Preparation and characterization of π-stacking quinodimethane oligothiophenes. Predicting semiconductor behavior and bandwidths from crystal structures and molecular orbital calculations

Article abstract of DOI:10.1021/ja0484597

A series of new quinodimethane-substituted terthiophene and quaterthiophene oligomers has been investigated for comparison with a previously studied quinoid oligothiophene that has demonstrated high mobilities and ambipolar transport behavior in thin-film transistor devices. Each new quinoidal thiophene derivative shows a reversible one-electron oxidation between 0.85 and 1.32 V, a quasi-reversible one-electron second oxidation between 1.37 and 1.96 V, and a reversible two-electron reduction between -0.05 and -0.23 V. The solution UV-vis-NIR spectrum of each compound is dominated by an intense (ε ? 100 000 M^-1 cm^-1) low energy π-π* transition that has a γ_max ranging between 648 and 790 nm. All X-ray crystal structures exhibit very planar quinoidal backbones and short intermolecular π-stacking distances (3.335-3.492 A). Structures exhibit a single π-stacking distance with parallel cofacial stacking (sulfur atoms of equivalent rings pointed in the same direction) or with alternating distances and antiparallel cofacial stacking (sulfur atoms of equivalent rings pointed in the opposite direction). Examples of the layered and herringbone-packing motifs are observed for both the parallel and the antiparallel cofacial stacking. Analysis of the X-ray structures and molecular orbital calculations indicates that all of these compounds have one-dimensional electronic band structures as a result of the π-stacking. For structures with a unique π-stacking distance, a simple geometric overlap parameter calculated from the shape of the molecule and the slip from perfect registry in the π-stack correlates well with the transfer integrals (f) calculated using molecular orbital theory. The calculated valence (633 meV) and conduction (834 meV) bandwidths for a quinoid quaterthiophene structure are similar to those calculated for the benchmark pentacene and indicate that both hole and electron mobilities could be significant.

Full text of DOI:10.1021/ja0484597

Published on Web 10/28/2004
Preparation and Characterization of π-Stacking
Quinodimethane Oligothiophenes. Predicting Semiconductor
Behavior and Bandwidths from Crystal Structures and
Molecular Orbital Calculations
Daron E. Janzen,† Michael W. Burand,† Paul C. Ewbank,† Ted M. Pappenfus,†
Hiroyuki Higuchi,‡ Demetrio A. da Silva Filho,§ Victor G. Young,†
Jean-Luc Bre´das,§ and Kent R. Mann*,†
Contribution from the Department of Chemistry, UniVersity of Minnesota, Minneapolis,
Minnesota 55455, Department of Chemistry, Toyama UniVersity, Toyama 930-8555, Japan, and
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332
Received March 17, 2004; E-mail: mann@chem.umn.edu
Abstract: A series of new quinodimethane-substituted terthiophene and quaterthiophene oligomers has
been investigated for comparison with a previously studied quinoid oligothiophene that has demonstrated
high mobilities and ambipolar transport behavior in thin-film transistor devices. Each new quinoidal thiophene
derivative shows a reversible one-electron oxidation between 0.85 and 1.32 V, a quasi-reversible one-
electron second oxidation between 1.37 and 1.96 V, and a reversible two-electron reduction between -0.05
and -0.23 V. The solution UV-vis-NIR spectrum of each compound is dominated by an intense (ꢀ =
100 000 M-1 cm-1) low energy π-π* transition that has a λmax ranging between 648 and 790 nm. All X-ray
crystal structures exhibit very planar quinoidal backbones and short intermolecular π-stacking distances
(3.335-3.492 Å). Structures exhibit a single π-stacking distance with parallel cofacial stacking (sulfur atoms
of equivalent rings pointed in the same direction) or with alternating distances and antiparallel cofacial
stacking (sulfur atoms of equivalent rings pointed in the opposite direction). Examples of the layered and
herringbone-packing motifs are observed for both the parallel and the antiparallel cofacial stacking. Analysis
of the X-ray structures and molecular orbital calculations indicates that all of these compounds have one-
dimensional electronic band structures as a result of the π-stacking. For structures with a unique π-stacking
distance, a simple geometric overlap parameter calculated from the shape of the molecule and the slip
from perfect registry in the π-stack correlates well with the transfer integrals (t) calculated using molecular
orbital theory. The calculated valence (633 meV) and conduction (834 meV) bandwidths for a quinoid
quaterthiophene structure are similar to those calculated for the benchmark pentacene and indicate that
both hole and electron mobilities could be significant.
Introduction
The inherent electronic and optical properties of these
materials derive from the extended π-system. Molecular struc-
ture, conformation, and the degree of overlap between molecules
determine the nature and magnitude of the semiconductor
properties. In general, the mechanism of charge transport3 in
these well-ordered soft materials can be summarized as bandlike
at lower temperatures, where the mobility is primarily deter-
mined by the valence and conduction bandwidths, and thermally
activated localized polaronic hopping at higher temperatures,
where geometry relaxation (vibrational) effects play a major
role. P-channel,4-6 n-channel,7-11 and, most recently, ambipo-
lar11 materials that can function as both have been described.
The preeminent (and somewhat enigmatic) organic p-channel
semiconducting material is pentacene. Pentacene and related
oligoacenes exhibit some of the highest carrier mobilities
measured in FET devices to date, even though the herringbone
2-D edge-to-face layered structure does not contain π-stacks.4
The critical edge-to-face interactions that produce the electronic
Polymeric and oligomeric materials have shown excellent
promise for use in organic electronic devices such as thin-film
field effect transistors (FETs) and light-emitting diodes (LEDs).1
Soluble organic semiconducting electronic materials have poten-
tial advantages over existing inorganic semiconductors including
processability and tunable electronic properties. Organic elec-
tronic materials are suitable for large area device manufacture
via printing2 because of mechanical flexibility and economical
manufacture. If the proper “soft” materials can be discovered,
processing techniques requiring expensive clean room and highly
specialized lithographic tools could be eliminated in many
applications.
† University of Minnesota.
‡ Toyama University.
§ Georgia Institute of Technology.
(1) (a) Dimitrakopoulos, C. D.; Malenfant, P. R. L. AdV. Mater. 2002, 14,
99-117. (b) Handbook of Oligo- and Polythiophenes; Fichou, D., Ed.;
Wiley-VCH: Weinheim, Germany, 1999.
(2) (a) Blanchet, G.; Rogers, J. J. Imaging Sci. Technol. 2003, 47, 296-303.
(b) Blanchet, G. B.; Loo, Y.-L.; Rogers, J. A.; Gao, F.; Fincher, C. R.
Appl. Phys. Lett. 2003, 82, 463-465.
(3) (a) Karl, N. Synth. Met. 2003, 133-134, 649-657. (b) Campbell, I. H.;
Smith, D. L. Int. J. High Speed Electron. Syst. 2001, 11, 585-615.
9
10.1021/ja0484597 CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 15295-15308
15295

