Quasi temperature independent electron mobility in hexagonal columnar mesophases of an H-bonded benzotristhiophene derivative

Article abstract of DOI:10.1021/cm902453z

Reported here are the unique properties of N,N0,N00-(3,4,5- tridodecyloxyphenyl)benzo[b,b0, b00]tristhiophene-2,20,200-tricarboxamide 3 as a new H-bonded discotic liquid crystal. Polarized optical microscopy and thermal analysis as well as variable temper

Full text of DOI:10.1021/cm902453z

1420 Chem. Mater. 2010, 22, 1420–1428
DOI:10.1021/cm902453z
Quasi Temperature Independent Electron Mobility in Hexagonal
Columnar Mesophases of an H-Bonded Benzotristhiophene Derivative
Andrey Demenev and S. Holger Eichhorn*
Department of Chemistry & Biochemistry, University of Windsor, 401 Sunset Avenue, Windsor,
Ontario N9B 3P4, Canada
Tyler Taerum and Dmitrii F. Perepichka*
Department of Chemistry, McGill University, 801 Sherbrooke Street West, Montreal H3A 2K6, Canada
Sameer Patwardhan, Ferdinand C. Grozema,* and Laurens D. A. Siebbeles
DelftChemTech, Delft University of Technology, Julianalaan 136, 2628 BL Delft, The Netherlands
Richard Klenkler
Xerox Research Centre Canada, 2660 Speakman Drive, Mississauga, Ontario L5K 2L1, Canada
Received August 10, 2009. Revised Manuscript Received December 11, 2009
Reported here are the unique properties of N,N0,N00-(3,4,5-tridodecyloxyphenyl)benzo[b,b0,
b00]tristhiophene-2,20,200-tricarboxamide 3 as a new H-bonded discotic liquid crystal. Polarized optical
microscopy and thermal analysis as well as variable temperature IR spectroscopy and X-ray diffraction
confirm the presence of two thermotropic H-bonded hexagonal columnar mesophases that cover a
temperature range from <-50 to 280 °C. Intermediate lyotropic mesophases of the highly viscous
material aid the alignment of the hexagonal columnar mesophases, which is essential for a detailed
structural characterization and applications. Solutions of 3in heptane at concentrations as low as 1 wt %
display isotropic organo-gel phases that consist of H-bonded networks of 3 but do not contain columnar
stacks. 2D-X-ray diffraction studies on aligned samples of the thermotropic hexagonal columnar
mesophases and DFT calculations on a tetramer of 3 reveal a helical columnar stacking of the individual
benzotristhiophene units. Charge carrier mobility measured by time-resolved microwave conductivity is
about 0.02 cm2 V-1 s-1 in both hexagonal columnar mesophases and quasi temperature independent
even across the phase transition between the two mesophases. The temperature independence is
explained by the interrelation between stacking distance and mutual rotation because of the persistent
intracolumnar H-bonds between amide groups. Half-life of the charge carriers, on the other hand,
drastically increases in the low temperature hexagonal columnar mesophase, which is most likely a result
of changing molecular dynamic and conformational states of the side chains. DFT calculations of the
frontier orbitals show that the benzotristhiophene core is the sole contributor to the LUMO but does not
contribute to the HOMO, whereas the trialkoxyaniline groups are the sole contributors to the HOMO.
This suggests that the observed combined mobility is that of electrons alone because no hole transport is
˚
expected to occur between trialkoxyaniline groups that are spaced apart by more than 4 A. Indeed, an
electron mobility of 2 ꢀ 10-3 cm2 V-1 s-1 but no transient signal for hole transport is obtained by time-
of-flight charge carrier mobility measurements on a multi domain sample of 3.
Introduction
the advantages generic to organic semiconductors, such
as low temperature solution processing and synthetically
adjustable properties, DLCs promise an important cap-
ability of self-healing and can be used in the formation of
highly anisotropic materials.
Two important limitations have hampered the broader
appeal of columnar liquid crystalline materials for or-
ganic electronics: 1. the formation of aligned mono-
domains of the required orientation and size may be
challenging;2 2. their electronic and other properties in
the liquid crystal phases strongly depend on temperature
Self-organized columnar stacks of aromatic com-
pounds such as discotic liquid crystals (DLCs)1,2 and
polyaromatic dendrons3 are an emerging class of organic
semiconductors. Intrinsic charge carrier mobility values
above 10-1 cm2 V-1 s-1 have been demonstrated in bulk
samples and along the columnar stacks of aligned mono-
domains.2,4 In addition to efficient charge transport and
*Corrresponding e-email: Eichhorn@uwindsor.ca (S.H.E.), Dmitrii.
perepichka@mcgill.ca (D.F.P.), F.C.Grozema@tudelft.nl (F.C.G.).
r
pubs.acs.org/cm
Published on Web 01/08/2010
2010 American Chemical Society

Article
Chem. Mater., Vol. 22, No. 4, 2010 1421
because of changing dynamic at the molecular level and
because of multiple phase transitions.5 Charge carrier
mobility, for example, decreases by a factor of 7 with a
temperature increase of only 20 °C in the hexagonal
columnar mesophase (Colh) of hexakis(hexylthio)-
triphenylene and decreases by an order of magnitude at
the phase transition from a higher ordered columnar to a
less ordered Colh mesophase.6,7
Charge carrier mobility generally decreases with de-
creasing order of the mesophases and large steps in
declining mobility are observed at phase transitions bet-
ween higher and lower ordered columnar mesophases
upon heating. A lesser dependence of mobility on tem-
perature is generally observed within the temperature
range of one mesophase although observed changes
largely vary with the types of materials and mesophases.
A decrease in charge carrier mobility with increasing
temperature has been observed in the Colh phases of most
triphenylene based discotics,6,8 although temperature
independent mobility values have also been reported.9
In contrast, a small increase in mobility is observed in the
Colh phases of many discotic phthalocyanines.7,10
intermolecular electronic interactions are most likely the
dominant source of temperature dependence of charge
carrier mobility in columnar mesophases, especially when
mobility is measured by TRMC that provides intrinsic
(short-range) values. This assumption is supported by the
fact that charge carrier mobility in many higher ordered
discotic columnar mesophases (e.g., helical and plastic
columnar mesophases) is rather temperature independent
because molecular motions are much more restricted than
in conventional columnar mesophases.7,12
A quantum chemical molecular description of the
dependence of charge carrier mobility on intermolecular
electronic interactions based on the Markus theory has
recently been provided by several groups.13,14 The theory
predicts an increase in hopping rate (charge carrier
mobility) with increasing transfer integral and decreasing
reorganization energy. The transfer integral, which is a
function of the overlap of frontier orbitals of adjacent
molecules, crucially depends on the relative distances,
positions, and orientations of stacked aromatic cores. An
increase in temperature usually leads to an increase of the
stacking distance and translational mobility of discotic
molecules that both cause a decrease in transfer integral.
