Specific interactions in discotic liquid crystals

Article abstract of DOI:10.1039/b208923h

A series of novel mesogens have been prepared by a five-fold Sonogashira reaction of terminal acetylenes with a functionalized pentabromophenol derivative. The corresponding side-chain substituted polymers were prepared by an analogous polymer substitution reaction. The mesogens differ in the nature of the substituents, i.e. CH₂, O, S, SO₂ and CONH groups, linking five hexyl tails to the core. A wide range of mesophases and corresponding transition temperatures have been detected, varying from low melting nematic phases to highly stable columnar phases. The wide spread in phase behaviour is described in terms of specific intermolecular interactions. The addition of planar electron deficient molecules resulted in the formation of charge transfer complexes. The observed stabilisation or destabilisation of the mesophases is explained by considering the complexation strength of the complex as well as steric factors.

Full text of DOI:10.1039/b208923h

View Article Online / Journal Homepage / Table of Contents for this issue
Specific interactions in discotic liquid crystals
Paul H. J. Kouwer,{ Wolter F. Jager, Wim J. Mijs and Stephen J. Picken*
Delft University of Technology, Polymer Materials and Engineering, Julianalaan 136, 2628
BL Delft, The Netherlands. E-mail: s.j.picken@tnw.tudelft.nl
Received 13th September 2002, Accepted 18th December 2002
First published as an Advance Article on the web 17th January 2003
A series of novel mesogens have been prepared by a five-fold Sonogashira reaction of terminal acetylenes with
a functionalized pentabromophenol derivative. The corresponding side-chain substituted polymers were
prepared by an analogous polymer substitution reaction. The mesogens differ in the nature of the substituents,
i.e. CH2, O, S, SO2 and CONH groups, linking five hexyl tails to the core. A wide range of mesophases and
corresponding transition temperatures have been detected, varying from low melting nematic phases to highly
stable columnar phases. The wide spread in phase behaviour is described in terms of specific intermolecular
interactions. The addition of planar electron deficient molecules resulted in the formation of charge transfer
complexes. The observed stabilisation or destabilisation of the mesophases is explained by considering the
complexation strength of the complex as well as steric factors.
the molecules are stacked into columns and the columns
Introduction
possess two-dimensional long-range positional order. The
intercolumnar order (hexagonal, rectangular, oblique) and
the intracolumnar order (ordered, disordered, tilted) are
variables that are indicated in suffixes.8 As an example, a
schematic presentation of the Colhd phase is shown in Fig. 1.
So-called plastic columnar (Colp) phases (not shown) may
appear from cooling columnar mesophases. They are char-
acterised by a reduced translational mobility, but the molecules
show unhindered in-plane rotational motions.9
The liquid crystalline properties of mesogens are strongly
related to their chemical structure. By changing the substitu-
tion pattern on a specific core a wide range of mesomorphic
properties can be obtained.3a The phase behaviour can also be
modified without chemically altering the mesogen, for example
by the introduction of various end-groups at the terminal
position(s) of the tails or spacers. A particularly strong effect is
observed when the mesogens are attached as side-groups to a
polymer chain.10 In general, the constrained topology of the
polymer backbone suppresses crystallisation and hence, side-
chain liquid crystalline polymers often show a much more
varied mesophase behaviour than their low molar mass
counterparts. Alternatively, the polymer chain may stimulate
the formation of mesophases that are not observed in the
monomers. In these systems, it is the backbone that effectively
induces the mesophases.7,10a Moreover, the phase behaviour
Supramolecular chemistry—‘‘the chemistry beyond the mole-
cule’’—has a large number of non-covalent interactions at
its disposal.1 These secondary interactions include forces like
(i) electrostatic interactions (ion–ion, ion–dipole, dipole–
dipole), (ii) hydrogen bonding, (iii) p–p stacking, (iv) van der
Waals forces (dispersion and induction forces) and (v)
hydrophobic or, more general, solvophobic effects. A combi-
nation of these non-covalent interactions is used in nature to
build complex structures that can show multiple levels of
organisation. The interest in self-organisation and self-
assembly in synthetic systems has grown considerably. This
is displayed in the numerous contributions discussing the
effects of the wide range of secondary interactions available.2
Liquid crystals are self-assembling by nature. In the
mesophase the molecules possess orientational and/or one-
or two-dimensional positional order, while retaining their
mobility. Secondary interactions can improve, or oppositely,
diminish the extent of organisation (i.e. extent of order) in these
materials. This makes liquid crystals ideal model compounds
with which to study the effects of various interactions on the
molecular organisation.
The relatively new class of disc-shaped liquid crystals are
becoming increasingly interesting, both from a scientific and an
application point of view.3 Currently, discotic liquid crystals
are applied as compensation layers in LCD technology4a,b
and interesting applications are expected from their (quasi)
one-dimensional conduction behaviour4c,d in photovoltaic
applications.
Disc-shaped liquid crystals organise into various meso-
phases, some of which are schematically depicted in Fig. 1. In
the nematic discotic (ND) phase,5 the mesogens only possess
orientational order and lack any positional order. The short
axes of the discs are aligned to the director n˜. Two other
nematic phases have been reported, the nematic columnar6
(NCol) and the nematic lateral7 (NL) phase. Both phases are
characterised by a locally well-defined organisation giving rise
to super-structures, which in turn show a nematic arrangement.
However, columnar phases are the most frequently observed
mesophases for disc-shaped mesogens. In the columnar phase,
Fig. 1 Schematic illustration of common mesophases observed for disc-
shaped liquid crystals: ND ~ nematic discotic and Colhd ~ columnar
hexagonal disordered. The molecules (in the ND phase) or assemblies of
molecules (Colhd phase) align to the director n˜.
{Current address: University of Hull, Department of Chemistry,
Cottingham Road, Hull, UK HU6 7RX. E-mail: p.h.j.kouwer@
hull.ac.uk.
458
J. Mater. Chem., 2003, 13, 458–469
This journal is # The Royal Society of Chemistry 2003
DOI: 10.1039/b208923h

View Article Online
can be tuned by variation of the degree of substitution along
the polymer backbone.11
shifts are reported in ppm relative to TMS. Molecular weights
were determined by gel-permeation chromatography (GPC) in
THF against narrow polystyrene standards. The thermal
properties of the materials were investigated by a Perkin
Elmer DSC 7 Differential Scanning Calorimeter (in a nitrogen
atmosphere) and a Jenapol optical polarizing microscope,
equipped with a Mettler FP82 HT hot stage and a Mettler
FP80 Central Processor. X-Ray diffraction analysis was
performed, using a Siemens Kristalloflex 710D X-ray generator
In this contribution, we present the synthesis and phase
behaviour of a series of novel mesogens with a pentakis-
(phenylethynyl)phenoxy core with various groups linking the
tails to the aromatic core, as well as some mesogen-substituted
polyacrylates, see Fig. 2. The strong effect of the linking groups
(CH2, O, S, SO2 and CONH) on the mesomorphic properties is
described in terms of various specific interactions between the
mesogens. Additional inter-disc interactions were introduced
by doping the mesogens with disc-shaped electron acceptors,
˚
with graphite monochromated Cu-Ka radiation (l ~ 1.54 A)
equipped with a Bruker HI-STAR area detector. The samples
were oriented in a magnetic field by a SmCo permanent magnet
with a field of about 1.5 T, and heated using a custom built
capillary heating element.
resulting in the formation of charge transfer complexes.12
A
distinctive advantage of such complexes is that by balancing
subtle interactions, such as steric and electronic effects, a wide
range of liquid crystalline properties can be obtained from only
a limited number of materials.7
4’-Heptylacetophenone (1a). Dichloromethane (150 mL)
was cooled to 0 uC and under vigorous stirring AlCl3 (22.3 g,
0.17 mol) was added. After 15 minutes, acetyl chloride (13.1 g,
0.17 mol) was added and the mixture was stirred until a clear
solution was obtained. Heptylbenzene (25.2 g, 0.13 mol) was
added dropwise. The mixture was stirred for 2 hours at 0 uC
and allowed to warm to room temperature. After TLC
indicated complete conversion, the reaction mixture was
poured into 300 mL 1 M HCl ice–water solution. CH2Cl2
was evaporated under reduced pressure and the aqueous
solution was extracted with diethyl ether (36). The combined
organic layers were washed with water and brine, dried over
MgSO4 and the solvent was evaporated, yielding a slightly
yellow oil (29.2 g, 99%) which was used without further
Experimental
Materials
All materials were used as purchased without further puri-
fication unless mentioned otherwise. Pyridine was distilled
from CaH2 and CH2Cl2 from P2O5. The synthesis of mesogens
CH3–D (laterally methyl substituted mesogen), the series of
corresponding side-chain polymers P(CH3–D) and 4CH2–D
(laterally pentyl substituted mesogen) have been described
before.10a,13,14 Poly(acryloyl chloride) (PAC) was prepared as
described previously.10a It was stored at 220 uC without
isolating the polymer from the reaction mixture. The crude
polymer mixture was used as such in the polymer analogous
reactions. Electron acceptors 2,4,7-trinitrofluoren-9-one (TNF)
and 2,4,7-trinitrofluoren-9-ylidenemalonitrile (DCM–TNF)
are commercially available.
1
purification. H NMR (60 MHz, CDCl3): d 7.4 and 7.1 (2 6
dd, 2 6 2H, CH aromatic), 2.6 (t, 2H, CH2Ph), 2.5 (s, 3H, CH3
acetyl), 1.2–1.9 (m, 10H, CH2), 0.9 (t, 3H, CH3 tail).
4’-Hexyloxyacetophenone (1b). Amixtureof4’-hydroxyaceto-
phenone (13.6 g, 0.1 mol), 1-bromohexane (22 g, 0.13 mol),
K2CO3 (26 g, 0.19 mol) and 150 mL DMF was stirred at 120 uC
until TLC indicated full conversion. The reaction mixture was
cooled and poured into a 500 mL 1 M HCl solution and
extracted with diethyl ether (36). The combined organic layers
Instrumental
Nuclear magnetic resonance (NMR) spectra were taken on a
Varian VXR 300 or VXR 400 MHz spectrometer. Chemical
Fig. 2 Low molar mass mesogens 6X–D and side-chain polymers P(6CH2–D) and P(6O–D).
J. Mater. Chem., 2003, 13, 458–469
459

