Phase transitions of 4′-n-hexacosyloxy-3′-nitrobiphenyl-4-carboxylic acid (ANBC-26): Two types of thermotropic cubic phases

Article abstract of DOI:10.1021/jp0018652

The phase transitions of 4′-n-hexacosyloxy-3′-nitrobiphenyl-4-carboxylic acid (ANBC-26, where `26' indicates the number of carbon atoms in the alkoxy group) were investigated by differential scanning calorimetric (DSC), polarizing optical microscopic, and X-ray diffraction (XRD) measurements. ANBC-26 is a new liquid crystalline material which exhibits successively two types of three-dimensionally ordered cubic mesophases on heating; the low-temperature cubic (CubI) phase has a space group of Im3m, and the high-temperature one (CubII phase) is of Ia3d type, identical with the cubic phase of shorter alkoxy chain homologues such as ANBC-16, and a cubic-cubic phase transition occurs at 435 K on DSC. Reversibility of the transitions between them was studied and the structures of those mesophases are discussed. The transformation from a one-dimensionally ordered smectic C (SmC) to the CubI phase and to the CubII phase can be interpreted qualitatively in terms of the increased curvature of the interface. The structured liquid I₁ phase, an intermediate phase between the CubII and a true isotropic liquid I₂ phases, has been also studied, indicating that this phase could be characterized by Ornstein-Zernike type short-range fluctuation, which means the presence of molecular clusters consisting of several ANBC-26 dimers in the temperature region.

Full text of DOI:10.1021/jp0018652

10196
J. Phys. Chem. B 2000, 104, 10196-10205
Phase Transitions of 4′-n-Hexacosyloxy-3′-nitrobiphenyl-4-carboxylic Acid (ANBC-26): Two
Types of Thermotropic Cubic Phases
Shoichi Kutsumizu,*,† Tatsuya Ichikawa,‡ Minoru Yamada,‡ Shuichi Nojima,§ and
Shinichi Yano‡
Instrumental Analysis Center, Gifu UniVersity, 1-1 Yanagido, Gifu 501-1193, Japan,
Department of Chemistry, Faculty of Engineering, Gifu UniVersity, 1-1 Yanagido, Gifu 501-1193, Japan,
and School of Materials Science, Japan AdVanced Institute of Science and Technology (JAIST), 1-1 Asahidai,
Tatsunokuchi, Nomi, Ishikawa 923-1292, Japan
ReceiVed: May 22, 2000; In Final Form: August 17, 2000
The phase transitions of 4′-n-hexacosyloxy-3′-nitrobiphenyl-4-carboxylic acid (ANBC-26, where “26” indicates
the number of carbon atoms in the alkoxy group) were investigated by differential scanning calorimetric
(DSC), polarizing optical microscopic, and X-ray diffraction (XRD) measurements. ANBC-26 is a new liquid
crystalline material which exhibits successively two types of three-dimensionally ordered cubic mesophases
on heating; the low-temperature cubic (CubI) phase has a space group of Im3m, and the high-temperature
one (CubII phase) is of Ia3d type, identical with the cubic phase of shorter alkoxy chain homologues such as
ANBC-16, and a cubic-cubic phase transition occurs at 435 K on DSC. Reversibility of the transitions between
them was studied and the structures of those mesophases are discussed. The transformation from a one-
dimensionally ordered smectic C (SmC) to the CubI phase and to the CubII phase can be interpreted qualitatively
in terms of the increased curvature of the interface. The structured liquid I1 phase, an intermediate phase
between the CubII and a true isotropic liquid I2 phases, has been also studied, indicating that this phase could
be characterized by Ornstein-Zernike type short-range fluctuation, which means the presence of molecular
clusters consisting of several ANBC-26 dimers in the temperature region.
1. Introduction
to the CubD phase (and vice versa) brings about dramatic
changes in physical properties in addition to the optical property,
and the storage modulus jumps about 1000 times at the transition
to show a value of the order of 107 dyn/cm2 in the CubD phase
region.20,26 ANBC-n has a relatively simple molecular structure
which consists of a biphenylcarboxylic acid core, a nitro group
at the side, and a long alkoxy tail, where the carboxylic acid
group is known to be hydrogen-bonded with other molecules
to dimerize.21,29 Such simple molecular structure would be
advantageous when the molecular organization within the cubic
lattice and the factors by which the phase formation is governed
are considered.
The molecular organizations in such mesophases are usually
investigated by X-ray diffraction measurements.9,10,14,16,18,23,28,30
For the n ) 16 and 18 homologues, the structure type of the
CubD phase has been already identified as space group
Ia3d.10,18,23 By aid of other physical measurements such as
NMR15,22 or dynamic viscoelastic measurements,20,26 it is now
accepted that the structure of the CubD phase is a bicontinuous
type, where the rod micelles are connected 3-by-3 to form two
sets of interwoven networks, according to Luzzati and Spect’s
rod-connectors scheme33 (see the upper part of Figure 6). A
somewhat surprising fact23 is that a 3′-cyano analogue with n
) 18 (4′-n-octadecyloxy-3′-cyanobiphenyl-4-carboxylic acid,
ACBC-18) shows a different type of structure with space group
Im3m, and the cubic structure is severely dependent on its
molecular structure.
Cubic mesophases are in the limelight in the field of soft
materials.1-5 They are formed in a wide range of chemical
systems, such as liquid crystals,2 biological lipid-water,3 and
block copolymers.4 The formation severely depends on tem-
perature, the molecular structure, and composition; phase
transitions between optically isotropic cubic phases and other
optically anisotropic mesophases with lamellar or hexagonal
columnar structure are still mysterious at several points.
Regardless of the diversity of chemical systems, it is generally
accepted that they consist of two incompatible components, e.g.,
aliphatic vs aromatic, apolar vs polar, or soft vs hard compo-
nents, but our understanding in this field is far from complete.
4′-n-Alkoxy-3′-nitrobiphenyl-4-carboxylic acids (designated
as ANBC-n, where n is the number of carbon atoms in the
alkoxy group, see Chart 1in the Experimental Section) have been
investigated by several researchers6-30 and are well-known for
forming a cubic phase labeled cubic D (CubD) in a certain
temperature range. The alkoxy chain length n is a key factor
for determining the phase behavior; when n g 16, the CubD
phase appears on both heating and cooling and its temperature
region becomes wider with increasing n (to n ) 26);6,7,11,19,29
) 15 is a critical length of the alkoxy chain, where the formation
of the CubD phase is severely affected by thermal cycles
used.26,29 The phase transition from a lamellar smectic C (SmC)
n
* Author to whom correspondence should be addressed. E-mail:
kutsu@cc.gifu-u.ac.jp.
Very recently, we have found that one ANBC homologue
having a very long alkoxy chain of n ) 26 successively forms
two types of cubic phases, one with Im3m symmetry and the
other with Ia3d, and the Im3m cubic phase is transformed into
† Instrumental Analysis Center, Gifu University.
‡ Department of Chemistry, Faculty of Engineering, Gifu University.
§ School of Materials Science, Japan Advanced Institute of Science and
Technology (JAIST).
