Molecular wires from contorted aromatic compounds

Article abstract of DOI:10.1002/anie.200502142

(Figure Presented) In a twist: A design strategy for molecule-based electronic materials using aromatic compounds with a nonplanar core is presented. A new class of hexabenzocoronene is put forth that has its core severely distorted out of planarity into a corrugated structure. When substituted with four alkoxy side chains, this material self-assembles into infinitely long columns through stacking (see picture) and acts as an active layer in field-effect transistors.

Full text of DOI:10.1002/anie.200502142

Communications
Self-Assembly
DOI: 10.1002/anie.200502142
Molecular Wires from Contorted Aromatic
Compounds**
Shengxiong Xiao, Matthew Myers, Qian Miao,
SØbastien Sanaur, Keliang Pang, Michael L. Steigerwald,
and Colin Nuckolls*
Dedicated to Professor Thomas J. Katz
The use of planar molecules is one unifying design feature in
essentially all the molecules that have been tested to date in
organic field-effect transistors.[1–7] The concept explored
herein is whether a nonplanar aromatic core could yield
efficacious electronic materials. One significant aspect of
using a nonplanar core is that the aromatic rings would be
forced to adopt new and unusual intermolecular contacts that
are typically unavailable to flat molecules. Moreover, these
informationally rich subunits could form the basis for self-
assembled and self-healing electronic materials if the core
could be bent into a shape that was self-complementary.
Herein, we detail the first studies of the self-assembly and
electrical properties of a new type of hexabenzocoronene
(HBC) 1, whose aromatic core is distorted away from
planarity by steric congestion in its proximal carbon atoms.
We became interested in this class of compound because it
combines structural elements of linear acenes, which are well-
[*] S. Xiao, M. Myers, Q. Miao, K. Pang, M. L. Steigerwald,
Prof. C. Nuckolls
Department of Chemistry
The Nanoscience Center, Columbia University
New York, NY 10027 (USA)
Fax: (+1)212-932-1289
E-mail: cn37@columbia.edu
S. Sanaur
Ecole Nationale SupØrieure des Mines de Saint Etienne
Centre MicroØlectronique de Provence—Georges Charpak
DØpartement Packaging et Supports Souples
Avenue des AnØmones, Quartier Saint Pierre
13541 Gardanne (France)
[**] We acknowledge primary financial supportfrom the Nanoscale
Science and Engineering Initiative of the National Science Foun-
dation (NSF Award Number CHE-0117752); the New York State
Office of Science, Technology, and Academic Research (NYSTAR);
and the Department of Energy, Nanoscience Initiative
(NSET 04ER46118). C.N. thanks the US National Science Founda-
tion (CAREER award DMR-02-37860), the American Chemical
Society (PRF type G 39263-G7), the Camille Dreyfus Teacher Scholar
Program (2004), and the Alfred P. Sloan Fellowship Program (2004).
We thank the MRSEC Program of the National Science Foundation
(Award Number DMR-0213574) and the New York State Office of
Science, Technology, and Academic Research (NYSTAR) for finan-
cial support for M.L.S. and the shared instrument facility. S.S.
acknowledges financial support from the Centre MicroØlectronique
de Provence—Georges Charpak.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
7390
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 7390 –7394

Angewandte
Chemie
known electronic materials,[2,5,8–11] with a disk-shaped and
potentially columnar liquid-crystalline core.[12–20] Methods for
the synthesis and derivatization of
1 are essentially
unknown.[21] Our important finding is that some derivatives
of this new class of compound self-organize into liquid-
crystalline phases composed of molecular stacks that orient
themselves parallel to the surface. Field-effect transistors
based on these materials show high charge-carrier mobilities,
high on/off ratios, and low turn-on voltages.
Figure 1. a) Schematic diagram of a field-effect transistor that was
fabricated by evaporating source and drain electrodes onto a spin-cast
film of 1c. b) Transconductance: the blue line is a fit of the linear
portion of the data points (the source–drain voltage VS–D was held at
ꢀ60 V). c) Transistor output for thin-film transistors from 1c.
devices prepared from 1c) and the on/off current ratios in the
device (106:1), are also very good. These values are the best
field-effect transistor properties achieved for a columnar
discotic material[16,17,26,27] (see below) and are good for spin-
cast films in general.[8,28–34] These results are even more
impressive because it is likely that these properties could be
improved by adjustment of critical parameters such as grain
size, deposition conditions, and direction of column growth.