View Article Online
were washed with water and brine, dried over MgSO4 and the
solvent was evaporated. Pure 3b was obtained by distillation
(Tb ~ 137–138 uC at 0.1 mbar) as a colourless oil. Yield: 17.9 g,
81%. 1H NMR (60 MHz, CDCl3): d 7.8 and 6.8 (2 6 dd, 2 6
2H, CH aromatic), 4.0 (t, 2H, CH2O), 2.5 (s, 3H, CH3 acetyl),
2.0–1.2 (m, 8H, CH2), 0.9 (t, 3H, CH3 tail).
2.97 (s, 1H, HCMC), 1.28–1.82 (m, 8H, CH2), 0.89 (t, 3H, CH3).
13C NMR (300 MHz, CDCl3): d 159.6, 133.6, 114.5, 113.9
(aromatic), 83.8 (HCMC-), 75.6 (HCMC-), 68.1 (CH2O), 31.6–
22.6 (CH2), 14.0 (CH3).
4-Hexylsulfanylphenylacetylene (3c). The procedure described
for the synthesis of 3b was used, yielding almost quantitatively, the
intermediate2c. 1HNMR(60MHz, CDCl3): d 10.2 (d, 1H, CHO),
7.3 and 7.7 (2 6 dd, 2 6 2H, CH phenyl), 6.7 (d, 1H, CH ethene),
3.0 (t, 2H, CH2O), 2.0–1.2 (m, 8H, CH2), 0.9 (t, 3H, CH3 tail).
Subsequent elimination under basic conditions yielded 3c. After
purification, as described for 3b, the acetylene was obtained as a
colourless liquid in 62% yield. 1H NMR (300 MHz, CDCl3): d 7.37
and 7.20 (2 6dd, 2 62H, CHaromatic), 3.06(s, 1H, HCMC), 2.90
4’-Hexylsulfanylacetophenone (1c). A mixture of 4’-fluoro-
acetophenone (27.6 g, 0.21 mol), hexane-1-thiol (24.8 g, 0.21 mol)
and K2CO3 (47 g, 0.42 mol) in dimethyl sulfoxide (150 mL) was
stirred overnight at 50 uC. After cooling, the mixture was
acidified with 20% HCl solution (250 mL) and extracted with
diethyl ether (26). The combined organic layers were washed
with a diluted HCl solution and brine, dried over MgSO4 and
evaporated until dry. The resulting light yellow oil, quickly
solidifying upon standing, was used without further purifica-
tion. 1H NMR (60 MHz, CDCl3): d 7.9 and 7.3 (2 6 dd, 2 6
2H, CH aromatic), 3.0 (t, 2H, CH2O), 2.6 (s, 3H, CH3 acetyl),
1.9–1.2 (m, 8H, CH2), 0.9 (t, 3H, CH3 tail).
(t, 2H, CH2S), 1.63–1.24 (2m, 8H, CH2), 0.88 (t, 3H, CH3). 13
C
NMR (300 MHz, CDCl3): d 138.9, 132.4 127.5, 118.8 (aromatic),
83.4 (HCMC-), 76.6 (HCMC-), 32.8–22.5 (CH2), 14.0 (CH3).
Heptanoic acid (4-trimethylsilanylethynylphenyl) amide (4). Hepta-
noyl chloride (10.0 g. 67 mmol) was added over 15 minutes to a
cooled (0 uC) mixture of 4-iodoaniline (14.7 g, 67 mmol),
triethylamine (10 mL) and dry THF (100 mL) under an argon
atmosphere. After stirring for 16 hours at room tempera-
ture Pd(II)Cl2(PPh3)2 (400 mg, 0.57 mmol), PPh3 (300 mg,
1.14 mmol), ethynyltrimethylsilane (8.8 g, 90 mmol), triethyl-
amine (10 mL) and piperidine (20 mL) were added to the
reaction mixture. The mixture was stirred for 1 hour and a
solution of CuI (200 mg, 1.1 mmol) and LiBr (1.0 g, 11 mmol)
in dry THF (5 mL) was added via syringe. After stirring for 24
hours at room temperature, the reaction mixture was poured
into diluted HCl and extracted with CH2Cl2 (36). The
combined organic layers were washed with a 0.5 M HCl
solution (26) and water (26) and the solvent was evaporated.
The crude product was obtained as a dark brown solid and was
directly used for the next step, without further purification.
4-Heptylphenylacetylene (3a). To PCl5 (29.2 g, 0.14 mol) was
added dropwise the crude 1a (29.2 g, 0.13 mol) under gentle
stirring at a rate such that the temperature of the reaction
mixture reached 70 uC. After the reaction had ceased, the
mixture was stirred overnight at room temperature. POCl3 was
removed under reduced pressure and DMSO (150 mL) was
slowly added, keeping the temperature below 30 uC. During
1 hour t-BuOK (31.4 g, 0.28 mol) was added in small portions
and after addition, the mixture was heated at 70 uC until TLC
indicated complete conversion. The dark reaction mixture was
poured into a 2 M HCl (200 mL) solution and extracted with
diethyl ether (36). The combined organic layers were washed
with water and brine, dried over MgSO4 and the solvent was
evaporated. After distillation (Tb ~ 98 uC at 0.4 mbar), pure
3a (15.5 g, 58%) was obtained as a colourless liquid. 1H NMR
(300 MHz, CDCl3): d 7.38 and 7.10 (2 6 dd, 2 6 2H, CH
aromatic), 3.00 (s, 1H, HCMC), 2.57 (t, 2H, CH2Ph), 1.58–
1.20 (m, 10H, CH2), 0.89 (t, 3H, CH3). 13C NMR (300 MHz,
CDCl3): d 149.3, 132.6, 128.4, 119.3 (aromatic), 83.9 (HCMC-),
76.5 (HCMC-), 35.9–22.7 (CH2), 14.1 (CH3).
Heptanoic acid (4-ethynylphenyl) amide (3d). To a solution
of 4 (all from the previous reaction) in methanol (100 mL)
was added a solution of NaOH (7.0 g, 175 mmol) in methanol
(250 mL). The reaction mixture was stirred at room
temperature until TLC indicated complete consumption of 4.
Half of the solvent was removed under reduced pressure and
the mixture was precipitated into water. The product was
filtered, washed with water, dried, dissolved in diethyl ether and
the solution was filtered over Al2O3. Pure 3d was obtained
after column chromatography (SiO2, eluent CH2Cl2) as a
white solid in an overall yield of 8.9 g, 58%. 1H NMR (CDCl3):
d 7.65 (s, 1H, amide), 7.50 and 7.20 (2 6 dd, 4H, CH aromatic),
3.04 (s, 1H, HCMC), 2.34 (t, 2H, CH2N), 1.75–1.20 (m, 8H,
CH2 aliphatic), 0.88 (t, 3H, CH3). 13C NMR (CDCl3): d
171.81 (CLO), 138.50, 132.90, 119.42, 117.54 (aromatic), 83.40,
76.75 (acetylene), 37.80, 31.54, 28.93, 25.54, 22.49, 14.02
(aliphatic).
4-Hexyloxyphenylacetylene (3b). Phosphorus oxychloride
(30 mL) was added dropwise to cold DMF (120 mL). The
solution was allowed to warm to room temperature, stirred for
15 minutes and cooled again to 0 uC. Crude 1b (20 g, 100 mmol)
dissolved in DMF (30 mL) was added over 15 minutes. The
mixture was heated to 50 uC for 1 hour and the warm solution
was slowly precipitated in a 20% aqueous sodium acetate
solution (500 mL). After cooling, the suspension was extracted
with ether (36). The combined organic layers were washed
with water (26) and brine and ether was evaporated. The
intermediate product 2b was obtained as a yellow oil and was
11-[Pentakis(4-heptylphenylethynyl)phenoxy]undecan-1-ol(6CH2–
D). A mixture of 3a (32 mmol), 11-(pentabromophenoxy)-
undecan-1-ol (3.2 mmol), Pd(II)Cl2(PPh3)2 (200 mg), PPh3
(150 mg) and triethylamine (50 mL) was stirred at 50 uC under
a continuous flow of Ar. After 1 hour the mixture was cooled to
room temperature and a solution of 100 mg CuI and 400 mg
LiBr in 2 mL THF was added. The reaction mixture was
refluxed for 12 hours in an argon atmosphere. The mixture was
cooled and added to a mixture of 400 mL hexane and 100 mL
ether and the solids were filtered. The dark brown solution was
cooled to 220 uC and the crude product was filtered from the
solution. Pure 6CH2–D was obtained by repeated column
chromatography (using silica gel and toluene : ether 40 : 1 as
eluent) and crystallisations from cold (–20 uC) hexane
(including rapid cold filtration to prevent melting of the
substance). Yield: 38% as a yellow solid, crystallised from the
1
directly used for the synthesis of 3b without purification. H
NMR (60 MHz, CDCl3): d 10.1 (d, 1H, CHO), 7.5 and 6.9 (2 6
dd, 2 6 2H, CH phenyl), 5.5 (d, 1H, CH ethene), 4.0 (t, 2H,
CH2O), 2.0–1.2 (m, 8H, CH2), 0.9 (t, 3H, CH3 tail). Crude 2b
was dissolved in refluxing dioxane (200 mL) and a 10% NaOH
solution (225 mL) with a catalytic amount of benzyltributyl-
ammonium chloride was added. The suspension was refluxed
for 40 minutes and dioxane was evaporated under reduced
pressure. The remaining mixture was extracted with ether (36)
and the combined organic layers washed with water (26) and
brine, dried over MgSO4 and the solvent evaporated. The
acetylene was purified by filtration over a 15 cm 300 mL plug of
silica gel with petroleum ether 40–60, yielding 9.7 g (48%
overall) of a colourless oil. 1H NMR (300 MHz, CDCl3): d 7.41
and 6.82 (2 6 dd, 2 6 2H, CH aromatic), 3.96 (t, 2H, CH2O),
460
J. Mater. Chem., 2003, 13, 458–469