10.1021/jp0018652 CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/13/2000

Phase Transitions of ANBC-26
J. Phys. Chem. B, Vol. 104, No. 44, 2000 10197
the Ia3d cubic phase at 435 K on heating.28 This is the first
example of thermotropic cubic-cubic phase transition in one-
component systems, although two examples have been reported
in lyotropic binary or ternary mixtures, where the phase be-
havior is a function of water content.3,31,32 For a fundamental
understanding of this transition, detailed investigation of the
phase structures is prerequisite, and in this work, the lamellar
SmC-to-Im3m cubic and the Im3m cubic-to-Ia3d cubic phase
transitions are greatly paid attention to. The results of thermo-
gravimetric and differential scanning calorimetric studies, optical
microscopic observation, and X-ray diffraction studies by using
a high-resolution small-angle X-ray scattering (SAXS) instru-
ment are presented, and the molecular organizations of the
Im3m-cubic (denoted as CubI in this paper), Ia3d-cubic (CubII),
and a structured liquid labeled I1, which is an intermediate phase
between the CubII and a true isotropic liquid (I2), are discussed.
CHART 1
Measurements. The IR spectra were recorded on a Perkin-
Elmer 1640 and a Perkin-Elmer system 2000 Fourier transform
IR spectrometer. The H NMR and MS spectra were recorded
1
on a JEOL JNM-R400 spectrometer and a Shimadzu GCMS
QP-1000 system, respectively. The phase transitions and thermal
stability were examined on a Seiko Denshi DSC-210 and TG/
DTA-300 interfaced to a TA data station (SSC 5000 system).
The measurements at a scanning rate of 5 K min-1 were
performed under a dry N2 flow of ca. 40 mL min-1 for DSC
and of ca. 200 mL min-1 for TG/DTA. The texture of each
mesophase was observed at a heating/cooling rate of 5 K min-1
by using a Nikon Optiphot-pol XTP-11 polarizing optical
microscope (POM) equipped with a Mettler FP-82 hot stage
and a Mettler FP80 central processor.
X-ray diffraction (XRD) patterns at elevated temperatures
were obtained for powder samples filled in thin glass capillaries
with 1.5 mm diameter (Hilgenberg Co.), which were placed in
a Mettler FP82HT hot stage, and the temperature was controlled
within (0.1 °C by a Mettler FP90 central processor. The
accuracy of the temperature was checked by using a calibrated
Fe-constantan thermocouple. The following two setups were
used.
(A) A MAC Science X-ray generator (M18XHF) was
operated with a copper target at 40 kV and 30 mA, and the Cu
KR radiation (λ ) 0.154 nm) was point-focused with Huxley-
Holms optics. The scattered X-ray was detected by a one-
dimensional position-sensitive proportional counter (PSPC) with
an effective length of 10 cm. The distance between the sample
and PSPC was about 40 cm, and the geometry was further
checked by a chicken tendon collagen, which gives a set of
sharp diffractions corresponding to 65.3 nm, and R-stearic acid
giving a set of diffractions of 3.962 nm. The accumulation time
for each measurement was 500-3600 s, depending on the
intensity and the quality needed.
2. Experimental Section
Preparation. Preparation of Hexacosyl Bromide (BA-26).34,35
First, a Grignard reagent (C16H33MgBr) was prepared in the
following manner. Magnesium turnings (0.04 mol) were covered
with a mixture of 1,2-dibromoethane (0.1 mL) and anhydrous
ethyl ether (30 mL) under a dry N2 gas atmosphere and stirring
was begun, to which a mixed solution of hexadecyl bromide
(0.04 mol) and anhydrous ethyl ether (20 mL) was added
dropwise. Next, the Grignard reagent prepared was added
dropwise to a stirred anhydrous tetrahydrofuran (THF) solution
(50 mL) of lithium tetrachlorocuprate(II) (0.0004 mol) and 1,10-
dibromodecane (0.04 mol) in an iced bath, where lithium
tetrachlorocuprate(II) was prepared by mixing CuCl2 (0.0004
mol) with LiCl (0.0008 mol) in anhydrous THF (20 mL). The
reaction solution was stirred for an additional 24 h, and after
that, water was added to remove the metal residue into the water
layer. Finally, the organic layer was filtered to give a white
solid of hexacosyl bromide (C26H53Br, BA-26), which was
recrystallized from ethanol several times until the melting
temperature reached a constant value of 56 °C.36 The purity
1
was checked by thin-layer chromatography (TLC), H nuclear
magnetic resonance (1H NMR), mass spectroscopy (MS), and
differential scanning calorimetry (DSC).
Preparation of ABC-26 and ANBC-26. 4′-n-Hexacosyloxy-
biphenyl-4-carboxylic acid (ABC-26) was prepared by reacting
BA-26 obtained above with 4′-hydroxybiphenyl-4-carboxylic
acid, and then nitration of ABC-26 gave ANBC-26. These two
processes were based on the established method of Gray et al.6,37
The final white powder product was purified by repeated
recrystallization from ethanol, and the purity was checked by
elemental analysis, TLC, 1H NMR, MS, and DSC. Yield, about
50%.
(B) A Rigaku R-AXIS IIC X-ray system was used and
operated with a copper target at 40 kV and 150 mA, and the
Cu KR radiation was collimated into the sample capillary. The
scattered X-ray was recorded on a two-dimensional imaging
plate (IP) detector with an effective area of 20 × 20 cm2, and
the sample-to-detector distance was 20 cm. The exposure time
for each measurement was 5-10 min.
Results
δH (400 MHz, solvent CDCl3, standard TMS): 0.88 (a, t,
3
3J ) 6.8 Hz, 3H), 1.25 (b, m, 44H), 1.50 (c, qn, J ) 6.9 Hz,
ANBC-26 synthesized was first studied by thermogravimetry
(TG) and was found to be thermally stable; no weight loss was
detected up to 500 K, which was the highest temperature used
in the present studies. Even at 550 K, about 80 K higher than
the clearing temperature, only a 0.3% loss was observed, much
less than that for ANBC-16 (1.4% loss at 550 K). Hence,
lengthening of the alkoxy chain from n ) 16 to 26 resulted in
a fair improvement of thermal stability of ANBC-n homologues,
which is of course very important for the detailed investigations
of ANBC-26.
2H), 1.87 (d, qn, 3J ) 7.6 Hz, 2H), 4.16 (e, t, 3J ) 6.5 Hz, 2H),
3
3
7.18 (f, d, Jfg ) 9.0 Hz, 1H), 7.67 (i, d, Jij ) 8.5 Hz, 2H),
3
4
7.79 (g, d,d, Jgf ) 8.8 Hz, Jgh ) 2.4 Hz, 1H), 8.12 (h, d,
4Jhg ) 2.4 Hz, 1H), 8.18 (j, d, Jji ) 8.5 Hz, 2H); m/z 624
3
(M+, 3.7%), 494 (2.8%), 259 (88.3%), 43 (100%). MS spectra
did not detect the mass fragment larger than 624 even when an
extremely concentrated solution was used, indicating no content
of longer alkoxy or alkyl chain byproducts than C26. The
elemental analysis for ANBC-26 was carried out by the
laboratory for Organic Elemental Microanalysis of Kyoto
University, giving a result well consistent with its formula:
Anal. Calcd for C26H53NO5: C, 75.12; H, 9.97; N, 2.25; O,
12.84. Found: C, 74.93; H, 9.99; N, 2.20; O, 12.86%.
Figure 1 shows the DSC trace of ANBC-26 at a heating/
cooling rate of 5 K min-1. Phase transition temperatures and
enthalpy changes are listed in Table 1. On first heating (1H),
ANBC-26 melts at 376 K into a smectic C (SmC) phase,

10198 J. Phys. Chem. B, Vol. 104, No. 44, 2000
Kutsumizu et al.