A R T I C L E S
Janzen et al.
properties of pentacene are apparently diminished by any
attempts at functionalization. In the extreme, substitution causes
transformation to a π-stacked structure.12 Derivatization is
necessarily accomplished at the edges of pentacene molecules,
disrupting the edge-to-face overlap. In contrast, molecules that
π-stack are more flexible. Substituents can be accommodated
by slippage, rotation, or expansion along the stacking axis while
maintaining some measure of intermolecular overlap.13 As the
π-overlap at perfect alignment can be large, even relatively large
deviations from perfect registry can in principle still give
materials with appreciable mobilities.
As it is a commonly held belief that strong intermolecular
interactions (such as π-stacking) can give materials with high
charge mobilities, many polymeric and oligomeric thiophene
systems have been investigated and found to be good p-channel
materials. For example, p-channel FETs constructed with defect-
free head-to-tail regioregular poly(3-hexylthiophene) (HT-P3HT)
(96% regioregularity) have shown excellent mobilities of 0.05-
0.1 cm2/(V s).5 The percent regioregularity and the method of
film deposition (cast vs spun) also affect the orientation of the
layered lamellae with respect to the substrate. The measured
mobility was highly anisotropic, with the mobility in the
π-stacking direction better by more than a factor of 100.
Because it can be difficult to determine X-ray structures of
amorphous or semicrystalline polymers such as HT-P3HT, the
study of well-characterized oligomeric systems should help
establish detailed structure-property relationships. A recent
example is bisdithienothiophene (BDT), a molecule with a short
π-stacking distance of 3.557 Å and good π-overlap.6 Calcula-
tions14 indicate BDT has significant interchain transfer integrals
(t) for both the HOMO and the LUMO (172 and 27 meV,
respectively), giving valence and conduction bandwidths (4t)
of 688 and 108 meV, respectively. The p-channel FET devices
constructed with BDT6 have mobilities of 0.02-0.05 cm2/(V
s), which lie within 2 orders of magnitude of the best pentacene
devices. Thus, both polymeric and oligomeric thiophenes are
viable alternatives to pentacene for p-channel materials and will
also offer the chance of property modification by derivatization.
Although thiophene derivatives are less studied as n-channel
materials,7 where intermolecular overlap is equally critical, the
N,N′-dialkylated 3,4,9,10-perylenetetracarboxylic acid diimides
(PTCDIs) are a particularly well-studied family of π-stacking
n-channel materials.8 Devices with mobilities as high as 0.6 cm2/
(V s) have been constructed using octyl disubstituted PTCDI
(C8PTCDI).9 Furthermore, it has been shown that an important
electronic property (the HOMO-LUMO optical band gap)
correlates with simple crystal packing parameters.13 Similar
correlations have not been described for quinoidal oligo-
thiophenes, and only recently have the geometric factors been
delineated for aromatic oligothiophenes.15
With these (and many other excellent studies as a backdrop),
we seek structural correlations for the observed properties of
the quinoidal oligothiophene DCMT (3′,4′-dibutyl-5,5′′-bis(di-
cyanomethylene)-5,5′′-dihydro-2,2′:5′,2′′-terthiophene).10,11,16,17
This family of molecules may be described electronically as
oxidized oligothiophene dications stabilized by malononitrile
anion capping groups. A spectroscopic comparison of DCMT
and its reduced forms with the corresponding aromatic olig-
othiophene and its oxidized forms confirms that the π-systems
are equivalent.16 Previously, we reported that vapor- and
solution-deposited films of DCMT function as n-channel
semiconductors with electron mobilities of 0.005 and 0.002 cm2/
(V s), respectively.10 More recently, devices constructed from
DCMT with different film morphologies demonstrate either
enhanced n-channel mobilities (0.2 cm2/(V s)) or ambipolar (n-
channel and p-channel depending on the sign of the gate bias)
behavior with lower mobilities (both hole and electron mobilities
< 10-4 cm2/(V s)).11 The unique electronic structure of the
quinoid thiophene molecules allows them to function as
simultaneous donors and acceptors. Their strong propensity to
π-stack and the availability of soluble n-hexyl-substituted
precursor oligomers make them ideal for systematic study of
how the solid-state structure affects semiconductor performance.
This work represents our efforts to further understand the effects
of alkyl substitution and oligomer length on the molecular and
solid-state properties of quinoidal oligothiophenes (Figure 1).
Analyzing the relationships between packing structure and solid-
state electronic structure can provide insight into the choice and
design of materials for good FET performance.
(4) (a) Campbell, R. B.; Robertson, J. M.; Trotter, J. Acta Crystallogr. 1961,
14, 705-711. (b) Lin, Y. Y.; Gundlach, D. J.; Nelson, S. F.; Jackson, T.
N. IEEE Electron DeVice Lett. 1997, 18, 606-608. (c) Butko, V. Y.; Chi,
X.; Lang, D. V.; Ramirez, A. P. Appl. Phys. Lett. 2003, 83, 4773-4775.
(d) Meng, H.; Bendikov, M.; Mitchell, G.; Helgeson, R.; Wudl, F.; Bao,
Z.; Siegrist, T.; Kloc, C.; Chem, C.-H. AdV. Mater. 2003, 15, 1090-1093.
(e) Butko, V. Y.; Chi, X.; Ramirez, A. P. Solid State Commun. 2003, 128,
431-434.
(5) Sirringhaus, H.; Brown, P. J.; Friend, R. H.; Nielsen, M. M.; Bechgaard,
K.; Langeveld-Voss, B. M. W.; Spiering, A. J. H.; Janssen, R. A. J.; Meijer,
E. W.; Herwig, P.; de Leeuw, D. M. Nature 1999, 401, 685-688.
(6) Li, X.-C.; Sirringhaus, H.; Garnier, F.; Holmes, A. B.; Moratti, S. C.; Feeder,
N.; Clegg, W.; Teat, S. J.; Friend, R. H. J. Am. Chem. Soc. 1998, 120,
2206-2207.
(7) (a) Facchetti, A.; Yoon, M.-H.; Stern, C. L.; Katz, H. E.; Marks, T. J.
Angew. Chem., Int. Ed. 2003, 42, 3900-3903. (b) Facchetti, A.; Mushrush,
M.; Katz, H. E.; Marks, T. J. AdV. Mater. 2003, 12, 33-38. (c) Sakamoto,
Y.; Komatsu, S.; Suzuki, T. J. Am. Chem. Soc. 2001, 123, 4643-4644. (d)
Yassar, A.; Demanze, F.; Jaafari, A.; El Idrissi, M.; Coupry, C. AdV. Funct.
Mater. 2002, 12, 699-708.
(8) Struijk, C. W.; Sieval, A. B.; Dakhorst, J. E. J.; van Dijk, M.; Kimkes, P.;
Koehorst, R. B. M.; Donker, H.; Schaafsma, T. J.; Picken, S. J.; van der
Craats, A. M.; Warman, J. M.; Zuilhof, H.; Sudholter, E. J. R. J. Am. Chem.
Soc. 2000, 122, 11057-11066.
(9) Malenfant, P. R.; Dimitrakopoulos, C. D.; Gelorme, J. D.; Kosbar, L. L.;
Graham, T. O.; Curioni, A.; Andreoni, W. Appl. Phys. Lett. 2002, 80, 2517-
2519.
(10) Pappenfus, T. M.; Chesterfield, R. J.; Frisbie, C. D.; Mann, K. R.; Casado,
J.; Raff, J. D.; Miller, L. L. J. Am. Chem. Soc. 2002, 124, 4184-4185.
(11) Chesterfield, R. J.; Newman, C. R.; Pappenfus, T. M.; Ewbank, P. C.;
Haukaas, M. H.; Mann, K. R.; Miller, L. L.; Frisbie, C. D. AdV. Mater.
2003, 15, 1278-1282.
(12) Haddon, R. C.; Chi, X.; Itkis, M. E.; Anthony, J. E.; Eaton, D. L.; Siegrist,
T.; Mattheus, C. C.; Palstra, T. T. M. J. Phys. Chem. B 2002, 106, 8288-
8292.
Experimental Section
General Considerations. All syntheses were carried out under an
inert atmosphere of Ar or N2. 1,2-Dimethoxyethane and N,N-dimeth-
ylformamide were distilled from Na/benzophenone under N2; CH2Cl2
was distilled from CaH2 under N2. Pd(PPh3)4 (Strem Chemicals, Inc.),
malononitrile (Acros Organics), N-iodosuccinimide, 3-hexylthiophene,
2,5-dibromothiophene, and 5,5′-dibromo-2,2′-bithiophene (Aldrich)
were used as received as were other reagents unless otherwise indicated.
2-Bromo-3-hexylthiophene was synthesized according to literature
(15) Curtis, M. D.; Cao, J.; Kampf, J. W. J. Am. Chem Soc. 2004, 126, 4318-
4328.
(16) Pappenfus, T. M.; Raff, J. D.; Hukkanen, E. J.; Burney, J. R.; Casado, J.;
Drew, S. M.; Miller, L. L.; Mann, K. R. J. Org. Chem. 2002, 67, 6015-
6024.
(17) Casado, J.; Miller, L. L.; Mann, K. R.; Pappenfus, T. M.; Higuchi, H.;
Orti, E.; Milian, B.; Pou-Amerigo, R.; Hernandez, V.; Lopez Navarrete, J.
T. J. Am. Chem. Soc. 2002, 124, 12380-12388.
(13) Klebe, G.; Graser, F.; Hadicke, E.; Berndt, J. Acta Crystallogr. 1989, B45,
69-77.
(14) Cornil, J.; Beljonne, D.; Calbert, J.-P.; Bredas, J.-L. AdV. Mater. 2001, 13,
1053-1067.
9
15296 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004

π
-Stacking Quinodimethane Oligothiophenes
A R T I C L E S
1
°C, 0.016 Torr. H NMR (300 MHz, CDCl3): δ 6.97 (s, 1H), 2.52 (t,
2H), 1.54 (t, 2H), 1.30 (m, 6H), 0.88 (t, 3H).
5,5-Dimethyl-2-(3-hexyl-2-thienyl)-1,3,2-dioxaborinane (D). Com-
pound D was synthesized from 2-bromo-3-hexylthiophene by the
reported method20 for the 3-octylthiophene analogue. D was obtained
as a colorless oil (6.59 g, 23.5 mmol, 67%). 1H NMR (300 MHz,
CDCl3): δ 7.44 (d, 1H, J ) 4.8), 7.01 (d, 1H, J ) 4.8), 3.77 (s, 4H),
2.89 (t, 2H), 1.59 (m, 2H), 1.33 (m, 6H), 1.04 (s, 9H), 0.90 (t, 3H).
HREIMS C15H25BO2S calcd, 280.1668; found, 280.1670 (M+).
3,4′-Dihexyl-5′-chloro-2,2′-bithiophene (E). Compound E was
synthesized by a previously reported method20 for the 3-octylthiophene
analogue from the starting materials D (3.196 g, 11.40 mmol) and a
4:1 mixture of C (4.03 g, 9.81 mmol) and B. The crude product was
dissolved in hexanes and filtered through a short pad of silica to remove
unreacted D and phosphanes (catalyst ligands). The filtrate was
concentrated, and the low-boiling components (including residual B
and C) were removed by Kugelrohr distillation (95 °C, 0.07 Torr) to
leave a yellow oil (3.39 g, 94%) that was sufficiently pure for
subsequent reaction. 1H NMR (300 MHz, CDCl3): δ 7.19 (d, 1H, J )
5.1), 6.94 (d, 1H, J ) 5.1), 6.83 (s, 1H), 2.74 (t, 2H), 2.61 (t, 2H),
1.65 (t, b, 4H), 1.37 (m, 12H), 0.93 (m, 6H). 13C NMR (300 MHz,
CDCl3): δ 139.8, 139.5, 132.5, 129.9, 126.5, 123.8, 31.6 (two peaks
separated by 0.02 ppm), 30.7, 29.5, 29.2, 29.1, 28.9, 28.0, 22.6, 14.1.
5′-Chloro-5-iodo-3,4′-dihexyl-2,2′-bithiophene (F). Compound F
was synthesized from E in 59% yield by a previously reported method20
Figure 1. Molecular structures of quinoidal terthiophenes.
procedures.18 5,5′′′-Bis(dicyanomethylene)-3,3′,4′′,3′′′-tetrahexyl-5,5′′′-
dihydro-2,2′:5′,2′′:5′′,2′′′-quaterthiophene (DCMQ) was synthesized by
a previously reported method.18 Malononitrile was purified by vacuum
distillation prior to use for the synthesis of Q3. 1H NMR spectra were
measured on a Varian Unity 300 (300 MHz) or Varian Inova 300 (300
MHz) instrument. The chemical shifts are reported in ppm, and the
coupling constants (J) are given in Hz. All spectra are referenced to
the residual proton peak of the solvent: chloroform, 7.27 ppm;
dichloromethane, 5.32 ppm; and 1,1,2,2-tetrachloroethane, 6.00 ppm;
for 13C spectra, the reference is the chloroform peak at 77.23 ppm.
Mass spectra were obtained on a Finnigan MAT 95 or VG 7070E-HF
mass spectrometer. Fast-atom bombardment (FAB) MS were acquired
with a m-nitrobenzyl alcohol/trifluoroacetic acid matrix. Elemental
analyses were obtained from Quantitative Technologies Inc., White-
house, NJ. Solution electronic spectra in dichloromethane solutions were
collected on a computer-controlled Cary 17 or Ocean Optics spectro-
photometer. Electrochemical experiments were performed with a BAS
100B electrochemical analyzer and a glassy carbon working electrode.
Solutions of electroactive compound (∼0.5 mM) were made in
dichloromethane with 0.1 M TBAPF6 as the supporting electrolyte.
Potentials are reported versus aqueous Ag/AgCl and are not corrected
for the junction potential. The E°′ value for the ferrocenium/ferrocene
couple under the conditions used in this study was +0.42 V.19
2-Chloro-3-hexylthiophene (A). Compound A was synthesized by
a slight modification of a previously reported method20 for the
3-octylthiophene analogue. Separation of starting material, desired
product, and the dichlorinated byproduct (B) was difficult to accomplish
by distillation. Thus, an excess of chlorinating agent was employed to
ensure that no 3-hexylthiophene remained in the product. The small
amount of dichlorinated byproduct (A:B ratio ) 5:1) formed using this
strategy was inert to subsequent reactions and could be removed at a
later stage. The colorless oil was distilled at 60 °C, 0.15 Torr. 1H NMR
(300 MHz, CDCl3): δ 7.02 (d, 1H, J ) 8.4), 6.80 (d, 1H, J ) 8.4),
2.57 (t, 2H), 1.57 (t, 2H), 1.31 (m, 6H), 0.89 (t, 3H).
1
for the 3-octylthiophene analogue. H NMR (300 MHz, CDCl3): δ
7.06 (s, H), 6.75 (s, H), 2.66 (t, 2H), 2.57 (t, 2H), 1.58 (m, 4H), 1.33
(m, 12H), 0.90 (m, 6H). 13C NMR (300 MHz, CDCl3): δ 141.9, 139.9,
138.7, 136.2, 131.3, 127.2, 124.9, 72.0, 31.8, 30.8, 29.7, 29.3, 29.1,
28.2, 22.8, 14.3 (two peaks separated by 0.02 ppm). HREIMS C20H28-
ClIS2 calcd, 494.0366; found, 494.0334 (M+).
5-Chloro-3′,3′′,4-trihexyl-2,2′:5′,2′′-terthiophene (G). Compound
G was synthesized from F in 38% yield by a previously reported
method20 for the 3-octylthiophene analogue. 1H NMR (300 MHz,
CDCl3): δ 7.17 (d, 1H, J ) 5.4), 6.93 (d, 1H, J ) 5.4), 6.93 (s, 1H),
6.82 (s, 1H), 2.78 (t, 2H), 2.71 (t, 2H), 2.58 (t, 2H), 1.64 (m, 6H), 1.33
(m, 18H), 0.91 (m, 9H). HREIMS C30H43ClS3 calcd, 534.2215; found,
534.2179 (M+).
2,5-Dibromo-3-hexylthiophene (H). This compound has been
previously prepared by a different procedure.18 In the dark, N-
bromosuccinimide (1.38 g, 7.76 mmol) was added in portions to
2-bromo-3-hexylthiophene (1.92 g, 7.76 mmol) in a solution of
chloroform and acetic acid (80 mL, 1:1 v/v) at 0 °C. The mixture was
allowed to warm to room temperature and stirred for 24 h. Water (50
mL) was added, and the organic layer was extracted with chloroform,
1
washed with 2.0 M KOH (2 × 200 mL), and dried with MgSO4. H
NMR spectroscopy of the resulting oil revealed the reaction was
incomplete, so the oil was redissolved in chloroform and acetic acid
(80 mL, 1:1 v/v) and additional N-bromosuccinimide (1.08 g, 6.07
mmol) was added. The mixture was refluxed for 1.5 h. Water (100
mL) was added, and the organic layer was extracted with chloroform,
washed with 2.0 M KOH (2 × 200 mL), and dried with MgSO4. The
solvent was removed via rotary evaporation, and the resulting dark oil
was filtered through a short pad of silica with hexanes. Removal of
the solvent by rotary evaporation yielded 2.21 g (87%) of H as a
2,5-Dichloro-3-hexylthiophene (B). Compound B was observed as
a byproduct in the synthesis of A (above). 1H NMR (300 MHz,
CDCl3): δ 6.65 (s, 1H), 2.51 (t, 2H), 1.57 (m, 2H), 1.31 (m, 6H), 0.89
(t, 3H).
2-Chloro-3-hexyl-5-iodothiophene (C). Compound C was synthe-
sized by a slight modification of a previously reported method20 for
the 3-octylthiophene analogue. The starting material was the 5:1 mixture
of A and B reported above. The product obtained was a 5:1 mixture of
C and B, recovered as a colorless oil by Kugelrohr distillation at 95
1
colorless oil. H NMR (300 MHz, CDCl3): δ 6.79 (s), 2.51 (t, J )
7.5), 1.56 (m), 1.30 (m), 0.90 (m).
3,3′′-Dihexyl-2,2′:5′,2′′-terthiophene (1).21 Mg turnings (1.33 g, 54.8
mmol) were suspended in 50 mL of diethyl ether. A solution of
2-bromo-3-hexylthiophene (10.3 g, 41.8 mmol) in 30 mL of diethyl
ether was added slowly to the reaction vessel via an addition funnel
over the course of 20 min. The resulting mixture was refluxed for 1.3
h. The solution was cannulated into an addition funnel connected to a
(18) Higuchi, H.; Nakayama, T.; Koyama, H.; Ojima, J.; Wada, T.; Sasabe, H.
Bull. Chem. Soc. Jpn. 1995, 68, 2363-2377.
(19) Graf, D. D.; Duan, R. G.; Campbell, J. P.; Miller, L. L.; Mann, K. R. J.
Am. Chem. Soc. 1997, 119, 5888-5899.
(20) Bidan, G.; De Nicola, A.; Enee, V.; Guillerez, S. Chem. Mater. 1998, 10,
1052-1058.
(21) Gallazzi, M. C.; Castellani, L.; Zerbi, G.; Sozzani, P. Synth. Met. 1991,
41, 495-498.
9
J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004 15297