Similarly strong effects on the transfer integral have
changes in rotational angle of stacked molecules because
the area of overlapping π-systems changes with mutual
rotation depending on the shape of the aromatic core.
H-bonded stacks of discotic molecules are ideal materi-
als for probing this model because the columnar stacking
distance as well as relative orientations and locations
are controlled by H-bonds and, consequently, well-
defined.15-17 Gearba et al. reported an increase in charge
carrier mobility and stacking distance with increasing
temperature in H-bonded columnar discotic mesophases
of hexaazatriphenylene.15 This counterintuitive experi-
mental finding is likely a result of conformationally
coupled changes in the stacking distance and mutual
rotation angle governed by H-bonding between amide
groups. An increase in stacking distance forces the amide
groups more out-of-plane to maintain their ideal
H-bonding distance and reduces the rotation angle bet-
ween adjacent cores. The reduction in rotational angle
apparently increases the intermolecular charge-transfer
integral more than the increase in stacking distance
Interpretation and comparison of these results is com-
plex because different experimental techniques have been
used, such as time-resolved microwave conductivity
(TRMC) and time-of-flight photoconductivity, and the
temperature dependence is not only affected by changing
intermolecular electronic interactions but also the charge
transport mechanism as well as the number and types of
defects in the discotic material.11 However, changing
(1) (a) Cammidge, A. N.; Bushby, R. J. In Handbook of Liquid
Crystals; Demus, D., Goodby, J., Gray, G., W., Spiess, H.-W., Vill,
V., Eds.; Wiley-VCH: Weinheim, 1998; Vol. 2B,
p 693. (b)
Chandrasekhar, S. In Handbook of Liquid Crystals; Demus, D.,
Goodby, J., Gray, G., W., Spiess, H.-W., Vill, V., Eds.; Wiley-VCH:
Weinheim, 1998; Vol. 2B, p 749. (c) Boden, N.; Movaghar, B. In
Handbook of Liquid Crystals; Demus, D., Goodby, J., Gray, G., W.,
Spiess, H.-W., Vill, V., Eds.; Wiley-VCH: Weinheim, 1998; Vol. 2B, p
781. (d) Wu, J.; Pisula, W.; M€ullen, K. Chem. Rev. 2007, 107, 718. (e)
Kato, T.; Yasuda, T.; Kamikawa, Y.; Yoshio, M. Chem. Commun. 2009,
729. (f) Kumar, S. Liq. Cryst. 2004, 31, 1037.
(2) (a) Laschat, S.; Baro, A.; Steinke, N.; Giesselmann, F.; Hagele, C.;
Scalia, G.; Judele, R.; Kapatsina, E.; Sauer, S.; Schreivogel, A.;
Tosoni, M. Angew. Chem., Int. Ed. 2007, 46, 4832. (b) Sergeyev, S.;
Pisula, W.; Geerts, Y. H. Chem. Soc. Rev. 2007, 36, 1902.
(3) Percec, V.; Glodde, M.; Bera, T. K.; Miura, Y.; Shiyanovskaya, I.;
Singer, K. D.; Balagurusamy, V. S. K.; Heiney, P. A.; Schnell, I.;
Rapp, A.; Spiess, H. W.; Hudson, S. D.; Duan, H. Nature 2002,
419, 384.
(4) Bushby, R. J.; Lozman, O. R. Curr. Opin. Solid State Mater. Sci.
2002, 6, 569.
(5) Shen, X.; Dong, R. Y.; Boden, N.; Bushby, R. J.; Martin, P. S.;
Wood, A. J. Chem. Phys. 1998, 108, 4324.
(12) van de Craats, A. M.; Schouten, P. G.; Warman, J. M. J. Jpn. Liq.
Cryst. Soc. 1998, 2, 12.
(13) (a) Cornil, J.; Lemaur, V.; Calbert, J. P.; Bredas, J. L. Adv. Mater.
2002, 14, 726. (b) Lemaur, V.; Da Silva Filho, D. A.; Coropceanu, V.;
Lehmann, M.; Geerts, Y.; Piris, J.; Debije, M. G.; Van de Craats, A. M.;
Senthilkumar, K.; Siebbeles, L. D. A.; Warman, J. M.; Bredas, J. L.;
Cornil, J. J. Am. Chem. Soc. 2004, 126, 3271. (c) Xinliang, F.; Marcon,
V.; Pisula, W.; Hansen, M. R.; Kirkpatrick, J.; Grozema, F.; Andrienko,
D.; Kremer, K.; M€ullen, K. Nat. Mater. 2009, 8, 421.
(14) Kirkpatrick, J.; Marcon, V.; Kremer, K.; Nelson, J.; Andrienko, D.
J. Chem. Phys. 2008, 129, 094506.
(15) Gearba, R. I.; Lehmann, M.; Levin, J.; Ivanov, D. A.; Koch, M. H.
J.; Barbera, J.; Debije, M. G.; Piris, J.; Geerts, Y. H. Adv. Mater.
2003, 15, 1614.
(6) Adam, D.; Schuhmacher, P.; Simmerer, J.; Haussling, L.;
Siemensmeyer, K.; Etzbach, K. H.; Ringsdorf, H.; Haarer, D.
Nature 1994, 371, 141.
(7) Warman, J. M.; van de Craats, A. M. Mol. Cryst. Liq. Cryst. 2003,
396, 41.
(8) Adam, D.; Schuhmacher, P.; Simmerer, J.; Haussling, L.; Paulus,
W.; Siemensmeyer, K.; Etzbach, K.-H.; Ringsdorf, H.; Haarer, D.
Adv. Mater. 1995, 7, 276.