I2
TABLE 1: Transition Temperatures (in K) and Transition Enthalpies (in kJ mol-1) of ANBC-26a
run
1H
Cr1/Cr′
Cr2/Cr
SmC
CubI
CubII
Hex
I1
•
359
[38.8]
•
376
[46.7]
349
[-47.1]
376
•
396
[2.2]
366
[-4.6]
395
•
435
[0.1]
•
467
[4.1]
465
[-1.4]
467
•
473
•
•
•
[6.9]
471
1C
2H
•
•
•
•
•
•
462
[-1.9]
•
•
•
[-6.9]
352
[-2.2]
•
•
436
[0.1]
472
[46.0]
[2.2]
[3.6]
[8.8]
a 1H, first heating; 1C, first cooling; 2H, second heating. Phases marked • appear on each run.
Figure 2. Photomicrograph of the Hex phase of ANBC-26 at 458.8
K, at which the CubII phase should appear on DSC because the
temperature region of the Hex phase is 462-465 K in the DSC cooling;
there exists a small discrepancy with temperature.
Figure 1. DSC traces measured at 5 K min-1 for ANBC-26. 1H, 2H,
and 1C mean the first and second heating, and first cooling runs,
respectively. Cr, crystalline; SmC, smectic C; Cub, cubic; Hex,
hexagonal columnar; I, isotropic liquid phases.
characterized by the appearance of a typical SmC schlieren
texture under crossed polarizers, and then undergoes a SmC-
to-cubic phase transition at 396 K, readily identified by slow
growth of completely black and thus optically isotropic domains
with straight edges. Initially, one type of cubic phases was
thought to range up to 467 K, but looking closer at the heating
trace, a faint endothermic peak is visible at 435 K, around at a
mid temperature of the cubic phase region. The enthalpy change
estimated (∼0.2 J g-1 ) 0.1 kJ mol-1) is barely beyond the
noise level of our DSC instrument used, and the peak was also
observed at 436 K on the second heating (2H). The XRD results
later described indicate that this is due to a cubic-to-cubic phase
transition, and hereafter, the low- and high-temperature cubic
phases are denoted as CubI and CubII, respectively, as shown
in the figure. At 467 K, a sharp endothermic peak is seen,
followed by a broad hump centered at 473 K. The XRD
measurements showed the destruction of the CubII lattice at
467 K, but there still remains a broad but distinct peak in the
small-angle region, suggesting the presence of some organization
between the two temperatures. The state of this temperature
region is thus assigned to a structured liquid, labeled I1 in the
figure. The structure of the I1 phase will be discussed later, in
the Discussion section (see Figure 9). At around 473 K, the
small-angle XRD peak virtually disappeared, indicating that the
sample changes into a genuine isotropic liquid (I2). On 2H scan,
the sample showed crystallization of residual glassy regions at
352 K, but was otherwise identical with 1H scan.
Figure 3. SAXS pattern of the Hex phase of ANBC-26 recorded at
465 K on cooling. The indexation (hk) are also shown.
(Hex) mesophase.38 The XRD pattern recorded at 465 K is
shown in Figure 3, where a strong diffraction peak at 1.31 nm-1
and two very weak peaks at 2.27 and 2.83 nm-1 are seen. Here,
q ) (4π/λ)sinθ, λ ) 0.154 nm, and 2θ ) scattering angle. The
ratios of the reciprocal spacings are x1:x3:x4, corresponding
to the indexation (hk) ) (10), (11), and (20) of the Hex phase
with a lattice parameter of a ) 5.52 nm. Below 462 K (457.9
K on POM observation; see the caption for Figure 2), the texture
suddenly turned to completely black, showing the formation of
a cubic phase, and on further cooling, no definite peaks were
observed until the sample shows a SmC phase at 366 K.
Consequently, the phase sequence of the cooling process is
hexagonal columnar-cubic-lamellar SmC in going from higher
temperature, reminiscent of the phase sequence often seen in
lyotropic liquid crystalline systems exhibiting a cubic phase,
although in the latter systems it is a function of the amount of
solvent.
On cooling (1C) from 498 K, there was observed a broad
exothermic peak centered at 471 K, and after that two sharp
exothermic peaks appear at 465 and 462 K. Between the two
temperatures, as shown in Figure 2, a pseudo-focal conic texture
was observed on POM, indicative of a hexagonal columnar

Phase Transitions of ANBC-26
J. Phys. Chem. B, Vol. 104, No. 44, 2000 10199
TABLE 2: X-ray Diffraction Data of Two Cubic Phases of
ANBC-26a
CubI at 404 K
space group Im3m,
a ) 20.02 ( 0.02 nm
CubII at 444 K
space group Ia3d,
a ) 12.30 ( 0.04 nm
hkl
110
dcalc
dobs
dcalc
dobs
14.155
10.009
8.172
7.077
6.330
5.779
5.350
5.004
4.718
4.476
4.268
4.086
3.926
3.655
abs
abs
5.020
4.347
abs
200
211
220
310
222
321
400
411
420
332
422
510/431
521
8.199
7.100
6.350
5.745
5.334
4.997
4.701
4.484
5.004
4.361
abs
3.286
3.074
abs
2.749
2.621
2.510
2.411*
2.245
3.09
2.75
2.65
4.080
3.928
3.667
2.39
a a (in nm) is the cubic cell parameter. dcalc and dobs (in nm) are the
calculated and observed spacings of the (hkl) small-angle reflections,
respectively; abs, systematic absences of reflections determined by each
space group; *, only the (431) reflection is observable.
Figure 4. SAXS patterns of two cubic phases of ANBC-26 recorded
at (a) 404 K and (b) 444 K, on heating. Corresponding two-dimensional
patterns are shown in the inset. Miller indices are also shown in the
figure.
Figure 4 shows XRD patterns in the small-angle region of
ANBC-26 detected by PSPC at two different temperatures below
and above the temperature at which the DSC faint peak was
observed on heating. The corresponding IP patterns are also
presented in the insets. Although it is not so obvious in the IP
patterns, the two PSPC patterns are clearly different. At 404 K
in Figure 4a, 11 peaks are observed with the following ratios
of reciprocal spacings: x3:x4:x5:x6:x7:x8:x9:x10:x12:
x13:x15. Since the ratios x7 and x15 are not compatible with
any cubic lattices, the sequence of numbers must be doubled.
Thus, the obtained ratios are x6:x8:x10:x12:x14:x16:x18:
x20:x24:x26:x30, which is characteristic of a body-centered
cubic symmetry, most probably Im3m, and thus the parameter
of the unit cell of the CubI phase is a ) 20.02 ( 0.02 nm.