A R T I C L E S
Janzen et al.
second flask, which contained a solution of 2,5-dibromothiophene (3.48
g, 14.4 mmol) and Ni(dppp)Cl2 (98 mg, 0.18 mmol) in 50 mL of diethyl
ether. After slow addition of the Grignard solution via an addition
funnel, the solution was refluxed for 1.25 h. The solution was cooled
to 0 °C, and 1 M HCl (50 mL) was added slowly. The organic layer
was extracted with diethyl ether (2 × 60 mL), washed with saturated
aqueous Na2CO3 (200 mL), and dried with MgSO4. The solvent was
removed via rotary evaporation to yield a brown oil which was filtered
through a short pad of silica, and then Kugelrohr distilled (203 °C, 1.5
g, 7.92 mmol) was added. Glacial acetic acid (1.35 mL, 23.8 mmol)
was slowly added dropwise via syringe. The ice bath was removed,
and the reaction mixture was allowed to stir at room temperature in
the dark. After 1 h, the solution was washed with saturated aqueous
Na2CO3 (2 × 50 mL) and dried with MgSO4. The solvent was removed
by rotary evaporation, and the resulting oil was purified using column
chromatography (silica gel/100% hexanes) to provide 1.38 g (57%) of
1
1I as a bright yellow solid. H NMR (300 MHz, CDCl3): δ 7.08 (s,
2H), 6.99 (s, 2H), 2.72 (t, J ) 7.5), 1.61 (m), 1.31 (m), 0.89 (m).
1
Torr) to yield 5.66 g (94%) of 1 as a yellow oil. H NMR (300 MHz,
HREIMS calcd, 667.9599; found, 667.9577 (M+). Anal. Calcd for
CDCl3): δ 7.19 (d, 2H, J ) 5.1), 7.06 (s, 2H), 6.95 (d, 2H, J ) 5.4),
2.79 (t, J ) 7.8), 1.66 (m), 1.33 (m), 0.89 (m). HREIMS C24H32S3
calcd, 416.1666; found, 416.1668 (M+).
C24H30I2S3: C, 43.12; H, 4.52. Found: C, 43.79; H, 4.51.
3,3′,3′′-Trihexyl-5,5′′-diiodo-2,2′:5′,2′′-terthiophene (2I). A solution
of 2 (1.50 g, 2.99 mmol) and 50 mL of dichloromethane was cooled to
0 °C, and N-iodosuccinimide (1.48 g, 6.59 mmol) was added. Glacial
acetic acid (1.12 mL, 19.8 mmol) was slowly added dropwise via
syringe. The ice bath was removed, and the reaction mixture was
allowed to stir at room temperature in the dark. The solvent was
removed via rotary evaporation, and the resulting oil was redissolved
in 30 mL of dichloromethane. After 1 h, the solution was washed with
saturated aqueous Na2CO3 (2 × 50 mL) and dried with MgSO4. The
solvent was removed by rotary evaporation, and the resulting oil was
purified by filtering through a short pad of silica with dichloromethane
3,3′,3′′-Trihexyl-2,2′:5′,2′′-terthiophene (2).18,22 Compound 2 was
prepared by two different methods.
Method A. A previously reported synthesis was followed;18 the only
change was that a 10-fold mol excess of 2-bromo-3-hexylthiophene
was used for the coupling reaction. The yield was 85%.
Method B. Mg turnings (740 mg, 30.4 mmol) were suspended in
50 mL of diethyl ether. A solution of 2-bromo-3-hexylthiophene (5.75
g, 23.3 mmol) in 30 mL of diethyl ether was added slowly to the
reaction vessel via an addition funnel over the course of 50 min. The
resulting mixture was refluxed for 2 h. The solution was cannulated to
an addition funnel connected to a second flask, which contained a
solution of 2,5-dibromo-3-hexylthiophene (2.62 g, 8.02 mmol) and Ni-
(dppp)Cl2 (55 mg, 0.10 mmol) in 50 mL diethyl ether. After slow
addition of the Grignard solution via addition funnel, the solution was
refluxed for 45 h. The solution was cooled to 0 °C, and 1 M HCl (30
mL) was added slowly, followed by the addition of water (30 mL).
The organic layer was extracted with diethyl ether (3 × 50 mL), washed
with saturated aqueous Na2CO3 (200 mL), and dried with MgSO4. The
solvent was removed via rotary evaporation to yield a brown oil, which
was filtered through a short pad of silica, and the lower-mass impurities
were removed by Kugelrohr vacuum distillation at 200 °C to yield 2.42
g (60%) of 2 as a yellow oil. 1H NMR (300 MHz, CDCl3): δ 7.31 (d,
1H, J ) 5.1), 7.16 (d, 1H, J ) 4.8), 6.98 (s, 1H), 6.98 (d, 1H, J )
4.8), 6.93 (d, 1H, J ) 5.1), 2.78 (t, J ) 8.1), 2.53 (m), 1.56 (m), 1.26
(m), 0.86 (m).
1
to provide 2.05 g (91%) of 2I as a yellow-brown oil. H NMR (300
MHz, CDCl3): δ 7.12 (s, 1H), 7.07 (s, 1H), 6.90 (s, 1H), 2.71 (m),
2.50 (m), 1.55 (m), 1.26 (m), 0.88 (m). HREIMS C30H42I2S3 calcd,
752.0538; found, 752.0534 (M+).
5,5′′-Dibromo-3,3′,3′′-trihexyl-2,2′:5′,2′′-terthiophene (2Br). N-
Bromosuccinimide was added in portions to 2 (300 mg, 0.60 mmol) in
a solution of chloroform and acetic acid (15 mL, 1:1 v/v) at room
temperature. After being stirred for 1 h, the reaction mixture was poured
into water and extracted with hexane. The extracts were shaken with
aqueous saturated Na2CO3, washed with brine, and dried with MgSO4.
The residue obtained after removal of the solvents was chromatographed
on silica gel (3.8 × 25 cm) with hexane to afford 315 mg (80%) of the
1
dibromo compound 2Br as a pale yellow oil. H NMR (300 MHz,
CDCl3): δ 6.93 (s, 1H, Th-H), 6.89 (s, 1H, Th-H), 6.88 (s, 1H, Th-
H), 2.69 (t, 2H, J ) 7.5, CH2C5H11), 2.50-2.45 (m, 4H, CH2C5H11),
1.61-1.26 (m, 24H, CH2-(CH2)4-CH3), 0.89-0.83 (m, 9H, CH3).
UV-vis (THF, nm (ꢀ/(M-1 cm-1)): 247 (7800), 318 (14 300). IR (neat,
cm-1): ν ) 2955, 2925, and 2855 (C-H). EI-MS: m/z, 656 (M+),
658 (M + 2+), and 660 (M + 4+) for C30H42Br2S3 based on 79Br. Anal.
3,4′,4′′-Trihexyl-2,2′:5′,2′′-terthiophene (3).22 A solution of G (1.00
g, 1.86 mmol) in 20 mL of THF under Ar was cooled to -78 °C, and
n-butyllithium (2.5 M in hexanes, 1.49 mL, 3.73 mmol) was added
dropwise over the course of 10 min. The solution was allowed to warm
to 0 °C, was quenched slowly with water (25 mL), and then hexanes
(25 mL) was added. After warming to room temperature, the organic
phase was washed with water (3 × 30 mL) and dried with MgSO4.
Rotary evaporation of the solvent yielded 750 mg (80%) of 3 as a yellow
Calcd for C30H42Br2S3: C, 54.71; H, 6.43. Found: C, 54.55; H, 6.60.
3,4′,4′′-Trihexyl-5,5′′-diiodo-2,2′:5′,2′′-terthiophene (3I). A solution
of 3 (500 mg, 0.998 mmol) in 30 mL of dichloromethane under Ar
was cooled to 0 °C, and N-iodosuccinimide (494 mg, 2.20 mmol) was
added. Glacial acetic acid (0.37 mL, 6.59 mmol) was slowly added
dropwise via syringe. The ice bath was removed, and the reaction
mixture was allowed to stir at room temperature in the dark. After 20
h, the solution was washed with saturated aqueous Na2CO3 (2 × 30
mL) and dried with MgSO4. The solvent was removed by rotary
evaporation, and the resulting oil was purified by filtering through a
short pad of silica with hexanes followed by dichloromethane to yield
1
oil. H NMR (300 MHz, CDCl3): δ 7.16 (d, 1H, J ) 5.1), 6.97 (d,
1H, J ) 1.2), 6.93 (s, 1H), 6.93 (d, 1H, J ) 5.4), 6.90 (d, 1H, J )
1.5), 2.75 (m), 2.62 (m), 1.65 (m), 1.33 (m), 0.89 (m). HREIMS
C30H44S3 calcd, 500.2605; found, 500.2605 (M+).
3,3′′′-Dihexyl-2,2′:5′,2′′:5′′,2′′′-quaterthiophene (4).23 4 was syn-
thesized in a manner similar to a literature procedure23 with the
following exceptions: a slight excess of Mg (1.3 equiv) was used,
reaction times were extended (31.5 h), and the product 4 was not
purified by column chromatography prior to use in the next synthesis.
1H NMR (300 MHz, CDCl3): δ 7.19 (d, 2H, J ) 5.1), 7.14 (d, 2H, J
) 3.9), 7.03 (d, 2H, J ) 3.6), 6.95 (d, 2H, J ) 5.4), 2.79 (t, J ) 7.8),
1.66 (m), 1.33 (m), 0.90 (m).
1
721 mg (96%) of 3I as a yellow oil. H NMR (300 MHz, CDCl3): δ
7.07 (s, 1H), 6.86 (s, 1H), 6.77 (s, 1H), 2.71 (m), 2.55 (m), 1.61 (m),
1.32 (m), 0.90 (m). HREIMS calcd, 752.0538; found, 752.0597 (M+).
Anal. Calcd for C30H42I2S3: C, 47.87; H, 5.62. Found: C, 47.05; H,
5.54.
3,3′′′-Dihexyl-5,5′′′-diiodo-2,2′:5′,2′′:5′′,2′′′-quaterthiophene (4I).
A solution of 4 (1.00 g, 2.00 mmol) in 20 mL of dichloromethane was
cooled to 0 °C, and N-iodosuccinimide (925 mg, 4.11 mmol) was added.
Glacial acetic acid (0.75 mL, 13.3 mmol) was slowly added dropwise
via syringe. The ice bath was removed, and the reaction mixture was
allowed to stir at room temperature. After 4 h, the solvent was removed
via rotary evaporation. The resulting solid was suspended in methanol
and filtered, and then purified using column chromatography (silica
gel/3:1 dichloromethane:hexanes) which yielded 873 mg (58%) of 4I
as a bright orange solid. 1H NMR (300 MHz, CDCl3): δ 7.11 (d, 2H,
3,3′′-Dihexyl-5,5′′-diiodo-2,2′:5′,2′′-terthiophene (1I).24 A solution
of 3,3′′-dihexyl-2,2′:5′,2′′-terthiophene (1.50 g, 3.60 mmol) in 60 mL
of dichloromethane was cooled to 0 °C, and N-iodosuccinimide (1.78
(22) Barbarella, G.; Bongini, A.; Zambianchi, M. Macromolecules 1994, 27,
3039-3045.
(23) Azumi, R.; Go¨tz, G.; Debaerdemaeker, T.; Ba¨uerle, P. Chem.-Eur. J. 2000,
6, 735-744.
(24) Kokubo, H.; Yamamoto, T. Macromol. Chem. Phys. 2001, 202, 1031-
1034.
9
15298 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004

