Synthesis, experimental and theoretical characterization, and field-effect transistor properties of a new class of dibenzothiophene derivatives: From linear to cyclic architectures

Article abstract of DOI:10.1039/c1jm13962b

We report the synthesis and characterization of a bis-dibenzothiophene cyclic dimer containing bis-ethylene linkages (DBT-CM) and of the corresponding mono-ethylene-linked 'linear' cis- and trans-isomers (cis- and trans-DBT-LM, respectively). The varied molecular architectures lead to notable differences both in terms of the solid-state packing and the molecular electronic and optical properties. X-ray crystallography reveals that the cyclic architecture of DBT-CM leads to a more densely packed stacking configuration that imparts stronger intermolecular electronic coupling for both hole and electron transport amongst adjacent molecules, while characterization of the thin-film morphology and crystallinity uncovers important temperature-dependent properties of the films as a function of the molecular architecture. Moreover, the redox, electronic structure, and optical properties of DBT-CM vary distinctly from those of its linear counterparts. The intramolecular reorganization energies for hole and electron transport for DBT-CM are markedly smaller than the linear counterparts, while the dispersion for the highest valence band (and the intermolecular electronic coupling for hole transport) is the largest for the series. The more favorable molecular packing/morphology characteristics and charge-transport properties (within the Marcus framework) of DBT-CM manifest themselves in thin-film field-effect transistor studies, where a field-effect hole-carrier mobility 0.026 cm² V^-1 s^-1 is measured, a value one-order-of-magnitude larger than either linear analog. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c1jm13962b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2012, 22, 1313
www.rsc.org/materials
PAPER
Synthesis, experimental and theoretical characterization, and field-effect
transistor properties of a new class of dibenzothiophene derivatives: From
linear to cyclic architectures†
a
a
b
b
bc
a
a
Yali Qiao, Zhongming Wei, Chad Risko, Hong Li, Jean-Luc Br ꢀe das,* Wei Xu* and Daoben Zhu*
Received 15th August 2011, Accepted 21st October 2011
DOI: 10.1039/c1jm13962b
We report the synthesis and characterization of a bis-dibenzothiophene cyclic dimer containing bis-
ethylene linkages (DBT-CM) and of the corresponding mono-ethylene-linked ‘linear’ cis- and trans-
isomers (cis- and trans-DBT-LM, respectively). The varied molecular architectures lead to notable
differences both in terms of the solid-state packing and the molecular electronic and optical properties.
X-ray crystallography reveals that the cyclic architecture of DBT-CM leads to a more densely packed
stacking configuration that imparts stronger intermolecular electronic coupling for both hole and
electron transport amongst adjacent molecules, while characterization of the thin-film morphology and
crystallinity uncovers important temperature-dependent properties of the films as a function of the
molecular architecture. Moreover, the redox, electronic structure, and optical properties of DBT-CM
vary distinctly from those of its linear counterparts. The intramolecular reorganization energies for
hole and electron transport for DBT-CM are markedly smaller than the linear counterparts, while the
dispersion for the highest valence band (and the intermolecular electronic coupling for hole transport)
is the largest for the series. The more favorable molecular packing/morphology characteristics and
charge-transport properties (within the Marcus framework) of DBT-CM manifest themselves in thin-
2
film field-effect transistor studies, where a field-effect hole-carrier mobility 0.026 cm V
ꢀ1 ꢀ1
s
is
measured, a value one-order-of-magnitude larger than either linear analog.
Introduction
charge-carrier mobilities that are comparable to those of tradi-
2
ꢀ1 ꢀ1
tional amorphous silicon-based materials, around 1 cm V s .
Progress in terms of molecular design has led to a number of
1
First reported in 1986, organic field-effect transistors (OFETs)
2–7
still garner extensive academic and industrial attention as they
offer a variety of advantages versus conventional inorganic
silicon-based transistors – e.g., relatively low-cost fabrication on
large-area and flexible substrates – and show potential for use in
a number of electronics applications, including radio frequency
organic electronic materials possessing charge-carrier mobilities
13–16
greater than that value being demonstrated.
In spite of this
progress, however, there remains a limited understanding of the
relationship between molecular structure and electrical transport
properties. Consequently, the design of new classes of organic
electronic materials for OFET applications remains a challenging
task.
8–12
identification (RFID) tags, smart cards, and flexible displays.