Table 2 summarizes the indexing. The pattern of the CubI phase
was arranged almost symmetrically with respect to the beam
stopper (not shown here), which well corresponds to the
observation of diffraction rings on the IP image in the inset.
On heating, the PSPC pattern abruptly changed at around 428
K, slightly lower than the corresponding temperature on DSC,
and at 444 K, in Figure 4b, two strong peaks with the ratio
x6:x8 and very weak peaks corresponding to the ratio x16:
x20:x22:x26 are seen. Although not shown in this figure, the
pattern was very often dissymmetric with respect to the beam
stopper. Correspondingly, the IP pattern recorded at the same
temperature consists of many spots, arranged into two rings
denoted as (211) and (220). These observations clearly reflect
an improved growth of CubII domains compared to the case of
the CubI phase in the glass capillary tube. Since ANBC-26 is
achiral, only a space group Ia3d matches the absence of
diffractions with lower indices (100), (110), (111), (200), and
(210). The cell parameter is a ) 12.30 ( 0.04 nm (see Table
2). The high-temperature CubII phase has the same symmetry
as the CubD phase of ANBC-1610 and -1818,23 previously
Figure 5. Two-dimensional SAXS patterns of ANBC-26 recorded at
(a) 462 K and (b) 383 K on cooling. In (b), Miller indices are also
shown.
reported. In fact, the PSPC pattern obtained for the CubII phase
of ANBC-26 is quite similar to that reported for the CubD phase
of ANBC-18.23 On the other hand, the low-temperature CubI
phase and the cubic phase of a 3′-cyano analogue, ACBC-18,23
are of the same type. Finally, it should be noted that the use of
a high-resolution SAXS equipment in the present work was
essential to the finding of two types of cubic phases in the cubic
phase temperature region of ANBC-26.
What is the type of the cubic phase on cooling? To clarify
this, we used the IP patterns because in the cubic temperature
region on cooling the PSPC patterns detected only a strong
reflection peak, as shown in Figure 3, from which it is difficult
to identify the type definitely. Figure 5 presents the IP patterns

10200 J. Phys. Chem. B, Vol. 104, No. 44, 2000
Kutsumizu et al.
for ANBC-26, while for ANBC-15 with a minimum length
of the alkoxy chain for showing a cubic phase, dLSmC/dT )
7 × 10-4 nm K-1, the smallest among the ANBC-n homologues
examined.39
(2) Ranc¸on and Charvolin40 studied the phase transitions of
hexaethylene glycol mono-n-dodecyl ether (C12EO6)/water
system and pointed out the existence of epitaxial relationships
between particular planes of three successive phases: the layer
of the lamellar phase, the (211) plane of the Ia3d cubic phase,
and the (10) plane of the hexagonal columnar phase. These three
planes are of highest density of matters in each of the three
phases, and the relations themselves demonstrate the direct
transformations of a highest density plane into other at the phase
transitions, when the temperature or the amount of water is
varied.
For ANBC-26, the SmC layer thickness is almost equal to
the (222) spacing of the CubI phase with Im3m symmetry at
the SmC-to-CubI phase transition on heating, suggesting the
existence of such an epitaxial relation. This is of course only a
suggestion because we did not use an oriented sample, but this
relation seems to have the following physical meaning. In the
Im3m cubic phase with a cell parameter a, one 6-by-6 connection
point is located at (000), the other being at (a/2,a/2,a/2), and
the distance between them is (x3)a/2, while the (222) spacing
is described as (x3)a/6. Here, the line connecting between (000)
and (a/2,a/2,a/2) is normal to the (222) planes. Hence, three
layers should exist between the two connection points. The
structure of the CubI phase will be discussed later in detail. On
the other hand, the (321) spacing versus temperature curve of
the CubI phase and the (211) spacing versus temperature curve
of the CubII phase show a discontinuity at the CubI-to-CubII
phase transition, where both planes are the highest intensity
plane of each.
Figure 6. Plots of d spacing versus temperature for ANBC-26; filled
and open symbols represent the data recorded on heating and cooling,
respectively. The model structures of SmC, CubI, and CubII phases
are shown in the upper part of this figure.
on cooling. The patterns on cooling are spot-like, rather than
in rings, reflecting a strong tendency for the formation of larger
domains compared to the case on heating. As the temperature
decreased, the spots sharpened and the number detected
increased, but the pattern itself was unchanged. At 383 K, in
Figure 5b, the pattern containing the largest number of spots
was obtained, and the indexing is consistent with the Ia3d
symmetry of the CubII phase. Thus, we concluded that the cubic
phase on cooling is the CubII phase alone.
To get insight into the transformation process along the
SmC-CubI (Im3m)-CubII (Ia3d) phase sequence on heating,
the d spacings of the diffraction peaks (filled symbols) are
plotted versus temperature in Figure 6, which also contains the
data on cooling (open symbols). The schematic models for the
three structures are depicted in the upper part of this figure.
The structures of two cubic phases are based on the models
proposed for lyotropic systems by Luzzati and co-workers,3,33
where white and black skeletons only depict the central lines
of the rods (or correspond to the core parts of the rod micelles),
not the rod micelles themselves. In ANBC-26, the aromatic part
and COOH group are considered to be at around the white and
black skeletons, and the aliphatic chain fills out the free space
between them.
(3) On cooling (open symbols in Figure 6), the sequence of
mesophases in ANBC-26 is the same in C12EO6/water system,
irrespective of the reverse order in the latter in the temperature
scale. However, in ANBC-26, a small but distinct discontinuity
is seen at the Hex-to-CubII phase transition, and at the CubII-
to-SmC phase transition temperature, then, a large discrepancy
was observed between the SmC layer thickness and CubII (211)
spacing. On the other hand, the temperature curve of the (211)
spacing on cooling gave a straight line down to the CubII-to-
SmC phase transition, superimposed with the curve on heating.
This also confirms the appearance of only one type in the cubic
region on cooling.
Figure 7 shows plots of the cubic cell parameters as a function
of temperature. What is noteworthy is that in the CubI phase
temperature range, the unit cell parameter a decreases with
increasing temperature, showing a negatiVe thermal volume
expansion ((1/a3)(∂(a3)/∂T) ) -2.3 × 10-3 K-1). This trend
continues but is greatly reduced (-6.6 × 10-4 K-1) when
temperature goes into the CubII phase temperature range. The
CubII phase on cooling also showed a negative thermal
expansion (-8.4 × 10-4 K-1). Bruce et al. reported such a
negative thermal volume expansion (-1.5 × 10-3 K-1) for the
Ia3d cubic phase of a silver(I)-alkoxystilbazole complex.41
Paleos et al. also reported that in dialkyl phosphate potassium
salts the thermal expansion of the Ia3d cubic phase is negative,
-1.2 × 10-3 K-1 for the octyl to -3.0 × 10-3 K-1 for the
octadecyl derivative.42 It seems that the negative thermal volume
expansion is characteristic of cubic mesophases. Provided that
the density is unchanged with temperature (and assumed to be
1), the negative thermal volume expansion in the case of ANBC-
26 implies that the number of molecules per unit cell decreases
Several interesting points are noted: (1) The layer thickness
of the SmC phase decreases as the temperature approaches
the SmC-to-CubI phase transition temperature. When the
SmC layer is transformed into the SmA layer, the tilt angle of
the molecular long axis against the layer normal usually goes
to zero and thus the layer thickness increases, as in the case
of ANBC-1439 (which shows no cubic phase). Therefore, the
“thinning” of the SmC layers may be characteristic of the
SmC phase that exhibits any cubic phase at the higher
temperature side. The largest magnitude of the decrease in layer
thickness (LSmC) was obtained (dLSmC/dT ) 1.6 × 10-2 nm K-1
)

Phase Transitions of ANBC-26
J. Phys. Chem. B, Vol. 104, No. 44, 2000 10201
stopped at 413 K in the CubI phase region and the cooling was
started, cycle 2, the CubI phase was maintained during the
cooling to 364 K, at which the SmC phase appeared. The CubI-
to-SmC phase transition temperature on this cooling is almost
the same with the CubII-to-SmC phase transition temperature
(366 K) on cooling in cycle 1; the transition enthalpy was 3.2
kJ/mol, 30% smaller than the CubII-to-SmC transition enthalpy
(4.6 kJ/mol). Moreover, annealing at a temperature in the CubI
phase for 1 h or 1 day did not change the XRD pattern. These
results conclude that the CubI phase is not a metastable phase.