π
-Stacking Quinodimethane Oligothiophenes
A R T I C L E S
J ) 3.9), 7.08 (s, 2H), 6.97 (d, 2H, J ) 3.6), 2.73 (t, J ) 7.8), 1.61
(m), 1.31 (m), 0.89 (t, J ) 6.6). UV-vis (CH2Cl2, nm (ꢀ/(M-1 cm-1)):
252 (18 000), 387 (37 000). HRFABMS calcd, 749.94764; found,
749.94692 (M+). Anal. Calcd for C28H32I2S4: C, 44.80; H, 4.30; I,
33.81. Found: C, 45.12; H, 4.13; I, 34.07.
5,5′′-Bis(dicyanomethylene)-3,3′′-dihexyl-5,5′′-dihydro-2,2′:5′,2′′-
60% in oil, 4.60 mmol) in 1,2-dimethoxyethane (20 mL) was added
malononitrile (126 mg, 1.90 mmol) at 0 °C under Ar. The reaction
mixture was removed from the ice bath and stirred for 10 min, and
then 3I (600 mg, 0.79 mmol) and Pd(PPh3)4 (73 mg, 0.063 mmol) were
added. The mixture was heated under reflux for 3.5 h. The resulting
dark red mixture was cooled to 0 °C, and 30 mL of saturated Br2/H2O
solution was added dropwise. The solution turned blue, and a precipitate
formed. The suspension was filtered through a short pad of diatoma-
ceous earth, washed with water (1 × 30 mL), and purified using column
chromatography (silica gel/100% dichloromethane) to yield 262 mg
terthiophene (Q1). To a suspension of sodium hydride (174 mg, 60%
in oil, 4.34 mmol) in 1,2-dimethoxyethane (25 mL) was added
malononitrile (119 mg, 1.80 mmol) at 0 °C under Ar. In a separate
round-bottom flask, diiodo compound 1I (500 mg, 0.748 mmol) was
added to Pd(PPh3)4 (69 mg, 0.060 mmol) and 25 mL of 1,2-
dimethoxyethane. The reaction mixture was degassed and heated to
reflux. The sodium hydride solution was cannulated into the solution
of 1I, and the resulting solution was heated under reflux for 2 h. The
resulting dark red mixture was cooled to 0 °C, and 25 mL of saturated
Br2/H2O solution was added dropwise. The solution turned dark blue,
and a precipitate formed. The suspension was filtered through a short
pad of diatomaceous earth, washed with water (2 × 15 mL), and
purified using column chromatography (silica gel/100% dichlo-
romethane) to yield 268 mg (68%) of Q1 as a metallic green solid. 1H
NMR (300 MHz, C2D2Cl4, 100 °C): δ 7.46 (s, 2H), 7.18 (s, 2H), 2.87
(t, J ) 7.8), 1.81 (quintet, J ) 7.5), 1.45 (m), 0.99 (t, J ) 6.9). UV-
vis (CH2Cl2, nm (ꢀ/(M-1 cm-1)): 546 sh, 594 sh, 648, 689 sh. HREIMS
calcd, 542.1633; found, 542.1631 (M+). Anal. Calcd for C30H30N4S3:
1
(53%) of Q3 as a metallic green solid. H NMR (300 MHz, C2D2Cl4,
100 °C): δ 7.34 (s, 1H), 7.22 (s, 1H), 7.13 (s, 1H), 2.96 (t, J ) 7.5),
2.85 (m), 1.78 (m), 1.49 (m), 0.99 (m). UV-vis (CH2Cl2, nm (ꢀ/(M-1
cm-1)): 537 sh, 581 sh, 640 sh, 672 (93 000), 768 sh. HREIMS calcd,
626.2572; found, 626.2596 (M+). Anal. Calcd for C36H42N4S3: C, 68.97;
H, 6.75; N, 8.94. Found: C, 68.44; H, 6.56; N, 8.70.
5,5′′′-Bis(dicyanomethylene)-3,3′′′-dihexyl-5,5′′′-dihydro-2,2′:5′,2′′:
5′′,2′′′-quaterthiophene (Q4). To a suspension of sodium hydride (93
mg, 60% in oil, 2.32 mmol) in 1,2-dimethoxyethane (15 mL) was added
malononitrile (64 mg, 0.96 mmol) at 0 °C under Ar. The reaction mix-
ture was removed from the ice bath, was stirred for 0.5 h, and then 4I
(300 mg, 0.40 mmol) and Pd(PPh3)4 (370 mg, 0.32 mmol) were added.
The mixture was heated under reflux for 3.5 h. The resulting dark red
mixture was cooled to 0 °C (via ice bath), and 20 mL of saturated
Br2/H2O solution was added dropwise. The solution turned blue, and a
precipitate formed. An additional 40 mL of water was added to the
suspension while stirring. The suspension was filtered and washed with
water (3 × 10 mL). The dark blue solid was purified using column
chromatography by sequentially eluting the column with 100% CH2-
Cl2, 5/95 ethyl acetate/CH2Cl2, and 10/90 ethyl acetate/CH2Cl2 to yield
68 mg (30%) of Q4 as a dark solid. UV-vis (CH2Cl2, nm (ꢀ/(M-1
cm-1)): 709 sh, 779 (190 000), 867 sh. HRFABMS calcd ([C34H32N4S4
+H]+), 625.1588; found, 625.1594. LRFABMS calcd ([C34H31-
BrN4S4+H]+), 704.8; found, 705.1; calcd ([C34H30Br2N4S4+H]+), 783.7;
found, 784.0. Anal. Calcd for C34H31.76Br0.24N4S4: C, 63.44; H, 4.97;
N, 8.71. Found: C, 63.53; H, 5.05; N, 8.24.
C, 66.38; H, 5.57; N, 10.32. Found: C, 66.42; H, 5.43; N, 10.07.
5,5′′-Bis(dicyanomethylene)-3,3′,3′′-trihexyl-5,5′′-dihydro-2,2′:
5′,2′′-terthiophene (Q2). Compound Q2 was prepared by two different
methods.
Method A.18,25,26 To a suspension of sodium hydride (89 mg, 60%
in oil, 2.2 mmol) in 1,2-dimethoxyethane (15 mL) was added
malononitrile (66 mg, 1.0 mmol) at 0 °C. The mixture was allowed to
warm to room temperature, and 2Br (264 mg, 0.40 mmol) and Pd-
(PPh3)4 were added successively. The reaction was heated at reflux for
4 h. The resulting mixture was extracted with chloroform, washed with
brine, and dried with MgSO4. The dark solid obtained after removal of
the solvents was chromatographed on silica gel (3.8 × 22 cm) with
hexane/chloroform (5:3) to afford 62 mg (26%) of the quinoid
compound Q2, which was recrystallized from hexane/chloroform as a
X-ray Structure Determinations. Relevant crystallographic data
are shown in Table 1. All crystals were grown by slow evaporation of
saturated solutions (Q1, CH3CN/CH2Cl2; Q2, CH2Cl2; Q3, CH2Cl2/n-
heptane; Q3tol, CH2Cl2/toluene; Q4, CH2Cl2). The data for the single-
crystal X-ray structures of Q2, Q3, and Q4 were collected at the APS
synchrotron sector 15-ID-C at Argonne National Laboratory due to the
extremely small size of at least one crystal dimension. In each case,
the crystals used for the synchrotron structure determinations were
obtained after months of attempts with different solvents and methods
of crystallization. Single crystals were attached to glass fibers and
mounted on the Bruker Kappa/SMART 6000 microdiffractometer for
data collection at 100 K using double diamond-monochromated Ag
KR radiation (λ ) 0.5500 Å). An initial set of cell constants was
calculated from reflections harvested from three sets of 65 frames
oriented such that orthogonal wedges of reciprocal space were surveyed.
Final cell constants were calculated from a minimum of 2033 strong
reflections from the actual data collection. Data were collected to the
extent of 2.0 hemispheres at a resolution of 0.80 Å using φ-scans.
Somewhat larger crystals were available for Q1 and Q3tol; the data
for these structure determinations were collected at the X-ray Crystal-
lographic Laboratory (Department of Chemistry, University of Minneso-
ta). Single crystals were attached to glass fibers and mounted on the
Siemens SMART system for data collection at 173 K using graphite-
monochromated Mo KR radiation (λ ) 0.71073 Å). An initial set of
cell constants was calculated from reflections harvested from three sets
of 20 frames oriented such that orthogonal wedges of reciprocal space
were surveyed. Final cell constants were calculated from a minimum
set of 765 strong reflections from the actual data collection. Data were
collected to the extent of 1.5-2.0 hemispheres at a resolution of 0.84
Å using φ-scans.
dark microcrystalline powder.
Method B. To a suspension of sodium hydride (464 mg, 60% in
oil, 11.6 mmol) in 1,2-dimethoxyethane (60 mL) was added malono-
nitrile (316 mg, 4.78 mmol) at 0 °C under Ar. In a separate round-
bottom flask, diiodo compound 2I (1.50 g, 1.99 mmol) was added to
Pd(PPh3)4 (184 mg, 0.159 mmol) and 50 mL of 1,2-dimethoxyethane
and heated to reflux under Ar. The sodium hydride solution was
cannulated into the solution of 2I, and the resulting solution was heated
under reflux for an additional 2 h. The resulting dark red mixture was
cooled to 0 °C, and 50 mL of saturated Br2/H2O solution was added
dropwise. The solution turned dark blue, and a precipitate formed. The
suspension was filtered through a frit, and the isolated solid was washed
with water (4 × 15 mL) to yield 1.51 g of crude solid. A 200 mg
portion of the crude solid was filtered through a short pad of silica
with dichloromethane, and then further purified using column chro-
matography (silica gel/100% dichloromethane) to yield 82 mg (50%)
of Q2 as a dark solid. 1H NMR (300 MHz, C2D2Cl4, 100 °C): δ 7.29
(s, 1H), 7.21 (s, 1H), 7.15 (s, 1H), 2.99 (t, J ) 7.5), 2.86 (t, J ) 7.8),
1.81 (m), 1.49 (m), 0.99 (m). UV-vis (CH2Cl2, nm): 669. UV-vis
(THF, nm (ꢀ/(M-1 cm-1)): 600 sh (42 700), 660 (94 300). EIMS 627
(M+), 628 (M + 1+), 629 (M + 2+). IR (KBr, cm-1) ν ) 2210 (Ct
N). Anal. Calcd for C36H42N4S3: C, 68.97; H, 6.75; N, 8.94. Found:
C, 68.71; H, 6.70; N, 8.72.
5,5′′-Bis(dicyanomethylene)-3,4′,4′′-trihexyl-5,5′′-dihydro-2,2′:
5′,2′′-terthiophene (Q3). To a suspension of sodium hydride (183 mg,
(25) Higuchi, H.; Yoshida, S.; Uraki, Y.; Ojima, J. Bull. Chem. Soc. Jpn. 1998,
71, 2229-2237.
(26) Yui, K.; Aso, Y.; Otsubo, T.; Ogura, F. Bull. Chem. Soc. Jpn. 1989, 62,
1539-1546.
9
J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004 15299

