenzene moieties, are tilting, the actual tilt will be
much larger than the calculated average values. As an example,
when only the azobenzene units are responsible for the observed
tilt in the CrX phase of 1, their tilt can be considered to be ꢄ18ꢀ.
Tilting of one of the two layers, or indeed of both of the layers,
is a straightforward means for the system to improve packing
density by reducing small discrepancies in the cross sections of
the discotic and calamitic layers. Therefore, this submolecular
cross sectional ratio should be considered a key parameter to
direct mesophase formation in such complex materials.
Notable is that in the systems where sub-layers of rods and
discs occur, on further cooling the rod containing sub-layers
organize first into higher ordered structures, whereas the disc rich
sub-layers remain in a less ordered structure. We believe that this
can be tuned by selecting proper disc–rod combinations.
Moreover, the combination of a uniaxial appearance in
microscopy experiments and a slightly decreasing layer spacing
with temperature suggests that all observed smectic phase have
a De Vries-type character. This is not surprising, as the tilt
directions of adjacent calamitic layers are expected to be
completely uncorrelated due to the presence of the layer disor-
dered discotics. It is noted that such behaviour is closely con-
nected to that of liquid crystal dimers forming bilayer SmA
phases.22
In short, we have presented three new members of the class of
shape-amphiphiles exhibiting rich mesophase behaviour. The
characterisation and phase assignment of the shape-amphiphiles
is still in its infancy, but the advantages of such materials are
apparent: combining an exceptionally high degree of organisation
at the 10 nm level with the advantages that liquid crystals offer.
Conclusions
We have presented three novel liquid crystalline shape-
amphiphiles based on azobenzenes and different discotic
moieties. With a strong nematogen, like those based on
pentakis(phenylethynyl)phenol, the shape-amphiphile exhibits
a nematic phase, where the discs and rods are homogeneously
mixed. This nematic phase is miscible with the nematic phases of
the precursors, albeit that the mixture with the discotic nema-
togen shows a strong suppression of the clearing temperature in
the phase diagram.
Acknowledgements
The authors acknowledge Prof. Stephen J. Picken for allowing
access to his SAXS setup and Prof. Goran Ungar for fruitful
1574 | J. Mater. Chem., 2009, 19, 1564–1575
This journal is ª The Royal Society of Chemistry 2009

Z. Q. He, J. J. Wang, Y. S. Wang and S. Ye, J. Mol. Liq., 2008,
138, 93–99; (i) S. K. Pal, S. Kumar and J. Seth, Liq. Cryst., 2008,
35, 521–525.
8 For instance read: (a) A. S. Matharu, S. Jeeva and P. S. Ramanujam,
Chem. Soc. Rev., 2007, 36, 1868–1880; (b) V. Shibaev, A. Bobrovsky
and N. Boiko, Prog. Polym. Sci., 2003, 28, 729–836; (c)
S. K. Yesodha, C. K. S. Pillai and N. Tsutsumi, Prog. Polym. Sci.,
2004, 29, 45–74.
9 N. Boden, R. J. Bushby and A. N. Cammidge, J. Am. Chem. Soc.,
1995, 117, 924.
10 P. H. J. Kouwer, W. F. Jager, W. J. Mijs and S. J. Picken,
Macromolecules, 2002, 35, 4322–4329.
discussions. Furthermore, they wish to thank the EU (RTN
‘‘LCDD’’ contract HPRN-CT2000-00016) and the Ramsay
Memorial Fellowship Trust (PHJK) for financial support.
Notes and references
1 For thermotropic liquid crystals: (a) K. Binnemans, Chem. Rev., 2005,
105, 4148–4204, and references therein; (b) P. H. J. Kouwer and
T. M. Swager, J. Am. Chem. Soc., 2007, 129, 14042–14052; (c)
K. Hoshino, M. Yoshio, T. Mukai, K. Kishimoto, H. Ohno and
T. Kato, J. Polym. Sci., Part A: Polym. Chem., 2003, 41, 3486–
3492; (d) M. Yoshio, T. Mukai, H. Ohno and T. Kato, J. Am.
Chem. Soc., 2004, 126, 994–995; (e) M. Yoshio, T. Kagata,
K. Hoshino, T. Mukai, H. Ohno and T. Kato, J. Am. Chem. Soc.,
2006, 128, 5570–5577.
11 P. G. De Gennes, J. Prost, The Physics of Liquid Crystals, 2nd ed.,
Oxford University Press, New York, 1993, p. 528.
12 P. H. J. Kouwer, W. F. Jager, W. J. Mijs and S. J. Picken,
Macromolecules, 2001, 34, 7582–7584.
13 In a previously reported shape-amphiphile, the (005) and (007)
reflections were missing, while in this material the (003), (006) and
(008) reflections were clearly seen.
14 OriginPro 7.0, OriginLabs Corp. Sharp peaks were fitted with
a Gaussian distribution, diffuse peaks with a Lorentzian.
15 A. De Vries, A. Ekachai and N. Spielberg, Mol. Cryst. Liq. Cryst.,
1979, 49, 143–152.