In cycle 3, the cycle was reversed at 448 K in the CubII phase,
and the CubII phase was observed on the cooling to 373 K. At
373 K, reheating was begun, and then, the transformation to
the SmC phase quickly took place. After that, the phase behavior
was identical with that in cycle 1, instead of the fact that the
CubI-to-CubII phase transition was located at a temperature
between 423.4 and 439.9 K by XRD but not detected by DSC.
In cycle 4, the heating was stopped at 448 K to start the cooling,
as in cycle 3, but unlike cycle 3, reheating was begun at 403
K, at which the CubI phase would be observed on heating.
However, the CubII phase was maintained on the reheating as
well as the preceding cooling, and even after annealing at 408
K (still in the CubI phase region on heating) for 1 h. These
results suggest an interesting fact that the CubI phase can be
formed from the SmC phase on heating, but the transformation
from CubII to CubI phase does not occur.
Figure 7. Temperature dependence of the unit cell parameters a of
the CubI and CubII phases; filled and open symbols represent the data
on heating and cooling, respectively. Vertical dashed line is the cubic-
cubic phase transition temperature determined by SAXS experiments.
Discussion
Structure of the CubII Phase with Ia3d Symmetry. For
the Ia3d cubic structure illustrated in the upper part of Figure
6, the diameter (D) and the length (L) of the constituent rods
and the number of molecules (N) per unit length of the rod are
approximately evaluated by the following equations:30,33
D ) (x5/4)a
L ) a/(x8)
(1)
(2)
N ) Fa3NA/24M
(3)
where a is the unit cell parameter, NA Avogadro’s number, M
molecular mass, and F the density (which is assumed to be 1
g/cm3 ()10-21 g/nm3)). The values of D, L, and N were
estimated for three ANBC-n homologues, ANBC-16, -18, and
-26, and are 5.82 nm, 3.68 nm, and 15.8 nm-1 for n ) 16 (at
457 K), and 6.04 nm, 3.82 nm, and 16.7 nm-1 for n ) 18 (at
455 K), and 6.88 nm, 4.35 nm, and 17.2 nm-1 for n ) 26 (at
449 K), respectively. D and L increase linearly with the alkoxy
chain length n, whereas N is almost constant with n.30 In the
wide-angle region of the XRD pattern,30 three cubic phases
showed a diffuse halo originating in a short-range correlation
of about 0.4 to 0.5 nm, corresponding to the average transverse
distance between ANBC-n molecules. It is thus visualized that
each slice of the rod with a thickness of ∼0.4 nm contains about
three ANBC-n dimers for n ) 16, 18, and 26. This molecular
picture within the rod micelles of the cubic phase was first
proposed by Guillon and Skoulios for n ) 16,45 and the present
work has revealed that their model is realized independent of n
for all ANBC-n homologues. Consequently, the structure of the
CubII phase and its molecular organization of ANBC-26 are
essentially the same as those of n ) 16 and 18, except for the
size aspect of the rod micelles, which is governed by the alkoxy
chain length n.
Figure 8. DSC traces (5 K min-1) of ANBC-26, recorded under various
thermal cycles. Cycle 1, heated to 498 K in the I2 phase, and cooled to
293 K; cycle 2, heated to 413 K in the CubI phase, and cooled to 293
K; cycle 3, heated to 448 K in the CubII phase, cooled to 373 K at
which the CubII phase was observed, and heated again to 498 K; cycle
4, heated to 448 K in the CubII phase, cooled to 403 K, a temperature
at which the CubI phase would be observed on heating, and heated
again to 498 K. Notations of phases are the same as in Figure 1.
with increasing temperature; for example, the number estimated
decreases by 6% from 396 to 419 K in the CubI phase region,
which can be said to reflect a fluid nature of the cubic phase.
In hexagonal columnar mesophases, a similar negative thermal
expansion was reported in the direction perpendicular to the
column,43,44 but not known whether the expansion is also
negative along the columnar direction.
A question arises whether the CubI phase is metastable. To
answer this question, we investigated the phase behavior of
ANBC-26 under different thermal cycles, by using DSC, POM,
and XRD. The result is shown in Figure 8. The cycle 1, heating
to and cooling from 498 K in the I2 phase region, is our routine
DSC measurement for ANBC-n homologues, the result of which
has already been described in Figure 1. When heating was
Structure of the I1 Phase. Figure 9 shows XRD patterns
(open circles) as a function of temperature in the vicinity of

10202 J. Phys. Chem. B, Vol. 104, No. 44, 2000
Kutsumizu et al.
“lattice melting”; they considered that the CubD phase is made
up of spherical micelles and leaves the remnants of the CubD
lattice or clusters of ANBC-n molecules in the next I1 phase
before the complete melting. The possibility of the existence
of such weak organization in the I1 phase was supported by the
XRD studies, which were first carried out by Gray et al. for n
) 1613 and recently by us for n ) 15, 18-20, 22, and 26 also.30
The XRD photos exhibit a sharp diffraction ring at a small angle,
corresponding to a spacing ∼30% shorter than the length of
ANBC-n dimer, together with a diffuse ring at a wide angle. In
the present work, the former small-angle ring was observed as
a halo at q ) q* in Figure 9, and the temperature dependence
of the correlation length has provided more clear evidence of a
weak organization in the I1 phase and its vanishing into the I2
phase.
In A-B diblock copolymer systems, a similar phenomenon
termed correlation hole effect has often been seen in a short
temperature interval above the order-disorder transition tem-
perature, where it is explained that each chain forms a globule
with a radius of gyration Rg, not so much entangled with other
chains, and thus the density fluctuations with the distance scale
larger than Rg are virtually reduced by the conformational
constraint, bringing about a scattering peak at q* (∼1/Rg).49-51
In the present low-molecular-mass system, the same type of
globules would be formed only by the aggregation of several
ANBC-n dimers, that is, the formation of molecular clusters.