A R T I C L E S
Janzen et al.
Q4
Table 1. X-ray Data Collection and Refinement
compound
Q1
Q2
Q3
Q3tol
formula
habit
color
lattice type
space group
a, Å
C
plate
metallic green
triclinic
P1h
7.951(3)
13.212(4)
13.936(4)
95.876(6)
98.489(6)
97.492(6)
1424.6(8)
2
542.76
1.265
173(2)
0.286
30H30N4S3
C
ribbon
aqua
triclinic
P1h
6.555(2)
14.858(7)
18.269(8)
93.214(16)
96.441(17)
94.512(17)
1758.8(13)
2
626.92
1.184
100(2)
0.130
36H42N4S3
C
plate
metallic green
triclinic
P1h
5.6513(9)
15.832(3)
19.096(3)
98.562(7)
90.409(7)
93.139(7)
1686.7(5)
2
36H42N4S3
C
rod
36H42N4S3‚C7H8
C34H30.8Br0.24S4
needle
metallic
monoclinic
P21/c
6.2902(10)
8.6138(11)
29.089(4)
90
94.301(6)
90
1571.7(4)
2
642.84
metallic green
triclinic
P1h
8.4670(14)
12.873(2)
19.364(5)
104.072(15)
102.560(14)
96.391(14)
1967.9(7)
2
719.05
1.213
173(2)
0.224
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
formula wt, g mol-1
Dc, g cm-3
temp (K)
µ, mm-1
F (000)
626.92
1.234
100 (1)
0.135
1.358
100(2)
0.340
670
572
668
668
768
θ range, deg
1.49-25.07
-9 e h e 9
-15 e k e 15
-16 e l e 16
14 280
0.88-19.58
-7 e h e 7
-17 e k e 17
0 e l e 21
30 111
2.56-20.18
-7 e h e 7
-19 e k e 19
0 e l e 23
22 100
1.12-25.05
-10 e h e 9
-15 e k e 12
-23 e l e 22
11 945
1.09-21.25
-7 e h e 8
-11 e k e 11
-37 e l e 38
18 073
index ranges
reflns collected
unique reflns
5062
5743
6804
6872
3588
(Rint ) 0.0745)
0.0488, 0.000
1.0000, 0.8903
5062/9/353
0.0503, 0.0947
0.1434, 0.1215
0.963
(Rint ) 0.0597)
0.0886, 6.2076
0.9987, 0.9923
5743/63/407
0.0789, 0.2150
0.1082, 0.2388
1.104
(Rint ) 0.0444)
0.0825, 1.2263
0.9997, 0.9866
6804/0/391
0.0466, 0.1295
0.0606, 0.1387
1.000
(Rint ) 0.0277)
0.0483, 0.9547
1.0000, 0.8602
6872/1/451
0.0560, 0.1175
0.0880, 0.1270
1.043
(Rint ) 0.0729)
0.0772, 3.2773
0.9966, 0.9668
3588/0/215
0.0725, 0.1943
0.0939, 0.2092
1.210
weighting factors,a a,b
max, min transmission
data/restraints/parameters
R1, wR2 (I > 2σ(I))
R1, wR2 (all data)
GOF
largest diff peak, hole, e Å-3
0.204, -0.367
1.135, -0.736
0.634, -0.423
0.413, -0.296
0.620, -0.459
2
2
2
a w ) [σ2(Fo ) + (aP)2 + (bP)]-1, where P ) (Fo + 2Fc )/3.
For all structures, the intensity data were corrected for absorption
and decay using SADABS.27 Space groups were determined on the
basis of systematic absences and intensity statistics. Direct-methods
solutions provided the positions of most non-hydrogen atoms. Several
full-matrix least-squares/difference Fourier cycles were performed to
locate the remaining non-hydrogen atoms. All calculations were
performed using the SHELXTL-V5.0 suite of programs28 on Pentium
computers. ORTEP diagrams that show the atomic numbering schemes
are shown in the Supporting Information.
Packing analysis parameters were measured using SHELXTL,
Mercury,29 Cerius2 v3.0,30 and Excel. All π-stacking distances were
measured in the following way: least-squares planes were calculated
from the atoms in the backbone of the quinoidal molecules (ter-
thiophenes using S1-3, N1-4, C1-18; quaterthiophene, S1-2, S1A-
2A, N1-2, C1-11, C1A-11A). The perpendicular distance between
adjacent planes was then calculated from the equations of the planes.
In all of the structures, the least-squares planes of the quinoidal
backbone are exactly parallel. The method of analyzing the distortions
of these structures from an ideal cofacial π-stack was adapted from
Curtis et al.15 The pitch and roll distances measure the translations along
the long and short molecular axes, respectively, from an ideal cofacial
stack. Pitch and roll distances were calculated from orthogonal atom
coordinates (SHELX) with the origin set at the geometric centroid of
the backbone atom positions.
interpreted as complete overlap of the molecules in adjacent planes
(pitch distance ) 0, roll distance ) 0). The calculated area overlaps
are relative to the specific molecule, so three and four ring systems
can be compared. For terthiophenes, the length of the rectangle was
defined as the distance between atoms C3 and C16, and the width was
calculated as the distance between the C6/C13 centroid and the C9/
C10 centroid. For the quaterthiophene, the length of the rectangle was
defined as the distance between atoms C3 and C3A, and the width
was calculated as the sum of the distances from the inversion center
(at the center of the molecule), perpendicular to the long molecular
axis, to atoms C6 and C10.
A “molecular volume” overlap was also calculated from the product
of the area overlap and a distance derived from the π-stacking axis.
The overlap in the stacking direction was calculated by subtracting the
π-stacking distance from 4 Å, an arbitrary distance at which the
interaction between adjacent layers was considered zero. An accepted
measure of the packing efficiency31 was also calculated and analyzed
for X-ray structural comparisons. The Kitaigorodskii packing coefficient
is defined by the following equation: CK ) (Z*Vc)/V, where CK is the
Kitaigorodskii packing coefficient, Z is the number of formula units in
the unit cell, Vc is the volume occupied by the atoms in the unit cell,
and V is the volume of the unit cell. The volume occupied by a molecule
was calculated within a van der Waals surface for each molecule using
Cerius2 software.
Additionally, a simple model to approximate the area overlap of
adjacent π-stacking molecules is introduced. Each molecule is treated
as a rectangle with the long molecular axis treated as the long axis of
the rectangle. The parallel spatial overlap of molecules in adjacent
planes is approximated by the area overlap of the molecular rectangles.
The calculated parallel area overlaps range from 0 to 1, where 1 is
Computational Details
As detailed in previous works, we use the semiempirical Hartree-
Fock INDO method (as developed by Zerner and co-workers for
spectroscopic purposes32) to compute the electronic structure of isolated
molecules and of supermolecular systems made of dimers or larger
molecular clusters. The atomic coordinates of isolated molecules and
(27) (a) Sheldrick, G. SADABS V.2.03; 2002. (b) Blessing, R. Acta Crystallogr.
1995, A51, 33-38.
(28) SHELXL V.6.1; Bruker AXS, Madison, WI, 2001.
(29) New software for searching the Cambridge Structural Database and
visualizing crystal structures. Bruno, I. J.; Cole, J. C.; Edgington, P. R.;
Kessler, M. K.; Macrae, C. F.; McCabe, P.; Pearson, J.; Taylor, R. Acta
Crystallogr. 2002, B58, 389-397.
(31) Kitaigorodskii, A. I. Organic Chemical Crystallography; Consultants Bureau
Enterprises: New York, 1961.
(32) (a) Ridley, J.; Zerner, M. C. Theor. Chim. Acta 1973, 32, 111-134. (b)
Zerner, M. C.; Loew, G. H.; Kichner, R. F.; Mueller-Westerhoff, U. T. J.
Am. Chem. Soc. 1980, 102, 589-599.
(30) Cerius2 V3.0 software; Molecular Simulations Inc., San Diego, CA.
9
15300 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004