In order to fully take advantage of their inherent attributes in
OFET applications, organic electronic materials should have
Macrocyclic molecules with fully extended p-conjugation have
been extensively studied in molecular materials chemistry,
17–27
a
Beijing National Laboratory for Molecular Sciences, Key Laboratory of
supramolecular chemistry, and nanotechnology applications.
Organic Solids, Institute of Chemistry, Chinese Academy of Sciences,
Beijing, 100080, P. R. China. E-mail: wxu@iccas.ac.cn; zhudb@iccas.ac.
cn; Fax: +86 10 62569349; Tel: +86 10 62639355
Compared to linear oligomers and polymers, cyclic p-conjugated
compounds possess interesting characteristics that are princi-
pally attributed to their specific architecture, which effectively
combines the well-defined structure of molecules/oligomers with
the absence of end-effect perturbations (typical of the conjugated
b
School of Chemistry and Biochemistry and Center for Organic Photonics
and Electronics, Georgia Institute of Technology, Atlanta, Georgia, 30332-
0
400, USA
c
Department of Chemistry, King Abdulaziz University, Jeddah, 21589,
Saudi Arabia
28
pathways of polymers). Studies focused on the use of cyclic
conjugated compounds in OFETs have been rather limited,
though, as devices made from these materials have rather modest
†
(
Electronic supplementary information (ESI) available: TGA data
Fig. S1), theoretical caculation data (Fig. S2, Table S1, S2, S3 and
29,30
S4), AFM images for linear isomers (Fig. S3), NMR data (Fig. S4, S5,
S6, S7 and S8). CCDC reference numbers 838083–838085. For ESI and
crystallographic data in CIF or other electronic format see DOI:
limited charge-carrier mobilities.
Recent investigations, however, reveal the potential for
molecular systems based on cyclic architectures as promising
10.1039/c1jm13962b
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 1313–1325 | 1313

View Article Online
3
1–35
materials for OFETs.
From a purely performance stand-
on vacuum-deposited films were successfully fabricated, and the
results are discussed in terms of the thin-film morphology and
crystallinity. OFETs based on cyclic DBT-CM display the largest
point, OFETs based on Langmuir–Blodgett films of cyclo[8]
ꢀ1 ꢀ1
2
pyrrole have large charge-carrier mobilities of 0.68 cm V s ,
31
2
ꢀ1 ꢀ1
among the largest reported for cyclic p-conjugated compounds.
Importantly, these expanded studies also allow for an improved
hole mobility of 0.026 cm V s , a value considerably larger
than those obtained for the two linear isomers. The results
suggest that cyclic DBT derivatives are interesting candidates
that warrant further scrutiny for OFET applications.
understanding of the relationship between molecular structure
33,35
and the charge-transport properties of cyclic architectures.
For instance, materials based on cyclic triphenylamine dimers
tested in OFET architectures have charge mobilities as large as
2
.015 cm V
ꢀ1 ꢀ1
Results and discussion
0
s
– among the best results for OFETs based on
any triarylamine derivative and 100 times larger than that of the
Synthesis and thermal characterization
33
corresponding linear structure under the same conditions. The
nature of the cyclic architecture provides both for favorable,
highly-ordered solid-state packing and a significant reduction in
the intramolecular reorganization energy, both of which are
major factors that facilitate efficient charge-carrier transport.
Dibenzothiophene (DBT) is a commercially available chalco-
genophene compound with a planar p-conjugated structure,
a feature desirable for inducing strong p–p intermolecular
interactions. Additionally, the presence of sulfur in the conju-
gated backbone offers the potential for short S/S contacts
that could provide additional channels for charge-carrier
The synthetic routes of the cyclic and linear dimers are presented
in Scheme 1. For the cyclic compound DBT-CM, the synthetic
33–35
strategy was the same as previously reported.
Commercially
available DBT was used as the starting material. DBT was
brominated, followed by lithiation and reaction with DMF to
afford 8-diformylibenzothiophene 2 in good yield. The cyclic
macromolecule was then successfully obtained with the
McMurry coupling reaction in much higher yields (16%) than
previous attempts, which is mainly attributed to the use of 1,4-
40
dioxane instead of THF as the reaction solvent. For the linear
dimers, the Wittig reaction was employed by using the precursors
necessary to obtain both the cis- and trans-ethylene linear
36,37
transport – e.g., various thiophene oligomers
and tetrathia-
38
fulvalene derivatives have lateral S/S interactions in the solid
state that result in interstack electronic coupling (in addition to
intrastack electronic coupling via p–p interactions), leading to
two-dimensional charge-carrier pathways. If such multiple
transport channels could be triggered in DBT-containing
systems, it is anticipated that molecular structures with efficient
charge-transport performance could be designed.