(9) Kreouzis, T.; Donovan, K. J.; Boden, N.; Bushby, R. J.; Lozman,
O. R.; Liu, Q. J. Chem. Phys. 2001, 114, 1797.
(10) Warman, J. M.; Kroeze, J. E.; Schouten, P. G.; van de Craats,
A. M. J. Porphyrins Phthalocyanines 2003, 7, 342.
(11) (a) Debije, M. G.; Piris, J.; de Haas, M. P.; Warman, J. M.;
(16) Paraschiv, I.; Giesbers, M.; van Lagen, B.; Grozema, F. C.;
Abellon, R. D.; Siebbeles, L. D. A.; Marcelis, A. T. M.; Zuilhof,
€
Tomovic, Z.; Simpson, C. D.; Watson, M. D.; Mullen, K. J. Am.
Chem. Soc. 2004, 126, 4641. (b) Iino, H.; Hanna, J.; Haarer, D. Phys.
Rev. B 2005, 72. (c) Iino, H.; Takayashiki, Y.; Hanna, J.; Bushby, R. J.;
Haarer, D. Appl. Phys. Lett. 2005, 87.
€
H.; Sudholter, E. J. R. Chem. Mater. 2006, 18, 968.
€
(17) Paraschiv, I.; Tomkinson, A.; Giesbers, M.; Sudholter, E. J. R.;
Zuilhof, H.; Marcelis, A. T. M. Liq. Cryst. 2007, 34, 1029.

1422 Chem. Mater., Vol. 22, No. 4, 2010
Demenev et al.
Scheme 1a
a Reagents and conditions: (i) NaOH, H2O/ethanol, reflux; (ii)
SOCl2, PhMe, reflux; (ii) 3,4,5-tridodecyloxyaniline, NEt3, THF, reflux.
reduces it, which is in line with theoretical findings.
Paraschiv et al. also reported an increase in charge carrier
mobility with increasing temperature for an H-bonded
columnar DLC, but in their material the amide groups are
decoupled from the columnar stacks of triphenylene cores
via aliphatic spacers.16 This decoupling improves solu-
bility and mesomorphism but complicates a correlation
between the orientation of amide groups and the packing
of discotic cores.
We report here quasi temperature independent electron
mobility across a phase transition of benzotristhiophene
trisphenylamide 3 as a new discotic molecule that has
three phenylamide groups directly linked to its core. The
observed temperature independence of intrinsic charge
carrier mobility is reasoned with two opposing effects that
compensate each other, the decrease of charge carrier
mobility due to an increase in stacking distance with
increasing temperature and the increase in charge carrier
mobility due to a decrease in rotation angle. We also
demonstrate the application of intermediate lyotropic
mesophases to overcome the notoriously difficult proces-
sing of H-bonded discotic materials into aligned fibres
and thin films.
Results and Discussion
Benzotristhiophene 3 was prepared from the earlier
described benzotristhiophene tricarboxylic acid 1,18 which
was converted to the acid chloride 2 and then reacted with
3,4,5-tridodecyloxyaniline to give the final product tria-
mide 3 as a pale yellow solid (Scheme 1). The identity and
Figure 1. a) Optical micrograph (760 ꢀ 360 μm, crossed polarizers) of a
free-standing film of 3 at 281 °C with fan-shaped textures at the interface
between isotropic (black) and LC areas; b) DSC graph of 3 (heating/
cooling at 10 °C/min, enthalpies given for transitions on cooling); c)
changes in wavenumbers with temperature of absorption maxima of
selected vibrational modes in IR.
the purity of the compound were established by 1H and 13
NMR, IR, HRMS, HPLC, and elemental analysis.
C
Polarized optical microscopy (POM) of compound 3
confirms the presence of a noncrystalline soft anisotropic
solid at room temperature that remains highly viscous
until it clears into its isotropic liquid at 279 °C. Fan-
shaped defect textures characteristic for columnar meso-
phases are observed only upon cooling from the isotropic
liquid at the boundary between the mesophase and
the isotropic liquid of a free-standing film (Figure 1).
Free-standing films of 1 mm diameter are surprisingly
stable in the isotropic liquid at 285 °C, which suggests
some H-bonding persists even at this temperature.
A high enthalpy of 79 kJ mol-1 for the transition
between mesophase and isotropic liquid was determined
by differential scanning calorimetry (DSC) and is addi-
tional evidence for the persistence of H-bonding over the
entire temperature range of the mesophase (Figure 1b).
For comparison, enthalpies for transitions between co-
lumnar discotic and isotropic liquid phases of discotic
(18) Taerum, T.; Lukoyanova, O.; Wylie, R.; Perepichka, D. F. Org.
Lett. 2009, 11, 3230.

Article
Chem. Mater., Vol. 22, No. 4, 2010 1423
materials free of intermolecular H-bonds are between
2-20 kJ mol-1 19
. Compound 3 undergoes only one other
phase transition at -8.9 °C (onset of peak on cooling)
that indicates the formation of a higher ordered solid
phase. No decomposition was observed upon heating to
290 °C, and the thermal stability of 3 determined by
thermal gravimetric analysis under He is 304 °C (onset
of weight loss).
Variable temperature IR measurements ultimately ver-
ified the persistence of H-bonds between amide groups
over the entire temperature range. Figure 1c shows the
changes in peak positions with temperature for the H-
bonding sensitive N-H and CdO (amide I) stretching
modes and the amide II mode. All values are typical for
H-bonded amide groups and vary only slightly with
temperature. No abrupt change in wavenumber is ob-
served across the phase transition between 0 and -20 °C.
Variable temperature XRD between -25 and 250 °C
reveals the presence of highly ordered hexagonal colum-
nar mesophases over the entire temperature range and
across the phase transition at about -10 °C. Analysis of
the diffraction patterns was best accomplished by collect-
ing 2D patterns of the Colh phase aligned in a fiber
(Figure 2a) because reflections around 2θ = 10° are
spatially separate while they overlap in the 1D diffraction
patterns. The small angle reflections orthogonal to the
long axis of the fiber indicate an alignment of the colum-
nar stacks along the fiber axis, and their relative recipro-
√ √ √ √
√
cal spacing of 1: 3: 4: 9: 12: 13 is characteristic of a
two-dimensional hexagonal columnar arrangement.