π
-Stacking Quinodimethane Oligothiophenes
A R T I C L E S
Table 2. Electrochemistry and UV-Vis Data for Quinoidal
Oligothiophenes
reduction
oxidation processes (volts) processes (volts)
compound rings alkyl chains
E1°
E2
°
E1°
(E2°
)
λ
max (nm)a
DCMTb
Q1b
3
3
3
3
4
4
4
2
2
3
3
2
4
4
1.27
1.32
1.27
1.23
0.99
f
1.96c
1.92c
1.86c
1.81c
1.55
f
-0.15d
-0.20d
-0.18d
-0.23d
0.06 (-0.09)
-0.07 (-0.18)
-0.086d
670
648
669
672
779
790g
790
Q2b
Q3b
Q4b
DCMQe
DCMQh
0.83
1.37
a Values from solution in CH2Cl2, unless otherwise indicated. b Scan rate
of 100 mV/s (ambient temp) in TBAPF6 at a glassy carbon working
electrode. c Irreversible, Epa value is given. d Two-electron process. e Scan
rate of 100 mV/s (ambient temp) in TBAClO4 at a platinum working
electrode, ref 18. f Value not reported. g Value in THF solution. h Scan rate
of 10 mV/s (ambient temp) in TBAPF6 at a glassy carbon working electrode,
ref 17.
Figure 2. Cyclic voltammograms of Q1 (solid line), Q2 (dashed line),
and Q3 (solid bold line). Scans are initiated from +0.50 V in the negative
direction.
near 1.9 V (not shown), and a reversible two-electron reduction
process near -0.2 V. The relative number of electrons trans-
ferred in these processes was confirmed by Osteryoung square
wave voltammetry. The first oxidative process generates a stable
cation radical by oxidation of the π-system, while the second
oxidative process generates a less stable dication. The two-
electron reduction process involves reduction of the π-system
to a dianion that is stabilized by the electron-withdrawing
dicyanomethylene groups. Slower or faster scan rates did not
resolve the two-electron process into separate one-electron
reductions. This suggests that the formal potentials for the first
and second reduction processes are inverted so that the one-
electron reduced species undergoes a facile disproportionation
reaction at the electrode surface.
The number and position of hexyl substituents subtly tune
the electrochemical properties of these molecules. In general,
the first oxidation is easier for molecules with three versus two
hexyl groups. If it is assumed that the electron-donating ability
of a butyl group is similar to a hexyl group, then the effect of
substitution on the oxidation potential can be attributed to both
the number and the position of alkyl chains. A comparison of
DCMT with Q1, both with two alkyl groups, indicates that the
location of the alkyl group influences the oxidation. The center-
ring ortho-dialkyl-substituted DCMT oxidizes at slightly lower
potential (1.27 vs 1.32 V for Q1). Alkylation of the center ring
of Q1 to give Q2 lowers the oxidation potential, making it equal
to DCMT. Monoalkylation of all rings in a head-to-tail fashion
(Q3) has the greatest effect and gives the lowest oxidation
potential of all of the terthiophene quinoids studied here.
Changes in the potential of the reduction process are also small
and not merely the reverse of the oxidation trend. In gen-
eral, the position of the alkyl substituent has a rather small effect;
increasing the number of rings from three to four produces a
much larger effect and reinforces the idea that the quinoidal
oligothiophenes are neutral-stabilized oligothiophene dications.
Extension of the π-system to a longer oligothiophene quinoid
system also greatly decreases the redox potentials, particularly
the oxidations. Q4 exhibits two sequential reversible one-
electron oxidation processes at E°′ ) 0.99 and 1.55 V and two
very closely spaced sequential reversible reduction processes
at E°′ ) 0.06 and -0.09 V. The relative number of electrons
transferred in the processes of Q4 was confirmed by Osteryoung
square wave voltammetry. The electrochemical behavior of Q4
is also quite similar to the related molecule DCMQ17,18 that
displays a two-electron reduction under experimental conditions
larger clusters were all taken from the experimental X-ray structures.
No geometry optimizations were performed.
The choice of the INDO/S Hamiltonian is driven by the fact that
INDO calculations have been shown to provide descriptions of the one-
electron structure of isolated and interacting conjugated molecules in
excellent agreement with corresponding experimental data14,33 and
theoretical data obtained at the ab initio level.34
Results and Discussion
Oligomer Synthesis. The precursors to the quinoid products
were prepared by iodination (compounds 1I-4I) or bromination
(2Br) of their respective parent oligothiophenes (compounds
1-4) with N-iodosuccinimide or N-bromosuccinimide in a 1:1
acetic acid/dichloromethane solution.18,25,26 Filtration through
silica gel removed succinimide and gave the halogenated
oligomers in good yields (57-96%). Of special interest is the
synthesis of the HT-regioregular oligomer 3. The syntheses of
the regioregular precursors (A-G) followed a general scheme
previously reported for n-octyl oligothiophene analogues.20
The title compounds Q1-Q4 (Figure 1) were prepared by
procedures previously developed for similar molecules.18,25,26
Palladium-catalyzed cross-coupling of sodium malonate to the
precursors affords a thienyl malononitrile that is deprotonated
in-situ by sodium hydride. The success of the reaction is
indicated by the formation of a dark orange fluorescent solution
of the doubly reduced quinoid dianion intermediate. Oxidation
of the dianion with excess bromine/water afforded the neutral
quinoid forms Q1-Q4. In the case of quaterthiophene Q4,
partial electrophilic substitution of the 4 and 4′′′ sites occurred
to give approximately 11.9% bromination, according to both
the X-ray and the elemental analysis data. The presence of
monobromo- and dibromo-substituted species was observed in
the mass spectrum of Q4. No evidence for bromination of Q1,
Q2, or Q3 was observed.
Electrochemical Properties. Cyclic voltammograms (CVs)
of oligomers Q1, Q2, Q3, and Q4 were measured in 0.1 M
TBAPF6 solutions in CH2Cl2. The electrochemical data of these
compounds as well as the related compounds 3′,4′-dibutyl-5,5′′-
bis(dicyanomethylene)-5,5′′-dihydro-2,2′:5,2′′-terthiophene(DCMT)
and DCMQ are summarized in Table 2. The CVs of Q1, Q2,
and Q3 (Figure 2) each exhibit a reversible one-electron
oxidation process near 1.3 V, a second less-reversible oxidation
(33) Cornil, J.; Calbert, J. P.; Bre´das, J. L. J. Am. Chem. Soc. 2001, 123, 1250-
1251.
(34) Newton, M. D. Int. J. Quantum Chem. 2000, 77, 255-263.
9
J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004 15301

A R T I C L E S
Janzen et al.
Figure 3. ORTEP diagrams of Q1, Q2, Q3, and Q4 (50% ellipsoids). Ellipsoids appear small for Q2, Q3, and Q4 as the structure data were acquired at
100 K. Q1 structure data were acquired at 173 K.
nearly identical17 to those used for Q1-Q4. The sensitivity of
these sequential reduction processes to supporting electrolyte
is also illustrated by DCMQ, which displays two separate one-
electron reductions at a platinum working electrode with
TBAClO4 as the supporting electrolyte. Finally, the two
additional electron-donating alkyl groups of DCMQ shift the
oxidations to less positive potentials and the reductions to more
positive potentials relative to Q4, resulting in a smaller
electrochemical band-gap for DCMQ. In summary, changing
the number of rings, and the position and number of the alkyl
group substituents, is an effective way to tune the redox
properties of these quinoidal oligothiophenes.
Electronic Spectra. Solution electronic spectral data for the
oligomers in this study were measured and are collected in Table
2. All of the oligomers have similar solution electronic spectra
that are dominated by an intense absorption (ꢀ = 100 000 M-1
cm-1) attributed to a π-π* transition of the highly conjugated
quinoidal π-system. The compound Q1 with only two hexyl
groups has a peak (λmax ) 648 nm) that is at slightly higher
energy than the three hexyl group analogues Q2 and Q3 (λmax
) 669 and 672 nm, respectively). In agreement with the
observed electrochemical measurements, the UV-vis spectrum
of the parent compound DCMT with two butyl groups on the
middle ring (λmax ) 670 nm)16 is more like the three hexyl
group-substituted molecules Q2 and Q3 than the end-ring-
substituted Q1. In Q4, the extension of the π-system by an
additional thiophene ring substantially lowers the energy of the
π-π* transition (λmax ) 779 nm). The related DCMQ17,18 (λmax
) 790 nm) electronic spectrum is very similar to that observed
for Q4, but slightly shifted to lower energy as is the electro-
chemical band-gap.
X-ray Structural Details. A brief summary of each indi-
vidual structure will be given followed by an analysis of the
packing and intermolecular forces that are common to all.
ORTEP diagrams are shown in Figure 3.
9
15302 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004

π
-Stacking Quinodimethane Oligothiophenes
A R T I C L E S
Figure 4. Stacking diagrams showing overlap in the direction parallel (left) and perpendicular (right) to the π-stacking for dimer structures Q1 (upper) and
Q3tol (lower).
The structure of Q1 displays a very planar bis-dicyanometh-
yleneterthiophene backbone with a mean deviation from planar-
ity of 0.0241 Å across all 25 atoms. Both of the inter-ring double
bond linkages adopt the transoid geometry. Disorder for one of
the hexyl chains was modeled over two positions with a 50.2/
49.8 ratio. The disordered hexyl group adopts one conformation
(C25′-C30′) that is virtually the same as the ordered hexyl
group (C19-C24), while the other disorder conformation (C25-
C30) has a kink in the hexyl group. The molecules pack (Figure
4) as π-dimers in a layered π-stacking arrangement with two
alternating distances of 3.335 and 3.455 Å.
The backbone of Q2 is quite planar with an average deviation
from a least-squares plane of 0.0494 Å. Both of the inter-ring
double bond linkages adopt the transoid geometry. The hexyl
group substituent on the center ring is disordered over two sites
with an occupancy ratio of 53.7/46.3. The disordered hexyl
group adopts two very similar conformations (C25-C30, C25′-
C30′) that are both essentially in the molecular plane of the
quinoidal terthiophene backbone and the end-ring hexyl groups
(C31-C36, C19-C24). In this case, the molecules pack (Figure
5) in layers with one unique π-stacking distance of 3.393 Å.
The structure of Q3 shows the backbone is rather flat, with
an average deviation from a least-squares plane of 0.0565 Å.
Both of the inter-ring double bond linkages adopt the transoid
geometry. One end-ring hexyl (C19-C24) is flat and lies in
the plane of the backbone. The center-ring hexyl (C25-C30)
and the other end-ring hexyl (C31-C36) are both kinked
substantially out of the molecular plane. Q3 packs (Figures 5,
6) in layers with one unique π-stacking distance of 3.492 Å.
When Q3 is crystallized from toluene solutions, the toluene
solvate structure (Q3tol) results. The structure of Q3tol contains
one molecule of toluene in addition to one Q3 molecule in the
asymmetric unit. This terthiophene backbone is the most planar
of all of the terthiophene structures reported here with a mean
deviation from a least-squares plane of 0.0235 Å. Both of the
inter-ring double bond linkages adopt the transoid geometry.
Two hexyl groups are located in the plane of the backbone,
while one (C31-C36) is kinked out of the plane. The toluene
molecule π-stacks nicely with Q3, forming a layered structure
similar to the unsolvated structure. The toluene molecule is
almost coplanar with the Q3 backbone (1.4° between least-
squares planes formed by Q3 and toluene); however, Q3tol
displays two unique Q3-Q3 distances (3.404, 3.465 Å) as well
as two unique Q3-toluene distances (Figure 4).
An inversion center lies at the middle of the bond connecting
the second and third thiophene rings of Q4; thus the asymmetric
unit is one-half of one molecule and all three inter-ring double
bond linkages adopt the transoid geometry. Q4 is very flat with
a mean deviation from a least-squares plane containing 30 atoms
(all non-hydrogen atoms except bromine and the hexyl groups)
of 0.0404 Å. The hexyl groups are disordered over two similar
positions with an occupancy ratio of 51.8/48.2. The bulk sample
from which these crystals were grown contains bromine
(elemental analysis, mass spectroscopy), and electron density
maps suggested that a partially occupied, symmetry related
bromine atom is present in the 4 and 4′′′ positions of the outer
thiophene rings. The bromine atom was allowed to refine as a
free variable, and the refined occupancy of 11.9% agrees well
9
J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004 15303