Strictly speaking, the I1 phase is not a liquid but should be
assigned reasonably to an intermediate phase leading to a true
liquid phase, just like Blue phases52 in cholesteric liquid
crystalline materials, and so the code “I” may not be appropriate.
However, it is worth noting that the reverse process from the I2
to the I1 phase on cooling exhibits an exothermic DSC peak
corresponding to the endothermic peak on heating (see Figure
1), and the I1 phase should not be discarded as a pretransitional
effect.
Figure 9. SAXS patterns (open circles) as a function of temperature
in the vicinity of the CubII-to-I1 phase transition temperature and above
on heating. The solid curves represent the fit by the Ornstein-Zernike
type function (see text), and the inset shows the temperature dependence
of the correlation length ê (filled circles) as obtained from the fits.
the CubII-to-I1 phase transition temperature and above. At the
CubII-to-I1 phase transition (at 467 K by DSC, Tc), a sharp (211)
reflection peak observed in the CubII temperature region
disappeared to leave a broad halo at around q ) 1.4 nm-1. As
temperature increases, this halo becomes broader, initially
rapidly in the temperature region of the I1 phase and into the I2
temperature region gradually. As shown as solid curves in the
figure, this halo was reproduced well by the following Ornstein-
Zernike type function,46,47
Structure of the CubI Phase with Im3m Symmetry. On
the basis of rod-connectors models illustrated in Figure 6, a
difference between Im3m and Ia3d cubic structures lies in the
way the constituent rods are connected to form a network, and
at first sight, the phase transition from the CubI phase with
symmetry Im3m to the CubII phase with symmetry Ia3d might
be regarded as a simple re-combination of those rods from
6-by-6 to 3-by-3 connections. With the same concept, the
number of connecting rods for the SmC phase may be regarded
as ∞, and 0 for the I1 phase, and thus, this concept seems to be
reasonable in that the number decreases successively along with
the phase sequence of SmC-CubI-CubII-I1 on heating. The
reality is, however, not so simple, and the size aspect of the
rods in the CubI phase is not the same as that in the CubII
phase.
I(q) ) I(q*)/[1+ê2(q - q*)2]
(4)
where ê is the correlation length associated with a short-range
fluctuation and q* is the q value at which the halo shows a
maximum. Here, the halo is so broad that instrumental smearing
is negligible. The inset plots the values of ê obtained from the
fits as a function of temperature. The plots are roughly divided
into two straight lines, which are crossed at 476 K, slightly
higher temperature than the I1-to-I2 phase transition temperature
determined by DSC (473 K). At temperatures in the I1 phase,
the value of ê is of the order of 6 nm, close to the diameter D
of rod micelles constructing the CubII lattice (described above),
which strongly suggests the existence of molecular clusters
probably containing several ANBC-n dimers in each. Into the
I2 phase region, then, the value of ê is around 3-4 nm, almost
corresponding to the distance between two neighboring ANBC
dimers, which is in good agreement with the fact that this phase
is a normal isotropic liquid.
As already mentioned, in the CubI phase with a lattice
constant of a, the distance between two 6-by-6 connection
points, for example, between (000) and (a/2,a/2,a/2), is very
close to three times SmC layer thickness. Here, one SmC layer
thickness corresponds to one dimer length of ANBC-26 whose
alkoxy chains would be thermally shrunk (or curled). It is,
therefore, suggested that in the CubI phase, each rod with a
diameter roughly corresponding to the ANBC-26 dimer length
is enrobed with an additional layer of ANBC-26 dimers, if we
wish to preserve the rod-connectors model. A similar molecular
organization picture has been proposed by Levelut and Clerc
for an Im3m cubic phase of ACBC-18.23 They postulated that
the location of methyl end groups constitutes three continuous
surfaces; one was described by the so-called infinite periodic
In connection with a broad hump observed by DSC above
the clearing temperature Tc, the assignment of the I1 phase of
ANBC-n homologues with shorter alkoxy chains such as n )
16 has long been attracted and disputed.11,13,19,21,22,24,25,48 For
example, Goodby et al.48 tried to explain this phase in terms of

Phase Transitions of ANBC-26
J. Phys. Chem. B, Vol. 104, No. 44, 2000 10203
The temperature variation of the alkyl chain area (A) on
cooling is monotonic and the obtained trend A(SmC) < A(CubII)
< A(Hex) in going from lower temperature is reasonable. Here,
the area in the SmC phase is denoted as A(SmC) and the areas
for other phases likewise. On heating, however, A(CubI) is
slightly (∼10%) larger than A(CubII) of the CubII phase that
appears at the high-temperature side, but considering the
situation on selecting the model for the Im3m-CubI phase, the
result may be regarded as an indication of A(CubI) ≈ A(CubII).
Hence, the formation of the CubI phase between the SmC and
CubII phases could be understood in terms of mean surface area
governed by thermal motion.
How do we understand the fact that the Im3m-CubI structure
can be formed from the preceding SmC layer structure on
heating but not from the Ia3d-CubII structure on cooling? In
thermotropic liquid crystalline systems, there have been several
reports revealing that the formation of cubic phases is strongly
affected by the preceding phase structure on thermal cycle.
Strontium stearate has an alkyl chain extending at both sides
of the -COOSrOOC- core part, very similar to the alkoxy
chains of ANBC-26 dimer. This compound is well-known for
exhibiting an Ia3d cubic phase (usually labeled Q) in a certain
temperature range on both heating and cooling,33 but it was
recently revealed that the phase formed on cooling from a
hexagonal columnar phase (labeled H) is more dynamically
disordered than that formed from a rhombohedral phase (labeled
R) on heating.60 In the ANBC-n homologue with n ) 15, where
the length of the alkoxy chain is critical for forming the CubD
phase with Ia3d symmetry, the appearance itself is strongly
affected by thermal cycles used.26 A hydrazine derivative, 1,2-
bis(4-n-octyloxybenzoyl)hydrazine, exhibits a cubic phase with
Ia3d on both heating and cooling and it was pointed out that
the formation of nuclei of the cubic structure in the preceding
phase would be important for the appearance of the phase.61
These dependences of cubic formation on the thermal cycle,
not seen for lyotropic liquid crystalline systems, may be char-
acteristic of thermotropic liquid crystalline systems, especially
one-chain systems with a long aliphatic chain at each terminal.
To understand such characteristic of one-chain systems and
confirm our hypothesis, the structural investigation of a binary
system of ANBC-14 and n-hexadecane is of much interest. As
expected from a report by Sorai, Saito, and their collaborators
for a binary system of ANBC-16 and n-tetradecane,24 in the
ANBC-14-n-hexadecane system, n-hexadecane as well as the
alkoxy chain of ANBC-14 act as a solvent to fill the remaining
space outside the core parts of rod micelles in the cubic structure
and contribute to the effective number n* of carbon atoms when
the system is regarded as a one-component system. In fact, it
has been found that the phase diagram with respect to n* is
very similar to that of pure ANBC-n system versus n. In the
binary system, the aliphatic component played by both n-
hexadecane and the alkoxy chain of ANBC-14 would occupy
the space more freely, without any conformational restriction;
the re-organization on varying temperature would be regarded
simply as a process of change in the interface such as a layer
deformation, not requiring to take the interactions within “layer”
into consideration. The investigation using XRD is in due course
in our laboratories and will be published elsewhere.