View Article Online
1
melt. H NMR (300 MHz, CDCl3): d 7.55–7.49 and 7.19–7.14
0.87 (t, 15H, CH3). 13C NMR (CDCl3): d 161.86 (inner phenyl,
C-O), 139.90, 139.61 (outer phenyl, C-SO2), 132.19, 132.12,
132.07 (CH, outer phenyl), 128.50, 128.47, 128.44 (CH, outer
phenyl), 128.18, 128.10, 127.93 (outer phenyl, attached to
ethynyl), 129.20 ,123.87, 120.43 (inner phenyl), 99.33, 99.28,
96.06 (inner ethynyl), 89.77, 89.41, 87.18 (outer ethynyl), 75.59
(CH2OPh), 62.95 (CH2OH), 56.26 (CH2SO2), 32.78–22.29
(CH2 tails and spacer), 13.91 (CH3 tails). Elemental analysis for
C87H108O12S5: calculated C 69.38%, H 7.23%, O 12.75%, S
10.64%, found C 69.49%, H 7.29%, S 10.45%.
(m, 2 6 10H, CH aromatic), 4.36 (t, 2H, CH2OPh), 3.61 (t, 2H,
CH2OH), 2.63 (t, 10H, CH2Ph), 1.98–1.28 (m, 68H, CH2 tails
and spacer), 0.89 (t, 15H, CH3). 13C NMR (300 MHz, CDCl3):
d 160.27 (inner phenyl, C-O), 144.01, 143.89, 143.70 (outer
phenyl, attached to alkyl), 131.79, 131.67, 131.61 (CH, outer
phenyl), 128.53 (CH, outer phenyl), 128.78, 124.09, 120.11
(inner phenyl), 120.75, 120.57, 120.52 (outer phenyl, attached
to ethynyl), 99.55, 99.39, 97.32 (inner ethynyl), 87.15, 86.66,
84.15 (outer ethynyl), 74.76 (CH2OPh), 63.09 (CH2OH), 36.03
(CH2Ph, tails), 32.9–22.7 (CH2 tails and spacer), 14.11 (CH3
tails). Elemental analysis for C92H118O2: calculated C 87.98%,
H 9.47%, O 2.55%, found C 87.38%, H 9.23%.
11-[Pentakis(4-heptanoylaminophenylethynyl)phenoxy]undecan-1-
ol (6Am–D). The procedure described for the synthesis of
6CH2–D was used with a reaction time of 8 hours. The reaction
mixture was diluted with CH2Cl2 and washed with a 0.5 M
HCl solution. Column separation (using silica gel and
CH2Cl2 : MeOH 20 : 1 as eluent) yielded a crude product in
y40% yield. A fraction was applied to subsequent column
separation (using silica gel and CH2Cl2 with an increasing
MeOH content of 2 to 5%), which yielded 6Am–D in 9% yield
as a yellow, fluorescent solid. Compound 6Am–D is hardly
soluble in CH2Cl2, (or CHCl3) and MeOH, but can be
dissolved easily in combinations of these solvents or polar
organic solvents, such as DMSO. 1H NMR (CDCl3 1 10%
methanol-d4): d 7.51–7.40 (m, 2 6 10H, CH aromatic), 4.32
(t, 2H, CH2OPh), 3.56 (t, 2H, CH2OH), 2.40 (t, 10H,
CH2CONH), 1.22–1.87 (m, 58H, CH2 tails and spacer), 0.90
(t, 15H, CH3). Due to strong aggregation, even in dilute
solutions, the 13C NMR is not as detailed as for the other
mesogens. 13C NMR (CDCl3 1 10% methanol-d4): d 173.5
(CLO), 138.8 (outer phenyl C-NH), 132.6–132.4 and 120.1
(outer phenyl CH), 118.9 (outer phenyl, attached to ethynyl),
99.1, 99.0, 96.2 (inner ethynyl), 87.8, 87.6, 84.6 (outer ethynyl),
77.9 (CH2OPh), 62.6 (CH2OH), 37.5–22.7 (CH2 spacer and
alkyl tails), 14.1 (CH3).
11-[Pentakis(4-hexyloxyphenylethynyl)phenoxy]undecan-1-ol
(6O–D). The procedure described for the synthesis of 6CH2–D
was used with a reaction time of 8 hours. After crude work up
(yield of y65%), pure 6O–D was obtained by repeated column
chromatography (using silica gel and toluene : ether 20 : 1 as
eluent) and crystallisation from ether : hexane (1 : 10). Yield
40% as a light yellow, bright fluorescent solid. 1H NMR
(300 MHz, CDCl3): d 7.54–7.49 and 6.88–6.84 (m, 2 6 10H,
CH aromatic), 4.34 (t, 2H, CH2OPh, spacer), 3.97 (t, 10H,
CH2OPh, tails), 3.61 (t, 2H, CH2OH), 1.95–1.24 (m, 58H, CH2
tails and spacer), 0.92 (t, 15H, CH3). 13C NMR (300 MHz,
CDCl3): d 159.84 (inner phenyl, C-O), 159.64, 159.53, 159.44
(outer phenyl, C-O), 133.28, 133.14, 133.10 (CH, outer phenyl),
128.40, 123.86, 119.86 (inner phenyl), 115.54, 115.37, 115.30
(outer phenyl, attached to ethynyl), 114.61 (CH, outer phenyl),
99.25, 99.01, 97.01 (inner ethynyl), 86.58, 86.05, 83.47 (outer
ethynyl), 74.61 (CH2OPh, spacer), 68.13 (CH2OPh, tails), 63.05
(CH2OH), 23.82–22.62 (CH2 tails and spacer), 14.05 (CH3
tails). Elemental analysis for C87H108O7: calculated C 82.55%,
H 8.60%, O 8.85%, found C 82.33%, H 8.24%.
11-[Pentakis(4-hexylsulfanylphenylethynyl)phenoxy]undecan-
1-ol (6S–D). The procedure described for the synthesis of
6CH2–D was used with a reaction time of 6 hours. After the
crude work up (yield 78%), pure 6S–D was obtained by
repeated column chromatography (using silica gel and
toluene : ether 20 : 1 as eluent) and crystallisation from
ether : hexane (1 : 10). Yield 22% as a yellow, highly fluorescent
Polymer analogous reaction. General procedure
Mesogens 4CH2–D, 6CH2–D or 6O–D (0.5 mmol), pyridine
(1 mL) and N,N-dimethylaminopyridine (catalytic amount)
was dissolved in dry CH2Cl2 (10 mL) under an argon
atmosphere. The crude PAC solution (0.3 mmol) was added
via syringe. The reaction proceeded for 24–72 hours and the
polymer was precipitated in methanol and reprecipitated in
methanol or acetone. After filtration and drying, the degree of
1
powder. H NMR (300 MHz, CDCl3): d 0.90 (t, 15H, CH3),
1.27–197 (m, 58H, CH2 tails and spacer), 2.97 (t, 10H, CH2S),
3.62 (t, 2H, CH2OH), 4.34 (t, 2H, CH2OPh), 7.21–7.28 and
7.42–7.49 (m, 2 6 10H, CH aromatic). 13C NMR (300 MHz,
CDCl3): d 160.20 (inner phenyl, C-O), 139.47, 139.23, 138.98
(outer phenyl, C-S), 132.02, 131.88, 127.47 (CH, outer phenyl),
128.40, 123.89, 119.39 (inner phenyl), 119.98, 119.79, 119.66
(outer phenyl, attached to ethynyl), 99.64, 99.27, 99.11 (inner
ethynyl), 88.05, 87.94, 87.07 (outer ethynyl), 74.80 (CH2OPh,
spacer), 63.08 (CH2OH), 22.55–32.84 (CH2 tails and spacer),
14.03 (CH3 tails). Elemental analysis for C87H108O2S5:
calculated C 77.63%, H 8.09%, O 2.38%, S 11.91%, found C,
77.43%, H, 8.27%, S, 11.64%.
1
substitution was determined by H NMR.
Results and discussion
Synthesis
The disc-shaped mesogens15 6CH2–D, 6O–D, 6S–D and
6Am–D have been prepared by a five-fold palladium catalysed
cross-coupling reaction of a pentabromophenol derivative
and the appropriate terminal acetylenes, similar to procedures
described before.10a,16 The synthesis of the terminal acetylenes
and the mesogens are outlined in Scheme 1. Although some of
the acetylenes are commercially available (though rather
expensive), we describe here the methods towards a larger
scale production (typically about 10 grams or more), which is
necessary for the synthesis of the discotic mesogens. The
preparation (and the liquid crystalline properties) of 4CH2–D
(with 5 butyl tails, linked by a methylene group to the aromatic
core)14,15 and CH3D (with 5 methyl substituents)13,15 have been
described before. Previous efforts to couple acetylenes with
electron withdrawing groups at the 4’-position never yielded
the desired product.17 Hence, 6SO2–D was prepared by
selective oxidation of 6S–D, using m-chloroperbenzoic acid
(MCPBA). Polymers P(4CH2–D), P(6CH2–D) and P(6O–D)
were prepared via a polymer analogous substitution reaction of
11-[Pentakis(4-hexylsulfonylphenylethynyl)phenoxy]undecan-
1-ol (6SO2–D). A solution of 6S–D (0.5 g, 0.37 mmol) in 25 ml
CH2Cl2 was cooled in an ice bath. m-Chloroperbenzoic acid
(1.25 g. 70% MCPBA, 13.5 eq.) was added over 3 minutes and
the solid slowly dissolved. The reaction mixture was stirred at
room temperature for 4.5 hours. The reaction was stopped by
washing the solution successively with an aqueous Na2SO3
solution, an aqueous NaHCO3 solution and pure water. After
drying over MgSO4 and evaporation of the solvent, pure
6SO2–D was obtained by crystallisation from methanol. Yield
1
0.51 g, 91% as a yellow, highly fluorescent powder. H NMR
(CDCl3): d 7.98–7.93 and 7.75–7.72 (m, 2 6 10H, CH
aromatic), 4.43 (t, 2H, CH2OPh), 3.62 (t, 2H, CH2OH), 3.13
(t, 10H, CH2SO2), 1.22–1.97 (m, 58H, CH2 tails and spacer),
J. Mater. Chem., 2003, 13, 458–469
461