A R T I C L E S
Janzen et al.
Figure 5. Stacking diagrams showing overlap in the direction parallel (left) and perpendicular (right) to the π-stacking for single-distance structures Q2
(upper), Q3 (middle), and Q4 (lower).
with the combustion analysis. The remaining 88.1% occupancy
was refined as a hydrogen atom in an ideal position for bonding
with a thiophene carbon. Q4 packs (Figure 6) in a herringbone
structure with one unique π-stacking distance of 3.358 Å.
Structural Comparisons. Perhaps the simplest way to
compare crystalline solids is to measure the packing efficiency.31
The Kitaigorodskii packing coefficient (CK) is calculated by
dividing the occupied molecular volume (calculated from a van
der Waals surface) by the total volume of the unit cell (calculated
from the unit cell parameters). The packing coefficient does
not distinguish between the types of interactions present but is
a general measure of the quality of all of the intermolecular
interactions present in a crystal lattice. The packing coefficients
of Q1-Q4 and other molecules of interest are compared in
Table 3. All molecules in the table pack with relatively good
efficiency. Aromatic molecules typically pack with coefficients
in the range of 0.6-0.8.31 The tightly packed structure of
graphite has a packing coefficient of 0.887.31 Although the three
single π-distance quinoid structures show a correlation between
the π-stacking distance and packing coefficient, the inclusion
of BDT (3.530 Å, CK ) 0.765) and C5PTCDI (3.405 Å, CK )
0.704)35 indicates that this simple correlation only holds for
molecules with very similar intermolecular interactions. An
anisotropic analysis of the intermolecular features divides the
forces into those along the stacking axis (π-π interactions) and
roughly perpendicular to the stacking axis (nonclassical hydrogen-
bonding interactions). Our discussion of these structures will
include the only other reported quinoidal oligothiophene
structure with three or more thiophene rings, DCMT.10,16
The Q1-Q4 structures reported here for the first time exhibit
packing arrangements with extensive π-stacking (Figures 4-6).
Structures Q1, Q2, Q3, and Q3tol all pack in π-stacked layers,
but Q4 (Figure 6) and DCMT pack as π-stacks arranged in the
γ motif, a layered herringbone structure, vide infra. At this point,
it is important to make the distinction between the herringbone
structure of pentacene and that of DCMT and Q4. Desiraju
(35) Kitaigorodskii packing coefficients calculated using the Cerius2 software
and the CIF files for BDT (CCDC data code NOBJET) and C5PTCDI
(CCDC data code DICMUX).
9
15304 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004

π
-Stacking Quinodimethane Oligothiophenes
A R T I C L E S
Figure 6. Packing diagrams showing the layered packing of Q3 (upper) and the layered herringbone packing of Q4 (lower).
Table 3. Comparison of Selected Structural Data and Calculations
dimer motifs (Figure 4) with two unique stacking distances. The
molecules in the stack alternate in an antiparallel fashion with
analogous thiophene rings pointing in the direction opposite to
that of molecules in adjacent layers. The antiparallel stacking
seems to be intimately related to the formation of dimers in
these structures. An additional difference between the dimer
structures and those with only one stacking distance is that the
dimer structures exhibit an oscillation of the pitch angle between
adjacent layers (pitch angle alternates between positive and
negative from the perpendicular stacking vector). This sawtooth
pattern is different from the stair-step pattern of the single-
distance stacking structures, in which the unique pitch angle
displaces the stack the same direction along the length of the
molecule.
HOMO
splittingb
(meV)
LUMO
splittingb
(meV)
π
-stacking
area
volume
a
compound
distances (Å)
overlap (%)
overlap (Å3)
CK
Q1
3.335
3.455
3.393
3.492
3.404
3.465
3.358
3.456
3.650
13.5
71.0
56.0
44.8
4.8
26.5
65.5
73.3
37.5
3.28
14.1
13.0
8.78
1.10
5.50
21.3
16.0
5.3
0.709
c
c
204
147
c
c
c
353
219
c
Q2
Q3
Q3tol
0.696
0.682
0.682
Q4
DCMT
0.724
0.685
317
c
c
417
c
c
a Kitaigorodskii packing coefficient. b Calculated from the crystal-
lographic atom positions for a dimer. c Not calculated.
The extreme flatness enforced by the quinoidal bonding motif
common to all of the molecules allows for very short π-π
stacking distances and significant π-overlap of adjacent mol-
ecules within a stack. The closest observed stacking distances
in these molecules range from 3.335 Å in Q1 to 3.492 Å in
Q3. The observed stacking distances in these structures are
almost all shorter than the previously determined DCMT
structure (3.456 and 3.650 Å). The stacking distances (shown
in Table 3) are very short relative to the few nonquinoidal
thiophene structures with π-stacking. For example, NC-3T-CN
(2,2′:5′,2′′-terthiophene-5,5′′-dicarbonitrile)37 stacks with π-π
distances of 3.65-3.70 Å; Me2t-BuSi-3T-Me2t-BuSi (5,5′′-bis-
(tert-butyldimethylsilyl)-2,2′:5′,2′′-terthiophene)38 stacks with a
and Gavezotti36 define four distinct crystal-packing arrangements
for aromatic hydrocarbons. Pentacene falls into the herringbone
motif where nearest neighbors are not parallel. The â motif with
a layered structure of parallel nearest neighbors describes the
packing of Q1, Q2, Q3, and Q3tol. Q4 and DCMT exhibit the
γ motif, which is a type of flattened out herringbone structure
with parallel nearest neighbors but with more than one orienta-
tion of layers. (An additional aromatic hydrocarbon motif
previously described (sandwich motif) is not observed in any
of the structures to be discussed here.) In the structures of Q2,
Q3, and Q4, a single unique π-stacking distance is observed,
and the molecules in adjacent layers (Figure 5) pack in a parallel
arrangement with the corresponding thiophene rings pointing
in the same direction as the layers above and below a given
molecule. The structures of Q3tol, Q1, and DCMT exhibit
(37) Barclay, T. M.; Cordes, A. W.; MacKinnon, C. D.; Oakley, R. T. Reed, R.
W. Chem. Mater. 1997, 9, 981-990.
(38) Barbarella, G.; Ostoja, P.; Maccagnani, P.; Pudova, O.; Antolini, L.;
Casarini, D.; Bongini, A. Chem. Mater. 1998, 10, 3683-3689.
(36) Desiraju, G. R.; Gavezotti, A. Acta Crystallogr. 1989, B45, 473-482.
9
J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004 15305

A R T I C L E S
Janzen et al.
π-π distance of 3.98 Å, and NO2-(3TBu2)-NO2 (3′,4′-di-n-
butyl-5,5′′-dinitro-2,2′:5′,2′′-terthiophene)16 stacks as a dimer
with one short distance (3.556 Å) and a longer distance (3.573
Å). Recently, tricyanovinyl-substituted terthiophenes39,40 have
also been shown to π-stack with distances of 3.535 and 3.528
Å for π-dimers in TCV-terthiophene (5-(tricyanovinyl)-2,2′:
5′,2′′-terthiophene)39 and 3.552 and 3.544 Å for π-dimers in
TCV-(3TBu2)-TCV (3′,4′-di-n-butyl-5,5′′-di(tricyanovinyl)-2,2′:
5′,2′′-terthiophene).40 A related heterosoric charge-transfer
complex bithiophene-TCNQ (2,2′-bithiophene tetracyanoquin-
odimethane)41 has a unique stacking distance of about 3.37 Å
with an almost parallel stacking motif (1.2° between adjacent
planes). The low-temperature crystal structure of [Ph-(3TBu2)-
Ph][PF6] (3,4′-di-n-butyl-5,5′′-diphenyl-2,2′:5′,2′′-terthiophene
hexafluorophosphate)19 has short π-stacking distances of 3.358
and 3.433 Å. Finally, for comparison, the interplanar spacing
in graphite is 3.35 Å,42 roughly equal to the sum of the van der
Waals thickness of adjacent layers.
These comparisons suggest that the short π-stacking distances
observed in these quinoidal thiophenes are more similar to those
found in charge-transfer complexes or the oxidized terthiophene
structures than to nonquinoidal terthiophene structures.1,15 The
donor-acceptor roles are predetermined in CT complexation
of two component systems. In the oxidized terthiophene
structures, the strong interactions between adjacent molecules
occur through π-dimerization of odd electron radical species.
In the structures Q1-Q4, the low reduction and oxidation
potentials of these molecules allow each molecule to simulta-
neously assume donor and acceptor properties. Additionally,
in the structure of the Q3 toluene solvate (Q3tol), two-
component donor-acceptor interactions are evident as toluene
functions as an additional donor in the π-stacking of Q3
molecules.
An interesting question is whether the π-stacking distance
correlates with the extent of π-overlap with adjacent molecules
or if the other packing forces determine the distance and hence
the overlap. Combining the pitch and roll distances with a
rectangular approximation of the molecular plane and π-system
of each molecule allows the projected area overlaps of molecules
in adjacent planes to be measured (Table 3). Inspection of the
area overlap versus π-stacking distance plot (Figure 7) reveals
a linear correlation between overlap and distance for Q4, Q2,
and Q3. However, for the dimerized structures, the better area
overlap does not always correlate with shorter π-π distances.
In the cases of Q1 and Q3tol, the longer inter-dimer stacking
distance has better parallel area overlap, while DCMT shows
the opposite behavior. In the three structures with only one
π-stacking distance, the shorter distance correlates with a larger
area overlap (Figures 5, 7), and we conclude that the π-area
overlap is correlated with the π-stacking distance in a simple
way for only the single distance structures.
Figure 7. Area overlap versus π-stacking distance comparison. The solid
line indicates the three molecules with one unique stacking distance.
between the π-systems of adjacent molecules is a better
comparison because it takes into account not only the slip of
the π-system, but also the stacking distance. The geometric
volume overlap considered here is the difference between 4 Å
(an arbitrary distance chosen for zero π-π interaction) minus
the π-stacking distance between adjacent molecules multiplied
by the area overlap. Calculated π-volume overlaps are shown
in Table 3. The DCMT compound that we have already studied
and shown to have relatively good electron mobility and a
significant hole mobility is typical of the series. Of importance
is the significantly larger value of this parameter for Q4, which
clearly has the best volume π-overlap in the series. Curiously,
pentacene has a π-overlap volume of zero, because there is no
parallel area overlap of the π-system in the structure. Although
these area and volume overlap comparisons are based on only
geometric considerations and do not take into account atomic
overlaps between carbon p-orbitals, we believe they will have
predictive value because they allow direct structural comparisons
of various polythiophene oligomers. In a later section of this
paper, we will show the correlation of the area and volume
overlap parameter with the more rigorous bandwidth calculations
based on molecular orbital calculations.
As the overall packing structure of these molecules is
determined by maximizing the contributions of the most
important molecular interactions, the intermolecular interactions
in the directions roughly perpendicular to the stacking axes have
been investigated. Nitriles have been shown to exert supramo-
lecular organization in compounds with halogens, chalcogens,
and C-H groups including thiophenes.37,43 Pairwise interactions
between nitrile groups and polarizable thiophene hydrogens
appear frequently in structures of Q1-Q4. We suggest that these
interactions are the important determiners of the geometry within
the layers, which, in turn, adjust interlayer steric interactions.
Thus, the π-stacking distance and the extent of slippage in the
pitch and roll angles are both affected, but not in a simple way.
A summary of the geometry and distances of these nitrile-
thiophene hydrogen interactions is given in Table 4. In Q1, each
molecule has two unique pairwise nitrile-thiophene hydrogen
interactions (N1-H5A, N4-H10A), utilizing an end-ring
hydrogen in one interaction and a center-ring hydrogen in the
A second way to compare the extent and magnitude of
intermolecular π-interactions along the stacking axis is to take
into account a “π-volume overlap”. A π-volume overlap
(39) Bader, M. M.; Custelcean, R.; Ward, M. D. Chem. Mater. 2003, 15, 616-
618.
(40) Pappenfus, T. M.; Burand, M. W.; Janzen, D. E.; Mann, K. R. Org. Lett.
2003, 5, 1535-1538.
(43) (a) Reddy, D. S.; Ovchinnikov, Y. E.; Shishkin, O. V.; Struchkov, Y. T.;
Desiraju, G. R. J. Am. Chem. Soc. 1996, 118, 4085-4089. (b) Cordes, A.
W.; Haddon, R. C.; Hicks, R. G.; Oakley, R. T.; Palstra, T. T. M. Inorg.
Chem. 1992, 31, 1802-1808. (c) Cordes, A. W.; Haddon, R. C.; Hicks, R.
G.; Kennepohl, D. K.; Oakley, R. T.; Palstra, T. T. M.; Schneemeyer, L.
F.; Scott, S. R.; Waszczak, J. V. Chem. Mater. 1993, 5, 820-825.
(41) Minxie, Q.; Heng, F.; Yong, C. Jiegou Huaxue (Chinese J. Struct. Chem.)
1986, 159.
(42) Cotton, F. A.; Wilkinson, G. AdVanced Inorganic Chemistry, 5th ed.;
Wiley: New York, 1988.
9
15306 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004