42
configurations. The compound cis-DBT-LM was initiated from
dibenzothiophene, following a similar synthetic procedure to
43
that mentioned above to produce 2-formyldibenzothiophene 4.
Subsequently, compound 4 was reduced and reacted with tri-
phenylphosphine hydrobromide to afford (dibenzothiophene-2-
44
methyl)triphenylphosphonium bromide 6.
Finally, the
Linear oligomers based on DBT are known, and OFETs
ethylene-bridged dimer cis-DBT-LM was successfully synthe-
sized using the typical Wittig reaction in 39% yield. Other
synthetic methods, for example, the McMurry reaction, were
also used. However, difficulties with separation and purification
39
fabricated with these materials show promising performance;
however, macrocyclic DBT molecules have yet to be explored.
Here, based on previous investigations on the synthesis of tri-
phenylamine and carbazole dimers, we report the synthesis of
a cyclic dibenzothiophene dimer bridged by two ethylene moie-
ties (DBT-CM in Fig. 1). Both cis- and trans- mono-ethylene-
linked ‘linear’ isomers (cis- and trans-DBT-LM in Fig. 1) were
also synthesized for direct comparison with the cyclic dimer. The
thermal, electrochemical, and optical properties were fully
characterized, and single-crystal X-ray diffraction structures for
all the three compounds were determined. OFET devices based
Fig. 1 Chemical structures of the cyclic and linear molecules: (Z, Z)-
cyclo[2](dibenzo[b,d]thiophen-2, 8-yl-ethylene) (DBT-CM), (Z)-1, 2-bis-
(dibenzo[b,d]thiophen-2-yl)ethane (cis-DBT-LM), (E)-1, 2-bis(dibenzo
[
b,d]thiophen-2-yl)ethene (trans-DBT-LM).
Scheme 1 Synthetic routes of the cyclic and linear molecules.
This journal is ª The Royal Society of Chemistry 2012
1
314 | J. Mater. Chem., 2012, 22, 1313–1325

View Article Online
made the overall yields very low. Typical Wittig reaction
conditions produced better results. In order to obtain the
compound trans-DBT-LM, the precursor diethyl dibenzothio-
phene-2-methylphosphonate 8 was synthesized from compound
6
through bromination and subsequent reaction with triethyl
phosphate. Finally, the trans-isomer was synthesized with satis-
factory yields (38%) using the improved Wittig reaction.
Notably, for the synthesis of the two linear isomers, both cis- and
trans-isomers are formed in each condensation reaction with
different proportions, but only one, which is the most abundant,
could be separated. Compounds DBT-CM and cis-DBT-LM
were soluble in common organic solvents, such as dichloro-
ꢀ1
methane and toluene, with a high solubility (>10 mg mL ), but
trans-DBT-LM had very low solubility in organic solvents (<1
ꢀ1
mg mL in chloroform or toluene at room temperature). All
compounds were characterized by HRMS, NMR spectroscopy,
and elemental analysis. Compound trans-DBT-LM could not be
13
characterized by C NMR spectroscopic analysis due to its poor
solubility.
The thermal stability of the cyclic and linear isomers was
investigated by thermal gravimetric analysis (TGA, see Fig. S1).
The TGAs reveal that the onset thermal decomposition
Fig. 2 X-ray crystal structure of compound DBT-CM: (a) Molecular
structure with 50% probability ellipsoids. (b) Herringbone pattern viewed
along the a-axis. (c) S/S short contacts viewed along the b-axis.
temperatures
– determined from the intersection of the
lines tangent to the isothermal mass-loss slope and initial base
ꢁ
ꢁ
line – are approximately 275 C for DBT-CM, 290 C for the cis-
ꢁ
structure analysis could be obtained only by very slow evapo-
isomer, and 345 C for the trans-isomer under nitrogen atmo-
ration of an extremely dilute CH
2
Cl
2
solution. The trans-isomer
/c space
sphere. These results indicate that the DBT-containing systems
crystallizes in a monoclinic system and belongs to the P 2
1
have good thermal stabilities.