View Article Online
Scheme 1 Synthesis of the terminal acetylenes and mesogens 6X–D.
mesogens 4CH2–D, 6CH2–D and 6O–D in the reactive Thermal analysis of the mesogens
poly(acryloyl chloride), by a similar procedure to that we
have described before,10a,11a see Scheme 2.
The thermal properties of the mesogens and polymers were
investigated by differential scanning calorimetry (DSC), see
Fig. 3, and optical polarising microscopy (OPM), see Fig. 4.
The results are summarised in Table 1.
Compound 6CH2–D showed a monotropic nematic phase at
low temperatures, easily recognised by the typical schlieren
texture and its low viscosity. In the corresponding polymer
P(6CH2–D), the mesophase disappeared completely and the
material turned out to be an amorphous polymer with a low
glass transition temperature. Compound 6O–D exhibited a
nematic phase over a temperature range of 47 uC. Thin samples
of 6O–D between two unprepared microscope slides quickly
Scheme 2 Synthesis of (co-)polymers P(4CH2–D), P(6CH2–D) and
P(6O–D).
462
J. Mater. Chem., 2003, 13, 458–469

View Article Online
Fig. 4 Optical textures of (a) the low birefringent homeotropic ND
phase of 6O–D at 80 uC (annealing for 30 s at 80 uC from the first
heating ramp); (b) the ND phase of P(6O–D) at 100 uC (annealing for
30 min at 100 uC upon cooling from the isotropic phase); (c) the Colhd
phase of 6S–D at 100 uC (upon cooling from the isotropic phase); (d)
the Colhd phase of 6SO2–D at 255 uC (freshly formed columnar phase
from the isotropic phase). All pictures were taken with crossed
polarisers, magnification 756.
Fig. 3 Normalised DSC thermograms of (a) 6CH2–D (monotropic
ND phase); (b) 6O–D; (c) P(6O–D)b (magnification 56); (d) 6S–D
and (e) 6SO2–D (magnification 106). For 6O–D, P(6O–D)b and
6S–D, the small transitions to the isotropic liquid have been magnified
(106).
Table 1 Thermal behaviour of the mesogens, measured with OPM and
DSCa Transition temperatures are shown in [uC] and latent heat values
(between the brackets) in [kJ?mol21
]
Materials
Thermal behaviourb
aligned homeotropically in the ND phase. Fig. 4a was recorded
30 seconds after heating the sample to 80 uC. Further annealing
(v5 minutes at 80 uC) resulted in almost complete homeotropic
alignment. The latent heat at the nematic to isotropic transition
is very small, as observed previously for similar materials.3a,c
Polymers P(6O–D) also showed ND phases with clearing
temperatures and latent heat values of the order of those of the
monomer 6O–D. Due to the presence of the polymer backbone,
crystallisation was suppressed and the polymers vitrified
around room temperature, resulting in the formation of a
nematic glass at low temperatures. Although room temperature
nematic phases for rod-shaped liquid crystals (optically
positive) are not uncommon, nematic phases occurring at
such low temperatures for disc-shaped mesogens (optically
negative) are very rare.3b,c,18 Chemically linking mesogens to a
polymer20a,26 proves to be an effective tool to reach low
temperature nematic discotic phases.19 It is anticipated that the
use of a more flexible backbone (like polyether or polysiloxane)
could lower the glass transition temperature even more and,
hence, real room temperature optically negative nematic phases
can be envisaged. As expected, the mesogens attached to the
polymer backbone did not show fast alignment between the
microscope slides. Even extensive annealing did not result in a
significant growth of the domain sizes (see Fig. 4b).
6CH2–D
6O–D
6S–D
6SO2–D
6Am–D
K
K
K
GCol
K
[39 (34)
71 (33)
56 (65)
115
ND 52 (1.2)]c
ND
Colhd
57 (36)
118 (0.15)
167 (0.21)
I
I
I
d
I/d
Colhd
w260
215 (16)
P(6CH2D)
P(6O–D)a
P(6O–D)b
GI
GN
GN
-15
107 (0.15)
113 (0.18)
I
I
I
25
24
ND
ND
aTransition temperatures are given in uC and latent heat values (in
parentheses) in kJ mol21 bAbbreviations: K ~ crystalline; GCol
.
~
glass with columnar phase frozen in; GN ~ glass with nematic phase
frozen in; GI ~ isotropic glass; Colhd ~ columnar hexagonal, disor-
dered; ND ~ nematic discotic; I ~ isotropic; d ~ decomposition
c
(prior to melting). Monotropic phase transition.
phase is observed. The phase behaviour of 6SO2–D was hard to
determine, since decomposition at very high temperatures
(w250 uC) precedes isotropization.20 In a rapid cooling
experiment from a freshly molten sample, the mesophase of
6SO2–D was identified by the observation of growing hexagons
from the isotropic phase, characteristic of a Colh mesophase.
Compound 6SO2–D did not show any melting or crystal-
lisation phenomena, but rather showed a glass transition at
lower temperatures. The penta-amide 6Am–D cleared at about
215 uC with some decomposition, but neither microscopy
experiments nor DSC analysis could distinguish between a
highly ordered columnar and an ill-defined crystalline phase at
lower temperatures.
Mesogen 6S–D shows a fan-shaped optical texture (see
Fig. 4c), characteristic of a columnar phase. DSC experiments
show two large transitions, the first is attributed to a crystal–
crystal transition and the second to a transition to a columnar
phase. At high temperatures, a small transition to the isotropic
J. Mater. Chem., 2003, 13, 458–469
463