Both sharpness of the (10) peak and the observation of
several higher order reflections indicate high lateral pack-
ing order.
Also of high persistence length is the intracolumnar
π-stacking because it gives rise to a strong and relatively
Figure 2. a) 2D diffractionpattern ofa drawn fiber of3 at25 °C. The blue
line indicates the orientation of the fiber (250 μm in diameter) and the
columnarstackswithregard totheX-ray beam. Intercolumnardiffraction
peaks are orthogonal to the fiber. b) Change of d-spacings with tempera-
ture of the (10) and (001) reflections in the XRD patterns.
˚
narrow wide-angle reflection at about 3.5 A that is
persistent over the entire temperature range. This reflec-
tion is perpendicular to the reflections of the hexagonal
lattice, which verifies an orientation of the benzotristhio-
pene cores orthogonal to the stacking axis. A circular
“halo” at about 4.5 nm confirms an equal distribution of
the aliphatic chains around the columnar stacks. Two
Both intra- and intercolumnar spacings increase with
increasing temperature as illustrated in Figure 2b for the
d-spacings of the (10) and (001) reflections. The lattice
parameter “a” of the Colh phase (a = d(10)/cos 30)
˚
˚
˚
broad reflections at 9 A situated at an angle of 34° ( 10 to
increases from 33.9 A at -25 °C to 37.5 A at 250 °C
and the intracolumnar spacing increases from 3.39 A to
˚
the long axis of the fiber are attributed to a periodic
packing of the trialkoxyphenyl moieties that form a
helical “coil” as proposed in the structural model shown
˚
3.57 A. Both increases appear to be linear within the high
temperature Colh phase, but about 35% of the increases
occur at the phase transition between the high and low
temperature Colh phases at about -10 °C. The spacing
˚
in Figure 3. The determined spacing of about 9 A
corresponds to the distance between phenyl rings of every
second benzotristhiopene in a helical stack and is illu-
strated by an arrow in Figure 3b. Similar reflections have
been observed in other helical columnar stacks.20
˚
between the phenyl groups increases by only 0.7 A over
the entire temperature range, and no change in the
angular position of these broad reflections is observed.
While an increase in intracolumnar spacing with tem-
perature is common in columnar discotic mesophases, both
decreases21 and increases15,16 have been reported for the
intercolumnar spacing, which emphasizes the complex re-
lationship between conformational changes, interdigitation,
(19) (a) Lelj, F.; Morelli, G.; Ricciardi, G.; Roviello, A.; Sirigu, A. Liq.
Cryst. 1992, 12, 941. (b) Eichhorn, S. H.; Bruce, D. W.; Guillon, D.;
Gallani, J. L.; Fischer, T.; Stumpe, J.; Geue, T. J. Mater. Chem. 2001,
11, 1576. (c) Foster, E. J.; Jones, R. B.; Lavigueur, C.; Williams, V. E. J.
Am. Chem. Soc. 2006, 128, 8569.
€
(20) Pisula, W.; Tomovic, Z.; Watson, M. D.; Mullen, K.; Kussmann,
J.; Ochsenfeld, C.; Metzroth, T.; Gauss, J. J. Phys. Chem. B 2007,
111, 7481.
(21) Engel, M. K.; Bassoul, P.; Bosio, L.; Lehmann, H.; Hanack, M.;
Simon, J. Liq. Cryst. 1993, 15, 709.

1424 Chem. Mater., Vol. 22, No. 4, 2010
Demenev et al.
˚
1.81 A). This creates a helical stack with a pitch of 9
molecules (9 ꢀ 40=360°), but, due to the C3 symmetry of
3, the unit cell of the stack is only 3 molecules long. We
note that the high symmetry of this arrangement is broken
˚
if the stacking distance is reduced to 3.39 A as observed by
XRD in the low-temperature phase. To maintain the
hydrogen bonding, the rotation of amide groups must
be reduced to ∼37°, which increases the mutual rotation
between benzotrithiophene cores. The model also implies
that the increase in stacking distance with temperature
observed by XRD must be accompanied by a decrease in
rotational displacement between adjacent benzotristhio-
phene units because of the persistent intermolecular
H-bonding between amide groups. This interdependency
between stacking distance and mutual rotation is pro-
posed to be one reason for the observed temperature
independent charge carrier mobility discussed below.
In the optimized geometry, the phenyl rings are only
slightly rotated out of the plane of the amide groups (∼5°)
˚
and interact with each other through H...H contacts (2.46 A)
forming a helical “coil” around the column. The distance
between the turns of the coil (which corresponds to the
distance between the centers of the Ph rings of each second
Figure 3. Optimized (B3LYP/3-21G*) structure of a tetramer of tris-
(phenylaminocarbonyl)benzotristhiophene (as a model of 3) in a chiral
columnar arrangement seen down the stack (a) and from the side (b). The
arrow in (b) indicates the periodic packing of phenyl rings that gives rise
to the broad reflection in XRD at 9 A at an angle of about 34° to the axis
of the stack.
˚
molecule) is ∼9Aand in good agreement with the XRD data.
˚
We note that due to a prochiral nature of benzotrithio-
phene, either chiral (RRRR is shown in Figure 3) or
racemic (RSRS is given in the SI) columnar stacks can be
formed by 3. However, neither the diffraction data nor
the calculated small difference in energy (<1 kcal/mol in
favor of the racemic tetramer) allow for a clear differ-
entiation between the two structures. Structurally, the
chiral and racemic stacks only differ in the mutual posi-
tions of sulfur atoms in the neighboring molecules, which,
nevertheless, might have a profound effect on charge
transport through the benzotristhiophene stacks.
and maximum space-filling of alkyl chains. However, an
increase in intercolumnar spacing with temperature has
been observed in all reported H-bonded Colh phases15,16
and other helical discotic LCs.20,22
Density functional theory calculations (B3LYP/3-
21G*) were performed on a tetramer23 of amide 3 without
alkoxy substituents to better understand the intermole-
cular interactions inside the columnar stacks (Figure 3).