ꢁ
group with unit-cell parameters of a ¼ 6.792(15) A, b ¼ 10.677(2)
ꢁ
ꢁ
ꢁ
A, c ¼ 12.729(3) A, b ¼ 97.719(4) , see Fig. 4. In the packing
diagram of trans-DBT-LM depicted in Fig. 4b, the molecules
stack in columns along the a-axis in a slipped face-to-face
manner. Close intermolecular packing is observed along the
Single crystal and molecular structures
Pale yellow, rod-like crystals of DBT-CM suitable for single-
crystal X-ray diffraction structure analysis were grown by slow
ꢁ
column as the interplanar distance is only 3.30 A (Fig. 4c).
evaporation of the CH
cyclic molecule crystallizes in a monoclinic structure and belongs
to the P 2
2 2
Cl solution at room temperature. The
Furthermore, intermolecular S/H–C hydrogen bonds link these
columns into a three-dimensional network.
1
/c space group with unit-cell parameters: a ¼ 12.339(3)
Regardless of the molecular architecture, each molecular
structure has a two-fold rotational axis, with the individual
ꢁ
ꢁ ꢁ
ꢁ
A, b ¼ 4.748(9) A, c ¼ 16.949(3) A, b ¼ 110.06(3) , see Fig. 2.
Slipped face-to-face stacking columns along the b-axis are
ꢁ
observed with an interplanar distance of 3.52 A between the
thiophene rings of the stacked DBT units (Fig. 2b). There do
ꢁ
exist relatively short S/S contacts with a distance of 3.48 A
between neighboring columns (Fig. 2c), a distance smaller than
ꢁ
twice the van der Waals radius of S atoms (3.70 A). The columns
are linked through these short contacts into a two-dimensional
layer in the a–b plane (Fig. 2c).
Colorless lamellar crystals of cis-DBT-LM suited for single-
crystal X-ray diffraction structure analysis were obtained by slow
evaporation of CH
2 2
Cl solution at room temperature. cis-DBT-
LM crystallizes in a monoclinic system, space group P 2/c, with
ꢁ
ꢁ
the unit cell parameters of a ¼ 18.924(4) A, b ¼ 6.2925(13) A, c ¼
ꢁ
ꢁ
7
.9164(16) A, b ¼ 99.48(3) , see Fig. 3. The cis-DBT-LM mole-
cules stack in two dimensional layers along the b–c plane
Fig. 3b), with each layer containing two DBT moieties that come
(
from neighboring cis-DBT-LM molecules (as the two DBT
moieties in a particular molecule stack into separate layers).
Within these layers, a herringbone packing pattern is observed
Fig. 3 X-ray crystal structure of compound cis-DBT-LM: (a) Molecular
structure with 50% probability ellipsoids. (b) Layer stacking viewed along
the b–c plane. (c) Herringbone pattern in one layer (one DBT moiety in
each molecule was omitted for clarity).
(
Fig. 3c).
As the solubility of the trans-isomer is poor in common
solvents, crystals suitable for single-crystal X-ray diffraction
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 1313–1325 | 1315

View Article Online
the same plane; the DBT units in DBT-CM are offset by a height
ꢁ
of 1.4 A (1.5 A DFT, defined by the interplanar distances
ꢁ
between the thiophene rings), while the dihedral angle about the
ꢁ
ethylene bridge between the DBT units in cis-DBT-LM is 52.7
ꢁ
(
52.3 DFT).
Redox properties
The empirical electrochemical and optical characteristics are
summarized in Table 1. Cyclic voltammetry (CV) and differ-
ential pulse voltammetry (DPV) measurements were carried out
on dichloromethane solutions of all three compounds and DBT
(
for reference) at room temperature. As shown in Fig. 5, the
three dimer compounds present irreversible oxidative peaks,
and no peaks along the negative potential sweep were observed
to allow further analysis concerning the reduction process. The
first oxidation onset for the cyclic, cis-, and trans-dimers are at
1
.31 V, 1.25 V and 1.10 V, respectively; each of these is at
considerably lower potential versus that for DBT (1.64 V).
Within the CV profile for DBT-CM, other oxidation peaks
appeared to be hidden within the broad feature, therefore DPV
was used to delineate any further oxidation peaks. At least two
definite oxidation peaks exist for each compound within the
DPV sweep area. The first two oxidation peaks lie at 1.34 V and
1.50 V for DBT-CM, 1.26 V and 1.60 V for cis-DBT-LM, and
1.12 V and 1.58 V for trans-DBT-LM. Onset oxidation poten-
tials from the DPV experiments are determined to be 1.25 V,
1.22 V, and 1.05 V for the cyclic, cis-, and trans-dimers,
respectively, and 1.58 V for DBT, trends that are in good
agreement with the CV-determined onsets. The solid-state
ionization potential energy was estimated from the onset
Fig. 4 X-ray crystal structure of compound trans-DBT-LM: (a)
Molecular structure with 50% probability ellipsoids. (b) Column stacking
viewed along a-axis. (c) p–p stacking in one column.