View Article Online
Table 2 X-Ray diffraction results
˚
d/A
Material
T/uC
Phase
ND
Miller indicesa
6O–D
90
(100)
Alkyl
(001)
20.8
5.2
4.4
P(6O–D)b
6S–Db
50
ND
(100)
Alkyl 1 (001)
20.5
4.5-5.0
100
Colhd
100
Alkyl
001
23.3
4.8
3.9
6SO2–Dc
180
Colhd
100
200
210
Alkyl
001
24.6
14.3
9.1
5.0
3.7
aNote that 6O–D and P(6O–D)b do not possess any long-range posi-
tional order in the nematic phases and hence Miller indices do not
apply. The Miller indices indicated (100) and (001) correspond to the
average lateral and planar distance between two adjacent discs,
respectively. bThe calculated lattice constants for 6S–D at 100 uC
are: a ~ b ~ 26.9 A, c ~ 3.9 A. cThe calculated lattice constants
˚
˚
˚
˚
for 6SO2–D at 180 uC are: a ~ b ~ 28.1 A, c ~ 3.7 A.
Scherrer equation: jhkl ~ 0.89l/(vhklcos hhkl), wherein l is the
wavelength, vhkl is the full width at half maximum of the
reflection and hhkl is the maximum of the reflection. Compound
6S–D shows a particularly low correlation length in the
˚
˚
intracolumnar direction (j100 ~ 110 A and j001 ~ 13.4 A), a
value that is typical for the extent of order we observe
in nematic mesophases.7 It should be noted that usually
Fig. 5 Normalised radially integrated scans of XRD patterns of (a)
6O–D at 90 uC (b) P(6O–D)b at 50 uC (c) 6S–D at 100 uC with
magnification of the wide-angle area (46) and (d) 6SO2–D at 140 uC
with magnification of the wide-angle area (46).
˚
much higher intracolumnar correlation lengths (50 to 100 A or
even higher) are observed for materials showing a columnar
phase.
X-Ray diffraction
The XRD pattern of 6SO2–D showed some narrowing of the
intercolumnar reflections as well as some higher order
reflections in the small angle region, the (110) and (210)
reflections. The intracolumnar reflection is shifted to higher
diffraction angle (i.e. shorter distances) and sharpened a bit
compared to 6S–D. Despite the increase in the intracolumnar
interactions indicated by the observed shift and narrowing, the
reflection is still characteristic of liquid-like behaviour within
the columns, i.e. a columnar disordered phase. The X-ray
pattern does not change when the material vitrifies at lower
temperatures. Table 3 shows the spacings d and the correlation
lengths j of 6SO2–D at various temperatures. The intra-
columnar spacing decreases with decreasing temperature, while
the intercolumnar distance remains constant. The increasing
correlations lengths of both reflections indicate a slight increase
in order, which is to be expected at lower temperatures. XRD
experiments of 6Am–D below 215 uC (not shown) revealed a
pattern typical of a crystalline phase, i.e. a multitude of sharp
reflections in the small angle region.
To gain more insight into the (local) molecular order in the
mesophases, the materials were subjected to powder X-ray
diffraction (XRD) measurements. Radially integrated scans
of the patterns are displayed in Fig. 5. The spacings d,
corresponding to the maximum of the diffraction angles h,
were calculated using the Bragg law, 2dsin h ~ nl, where n is an
integer and l is the wavelength: l ~ 1.54 A. The results are
summarised in Table 2.
Compound 6O–D showed the typical pattern of the low
ordered ND phase. Two diffuse reflections are observed,
attributed to the lateral and planar distance between two
˚
˚
adjacent discs, as well as a broad non-oriented halo at 4.5–5 A,
which is attributed to (non-oriented) alkyl chains. Due to
orientation of the mesogen in the applied magnetic field, it was
possible to deconvolute the contributions of the aligned planar
inter-disc reflections and the isotropic alkyl reflections. The
integrated pattern for the macromolecular equivalent P(6O–
D)b looks nearly identical apart from some further peak
broadening. However, the polymer showed hardly any
alignment in the magnetic field and deconvolution was not
possible.
The non-aligned columnar phase of 6S–D showed a much
sharper lateral, intercolumnar (100) reflection, indicating a
well-defined two-dimensional columnar organisation. The
intracolumnar (001) reflection is still very diffuse, which
indicates true liquid-like order within the columns, charac-
teristic of columnar disordered (Colxd) phases. To quantify
the extent of organisation, the correlation length j can be
calculated. The correlation length jhkl is calculated from
the peak width of the appropriate reflection after fitting the
reflections with a Lorentzian distribution function, using the
Table 3 Temperature dependence of the X-ray diffraction character-
istics of 6SO2–D in the intercolumnar (100) and intracolumnar (001)
directionsa
˚
d100/A
˚
j100/A
˚
d001/A
˚
j001/A
T/uC
180
140
100
3.7
3.7
3.7
17
20
22
24.6
24.3
24.1
121
123
126
aThe diffraction pattern of 6SO2–D at 180 uC was recorded in the
Colhd phase, at 100 uC in the GCol phase and the pattern at 140 uC
was recorded at the transition between the two phases.
464
J. Mater. Chem., 2003, 13, 458–469

View Article Online
Fig. 7 Clearing temperature TNI as a function of the rigid mesogen
fraction [M] for various alkyl-substituted mesogens. Both side-chain
polymers and low molar mass mesogens are shown in the graph. The
solid line21 represents the power-law equation: TNI ~ 695[M]0.98
.
show points under the line in Fig. 7. We believe that the
hydrogen bonds, introduced by the terminal hydroxy groups
present in the ‘monomers’ cancel out the effects of linking of
multiple mesogens (loss in entropy), and hence, matching phase
transition temperatures are observed.
Fig. 6 Alkyl-substituted mesogens and polymers.
Intermolecular interactions
Alkoxy-substituted mesogens. As a rule of the thumb, for
disc-shaped and rod-shaped liquid crystals, the alkoxy-
substituted mesogens show an increase in both the clearing
temperature and the melting temperature compared to their
alkyl-substituted counterparts. The introduction of oxygen in
the linking group results in an increased electron density in
the aromatic core. The corresponding enhancement of the
The large differences in the phase behaviour of the mesogens
can be ascribed to the differences in molecular structure. We
will discuss the observed phase behaviour in terms of
intermolecular interactions, starting with the low ordered
alkyl-substituted mesogens through to the highly ordered
sulfone- and amide-substituted molecules.
attractive p–p interactions results in
temperature of 6O–D compared to 6CH2–D.
a higher clearing
Alkyl-substituted mesogens. In a previous publication, we
reported the phase behaviour of methyl-substituted mesogens
(CH3–D), its corresponding side-chain polyacrylates and a
series of copolymers with different alkyl acrylates,11a see
polymers P(CH3–D) in Fig. 6. A master phase diagram was
Alkylthio substituted mesogens. Replacement of oxygen by
sulfur in the substituents of discotic liquid crystals usually
results in similar phase behaviour with a slight decrease in the
transition temperatures, as was observed for the series of
triphenylenes,23 tricycloquinazolines24 and phthalocyanines.25
However, we found a columnar mesophase with a higher
clearing temperature for 6S–D, instead of the ND phase found
for 6O–D, which seems to be in a marked contrast to the results
of the series quoted above. The unexpected formation of this
columnar phase could be attributed to packing effects, due to
the larger dimensions of the sulfur atom compared to oxygen
and may be reinforced by p–p interactions involving the sulfur
d-orbital. It is clear that the effects of subtle changes in the
molecular geometry on mesophase formation are significant.
Peculiarly, the latent heat for the Colhd to I transition of 6S–D
is very small, at least a factor of 10 larger would be expected.26
At a phase transition the total amount of free energy is
constant: DG ~ DH 2 TDS ~ 0, so when DH is small, either
the order in the columnar phase is extraordinarily low, or, more
likely, the extent of local organisation is still substantial in the
isotropic phase! Unfortunately, reliable XRD measurements in
the isotropic phase could not be obtained, because of severe
thermal degradation of the mesogen.
constructed which showed a clear relation between the NCol
–
ND–I and the NCol–I transitions and the (weight) fraction of
rigid mesogen in the different polymers. The transition
temperatures were determined by simple exponential fits and
were valid for all co-polymers, irrespective of the length and the
degree of substitution of the dangling alkyl groups along the
polymer backbone.11b In the heptyl-substituted mesogen
6CH2–D, the well-known pentyl-substituted mesogen 4CH2–
D as well as their corresponding side-chain polymers, the
fraction of rigid mesogens is further reduced by the laterally
attached tails. When their ND–I transition temperatures are
added to the master phase diagram, Fig. 7 is obtained.
The clearing temperatures of all alkyl-substituted mesogens
of the type nCH2–D (where n represents the lengths of the
laterally substituted alkyl tails15) can be determined from the
drawn curve. This means that the transition temperatures for
these series of mesogens can be tuned conveniently, either by
attaching long tails to the mesogens or by ‘co-polymerisation’
of a mesogen with different alkyl acrylates. In other words,
regardless of the length and the position (attached to the
mesogen or to the polymer backbone) of any alkyl group, the
N
D to isotropic transition temperature can simply be predicted,
Alkylsulfonyl and alkylamide-substituted mesogens. Sulfone27
and amide28 functional groups are rarely used in liquid crystals,
since they have a tendency to form highly ordered, crystalline
materials. The high clearing temperatures of 6SO2–D and 6Am–D
strongly indicate that the sulfone and amide groups enhance the
intermolecularnon-covalentinteractionsbetweenthemesogens.It
is anticipated that strong attractive interactions are directed in the
intracolumnar direction, since the functional groups are confined
between the core and the tails. This is schematically illustrated in
just by calculating its rigid mesogen fraction [M]. It is
remarkable to see that this relation holds over a temperature
range of 200 uC.
Interestingly, the clearing temperatures of the low molar
mass mesogens lie on the same line as the polymers. However,
the true monomeric species (the propionyl esters of the
appropriate mesogens) show a slight decrease in the clearing
temperatures, as would be expected.22 Accordingly, they would
J. Mater. Chem., 2003, 13, 458–469
465