The B3LYP hybrid density functional has been earlier
used to successfully describe columnar arrangements of
hydrogen-bonded amides of trimesic acid.24 As expected,
the optimized geometry shows a π-stacking arrangement
of the molecules reinforced by hydrogen bonding between
the amide fragments of the adjacent molecules. The
Other unusual features of this material that may affect
charge transport are the locations of frontier orbitals
depicted in Figure 4. The trialkoxyaniline groups con-
tribute to 100% to the HOMO of 3, whereas the benzo-
tristhiophene core contributes 100% to the LUMO,
which suggests that hole transport must be facilitated
through the trialkoxyaniline groups and electron trans-
port through the benzotristhiophene cores.
˚
calculated stacking distance of 3.59 A is larger than the
˚
3.39 A observed by XRD in the low-temperature phase
even though a similar stacking distance was observed in
25
˚
the high-temperature phase (3.57 A).
The contribution of the trialkoxyaniline groups to the
HOMO is supported by cyclic voltammetry (CV) and
differential pulse voltammetry (DPV) measurements. Tri-
All amide groups in the model rotate out of the
thiophene plane by 40° and the adjacent molecules rotate
by an equivalent angle of 40° vs the stacking axis to
engage in hydrogen bonding interactions (N...H distance
amide 3 displays an irreversible oxidation wave at Eox
=
1.07 V vs Ag/AgCl (DPV) in benzonitrile solution that
corresponds to the formation of the radical cation on the
trialkoxyaniline groups rather than the benzotristhiophene
core. Eox of the unsubstituted benzotristhiophene has been
reported to be 1.39 V18 and is increased to 1.86 V by the
electron withdrawing carboethoxy substituents in a triester
of 1. A value similar to that of triester of 1 (X=OEt) is
expected for the benzotristhiophene triamide core.26
(22) Vyklicky, L.; Eichhorn, S. H.; Katz, T. J. Chem. Mater. 2003, 15,
3594.
(23) The distortion of the benzotrithiophene core in the top and bottom
molecules (Figure 3b) is an “edge effect”. All distance were
measured between the inner two molecules where this effect is less
apparent. Unfortunately, we presently can not model longer
oligomers of this compound or periodic boundary conditions.
ꢀ
(24) Rochefort, A.; Bayard, E.; Hadj-Messaoud, S. Adv. Mater. 2007,
19, 1992.
(25) An overestimation of the length of π,π interactions by B3LYP has
previously been reported and reasoned with an inaccurate descrip-
(26) Hammett’s constants σp of CO2Et and CONHMe substituents are
0.45 and 0.36, respectively . Hanch, C.; Leo, A.; Taft, R. W. Chem.
Rev. 1991, 91, 165.
ꢀ
tion of dispersive interactions Kristyan, S.; Pulay, P. Chem. Phys.
Lett. 1994, 229, 175.

Article
Chem. Mater., Vol. 22, No. 4, 2010 1425
Figure 5. Charge carrier mobility (left axis, solid line, squares) and
conductivity lifetime (right axis, dotted line, triangles) of compound 3
as a function of temperature.
transition is a mix of HOMO, HOMO-1, HOMO-2
(which are all close in energy) with LUMO and
LUMOþ1 orbital interactions and has a charge-transfer
character.
The intrinsic mobility of charges in compound 3 was
measured by time-resolved microwave conductivity
(TRMC) measurements28 at different temperatures
(Figure 5). A quasi constant mobility of about 0.02 cm2
V-1 s-1 for the sum of charge carriers was obtained
between -50 and 50 °C and, most remarkably, no abrupt
change in the charge carrier mobility and its depend-
ence on temperature was observed at the phase transition
(-10 °C). While the mobility is not affected by the phase
transition, the conductivity lifetime clearly changes and
half-life time increases from about 50 ns in the high
temperature Colh phase to almost 500 ns in the low
temperature Colh phase (Figure 5). Consequently, charge
recombination and/or trapping in the low temperature
Colh phase must be less efficient than in the high tem-
perature Colh phase.
Figure 4. Contributions to frontier orbitals of 3 calculated by DFT
(B3LYP/3-21G*; methyl for dodecyl groups); degenerated HOMO þ
HOMO-1 at -5.51 eV (top) and degenerated LUMO þ LUMOþ1
at -2.06 eV (bottom).
Furthermore an EHOMO of -5.4 eV is obtained based on
E
ox of 3using ferrocene as internal standard (-4.8 - Eox (vs
Fc/Fcþ)),27 which is in excellent agreement with the calcu-
lated EHOMO of -5.51 eV. The calculated ELUMO is rather
high (-2.06 eV). A corresponding electrochemical reduc-
tion would be expected at -2.3 V vs Ag/AgCl, which is
outside the electrochemical transparency window of the
electrolyte and was not observed.
The observed mobility of 0.02 cm2 V-1 s-1 is compar-
able to values reported for similarly sized discotic
alkoxy-substituted triphenylenes even though higher
values have been observed in discotic materials with
larger aromatic cores.29 Also, H-bonded columnar dis-
cotic materials of similar core size have shown higher
mobility15,16 than 3. Considering the results of DFT
calculations, it is likely that only electron mobility
contributes to the combined mobility measured for 3
by TRMC, which would explain the comparably low
value for the combined mobility but suggests a high
value for electron mobility. Hole transport through the
trialkoxyaniline groups is not feasible because their
UV-vis absorption spectra of 3 in THF solution and as
thin film on quartz (Figure S12 in the SI) are very similar,
and the only evidence of electronic interactions between
the molecules in the solid film is a slight broadening of the
absorptions and changes in relative intensities. The op-
tical gaps of 3 in solution and as thin film were calculated
to 3.18 and 3.09 eV, respectively, based on the onset of the
longest wavelength absorption. The optical gap is in
reasonably good agreement with the calculated HOMO-
-LUMO gap of 3.45 eV. In fact, time-dependent DFT
(TD-DFT) calculations (at the same level as the structure
optimization) predict the lowest energy transition at 3.11
eV (398 nm) in the planar configuration (gas-phase
minimum for individual molecule) and 3.32 eV (373 nm)
for the molecule with arylamide groups twisted out-of-
plane (minimum for stacked molecules). In both cases, the
˚
packing distance is larger than 4 A and only the trialk-
oxyaniline groups but not the benzotristhiophene core
(28) Warman, J. M.; de Haas, M. P.; Dicker, G.; Grozema, F. C.; Piris,
J.; Debije, M. G. Chem. Mater. 2004, 16, 4600.