16 We have not been able to assign the XRD patterns using different
phases. An alternative would deviate from the lamellar structure
that is evident from the optical textures. Commensurate smectic
structures, however, may be responsible for highlighting the even
reflections at the cost of the odd layer reflections.
17 For rod–rod dimers, these experiments are difficult to perform.
Resonant X-ray scattering experiments have been performed, but
they require specific (heavy) elements in the mesogen.
18 For bolamphiphiles, Tschierske and co-workers have developed
a nomenclature that considers different levels of organisation in
different dimensions: [X. Cheng, M. Prehm, M. K. Das, J. Kain,
U. Baumeister, S. Diele, D. Leine, A. Blume and C. Tschierske,
J. Am. Chem. Soc., 2003, 125, 10977–10996]. This is, however, not
completely equivalent with the shape-amphiphiles that we are
studying in the sense that in our systems differences in the
organization of alternating layers exists and, hence, a ‘simple’
smectic phase can be composed of a multitude of combinations of
differently ordered layered phases.
2 Self-assembly of (ionic) surfactants in solution can result in the
formation of lyotropic mesomorphism.
3 (a) C. Tschierske, Curr. Opin. Colloid Interface Sci., 2002, 7, 355–370;
(b) S. Laschat, A. Baro, N. Steinke, F. Giesselmann, C. Hagele,
G. Scalia, R. Judele, E. Kapatsina, S. Sauer, A. Schreivogel and
M. Tosoni, Angew, Chem Int, Ed., 2007, 46, 4832–4887; (c)
S. Kumar, Chem. Soc. Rev., 2006, 35, 83–109; (d) T. Kato,
N. Mizoshita and K. Kishimoto, Angew. Chem Int. Ed., 2006, 45,
38–68.
4 (a) P. Lo Nostro, Curr. Opin. Colloid Interface Sci., 2003, 8, 223–226;
(b) W. Mahler, D. Guillon and A. Skoulios, Mol. Cryst. Liq. Cryst.,
1985, 2, 111; (c) P. Marczuk, P. Lang and M. Moller, Colloids
Surf., A, 2000, 163, 103–113; (d) J. F. Rabolt, T. P. Russell and
R. J. Twieg, Macromolecules, 1984, 17, 2786–2794; (e) T. P. Russell,
J. F. Rabolt, R. J. Twieg, R. L. Siemens and B. L. Farmer,
Macromolecules, 1986, 19, 1135–1143.
5 R. W. Date and D. W. Bruce, J. Am. Chem. Soc., 2003, 125, 9012–
9013.
6 (a) P. H. J. Kouwer and G. H. Mehl, Angew. Chem., Int. Ed., 2003, 42,
6015–6018; (b) P. H. J. Kouwer and G. H. Mehl, J. Am. Chem. Soc.,
2003, 125, 11172–11173; (c) P. H. J. Kouwer and G. H. Mehl, Mol.
Cryst. Liq. Cryst., 2003, 397, 301–316; (d) P. H. J. Kouwer,
J. Pourzand and G. H. Mehl, Chem. Commun., 2004, 66–67; (e)
P. H. J. Kouwer, C. J. Welch, G. Mcrobbie, B. J. Dodds, L. Priest
and G. H. Mehl, J. Mater. Chem., 2004, 14, 1798–1803.
19 P. H. J. Kouwer, S. J. Picken and G. H. Mehl, J. Mater. Chem., 2007,
17, 4196–4203.
20 J. P. F. Lagerwall and F. Giesselmann, ChemPhysChem, 2006, 7, 20–
45.
21 C. V. Yelamaggad, I. S. Shashikala, V. P. Tamilenthi, D. S. S. Rao,
G. G. Nair and S. K. Prasad, J. Mater. Chem., 2008, 18, 2096–2103.
22 Recent reviews of liquid crystalline dimers: (a) C. T. Imrie and
P. A. Henderson, Chem. Soc. Rev., 2007, 36, 2096–2124; (b)
C. V. Yelamaggad, G. Shanker, U. S. Hiremath and S. K. Prasad,
J. Mater. Chem., 2008, 18, 2927–2949.
7 (a) G. Zucchi, B. Donnio and Y. H. Geertz, Chem Mater., 2005, 17,
4273–4277; (b) D. Apreutesei and G. Mehl, Mol. Cryst. Liq. Cryst.,
2006, 449, 107–115; (c) D. Apreutesei and G. H. Mehl, Chem.
Commun., 2006, 609–611; (d) C. T. Imrie, Z. B. Lu, S. J. Picken and
Z. Yildirim, Chem. Commun., 2007, 1245–1247; (e) K. U. Jeong,
A. J. Jing, B. Mansdorf, M. J. Graham, D. K. Yang, F. W. Harris
and S. Z. D. Cheng, Chem. Mater., 2007, 19, 2921–2923; (f)
K. U. Jeong, A. J. Jing, B. Monsdorf, M. J. Graham, F. W. Harris
and S. Z. D. Cheng, J. Phys. Chem. B, 2007, 111, 767–777; (g)
S. Romano, Phys. Lett. A, 2004, 333, 110–119; (h) C. X. Zhang,
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1564–1575 | 1575