Encouraging electrical properties are only developed in
the more substituted derivatives of 1, thus indicating self-
organization. One of the hallmarks of many self-assembled
systems is the presence of liquid-crystalline phases. Each of
these compounds were investigated by DSC, polarized light
microscopy, and X-ray diffraction to determine whether
mesophases are present (the DSC trace for 1c is shown in
Figure 2a). An extra transition that indicates the presence of
an intermediate phase can be seen from both the heating and
cooling cycles for 1c. The microscopy and X-ray diffraction
studies of 1c (see below) indicate that it is a liquid-crystalline
phase between 91 and 2858C, whereas above these temper-
atures it melts to form an isotropic liquid. The enthalpy
values, particularly going from the liquid-crystalline phase to
the isotropic liquid phase, are larger than those typically
observed for discotic liquid crystals.[12] This difference implies
that 1c associates strongly in this intermediate phase, possibly
because it is not a flat disk.[35] DSC studies do not show any
evidence for a liquid-crystalline phase for 1a and 1b; they
simply melt, 1a at 5168C[21] and 1b at 2208C. The appearance
of a mesophase correlates with 1c having good electrical
properties and indicates the potential of self-organized
materials for electronic applications.
Compounds 1a–c[22] are yellow-orange solids that are
readily soluble in a number of common organic solvents; for
example, more than 200 mg of 1c can be dissolved in 1 mL of
CHCl3. Furthermore, these solutions have an intense optical
emission in the visible region. The thermal and oxidative
stabilities of these compounds are also remarkable, as no loss
of color is observed when oxygen is bubbled through solutions
of 1a–c over a period of a few hours under ambient light.
Under identical conditions, solutions of pentacene lose their
color within a few minutes through photooxidation.[23] Differ-
ential scanning calorimetry (DSC) showed no decomposition
of 1a–c up to temperatures well above 3208C. Uniform films
(approximately 100-nm thick) could be formed by spin-
casting 1c from 1,2-dichloroethane or CHCl3. In contrast,
films formed from 1a and 1b were granular in appearance and
of a low quality. The ability to form good films of 1c is a
harbinger of the electrical properties described below.
Field-effect transistors were fabricated on spin-cast films
of 1a–c (Figure 1a shows the structure of the resulting
devices). Transistors fabricated from 1c have good electrical
characteristics, whereas those prepared from 1a and 1b
showed no useful properties. We deposited Au source and
drain electrodes by thermal evaporation onto the spin-cast
films through a metal-shadow mask. The channel length was
75 mm and the electrode width was 2 mm, and the trans-
conductance and transistor output are shown in Figure 1b,c.
The mobility (0.02 cm2 Vꢀ1 sꢀ1) shown is calculated from the
linear portion of the data in Figure 1b[24] and uses a
capacitance of 11.3 nFcmꢀ2 for the gate dielectric layer
(300 nm of SiO2 and a monolayer of octadecyltrichloro-
silane).[11] The gate dielectric capacitance comes from meas-
urements made over a range of frequencies[11] and is as
expected for a 300-nm-thick layer of silicon oxide with few
defects.[25] Other critical parameters, such as the threshold
voltage for the device to turn on (as low as ꢀ3 V for some
The structure of this phase was probed by powder X-ray
diffraction studies and was found to be a hexagonally ordered
columnar liquid-crystalline phase for 1c. In the X-ray
diffraction studies, 1c was heated (in a Lindemann capillary
tube) to an isotropic liquid (2958C) and cooled into its
mesophase. The data shown in Figure 2b were collected upon
cooling to 1208C. The diffractogram is dominated by an
intense reflection characteristic of d ꢁ 26 Š. The only other
discernible features of the diffractogram are two very weak,
Angew. Chem. Int. Ed. 2005, 44, 7390 –7394ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7391

Communications
Figure 2. a) DSC trace of 1c (scan rate=58Cminꢀ1) showing a mesophase. b) X-ray diffraction studies of 1c cooled into its mesophase from 295
to 1208C and a hexagonal lattice of columns. c) Polarized light micrograph of 1c cooled into its mesophase which shows a planar alignment of
columns that are extinguished when aligned with the polarizer or analyzer axes. The cooling rate was 58Cminꢀ1, and the exposure time was the
same for each picture. d) Schematic diagram of columns aligned parallel to the surface.
higher-order reflections (d = 15 and 13 Š) that allow the
lattice to be indexed to a hexagonal arrangement of columns.