(29) Warman, J. M.; van de Craats, A. M. Mol. Cryst. Liq. Cryst. 2003,
396, 41.
(27) Seguy, I.; Jolinat, P.; Destruel, P.; Farenc, J.; Mamy, R.; Bock, H.;
Ip, J.; Nguyen, T. P. J. Appl. Phys. 2001, 89, 5442.

1426 Chem. Mater., Vol. 22, No. 4, 2010
Demenev et al.
contribute to the highest occupied orbitals (HOMO to
HOMO-3) of 3 as outlined above.
A distinction between mobilities of positive and nega-
tive charge carriers is not possible by TRMC but by time-
of-flight (TOF) charge carrier mobility measurements.
Indeed, time-of-flight charge carrier mobility measure-
ments on 25 μm thick multidomain samples of 3 sand-
wiched between semitransparent Al electrodes on glass
slides revealed a fairly nondispersive transport for elec-
trons with a calculated mobility of up to 2 ꢀ 10-3 cm2
V-1 s-1 but no observable transient signal for holes. The
macroscopic (TOF) electron mobility across a 25 μm
multidomain sample is expectedly lower than the com-
bined intrinsic mobility measured by TRMC but only by
a factor of 10. In fact, this is a surprisingly high macro-
scopic electron mobility considering that the measure-
ments were taken in air on samples that are not ultra pure
and with a compound that has a high ELUMO. Both, the
absence of a transient signal for holes and the relatively
high mobility of electrons in the TOF measurements
confirm that the combined mobility measured by TRMC
is solely due to electrons.
Figure 6. Opticalmicrographs(35ꢀ 42μm) ofa shearalignedthin filmof
3 on glass with shear direction (arrow) parallel to one of the crossed
polarizers (left) and after turningthe sample ccw by10° (right). Alignment
was achieved in the lyotropic phase of 3 in heptane followed by evapo-
ration of the solvent.
than a decrease in packing distances with decreasing
temperature. Decreasing intercolumnar spacing, how-
ever, should increase the rate of recombination.
The high temperature Colh phase of 3 fulfills all com-
mon criteria of a liquid crystalline phase but it is not
liquidlike. In fact, the material is so highly viscous that no
flow can be mechanically induced into a free-standing
film at 275 °C. This lack of mobility impedes the proces-
sing into aligned monodomains that is important for
applications as an organic semiconductor in thin-film
electronics. We have previously demonstrated the appli-
cation of intermediate lyotropic mesophases for the align-
ment of highly viscous discotic phthalocyanines32 and
report here that the same methodology is applicable to
H-bonded discotic LCs. Triamide 3 was aligned in both
fibers (Figure 2a) and thin films (Figure 6) by extrusion
from lyotropic solutions or by mechanical shear align-
ment of lyotropic thin films on substrates, respectively.
Subsequent evaporation of the solvent (usually
n-heptane) at elevated temperatures under vacuum gene-
rates the aligned solvent free (thermotropic) mesophases
of 3 without significant disturbance of the alignment
(Figure 6).
Lyotropic Colh mesophases of 3 in n-heptane are best
prepared at elevated temperatures (50-60 °C) by mixing
equal masses of 3 and the solvent to avoid the formation
of biphasic mixtures at higher solvent contents. In con-
trast, isotropic organo-gels of 3 in n-heptane are obtained
at concentrations as low as 1% wt. IR spectroscopy of
the gel confirms that all amide groups are involved in
H-bonding, but no columnar stacks are detected by
XRD. Consequently, isotropic H-bonded networks of 3
must have been formed that contain no or only very short
columnar aggregates. A detailed investigation of the
lyotropic mesomorphism and gel formation is the subject
of a separate study.
The absence of an abrupt change in mobility at a phase
transition has been reported before for other meso-
morphic materials of high structural order because the
structural changes to the columnar stacking are minimal
at the phase transition.15,30 The estimated persistence
length of the columnar stacks of 3 determined by the
Scherrer equation based on the line broadening of the
˚
(001) diffraction peak is 105 A and practically constant
between -25 and 70 °C. Still surprising is that a decrease
in stacking distance by 0.1 A in 3 does not result in a
˚
measurable increase of the mobility. The most probable
explanation is that an increase in charge transfer integral
due to tighter packing of the stack at lower temperatures
is compensated by a decrease of charge transfer integral
due to an increase in mutual rotation angle between
neighboring molecules (see discussion above). Increasing
mobility due to thermal activation, however, may also
help compensate for a decreasing mobility with increasing
stacking distance.14
So, what causes the large change in lifetime if the
structural changes are minimal? Warman et al. proposed
that charge carrier recombination in highly ordered dis-
cotic columnar phases at time scales >5 ns is dominated
by recombination of charge carriers on separate columns
by intercolumnar electron tunnelling through the layer of
aliphatic chains.31 The authors also showed that the
tunnelling process critically depends on the packing
structure of the aliphatic chains among other factors.
We propose here that changes to the dynamic and con-
formational states of the side chains are mainly respon-
sible for the large increase in lifetime with decreasing
temperature in the low temperature Colh phase of 3
because no changes to the core packing are detected other
Conclusion
Incorporation of three amide groups in 3 is sufficient
for holding the molecules together in columnar stacks up
to a temperature of 280 °C and suppresses the formation
(30) van de Craats, A. M.; Warman, J. M.; Schlichting, P.; Rohr, U.;
€
Geerts, Y.; Mullen, K. Synth. Met. 1999, 102, 1550.
(31) Warman, J. M.; Piris, J.; Pisula, W.; Kastler, M.; Wasserfallen, D.;
€
(32) Eichhorn, S. H.; Bruce, D. W.; Wohrle, D. Adv. Mater. 1998, 10,
€
Mullen, K. J. Am. Chem. Soc. 2005, 127, 14257.
419.