View Article Online
Fig. 9 Electron acceptors: 2,4,7-trinitrofluoren-9-one, TNF (left) and
2,4,7-trinitrofluoren-9-ylidenemalonitrile, TNF–DCM (right).
Charge transfer interactions
Doping nCH2–D, 6O–D and 6S–D with electron acceptors
TNF and TNF–DCM (see Fig. 9) gave rise to the formation of
strongly coloured charge transfer (CT) complexes.12 Their
phase behaviour is summarised in Table 4. In complexes of
6CH2–D or 4CH2–D with TNF or TNF–DCM enantiotropic
columnar hexagonal (Colh) phases were induced. Complexes
with the more electron deficient TNF–DCM showed higher
clearing temperatures. The formation of CT complexes also
suppresses crystallisation; instead a second order transition to
the columnar plastic (Colp) phase is observed at much lower
temperatures, leading to a wide temperature range for the
columnar mesophase. It should be mentioned that the most
stable complexes ( those with the maximum clearing tempera-
ture in the phase diagram) were found for complexes of 4CH2–
D and 6CH2–D with an excess of the acceptor. This is in
contrast to what was observed previously for similar materials,
which showed optimum complexation conditions at equimolar
ratios7,19 However, for other CT complexes optimum complex-
ing conditions were reported for mixtures with an excess of the
donor.12i–d,19
CT complexes of 6O–D and 6S–D with TNF showed lower
clearing temperatures than the pure mesogens. Complexes of
6O–D with various amounts of TNF were prepared and
investigated and a schematic phase diagram was constructed30
(see Fig. 10). When an attractive electronic interaction between
two species in a 1 : 1 ratio occurs, then optimum complexation
conditions (expressed as a local maximum in the clearing
temperature in the phase diagram) are likely to be found at
equimolar ratios in the phase diagram. We have several reasons
to believe that complex formation takes place at equimolar
ratios, despite the absence of the local maximum in the clearing
temperatures. The first indication is provided by UV studies in
dilute solution for which we indeed find a maximum in the
Fig. 8 Schematic representation of the attractive forces due to: (a)
dipole–dipole interactions in 6SO2–D and (b) hydrogen bonding in
6Am–D. Note that non-hydrogen-substituted benzene at the left-hand
side represents the central benzene group of the core and that for the
sake of clarity only one of the five substituted phenyl-ethynyl groups is
shown. Recent calculations have elucidated that a 45u out-of-plane
˚
twist of the amide groups is sufficient to efficiently stabilise a 3.8 A
distance between the aromatic cores.28b The pictures are obtained,
using CS Chem3D Ultra1, supplied by CambridgeSoft.com.
Fig. 8, which shows the central aromatic core (on the left side),
substituted with one of the five phenylethynyl groups and the
appropriate tail. When two neighbouring discs are in the correct
positions and conformations (multiple) attractive interactions can
occur:dipole–dipoleinteractionsfor 6SO2–Dandhydrogenbonds
for 6Am–D. The intracolumnar interactions promote a columnar
organisation with a well-developed two-dimensionally ordered
columnar structure, that is easily recognised by the sharp
intercolumnar reflection (100) and the higher order reflections
observed in the XRD pattern of 6SO2–D (as well as in the XRD
pattern of the crystalline phase of 6Am–D).
Despite the presence of intracolumnar attractive forces, a
considerable amount of disorder is found to exist from the
results of the XRD experiments (indicated by the diffuse (001)
reflections), which, in our opinion, can be attributed to two
major causes. (i) A ‘random’ organisation of the linking
groups, e.g. sulfone groups with dipoles pointing in the same
direction (i.e. up or down with respect to the disc-plane) in a
large stack of sulfone groups contribute to a high intra-
columnar organisation. Alternatively, if the dipole moment of
sulfone groups of neighbouring mesogens point up or down in
the stack irregularly, this will give rise to a defect structure,
causing disruption of the intracolumnar order. (ii) The presence
of the flexible spacer, anchored at the central phenyl group in
the core, prevents efficient packing of the mesogens into
columns and, hence, further diminishes the intracolumnar
order in the material.
Table 4 Thermal behaviour of the charge transfer complexes of
mesogens nCH2–D, 6O–D and 6S–D with TNF and TNF–DCM,
measured by OPM and DSCa
Phase behaviour at equimolar
concentrationsb
Mesogen Acceptor
4CH2–D TNFc d
Colp 15
TNF–DCM Colp 16
Colh 110 (3.2)
Colh 173 (3.4)
I
I
6CH2–D TNFe
Colp 15 Colh 72 (1.3)
TNF–DCM Colx 28 (3.0) Colh 95 (0.3)
I
I
6O–D
TNF
TNF–DCM Colp 50
K
64 (21)
ND 94 (0.4)
Colh 91 (8.0) ND 128 (0.4) I
I
The multiple dipolar interactions in 6SO2–D resulted in the
formation of a columnar mesophase that was stable until
degradation at high temperatures. Apparently, the aggregation
process in the multi-hydrogen bonded structure in 6Am–D is
more effective, resulting in a somewhat disordered crystalline
phase.29 A clear illustration of the extent of aggregation in
6Am–D is that upon dissolving small amounts of 6Am–D in
chloroform, a viscous gel is obtained. The addition of 10% of
methanol breaks-up the hydrogen bonds and yields a low
viscous solution, although the broad NMR signals still indicate
substantial interactions.
6S–D
TNF
Colp
4
Colh 83 (0.6) I
aTransition temperatures are given in uC and latent heat values (in
parentheses) in J g21 bK ~ crystalline; Colp ~ plastic columnar;
.
Colh ~ columnar hexagonal; ND ~ nematic discotic; I ~ isotropic.
cThe most stable CT complexes of 4CH2–D?TNF were observed at a
TNF fraction of 0.6: Colp 15 Colh 126 (5.0) I. dThe thermodynamic
results of complex 4CH2–D?TNF have been published before.14 eThe
most stable CT of 6CH2–D?TNF complexes were observed at a TNF
fraction of 0.65: Colp 15 Colh 93 (3.1) I.
466
J. Mater. Chem., 2003, 13, 458–469