π
-Stacking Quinodimethane Oligothiophenes
A R T I C L E S
Figure 8. Intermolecular thiophene hydrogen-nitrile interactions in the structure of Q1.
Table 4. Summary of Distances and Geometry of
correctly predicts the order of the electronic coupling in the
π-stacks for the single distance structures considered here.
Detailed MO calculations were performed to understand the
intermolecular interactions in the solid-state structures of the
quinoidal oligothiophenes in this study. The calculations also
provide a means of comparing existing calculations of molecular
structures for which TFT devices have been constructed.
Calculations were performed only on the structures with one
unique π-stacking distance as the dimer structures represent a
different type of intermolecular interaction between adjacent
molecules. The methodology used to explore nearest-neighbor
interactions involved the calculation of the HOMO-LUMO
band splittings of a molecule and each neighbor in the
experimentally determined X-ray structure to a distance where
interactions were negligible. One way to quantify the electronic
coupling between adjacent molecules has been defined as the
transfer integral (t), which is approximately one-half the HOMO/
HOMO and LUMO/LUMO splittings for a dimer and is
different for the HOMO and the LUMO. Bandwidths for both
the HOMO and the LUMO in an extended one-dimensional
stack are 4t according to the tight binding approximation.45
For the case of Q4, the only significant interactions calculated
were between adjacent molecules within a π-stack. Using the
tight binding approximation, we can approximate the bandwidths
of an infinite stack for the HOMO as 633 meV and the LUMO
as 834 meV. The evolution of the bandwidth as a function of
the number of molecular units in a π-stack linearly correlates
with the function cos(π/(N + 1)) (where N ) number of
interacting oligomers in the cluster), which validates the use of
the tight binding approximation (Figure 9). For comparison,
similar calculations for bisdithienothiophene (BDT),46 which
packs in a γ motif with a π-stacking distance of 3.557 Å but
exhibits a large roll distortion which causes poor area overlap,
have given HOMO and LUMO bandwidths of 688 and 108
meV, respectively. BDT only has significant HOMO-LUMO
splittings along the π-stacking direction. Bandwidth calculations
of pentacene,46 which packs in a herringbone motif with no
Nitrile-Thiophene Hydrogen Interactions
donor acceptor distance distance
− −N angle CN−H angle
VDWa CH
compound atom
atom
(Å)
(Å)
(deg)
(deg)
Q1
Q1
Q2
Q3
Q3tol
Q4
N1 H5A
N4 H10A
N3 H14A
N1 H5A
N2 H9A
N2 H10A
2.538
2.539
2.541
2.866
2.716
2.597
-0.212
-0.211
-0.209
0.116
-0.034
-0.153
169.2
139.7
175.5
173.2
169.0
148.0
163.9
157.3
169.2
168.5
157.6
115.8
a VDW is the sum of the van der Waals radii of the donor and acceptor
atoms.
other. Figure 8 shows the nitrile-thiophene hydrogen interac-
tions of Q1. Q2 only has one nitrile end-ring thiophene hydrogen
interaction (N3-H14A). Q3 has a similar interaction at a longer
distance (N1-H5A). Q3tol has a middle-ring hydrogen interac-
tion with a nitrile (N2-H9A). Q4 has an inner-ring hydrogen
interaction (N2-H10A). In DCMT, interactions between nitrile
groups and thiophene hydrogens are present, but they occur
between molecules in π-stacks that are not parallel.
Although the nitrile-thiophene hydrogen interactions are
important intermolecular interactions for packing determination,
additional intermolecular interactions evidenced by close con-
tacts were discovered. Only those contacts with reasonable
geometry considerations are noted here. Close parallel nitriles
are present in a plane of the structure of Q2 (C1-N1) and are
only 3.173 Å apart. A close S1-H30D approach of 2.817 Å is
present in the structure of Q1. A close S2-N2 distance of 3.304
Å is found in the structure of Q4. In all structures, there are no
S-thiophene hydrogen interactions. Numerous close contacts
also exist between alkyl groups in each structure. While
association of the alkyl chains occurs in all of these structures,
these interactions are considered to be weak in comparison with
the forces driving π-stacking and the nonclassical hydrogen
bonding of the nitriles.
Bandwidth Calculations. Parallel area overlap calculations
approximate the interaction between adjacent molecules but do
not capture the shape of the molecular orbitals;44 however, MO
calculations that are described below show that the area overlap
(45) Bre´das, J.-L.; Calbert, J. P.; da Silva Filho, D. A.; Cornil, J. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 5804-5809.
(46) Cornil, J.; Calbert, J. P.; Beljonne, D.; Silbey, R.; Bre´das, J.-L. AdV. Mater.
2000, 12, 978-983.
(44) Kazmaier, P. M.; Hoffman, R. J. Am. Chem. Soc. 1994, 116, 9684-9691.
9
J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004 15307

A R T I C L E S
Janzen et al.
solid state; the values are summarized in Table 3. The largest
HOMO-LUMO splittings are observed for Q4, with Q2 as next
largest. It is interesting to note that for this series the stacking
distances, area overlaps, and calculated band splittings are
consistent with one another.
Conclusions
The synthesis and characterization of quinodimethane-
substituted oligomers are reported. The electrochemistry of these
materials indicates their suitability for both p and n biasing in
FETs. The structural features of these materials are favorable
for efficient transport. Together, these features indicate DCMT
derivatives are good candidates for ambipolar semiconductor
materials. Preliminary evidence supports this conclusion.11 Close
π-stacking, large area overlap, and significant intermolecular
electronic coupling, as calculations indicate, correlate for
structures with a unique π-stacking distance and parallel cofacial
arrangement. Calculations suggest that these compounds could
exhibit hole and electron mobilities in devices that could rival
those of pentacene if the proper thin film morphology and other
important device parameters are also favorable.
Figure 9. Calculated bandwidth evolution of Q4. Cluster size increases to
the right.
obvious π-stacking, but with significant edge-to-face interac-
tions, indicate significant transfer integrals in two different
directions, yielding two-dimensional electronic coupling char-
acter. The total HOMO and LUMO bandwidths calculated are
608 and 588 meV, respectively. Although the Q4 calculation
shows distinctly one-dimensional charge transport character
(along the π-stack), the HOMO bandwidth calculated for Q4
is similar to that of pentacene and the LUMO bandwidth is
considerably larger.
In the case of Q2 and Q3, the strongest pairwise interactions
are again between adjacent molecules within a π-stack, but
significant interactions were also found beyond nearest neigh-
bors within a π-stack and for some of the nearest neighbors in
adjacent π-stacks. Particularly in the case of Q3, the parallel
stacking of molecules in a given π-stack causes the alignment
of the large (6.5 D, according to an INDO calculation) molecular
dipole moments for all molecules in the stack. These longer-
range dipole-dipole interactions cause a breakdown of the
nearest-neighbor approximation. When a dimer is extracted from
the crystal structure to compute the orbital splittings, the two
molecules in the dimer will no longer be equivalent. In this
case, the HOMO-LUMO splittings do not simply correspond
to twice the transfer integral;33,47 as a result, we do not provide
bandwidths for Q2 and Q3.
Acknowledgment. This work was supported primarily by the
MRSEC Program of the National Science Foundation under
Award Number DMR-0212302. P.C.E. would like to acknowl-
edge Professor Larry Miller for the generous donation of his
lab space and many helpful discussions. D.E.J. would like to
acknowledge the University of Minnesota Department of
Chemistry for financial support through a Block Grant Fellow-
ship as well as the University of Minnesota Graduate School
for support through a Doctoral Dissertation Research Grant. We
would like to acknowledge the X-ray crystallographic laboratory
in the University of Minnesota Department of Chemistry.
ChemMatCARS Sector 15 is principally supported by the
National Science Foundation/Department of Energy under grant
number CHE0087817 and by the Illinois board of higher
education. The Advanced Photon Source is supported by the
U.S. Department of Energy, Basic Energy Sciences, Office of
Science, under Contract No. W-31-109-Eng-38.
The HOMO-LUMO splittings (2t) calculated for dimers in
the Q2, Q3, and Q4 structures allow a comparison of the
intermolecular interaction between adjacent molecules in the
Supporting Information Available: ORTEP diagrams and
crystallographic data in CIF format. This material is available
free of charge via the Internet at http://pubs.acs.org.
(47) Senthilkumar, K.; Grozema, F. C.; Bickelhaupt, F. M.; Siebbeles, L. D. A.
J. Chem. Phys. 2003, 119, 9809-9817.
JA0484597
9
15308 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004