Article
Chem. Mater., Vol. 22, No. 4, 2010 1427
of crystalline phases in favor of a low temperature Colh
mesophase. The specific design of 3 ensures the formation
of an isotropic liquid below its decomposition tempera-
ture and the formation of lyotropic mesophases at high
concentrations as well as isotropic organo-gels at low
concentrations in alkanes. Both, formation of an isotropic
liquid and lyotropic mesomorphism, are essential for the
preparation of uniaxially aligned fibers and films of 3 and
its potential use in devices in general. The intracolumnar
H-bonding between molecules of 3 in the hexagonal co-
lumnar mesophases necessitates a helical stacking of the
benzotristhiophene units and couples the observed increase
in stacking distance with increasing temperature with a
decrease in mutual rotation of adjacent benzotristhiophene
units. It is proposed here that this interdependence of
stacking distance and mutual rotation angle together with
the absence of major changes to the columnar structure are
responsible for the observed quasi temperature indepen-
dent charge carrier mobility. The combined charge mobility
of 0.02 cm2 V-1 s-1 measured by TRMC is modest for
H-bonded Colh phases but it originates from electrons
alone as confirmed by TOF charge carrier mobility mea-
surements. This unusual behavior (for oligothiophene ma-
terials) is a consequence of the specific topology of frontier
orbitals in 3: the LUMOs are localized on the benzotrithio-
phene unit and form electron channels in the columnar
stacks, while the HOMOs are localized on the pendant
trialkoxyphenyl substituents that are spaced too far apart
to form channels for hole transport. We expect that chang-
ing the location of the carbonyl and amino group in the
amide linker (NH attached to benzotristhiophene) would
localize both HOMO and LUMO on the benzotristhio-
phene core. This compound should show high hole mobi-
lity but probably lower electron mobility.
removed in vacuum. An off-white solid was obtained as crude
product that was washed with acetone, hexane, and ether and
finally recrystallized from THF/methanol to give 0.25 g (60%)
of analytically pure compound 3 as a pale yellow solid. 1H NMR
(500 MHz, pyridine-d6, δ): 11.15 (s, 3H, NH), 7.6 (3H, thio-
phene overlapping with Py), 4.95 (Ph overlapping with H2O),
4.26 (t, J=6.7 Hz, 6H, OCH2), 3.95 (t, J=6.3 Hz, 12H, OCH2),
1.90 (m, 6H, CH2), 1.77 (m, 12H, CH2), 1.63 (m, 6H, CH2), 1.48
(m, 12H, CH2), 1.4-1.2 (m, 144H, CH2), 0.87 (t, J=6.0 Hz, 27H,
CH3); (300 MHz, THF-d8, δ): 9.28 (broad s, 3H, NH), 7.92
(broad s, 3H, thiophene), 6.80 (broad s, 6H, Ph), 3.22 (broad s,
18H, OCH2), 1.6-1.4 (m, 18H, CH2 partially overlapping with
THF), 1.3-0.9 (m, 162H, CH2), 0.55 (t but not fully resolved,
27H, CH3); 13C NMR (75 MHz, THF-d8, δ): 160.2 (CdO),
153.93, 142.2, 136.6, 136.0, 135.21, 132.25, 122.8, 99.9, 73.7,
69.6, 32.9, 31.5, 30.7, 30.6, 30.4, 27.3, 27.2, 25.9, 25.6, 25.3, 25.0,
14.3; IR (KBr): ν=3270 (m, ν(N-H)), 3139 (w), 3079 (w), 2923
(s, ν(C-H)), 2852 (s, ν(C-H)), 1631 (s, ν(CdO)amide I), 1608 (s,
ν(CdC)arom.), 1545 (s, amide II), 1505 (s), 1464 (s), 1424 (s), 1382
(m), 1357(m), 1306 (m), 1273 (m), 1233 (s, ν(C-O-C)), 1114
(s,ν(C-O-C)), 1007 (m), 838 (m), 723 (m), 624 (w); UV-vis
(THF): λmax (ε)=332 (13000), 267 nm (5400); HRMS (MALDI-
ToF, m/z): [M þ H]þ calcd for C141H237N3O12S3, 2261.73;
found, 2261.89. Anal. Calcd for C141H237N3O12S3: C 74.85; H
10.56; N 1.86; found: C 75.02, H 10.46, N 1.59.
Spectroscopy and Spectrometry. UV-vis spectra of solutions
in THF (spectroscopic grade) were recorded on a Varian Cary
50 conc. UV-visible spectrophotometer. 1H NMR and 13C
NMR spectra were run on Bruker NMR spectrometers (DRX
500 MHz, DPX 300 MHz, and DPX 300 MHz with autotune).
Deuterated chloroform, THF, and pyridine were used as sol-
vents, and their residual proton signals functioned as reference
signals. Multiplicities of the peaks are given as s=singlet, d=
doublet, t=triplet, m=multiplet. Coupling constants for first-
order spin systems are given in Hz. Data are presented in the
following order (multiplicity, integration, coupling constant).
Fourier Transform Infrared spectra (FT-IR) were obtained on a
Bruker Vector 22. Relative peak intensities in IR are abbreviated
as vs=very strong, s=strong, m=medium, w=weak, br=
broad. Liquid samples were recorded as films on potassium
bromide plates, and solid samples were run as potassium
bromide pellets. Mass spectrometry measurements were per-
formed by Kirk Green at the Regional Center for Mass Spectro-
metry, and MALDI MS data were obtained on a Waters/
Micromass Micro MX.
Experimental Section
Synthesis of Benzo[B,b0,b00]tristhiophene-2,20,200-tricarbonyl
Trichloride (2). A mixture of benzotristhiophene tricarboxylic
acid 118 (0.030 g, 0.079 mmol) in 5 mL of dry toluene was stirred
at reflux for 12 h and then cooled to room temperature. Thionyl
chloride (0.094 g, 0.06 mL, 0.792 mmol) was added, and the
reaction mixture was stirred at reflux for 72 h. Excess of SOCl2
was removed in vacuum, and the residue was washed with
toluene (2 ꢀ 5 mL). The product was dried in vacuum to give
0.033 g (92%) of a white solid. mp: >320 °C; IR (KBr) (cm-1):
1723, 1542, 1494, 1350, 1182, 1096, 875, 842, 790, 663. 1H NMR
(DMSO-d6) (ppm): 8.45 (3 β-CH of thiophene); 13C NMR
(DMSO-d6) (ppm): 162.8 (CdO), 136.7, 135.2, 130.5, 127.8.