View Article Online
It was expected that the (weakly) electron deficient 6SO2–D
would form stable charge transfer complexes with 6O–D and
6S–D, since the molecules have similar molecular structure and
dimensions. However, no CT interactions were observed for
mixtures of 6SO2–D with either 6O–D or 6S–D.32 While 6O–
D?6SO2–D macroscopically demixed, mixtures of 6S–D with
6SO2–D showed a linear mixing behaviour with clearing
temperatures in between those of the two constituents.33 As
was expected from the electron deficient nature of 6SO2–D, it
does not form CT complexes with the electron acceptors TNF
or TNF–DCM, and is inmiscible in all ratios.
In our experience, several factors determine the degree of
stabilisation of mesophases via charge transfer complexation.
The first important parameter is the strength of the donor–
acceptor complex (DHDA). However, this has to be compared
to the sum of donor–donor and acceptor–acceptor interactions
that are broken because of the complex formation. This can be
summarised as an effective complexation enthalpy: DHeff
~
DHAD 2 ½(DHDD 1 DHAA). It is stressed that in this
expression all relevant contributions must be taken into
account, including CT energies, van der Waals and quadru-
polar forces. When these electronic factors play a dominant
role, CT complexes with a maximum clearing temperature are
expected at exactly equimolar ratios.
Another important factor to be considered in this process
comprises the steric (packing) effects. For our complexes two
important steric contributions can be distinguished: (i) the long
spacer that is anchored at the centre of the core;34 and (ii) the
misfit in dimensions of the donor and the acceptor.35 Both
effects give rise to spatial distortion in a columnar organisation.
For example, complexing a large donor with a small acceptor in
an alternating columnar arrangement of the donor and
acceptor would result in a large amount of free volume in a
local stack, which will cause a strong destabilisation of the
mesophase.
Quantification of the enthalpic and steric (entropic) effects
will increase the understanding of these complex materials.
However, we will only qualitatively explain the thermodynamic
results of the CT complexes: The enthalpic effects are
responsible for the observed stabilisation of the mesophase.
Alternatively, the CT complexes discussed here, exhibit a
destabilising effect on the mesophase formation originating
from steric effects. Balancing these effects results in an overall
stabilisation or destabilisation. For example, in 6O–D?TNF the
steric destabilising forces dominate, whereas in the complex
with TNF–DCM (a stronger acceptor of a similar size to TNF)
Fig. 10 (a) Phase diagram of 6O–D and TNF. The phase diagram is
constructed from studying mixtures by OPM and DSC as well as by
studying contact samples (OPM) between the two components. (b)
Change in entropy at the clearing point as a function of the TNF
fraction. Note that with increasing TNF concentration, the clearing
temperature TNI decreases, while the change in entropy increases.
absorption band for 1 : 1 ratios (the calculated complexation
constant Kc and the activation energy DH are 16.6 l mol21 and
,
213.7 kJ mol21 respectively).31 However, we are not
convinced that these systems will behave similarly in dilute
solution and in the bulk. A much more convincing indication is
the absence of biphasic areas at exact equimolar ratios in the
phase diagram. In fact, since the CT complex effectively acts as
a new species7,19 it cannot show broad biphasic areas and ,
therefore, it has to appear at equimolar ratios.
The left-hand side of the phase diagram shows the more or
less linear mixing behaviour between the pure 6O–D and the
6O–D?TNF complex. Increasing the TNF content causes a
gradual decrease in both the melting and the clearing
temperature compared to pure 6O–D. The corresponding
change in entropy for the ND to I transition increases nearly
linearly with increasing TNF concentration. Hence, while the
complexes show an increase in intermolecular interactions
(increase in DH at the clearing temperature), other effects
destabilising the mesophase dominate. Complexes with more
than 60 mol% TNF showed macroscopic demixing of the
complex (dark-red coloured) and pure TNF (pale yellow). Even
in the isotropic phase no mixing behaviour was observed.
Application of the stronger acceptor, TNF–DCM, to 6O–D
resulted in stabilisation of the nematic mesophase, i.e. a higher
clearing temperature in combination with a larger latent heat.
In the equimolar complex, the ND phase cleared at 128 uC (with
DS ~ 0.72 J mol21 K21). In addition a columnar phase was
observed at low temperatures.
stabilisation is observed, or in energy terms, DHAD,TNF–DCM
w
DHAD,TNF, while DSAD,TNF–DCM # DSAD,TNF. Clearly, in the
CT complexes of the alkyl-substituted mesogens, the stabilising
enthalpic forces play a dominant role, much more important
than in the complexes of 6O–D and 6S–D. This can be easily
understood when it is realised that the important parameter is
the gain in enthalpy, represented by DHeff, which is expected to
be much higher for the weakly interacting alkyl-substituted
mesogens than for their alkoxy- or alkylthio-substituted
analogues. Similarly, the dipole–dipole interactions of 6SO2–
D are so strong (large DHAA) that no complex is formed with
6O–D and the mixture macroscopically phase separates.
Conclusions
CT complexes of 6S–D with TNF showed similar behaviour.
At equimolar concentrations a large decrease in the clearing
temperature to 83 uC was observed (with an increase in the
latent heat at the transition). Crystallisation was also
suppressed and a glass transition was found below room
temperature, resulting in an overall shift of the temperature
window for the columnar phase from 56–167 uC for the pure
6S–D to 4–83 uC for the 6–D?TNF complex.
A
series of novel discotic liquid crystals with varying
substituents, which link the flexible spacer to the aromatic
core have been synthesised and investigated. The nature of this
linking group gives rise to a number of intermolecular
interactions, which are supplementary to the interactions
already present in (discotic) liquid crystals. The five-fold
presence of the linking group assures a strong effect on the
molecular organisation and, hence a wide spread in the
J. Mater. Chem., 2003, 13, 458–469
467

View Article Online
had sufficient X-ray diffraction evidence to choose between one of
the two extremes. The suffix was omitted when the diffraction
peaks showed intermediate levels of intracolumnar order.
(a) B. Glu¨sen, A. Kettner and J. H. Wendorff, Mol. Cryst. Liq.
Cryst., 1997, 303, 115–120; (b) In general, a transition from the
Colp to a Colh or Colr phase is characterised by a second order
phase transition, visible in the DSC diagram. The X-ray
diffractogram shows mixed reflections, indicating a three-dimen-
sional positional organisation. Despite the increased viscosity, the
sample remains deformable during OPM experiments.
mesomorphic properties is observed, ranging from room
temperature nematics up to columnar phases that are stable
over temperatures of 250 uC.
9
In mesogens with alkyl- and alkoxy-substituents, p–p
interactions and solvophobic effects play an important role
in the mesophase formation. The series of alkyl-substituted
mesogens (nCH2–D) show particularly interesting results, since
their clearing temperature is solely determined by the fraction
of rigid mesogen. In addition, attachment of the mesogens to a
polyacrylate backbone suppresses the crystallisation process,
resulting in room temperature nematic discotic mesophases for
P(6O–D). It should be realised that an important drawback of
these systems is the reduced mobility of the mesogens. This can
be overcome by the application of low molar mass multipodes
with a well-defined core instead.36 For hexylthio (6S–D),
hexylsulfonyl (6SO2–D) and hexylamido (6Am–D) substituted
mesogens, additional interactions have a major influence on the
phase behaviour. Packing effects, dipole–dipole interactions
and H-bonds, respectively, stabilise the formation of columnar
phases. Eventually, for the penta-amide (6Am–D) the strong
interactions result in the disappearance of the mesophase and a
crystalline phase was observed instead.
The use of charge transfer complexing agents offers an easy
and versatile tool to manipulate the liquid crystalline proper-
ties. Often the CT interactions give rise to stabilisation of the
mesophase (due to enthalpic effects), but in some cases,
presented here, the steric effects (entropic) dominate, which
resulted in an overall destabilisation of the observed meso-
phases. It is apparent that ultimate CT stabilisation can be
achieved by increasing the donor–acceptor interactions or by
diminishing the steric factors, as discussed above.
10 For reports on polymers bearing disc-shaped mesogens as a side-
chain, see: (a) P. H. J. Kouwer, W. F. Jager, W. J. Mijs and
S. J. Picken, Macromolecules, 2000, 33, 4336–4342; (b) D. Stewart,
G. S. McHattie and C. T. Imrie, J. Mater. Chem., 1998, 8, 47–51;
(c) M. Weck, B. Mohr, B. R. Maughon and R. H. Grubbs,
Macromolecules, 1997, 30, 6430–6437; (d) N. Boden, R. J. Bushby
and Z. B. Lu, Liq. Cryst., 1998, 25, 47–58; (e) W. Kreuder and
H. Ringsdorf, Makromol. Chem., Rapid Commun., 1983, 4, 807–
815; (f) for reports on polymers bearing disc-shaped mesogens in
the main-chain, see: O. Herrmann-Scho¨nherr, J. H. Wendorff,
W. Kreuder and H. Ringsdorf, Makromol. Chem., Rapid
Commun., 1986, 7, 97–101; (g) G. Wenz, Makromol. Chem.,
Rapid Commun., 1985, 6, 577–584; (h) N. Boden, R. J. Bushby,
A. N. Cammidge and G. Headdock, J. Mater. Chem., 1995, 5,
2275–2281; (i) for reports on macromolecular networks bearing
disc-shaped mesogens, see: C. D. Favre-Nicolin and J. Lub,
Macromolecules, 1996, 29, 6143–6149; (j) S. Disch, H. Finkelmann,
H. Ringsdorf and P. Schuhmacher, Macromolecules, 1995, 28,
2424–2428; (k) C. D. Braun and J. Lub, Liq. Cryst., 1999, 26,
1501–1509.
11 (a) P. H. J. Kouwer, J. Gast, W. F. Jager, W. J. Mijs and
S. J. Picken, Mol. Cryst. Liq. Cryst., 2001, 364, 225–234;
(b) P. H. J. Kouwer, W. J. Mijs, W. F. Jager and S. J. Picken,
J. Am. Chem. Soc., 2001, 123, 4645–4646.
12 (a) Complexes of an electron-rich donor with an electron rich
acceptor, showing an induced absorption band, are often referred
to as charge transfer complexes. However, it should be stressed
that the attractive interaction between these molecules is rather a
combination of van der Waals, quadrupolar and charge transfer
interactions. For the sake of clarity, in this contribution the name
charge transfer complex is used to describe the total sum of
interactions. For a review on charge transfer complexing in
discotic liquid crystals read; (b) K. Praefcke and D. Singer,
Handbook of Liquid Crystals, ed. D. Demus, J. W. Goodby,
G. W. Gray, H. W. Spiess and V. Vill, Wiley-VCH, New York,
1998, part 2B, pp. 945–967; (c) K. Praefcke and J. D. Holbrey,
J. Inclusion Phenom., 1996, 24, 19–41; (d) H. Ringsdorf and
R. Wu¨stefeld, Philos. Trans. R. Soc. London, 1990, A 330, 95–108.
13 D. Janietz, Chem. Commun., 1996, 713–714.
References and notes
1
J. M. Lehn, Supramolecular Chemistry, Wiley-VCH, New York,
1995.
2
(a) D. S. Lawrence, T. Jiang and M. Levett, Chem. Rev., 1995, 95,
2229–2260; (b) D. Philip and J. F. Stoddart, Angew. Chem., Int.
Ed. Engl., 1996, 35, 1154–1196.
3
For a review on discotic liquid crystals read: (a) A. N. Cammidge
and R. J. Bushby, in Handbook of Liquid Crystals, ed. D. Demus,
J. W. Goodby, G. W. Gray, H. W. Spiess and V. Vill, Wiley-VCH,
New York, 1998, part 2B, pp. 693–748; (b) S. Chandrasekhar, in
Handbook of liquid crystals, ed. D. Demus, J. W. Goodby,
G. W. Gray, H. W. Spiess and V. Vill, Wiley-VCH, New York,
1998, part 2B, pp. 749–780; (c) for a review on discotic liquid
crystals, showing a nematic phase, see: K. Praefckein Physical
Properties of Liquid Crystals: Nematics, ed. D. A. Dunmur,
A. Fukuda and R. R. Luckhurst, INSPEC, London, 2001, pp. 17–
35.
14 D. Janietz, K. Praefcke and D. Singer, Liq. Cryst., 1993, 13, 247–
253.
15 The materials are abbreviated according to the length of the alkyl
group (n) and the nature of the linking group (X), i.e. nX–D, where
X ~ Am for the amide substituted mesogen.
16 B. Kohne and K. Praefcke, Chimia, 1987, 41, 196–198.
17 During the five-fold coupling reaction, the phenylacetylene with an
electron withdrawing group is consumed more rapidly than the
corresponding acetylene with an electron donating group,
although no traces of product could be detected. Since the
reaction proceeds in an argon atmosphere, oxidative dimerisation
can be excluded. NMR analysis indicates a complex mixture of
aromatic groups.
18 (a) Most rod-shaped mesogens show a positive birefringence (n ~
ne 2 no w 0). For disk-shaped liquid crystals, the birefringence is
usually negative, making discotics very suitable for application as
compensation layers in display technology; (b) only recently, the
first example of a room temperature nematic phase obtained from
disk-shaped liquid crystals was observed in a discotic liquid crystal
based on a hexakis(phenylethynyl)benzene core and substituted
with rac-4-methylnonyl tails, resulting in a broad distribution of
stereoisomers: S. Kumar and S. K. Varhney, Angew. Chem., Int.
Ed., 2000, 39, 3140–3142.
19 (a) P. H. J. Kouwer, O. van den Berg, W. F. Jager, W. J. Mijs
and S. J. Picken, Macromolecules, 2002, 35, 2576–2582;
(b) P. H. J. Kouwer, W. F. Jager, W. J. Mijs and S. J. Picken,
Macromolecules, 2002, 35, 4322–4329.
20 The forming optical texture was observed after quickly heating the
sample to 300 uC followed by cooling to 250 uC. Despite the
degradation process (especially on the edges of the sample between
the microscope slides) growing hexagons were clearly observed,
4
(a) H. Mori, Y. Itoh, Y. Nishiura, T. Nakamura and
Y. Shinagawa, Jpn. J. Appl. Phys., 1997, 36, 143–147;
(b) K. Kamada, J. Watanabe, K. Arakawa and H. Kozono,
Fuji Photo Film Co. Ltd., 1995, Eur. Pat. Appl.646829 A1;
(c) D. Adam, P. Schuhmacher, J. Simmerer, L. Haussling,
K. Siemensmeyer, K. H. Etzbach, H. Ringsdorf and D. Haarer,
Nature, 1994, 371, 141–143; (d) A. N. Cammidge and R. J. Bushby,
Handbook of Liquid Crystals, ed. D. Demus, J. W. Goodby,
G. W. Gray, H. W. Spiess and V. Vill, Wiley-VCH, New York,
1998, pp. 693–748.
5
6
N. H. Tinh, C. Destrade and H. Gasparoux, Phys. Lett., 1979,
72A, 25.
(a) H. Ringsdorf, R. Wu¨stefeld, E. Zerta, M. Ebert and
J. H. Wendorff, Angew. Chem., Int. ed. Engl., 1989, 28, 914–
918; (b) K. Praefcke, D. Singer, B. Kohne, M. Ebert, A. Liebmann
and J. H. Wendorff, Liq. Cryst., 1991, 10, 147–159.
7
8
P. H. J. Kouwer, W. F. Jager, W. J. Mijs and S. J. Picken,
Macromolecules, 2001, 34, 7582–7584.
(a) J. W. Goodby and G. W. Gray, Handbook of Liquid Crystals,
ed. D. Demus, J. W. Goodby, G. W. Gray, H. W. Spiess and
V. Vill, Wiley-VCH, New York, 1998, part 1, pp. 17–24; (b) the
suffices ‘o’ and ‘d’ for ‘ordered’ and ‘disordered’ reflect the extent
of intracolumnar organisation of the columnar phase with the two
extremes of ‘o’ for long-range positional order and ‘d’ for one-
dimensional liquid-like order. We have used the suffix when we
468
J. Mater. Chem., 2003, 13, 458–469