The lack of intensity in these higher-order reflections
indicates that the columns are not well correlated with each
other, which is likely to be a result of the subunits not having a
perfect disk shape, as is implied by the molecular model in
Figure 3. The fourfold substitution would signal that at some
level there is a competition between rectangular and hex-
agonal lattice formation.
birefringence of the domains is extinguished when the column
axes are aligned with the polarizer or analyzer (Figure 2c)
and is maximally bright when the stage is rotated by 458. The
micrograph is characteristic of a planar arrangement[36] of
columns in a discotic mesophase (shown schematically in
Figure 2d). This arrangement of columns is important
because the current flow in thin-film transistors is parallel
to the surface and, in this case, runs along the column axis.
Strong aggregation of 1c occurs in solution, thus further
indicating its tendency to self-associate. NMR, UV/Vis, and
fluorescence spectroscopic techniques were used to monitor
this aggregation in solution. All the resonances from the
protons of the aromatic ring shifted upfield as the solutions of
1c were concentrated. For example, the ring protons of 1c
shift upfield by approximately Dd = 0.13 ppm in CDCl3 as the
concentration is increased from 6.3 to 126 mgmLꢀ1. This shift
is a result of shielding from the ring current of neighboring
aromatic compounds within a cofacial stack.[37–39] This face-to-
face association of molecules can be further probed by UV/
Vis absorption and fluorescence emission spectroscopy. In
both cases, the maxima are red-shifted as the solutions
become more concentrated. This behavior is not unique as
there are many discotic materials that also aggregate in
solution to form isolated molecular columns which show
similar shifts in their UV/Vis and fluorescence spectra.[40–43]
The formation of the columns from the interlocking of
gearlike molecular subunits is the model that develops from
these spectroscopic data.
Figure 3. Molecular model of 1c with its side chains fully extended.
Polarized light microscopy revealed that these materials
tend to orient their columns parallel to the surface. The
micrograph of a sample cooled to just below 2788C and
captured as the mesophase formed is shown in Figure 2c. The
When films of these materials are spin-cast onto trans-
parent substrates, birefringent domains form that have the
same extinction as that described for the bulk films in
7392
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 7390 –7394

Angewandte
Chemie
ꢀ
Figure 2c, thus indicating that the films also have their
columns aligned parallel to the substrate. Aggregation in
solution that causes the formation of columns, which can then
be deposited by spin-casting, is likely to be responsible for the
good electrical properties of 1c.
Crystals of the parent HBC 1a were grown to investigate
the underlying molecular structure of the HBCs.[44] The
crystals were grown from 1,2,4-trichlorobenzene and shown
to include two disordered solvent molecules oriented over the
coronene core. The molecular structure of this HBC moiety is
shown in Figure 4a–d. The three intersecting pentacene
whereas the Cb Cc bond is relatively long. A strong contri-
bution from the radialene resonance would explain these
unusual bond lengths.
In summary, this study puts forth a new design strategy for
molecule-based electronic materials in which a nonplanar
rigid core is used. We prepared a new class of hexabenzocor-
onene that is readily substituted to form columnar liquid-
crystalline phases, and these columnar materials have good
electrical properties in thin-film transistors. Structures with
nonplanar aromatic units, such as 1, may be able to achieve
much higher mobilities than those available to flat hydro-
carbons because the p surfaces of contorted molecules can
approach each other and arrange themselves in very different
ways. Moreover, these molecules make interesting starting
materials for the synthesis of other topologically interesting
hydrocarbons by oxidative closure of the proximal carbon
atoms to form five-membered rings.
Received: June 20, 2005
Published online: September 20, 2005
Keywords: liquid crystals · molecular electronics ·
.
nanostructures · self-assembly
Figure 4. Crystal structure of 1a grown from 1,2,4-trichlorobenzene.
Hydrogen atoms have been removed for clarity. a) Face-on view with
[1] C. D. Dimitrakopoulos, P. R. L. Malenfant, Adv. Mater. 2002, 14,
99.
[2] M. Bendikov, F. Wudl, D. F. Perepichka, Chem. Rev. 2004, 104,
4891.