Figure 10. Temperature dependence of alkyl chain area in nm2; filled
and open symbols represent the data recorded on heating and cooling,
respectively. Sections shown in the upper and lower parts represent
the phase transition temperatures on heating and cooling, respectively.
minimal surface (IPMS)53-56 of type P (P-surface), and the other
two were constructed by a space foliation,57 where the latter
two correspond to inner surfaces of the doubled circular rods
and sandwich the former P-surface. The total interface area (S),
which is the sum of the areas of three surfaces, was computed
to result in the relation S ) 5.39a2, whereas the corresponding
relation for an Ia3d cubic IPMS (G-surface) is S ) 2.453 21/3a2
) 3.091a2.53,56 Provided that all the surfaces are continuous
everywhere in the unit cell as usually postulated for IPMS, the
minimum layer thickness encountered within their cubic cell
model is a/6 () 3.3 nm), which seems too narrow compared
with the ANBC-26 dimer length (8.5 nm) and the SmC layer
thickness (∼6 nm), even if the “curling” of the alkoxy chains
is taken into consideration. At this point, their model is
somewhat questionable, and also in our model, it is more
probable that the diameter of the rod is not constant but
modulated in the longitudinal direction. Probably, the real cubic
phase has some defects at which the surface is discontinuous.
Anyway, if we notice that the CubI phase still has some regions
of lamellar-like (but not lamellar) structure in the outside layer
of the doubled circular rods, it is reasonable that the Im3m-
CubI phase is an intermediate structure between the SmC
lamellar and the Ia3d-CubII structures. Qualitatively, this implies
that the average curvature of the methyl end surfaces of the
Im3m-CubI phase lies between those of the SmC (which is 0)
and the Ia3d-CubII phases (see also Figure 11).
Phase Transitions from the SmC to the CubI and from
the CubI to the CubII Phase. With analogy of the formation
of lyotropic cubic phases, one driving force to deform the SmC
lamellar structure might be considered thermal motion of alkyl
chain ends, which increases the actual bulkiness with increasing
temperature.5 The alkyl chain area (A) is a parameter reflecting
such thermal motion, defined by Levelut and Fang18 as
A ) S/(N0/2)
(5)
where S is the total surface area covered by methyl end groups
and N0 is the number of the methyl groups present at both sides
of the surface; such surface is called mean surface (between
two interfacial surfaces) by Charvolin,58 or neutral surface by
Caffrey et al.59 Figure 10 shows plots of the estimated alkyl
chain area as a function of temperature, where the relation S )
5.39a2 was used for the estimation in the CubI phase region,
although this model is actually one option for the Im3m-CubI
phase.
Concluding Remarks
Despite the very simple chemical structure of ANBC-26, the
polymorphism is very unique and includes two cubic meso-
phases with Im3m and Ia3d symmetries on heating and a weak
organization made up of the Ia3d lattice remnants in the I1 phase.

10204 J. Phys. Chem. B, Vol. 104, No. 44, 2000
Kutsumizu et al.
Figure 11. Schematic phase behavior of ANBC-26.
The reverse transition from the Ia3d-CubII to the Im3m-CubI
phase could not be observed on cooling, and instead, a Hex
phase newly appears above the temperature region of the Ia3d-
CubII phase. Figure 11 summarizes the change of molecular
organization of ANBC-26 on varying temperature (T).
The structure of the Ia3d-CubII phase is depicted as two sets
of interwoven networks of rod micelles each of which is
connected 3-by-3. The diameter of the rod micelles roughly
corresponds to the length of ANBC-26 dimer, and the molecular
organization is well understood by the picture proposed by
Guillon and Skoulios.
The structure of the Im3m-CubI phase remains to be further
reflected on, but at this stage, it is considered that this cubic
phase has also a paired network structure of rod micelles, where
the rods are connected 6-by-6 and each rod is a doubled circular
rod, that is, a rod with a diameter roughly corresponding to the
dimer length of ANBC-26 is enrobed with an additional layer
of ANBC-26 dimers. The outer layer may be regarded as more
lamellar-like than the inner rod, and in this respect, the Im3m-
CubI phase is an intermediate structure between the SmC
lamellar and Ia3d-CubII phases. In other words, qualitatively,
the average curvature of the methyl end surfaces of the Im3m-
CubI phase lies between those of the SmC and the Ia3d-CubII
phases.
Acknowledgment. We thank Dr. Satoshi Tanimoto at JAIST
for his kind help with the X-ray experiment using IP, and Prof.
Keiichi Moriya at Gifu University for his valuable experimental
suggestions. Thanks are also due to Ms. Yumiko Murase and
Eri Kumita at Instrumental Analysis Center, Gifu University,
for the measurements of MS spectra. This work was partly
supported by the Research Foundation for the Electrotechnology
of Chubu (R-11123). We are also grateful to the two reviewers
for their careful reading of the manuscript and for their
constructive criticism.
References and Notes
(1) de Gennes, P.-G. Angew. Chem., Int. ed. Engl. 1992, 31, 842.
(2) Diele, S.; Go¨ring, P. In Handbook of Liquid Crystals; Demus, D.,
Goodby, J., Gray, G. W., Spiess, H.-W., Vill, V., Eds; Wiley-VCH:
Weinheim, 1998; Vol. 2B, pp 887-900.
(3) Mariani, P.; Luzzati, V.; Delacroix, H. J. Mol. Biol. 1988, 204,
165.
(4) Bates, F. S.; Fredrickson, G. H. Annu. ReV. Phys. Chem. 1990, 41,
525.
(5) Israelachvili, J. N. Intermolecular and Surface Forces; Academic
Press: London, 1985.
(6) Gray, G. W.; Jones, B.; Marson, F. J. Chem. Soc. 1957, 393.
(7) Demus, D.; Kunicke, G.; Neelsen, J.; Sackmann, H. Z. Naturforsch.
1968, 23a, 84.
(8) Pelzl, G.; Sackmann, H. Symp. Chem. Soc., Faraday DiVision 1971,
5, 68.
It is worth noting that high-resolution SAXS experiments are
essential for distinguishing different types of cubic phases, and
thus, the cubic region of ANBC-n homologues other than
ANBC-26 may include more than one type of cubic phases,
which is currently being reinvestigated in our laboratories.
We have also shown in this paper that the I1 phase could be
characterized by Ornstein-Zernike type short-range fluctuation,
indicating the presence of molecular clusters consisting of
several ANBC-26 dimers in this phase. At this point, this phase
is an intermediate state into a true isotropic liquid I2.
In ANBC-n homologues, the simple molecular shape would
be advantageous when more detailed theoretical considerations
are made, and thus we anticipate that our findings in the present
work will lead to more complete understanding of cubic
molecular organization and its transformation to and from other
types, especially in thermotropic liquid crystalline systems.
(9) Diele, S.; Brand, P.; Sackmann, H. Mol. Cryst. Liq. Cryst. 1972,
17, 163.
(10) Tardieu, A.; Billard, J. J. Phys. (Paris) Coll. 1976, 37, C3-79.