View Article Online
indicative of a columnar hexagonal mesophase. By this method a
clearing temperature of 280 uC was estimated.
21 The drawn curve was taken from a previous publication (see ref.
11b) and was calculated from fitting the series of alkyl-substituted
polymers to a power-law.
22 C. T. Imrie and G. R. Luckhurst, Handbook of Liquid Crystals, ed.
D. Demus, J. W. Goodby, G. W. Gray, H. W. Spiess and V. Vill,
Wiley-VCH, New York, 1998, part 2B, pp. 801–833.
23 (a) C. Destrade, M. C. Mondon and J. Maltheˆte, J. Phys. Paris C3,
1979, 40, 17–21; (b) B. Kohne, W. Poules and K. Praefcke, Chem.
Z., 1987, 108, 408; (c) K. Praefcke, B. Kohne, W. Poules and
E. Poetsch, Merck GmbH, Germany, 1983, Ger. Pat. 3346980A1; .
24 (a) S. Kumar, E. J. Wachtel and E. Keinan, J. Org. Chem., 1993,
58, 3821–3827; (b) E. Keinan, S. Kumar, S. P. Singh, R. Ghirlando
and E. J. Wachtel, Liq. Cryst., 1992, 11, 157–173.
H-bonds, e.g. by preparing the meta-amide-substituted derivative
of 6Am–D. It was anticipated that the number of hydrogen bonds
was strongly reduced by assuming a random orientation of the
amide group. Indeed, the material obtained showed a stable
columnar mesophase, but unfortunately, we were not able to
purify the material to the high level that is required for a good
analysis of the liquid crystalline properties.
30 Complexes with a TNF mol fraction higher than 0.6 were unstable
and show macroscopic phase separation. In contact samples
between 6O–D and TNF, as studied with OPM, it could clearly be
observed that TNF was immiscible with the CT complex.
31 M. W. C. P. Franse, unpublished results.
32 The formation of charge transfer complexes between materials
based on the pentakis(phenylethnyl)phenoxy core and TNF
derivatives is easily recognised by the formation of a charge
transfer band in the absorption spectrum of the complex. The
complexes are usually orange (weak complex) to dark red (strong
complex) compared to (pale) yellow constituents.
25 (a) C. Piechocki, J. Simon, A. Skoulios, D. Guillon and P. Weber,
J. Am. Chem. Soc., 1982, 104, 5245–5247; (b) D. Guillon,
A. Skoulios, C. Piechocki, J. Simon and P. Weber, Mol. Cryst.
Liq. Cryst., 1983, 100, 275–284; (c) I. Cho and Y. Lim, Mol. Cryst.
Liq. Cryst., 1988, 154, 9–26.
33 This is proof that the mesophases observed in 6S–D and 6SO2–D
are identical, i.e. the Colhd phase.
26 In
a
recent publication (ref. 19a) we described some very
34 For studies on the length of spacer, originating from the central
core of the disc, see: (a) K. Praefcke, D. Singer, M. Langner,
B. Kohne, M. Ebert, A. Liebmann and J. H. Wendorff, Mol.
Cryst. Liq. Cryst., 1992, 215, 121–126; (b) H. Bengs, O. Karthaus,
H. Ringsdorf, C. Baehr, M. Ebert and J. H. Wendorff, Liq. Cryst.,
1991, 10, 161–168.
35 For studies of CT complexes with acceptors similar in size, but
different in electron affinity, see: (a) R. Wu¨stefeld, Ph.D. Thesis,
Johannes Gutenberg Universita¨t, Mainz, Germany 1990;
(b) M. Ebert, Ph.D. Thesis, Technische Hochschule Darmstadt,
Darmstadt, Germany, 1990.
disordered columnar materials, as determined by XRD, and
even these exhibited quite ‘normal’ latent heat values.
27 The exception is the series of six-fold substituted alkylsulfonyl-
benzenes in: K. Praefcke, W. Poules, B. Schleube, R. Poupko and
Z. Luz, Z. Naturforsch. B, 1984, 39, 950. No liquid crystalline
properties were detected in alkylsulfonyl substituted triphenylenes.
28 The use of amide bonds in combination with steric repulsion can
yield highly ordered columnar phases, which may be applied as
molecular wires, for example: (b) M. L. Bushey, A. Hwang,
P. W. Stephens and C. Nuckolls, J. Am. Chem. Soc., 2001, 123,
8157–8158 and references therein.
36 R. Elsa¨sser, G. H. Mehl, J. W. Goodby and M. Veith, Angew.
Chem., Int. Ed., 2001, 40, 2688–2690 and references therein.
29 Some experiments have been performed to decrease the number of
J. Mater. Chem., 2003, 13, 458–469
469