ꢀ
marked pivot points (magenta circles). The Ca Cb bond is relatively
ꢀ
shortand the C
Cc bond is relatively long. Side views are given in (b)
b
and (c). d) Side view of one of the acene segments extracted from the
crystal structure. The other atoms of the HBC are hidden from view.
[3] F. Garnier, Chem. Phys. 1998, 227, 253.
[4] G. Horowitz, J. Mater. Chem. 1999, 9, 2021.
[5] H. E. Katz, Z. Bao, S. L. Gilat, Acc. Chem. Res. 2001, 34, 359.
[6] D. J. Gundlach, Y. Y. Lin, T. N. Jackson, D. G. Schlom, Appl.
Phys. Lett. 1997, 71, 3853.
[7] C. Reese, M. Roberts, M.-m. Ling, Z. Bao, Mater. Today 2004, 7,
20.
[8] M. M. Payne, S. R. Parkin, J. E. Anthony, C.-C. Kuo, T. N.
Jackson, J. Am. Chem. Soc. 2005, 127, 4986.
[9] C. D. Sheraw, T. N. Jackson, D. L. Eaton, J. E. Anthony, Adv.
Mater. 2003, 15, 2009.
subunits contort into a zigzag conformation. This bending of
the acene core is visualized most clearly in Figure 4d, in which
portions of the structure have been suppressed to show only
one pentacene subunit of the HBC. Essentially all of the
bending is concentrated in pairs of carbon atoms that act as
pivot points (these are highlighted with magenta circles in
Figure 4a). A comparison of some of the bond lengths in this
HBC with those from the pentacene crystal structure[45] is
given in Table 1. One of the most interesting features is that
[10] H. Meng, M. Bendikov, G. Mitchell, R. Helgeson, F. Wudl, Z.
Bao, T. Siegrist, C. Kloc, C.-H. Chen, Adv. Mater. 2003, 15, 1090.
[11] Q. Miao, M. Lefenfeld, T.-Q. Nguyen, T. Siegrist, C. Kloc, C.
Nuckolls, Adv. Mater. 2005, 17, 407.
ꢀ
the Ca Cb bond is short in the HBC relative to the pentacene,
[12] S. Chandrasekhar, Handb. Liq. Cryst. 1998, 2B, 749.
[13] S. Chandrasekhar, S. K. Prasad, Contemp. Phys. 1999, 40, 237.
[14] D. Guillon, Struct. Bonding (Berlin) 1999, 95, 41.
[15] V. S. Iyer, M. Wehmeier, J. D. Brand, M. A. Keegstra, K. Müllen,
Angew. Chem. 1997, 109, 1676; Angew. Chem. Int. Ed. Engl.
1997, 36, 1604.
[16] W. Pisula, A. Menon, M. Stepputat, I. Lieberwirth, U. Kolb, A.
Tracz, H. Sirringhaus, T. Pakula, K. Müllen, Adv. Mater. 2005, 17,
684.
[17] A. M. van de Craats, N. Stutzmann, O. Bunk, M. M. Nielsen, M.
Watson, K. Müllen, H. D. Chanzy, H. Sirringhaus, R. H. Friend,
Adv. Mater. 2003, 15, 495.
[18] A. M. van de Craats, M. P. de Haas, J. M. Warman, Synth. Met.
1997, 86, 2125.
ꢀ
Table 1: Comparison of C C bond lengths between 1a and pentacene.
ꢀ
C C bond length [ꢀ]
1a[a]
Pentacene[b]
ꢀ
Ca Ca’
ꢀ
Ca Ca
–
1.44
1.46
1.42
1.38
1.45
1.43
1.35
1.43
1.43
1.40
1.45
1.42
1.42
1.38
1.42
ꢀ
Ca Cb
ꢀ
Cb Cc
ꢀ
Cc Cc
ꢀ
Cc Cd
ꢀ
Cd Ce
ꢀ
Ce Ce
[a] The atom labeling scheme is shown in Figure 4a. [b] The atom
labeling scheme of the pentacene skeleton is shown below.
[19] N. Boden, R. J. Bushby, J. Clements, B. Movaghar, J. Mater.
Chem. 1999, 9, 2081.
[20] Z. Wang, F. Dötz, V. Enkelmann, K. Müllen, Angew. Chem. Int.