Synthesis of N,N0,N00-(3,4,5-Tridodecyloxyphenyl)benzo[b,b0,
b00]tristhiophene-2,20,200-tricarboxamide (3). To a suspension of
acid chloride 2 (0.081 g, 0.187 mmol) in dry THF (30 mL) was
added 3,4,5-tridodecyloxyaniline33 (0.60 g, 0.93 mmol) and
triethyl amine (0.11 mL, 0.81 mmol). The mixture was refluxed
for 15 h, precipitated salt was filtered off, and the solvent was
Mesomorphism. Polarized light microscopy was performed
on an Olympus TPM51 polarized light microscope that is
equipped with a Linkam variable temperature stage HCS410
and digital photographic imaging system (DITO1). Calori-
metric studies were performed on a Mettler Toledo DSC 822e,
and thermal gravimetric analysis was performed on a Mettler
Toledo TGA SDTA 851e. Helium (99.99%) was used to purge
the system at a flow rate of 60 mL/min. Samples were held at
30 °C for 30 min before heated to 550 °C at a rate of 10 °C/min.
All samples were run in aluminum crucibles. XRD measure-
ments were run on a Bruker D8 Discover diffractometer
equipped with a Hi-Star area detector and GADDS software
package. The tube is operated at 40 kV and 40 mA and Cu KR1
˚
radiation (λ = 1.54187 A) with an initial beam diameter of
0.5 mm is used. Compound 3 was studied as self-supported bulk
material and aligned fibers. Sample detector distances were
varied between 15.0 and 9.0 cm. A modified Instec hot and cold
stage HCS 402 operated via controllers STC 200 and LN2-P (for
below ambient temperatures) was used for variable temperature
(33) (a) Lincker, F.; Bourgun, P.; Masson, P.; Didier, P.; Guidoni, L.;
Bigot, J.-Y.; Nicoud, J.-F.; Donnio, B.; Guillon, D. Org. Lett. 2005,
7, 1505. (b) Percec, V.; Aqad, E.; Peterca, M.; Rudick, J. G.; Lemon, L.;
Ronda, J. C.; De, B. B.; Heiney, P. A.; Meijer, E. W. J. Am. Chem. Soc.
2006, 128, 16365. (c) Yelamaggad, C. V.; Achalkumar, A. S.; Rao, D. S.
S.; Prasad, S. K. J. Org. Chem. 2007, 72, 8308.

1428 Chem. Mater., Vol. 22, No. 4, 2010
Demenev et al.
XRD and IR measurements. Free standing films were obtained
by spreading compound 3 over sharp edged holes of 1-2 mm
diameter in metal plates at a temperature of 250 °C.
Time-of-flight (TOF) measurements were preformed on 25 μm
thick layers of compound 3 sandwiched between semitranspa-
rent Al electrodes on glass substrates. The TOF setup and the
measurement method were similar to that described by Melnyk
and Pai.34 All measurements took place in air under ambient
conditions (22 °C and 30% relative humidity). A Spectra-
Physics (VSL-337ND-S) nitrogen laser emitting a 4 ns pulse at
337 nm served as the excitation light source. At this wavelength
the Beer-Lambert attenuation coefficient is about 11 μm-1 in
the layer of 3. Thus, 90% of the laser pulse is absorbed within the
first 0.5 μm of the material. The thickness of this 90% absorp-
tion region is much thinner than the overall film thickness of
the layer. Measurements were performed at an electric field of
30 V/μm, and the measurement circuit resistance was adjusted to
ensure that the RC time constant was much less than the charge
carrier transit time. The resultant transient signals were re-
corded on a GHz storage oscilloscope (Tektronics TDS
580D). Typically 10-12 measurements were taken under the
same conditions and then averaged.
Electrochemistry. Electrochemical characterization was per-
formed with an Epsilon potentiostat in benzonitrile solution
(HPLC grade) using Bu4NPF6 (0.1 M) as electrolyte. A Pt disk,
Pt wire, and Ag/AgCl electrode were used as working, auxiliary
and reference electrodes, respectively. Ferrocene was added as
an internal standard at the end of the experiments, showing an
E0 of þ0.45 V. Because of the relatively low solubility and
ox
strong aggregation of triamide 3, its electrochemical characteri-
zation by CV and DPV was performed in benzonitrile at 65 °C.
Pulse-radiolysis time-resolved conductivity (PR-TRMC) mea-
surements were performed on solid samples (30 mg) in a perspex
container. The sample was placed in microwave cells consisting
of a rectangular waveguide with inner dimensions of 3.55 ꢀ
7.00 mm2 that was short-circuited with a metal end plate. The
materials were uniformly ionized with a nanosec pulse of
3 MeV electrons from a Van de Graff accelerator. The energy
absorbed by the sample (the radiation dose, D) was accurately
known from dosimetry and leads to the formation of charge
carrier pairs with concentrations of ∼1021 m-3. If the charge
carriers formed by ionization are mobile, the sample conduc-
tivity will increase upon irradiation. The change in conductiv-
ity of the sample was measured as a function of time by
monitoring the decrease in microwave power reflected by the
cell. The conductivity increases during the irradiation pulse
due to the formation of mobile charge carriers. After the pulse,
the conductivity decays because of trapping and/or recombi-
nation of positive and negative charges. From the value of the
conductivity at the end of the pulse, an estimate of the charge
carrier mobility may be obtained if the concentration of charge
carriers generated by irradiation is known. Accurate estimates
of the charge carrier concentrations generated by pulse
radiolysis can be made for discotic materials as described
previously.29 The PR-TRMC technique is described more
extensively in a recent review.28
Acknowledgment. A.D. and S.H.E. are grateful for finan-
cial support from NSERC of Canada, CFI/OIT, OCE, and
Xerox Corporation. T.T. and D.F.P. acknowledge support
from NSERC of Canada, Centre for Self-Assembled
Chemical Structures and CFI/LOF. F.C.G. acknowledges
a VENI grant from the Dutch Foundation for Scientific
Research (NWO).
Supporting Information Available: Spectra, chromatogram,
electrochemistry, variable temperature diffraction and IR data,
thermal gravimetric analysis, gel phase, calculated orbitals, and
racemic columnar stack as well as TOF data. This material is
available free of charge via the Internet at http://pubs.acs.org.
(34) Melnyk, A. R.; Pai, D. M. In Physical Methods of Chemistry Series,
2nd ed.; Rossiter, B. W., Baetzold, R. C., Eds.; John Wiley & Sons, Inc.:
New York, 1993; Chapter 5, Vol. VIII, pp 321-386.