(11) Demus, D.; Marzotko, D.; Sharma, N. K.; Wiegeleben, A. Krist.
Tech. 1980, 15, 331.
(12) Gray, G. W.; Goodby, J. W. Smectic Liquid Crystals; Leonard
Hill: Glasgow, 1984; pp 68-81, including also earlier references on
thermotropic cubic phases.
(13) Gray, G. W. In Zehn Arbeiten uber Flu¨ssige Kristalle; Kongress-
und Tagung-berichte der Martin-Luther-Universita¨t; Halle-Wittenberg, 1986;
pp 22-42.
(14) Etherington, G.; Leadbetter, A. J.; Wang, X. J.; Gray, G. W.;
Tajbakhsh, A. Liq. Cryst. 1986, 1, 209.
(15) Ukleja, P.; Siatkowski, R. E.; Neubert, M. E. Phys. ReV. 1988, A38,
4815.
(16) Etherington, G.; Langley, A. J.; Leadbetter, A. J.; Wang, X. J. Liq.
Cryst. 1988, 3, 155.
(17) Billard, J.; Zimmermann, H.; Poupko, R.; Luz, Z. J. Phys. (Paris)
1989, 50, 539, including references on other types of thermotropic cubic
phases.

Phase Transitions of ANBC-26
J. Phys. Chem. B, Vol. 104, No. 44, 2000 10205
(18) Levelut, A.-M.; Fang, Y. Coll. Phys., Coll. 1990, 51, C7-229.
(19) Kutsumizu, S.; Yamada, M.; Yano, S. Liq. Cryst. 1994, 16, 1109.
(20) Yamaguchi, T.; Yamada, M.; Kutsumizu, S.; Yano, S. Chem. Phys.
Lett. 1995, 240, 105.
(21) Kutsumizu, S.; Kato, R.; Yamada, M.; Yano, S. J. Phys. Chem., B
1997, 101, 10666.
(22) Tansho, M.; Onoda, Y.; Kato, R.; Kutsumizu, S.; Yano, S. Liq.
Cryst. 1998, 24, 525.
(23) Levelut, A.-M.; Clerc, M. Liq. Cryst. 1998, 24, 105.
(24) Saito, K.; Sato, A.; Sorai, M. Liq. Cryst. 1998, 25, 525.
(25) Sato, A.; Saito, K.; Sorai, M. Liq. Cryst. 1999, 26, 341.
(26) Kutsumizu, S.; Yamaguchi, T.; Kato, R.; Yano, S. Liq. Cryst. 1999,
26, 567.
(27) Sato, A.; Yamamura, Y.; Saito, K.; Sorai, M. Liq. Cryst. 1999, 26,
1185.
(28) Kutsumizu, S.; Ichikawa, T.; Nojima, S.; Yano, S. Chem. Commun.
1999, 1181.
(29) Kutsumizu, S.; Yamaguchi, T.; Kato, R.; Ichikawa, T.; Yano, S.
Mol. Cryst. Liq. Cryst. 1999, 330, 359.
(30) Kutsumizu, S.; Kobayashi, H.; Nakamura, N.; Ichikawa, T.; Yano,
S.; Nojima, S. Mol. Cryst. Liq. Cryst., in press.
(42) Paleos, C. M.; Kardassi, D.; Tsiourvas, D.; Skoulios, A. Liq. Cryst.
1998, 25, 267.
(43) Zamir, S.; Singer, D.; Spielberg, N.; Wachtel, E. J.; Zimmermann,
H.; Poupko, R.; Luz, Z. Liq. Cryst. 1996, 21, 39.
(44) Artzner, F.; Veber, M.; Clerc, M.; Levelut, A.-M. Liq. Cryst. 1997,
23, 27.
(45) Guillon, D.; Skoulios, A. Europhys. Lett. 1987, 3, 79.
(46) Feigin, L. A.; Svergun, D. I. Structure Analysis by Small-Angle
X-ray and Neutron Scattering; Plenum Press: New York, 1987; p 243.
(47) de la Cruz, M. O.; Sanchez, I. C. Macromolecules 1986, 19, 2501.
(48) Goodby, J. W.; Dummur, D. A.; Collings, P. J. Liq. Cryst. 1995,
19, 703.
(49) de Gennes, P.-G. Scaling Concepts in Polymer Physics; Cornell
University Press: Ithaca, New York, 1979.
(50) Nojima, S.; Roe, R.-J. Macromolecules 1987, 20, 1866.
(51) Nojima, S.; Kato, K.; Yamamoto, S.; Ashida, T. Macromolecules
1992, 25, 2237.
(52) For example, Goodby, J. W. In Handbook of Liquid Crystals;
Demus, D., Goodby, J., Gray, G. W., Spiess, H.-W., Vill, V., Eds; Wiley-
VCH: Weinheim, 1998; Vol. 1, pp 115-132.
(53) Schoen, A. H. NASA Tech. Note 1970, No. D-5541.
(54) Andersson, S.; Hyde, S. T.; Larsson, K.; Lidin, S. Chem. ReV. 1988,
88, 221.
(31) Longley, W.; McIntosh, T. J. Nature 1983, 303, 612.
(32) Larsson, K. Nature 1983, 304, 664.
(33) Luzzati, V.; Spegt, P. A. Nature 1967, 215, 701.
(34) Tamura, M.; Kochi, J. Synthesis 1971, 303.
(35) Friedman, L.; Shani, A. J. Am. Chem. Soc. 1974, 96, 7101.
(36) Nakamura, N.; Watanabe, A. Yukagaku 1976, 25, 485 (in Japanese).
(37) Gray, G. W.; Hartley, J. B.; Jones, B. J. Chem. Soc. 1955, 1412.
(38) Destrade, C.; Foucher, P.; Gasparoux, H.; Nguyen, H. T.; Levelut,
A.-M.; Malthete, J. Mol. Cryst. Liq. Cryst. 1984, 106, 121.
(39) Kutsumizu, S.; Ichikawa, T.; Nojima, S.; Yano, S. Unpublished
results.
(55) Hyde, S. T.; Andersson, S.; Ericsson, B.; Larsson, K. Z. Kristallogr.
1984, 168, 213.
(56) Clerc, M.; Dubois-Violette, E. J. Phys. II France 1994, 4, 275.
(57) Fogden, A.; Lidin, S. J. Chem. Soc., Faraday Trans. 1994, 90, 3423.
(58) Charvolin, J. Mol. Cryst. Liq. Cryst. 1991, 198, 145.
(59) Chung, H.; Caffrey, M. Biophys. J. 1994, 66, 377.
(60) Nishino, Y.; Yano, S.; Tansho, M.; Yamaguchi, T. Chem. Phys.
Lett. 1998, 296, 408.
(40) Ranc¸on, Y.; Charvolin, J. J. Phys. Chem. 1988, 92, 2646.
(41) Donnio, B.; Heinrich, B.; Gulik-Krzywicki, T.; Delacroix, H.;
Guillon, D.; Bruce, D. W. Chem. Mater. 1997, 91, 2951.
(61) Go¨ring, P.; Diele, S.; Fisher, S.; Wiegeleben, A.; Pelzl, G.;
Stegemeyer, H.; Thyen, W. Liq. Cryst. 1988, 25, 467.