Ed. 2005, 44, 1247.
[21] E. Clar, J. F. Stephen, Tetrahedron 1965, 21, 467.
Angew. Chem. Int. Ed. 2005, 44, 7390 –7394ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7393

Communications
[22] Details for the synthesis of 1a–c are contained in the Supporting
Information.
[23] A. Maliakal, K. Raghavachari, H. Katz, E. Chandross, T.
Siegrist, Chem. Mater. 2004, 16, 4980.
1/2
[24] The mobility is calculated by plotting j IDS
j
with j VG j and
using the equation: IDS = (mWCi/2L)(VGꢀV0)2 with W= elec-
trode width and L = channel length; the source–drain voltage is
set at ꢀ60 V, which is in the saturated region of the current–
voltage curves shown in Figure 1c.
[25] S. M. Sze,
Physics of Semiconductor Devices, Wiley-Inter-
science, New York, 1969.
[26] N. Boden, B. Movaghar, Handb. Liq. Cryst. 1998, 2B, 781.
[27] N. Boden, R. J. Bushby, O. R. Lozman, Z. Lu, A. McNeill, B.
Movaghar, K. Donovan, T. Kreouzis, Mol. Cryst. Liq. Cryst.
2004, 410, 541.
[28] M. Mushrush, A. Facchetti, M. Lefenfeld, H. E. Katz, T. J.
Marks, J. Am. Chem. Soc. 2003, 125, 9414.
[29] A. Facchetti, J. Letizia, M. H. Yoon, M. Mushrush, H. E. Katz,
T. J. Marks, Chem. Mater. 2004, 16, 4715.
[30] A. R. Murphy, J. M. J. Frechet, P. Chang, J. Lee, V. Subramanian,
J. Am. Chem. Soc. 2004, 126, 1596.
[31] H. E. Katz, W. Li, A. J. Lovinger, J. Laquindanum, Synth. Met.
1999, 102, 897.
[32] J. G. Laquindanum, H. E. Katz, A. J. Lovinger, J. Am. Chem.
Soc. 1998, 120, 664.
[33] P. T. Herwig, K. Müllen, Adv. Mater. 1999, 11, 480.
[34] M. Mas-Torrent, P. Hadley, S. T. Bromley, X. Ribas, J. Tarres, M.
Mas, E. Molins, J. Veciana, C. Rovira, J. Am. Chem. Soc. 2004,
126, 8546.
[35] A. M. Van de Craats, J. M. Warman, K. Müllen, Y. Geerts, J. D.
Brand, Adv. Mater. 1998, 10, 36.
[36] “Low Molecular Weight Liquid Crystals II”: D. Demus, J.
Goodby, G. W. Gray, H. W. Spiess, V. Vill, Handb. Liq. Cryst.
1998, 2B.
[37] R. J. Abraham, S. C. M. Fell, K. M. Smith, Org. Magn. Reso-
nance 1977, 9, 367.
[38] Y. Hamuro, S. J. Geib, A. D. Hamilton, J. Am. Chem. Soc. 1997,
119, 10587.
[39] A. S. Shetty, J. Zhang, J. S. Moore, J. Am. Chem. Soc. 1996, 118,
1019.
[40] T.-Q. Nguyen, R. Martel, P. Avouris, M. L. Bushey, L. Brus, C.
Nuckolls, J. Am. Chem. Soc. 2004, 126, 5234.
[41] O. Braitbart, R. Sasson, A. Weinreb, Mol. Cryst. Liq. Cryst. 1988,
159, 233.
[42] C. Nuckolls, T. J. Katz, L. Castellanos, J. Am. Chem. Soc. 1996,
118, 3767.
[43] U. Rohr, P. Schilichting, A. Böhm, M. Gross, K. Meerholz, C.
Bräuchle, K. Müllen, Angew. Chem. 1998, 110, 1463; Angew.
Chem. Int. Ed. 1998, 37, 1434.
[44] CCDC-275657 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from
the Cambridge Crystallographic Data Centre via www.ccdc.
cam.ac.uk/data_request/cif.
[45] C. C. Mattheus, A. B. Dros, J. Baas, A. Meetsma, J. L. de Boer,
T. T. M. Palstra, Acta Crystallogr. Sect. C 2001, 57, 939.
7394
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 7390 –7394