molecular symmetries being C2h for DBT-CM and trans-DBT-
2
LM and C for cis-DBT-LM. Density functional theory (DFT)
calculations of the neutral, ground-state structures were carried
out to directly compare with those derived by X-ray crystallog-
raphy; select DFT (B3LYP/6-31G**)-calculated geometric
parameters for the neutral, radical-cation, and radical-anion
states of DBT-CM, cis-DBT-LM, and trans-DBT-LM are given
in Table S1 (bond numbering scheme shown in Fig. S2) and
compared with crystallographic bond lengths. In general, there is
good agreement between experiment and theory, with differences
in carbon–carbon bond lengths for the various molecular struc-
ss
oxidation potential by using the relationship IP ¼ꢀ(4.40 +
onset
ox
45
)eV. The IP values are relatively large, on the order of
ss
E
5.5–5.6 eV, which should indicate good stability against oxygen
ꢁ
tures of <0.01 A; the largest deviation occurs for the description
46
under ambient conditions. However, such large IP energies
ss
of the carbon–sulfur bonds, with the DFT results providing
could be problematic with respect to OFET applications as they
ss
might lead to large hole-injection barriers (for example, the IP
ꢁ
a slightly longer (by 0.015 A–0.017 A) bond length.
ꢁ
The presence of the bis-ethylene linkers and steric interactions
between neighboring hydrogen atoms in DBT-CM induces
considerable strain within the molecule as compared to its linear
counterparts. As such, the ethylene bonds in DBT-CM are longer
of sexithienyl is 5.2 eV, which matches well with the work
function of the Au electrode and thus provides for a much
smaller hole-injection barrier).
Adiabatic/vertical ionization potential and electron affinity
energies and HOMO and LUMO energies as determined at the
B3LYP/6-31G(d,p) level are given in Table S2; data for DBT are
provided for reference. The bridged dimer systems show
a marked reduction in the ionization potential versus the parent
DBT structure (>1 eV), a result in agreement with the electro-
chemical data. The calculated ionization potentials suggest that
the cyclic and trans-linear forms are more easily oxidized than the
cis-isomer, though the energetic differences are small. This differs
slightly from the electrochemical analysis, which indicates that
the trans-isomer is more readily oxidized than both the cyclic and
cis-linear isomers (which are oxidized at nearly the same poten-
tial). The very limited potential/energetic ranges observed in the
electrochemical/theoretical data (0.20 V range in electrochemical
results and 0.11 eV range in the adiabatic ionization potentials)
indicate that all three compounds are in fact oxidized at similar
energies, with polarization effects from the solvent environment
inducing slight changes to the relative order. The electron affinity
energies indicate that the reduction of DBT-CM (ꢀ0.47 eV)
ꢁ ꢁ
(
1.352 A/1.358 A, X-ray/DFT) versus the linear systems (1.335
ꢁ
ꢁ
ꢁ
ꢁ
A/1.350 A for cis-DBT-LM and 1.334 A/1.349 A for trans-DBT-
LM). The longer ethylene bond in the strained vinylene linkages
in DBT-CM decreases the degree of bond-length alternation
(
BLA) within this segment versus the linear structures: 0.118
ꢁ
A/0.113 A (X-ray/DFT) for DBT-CM, 0.137 A/0.124 A for cis-
ꢁ
ꢁ
ꢁ
ꢁ
DBT-LM, and 0.128 A/0.116 A for trans-DBT-CM.
ꢁ
Additional structural differences across the molecular series
are induced due to the varying steric interactions caused by the
modified bonding patterns about the ethylene linkers. First, the
individual DBT units in the linear analogs are planar, while
the DBT units in the more strained DBT-CM structure adopt
a slightly curved geometry with bends (dihedral angle) of about
ꢁ ꢁ ꢁ
4
.7 and 6.7 (7.1 for both sides, DFT) between the plane of the
five-membered thiophene ring and the individual planes of the
fused six-membered benzene rings. Moreover, the molecular
structure of trans-DBT-LM displays a nearly coplanar geometry,
while the DBT units in cis-DBT-LM and DBT-CM do not lie in
1
316 | J. Mater. Chem., 2012, 22, 1313–1325
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Table 1 Optical and electrochemical properties of cyclic and linear molecules
CV
DPV
UV-vis
d
e
Soln.
Film
onset
b
1
a
onset
ox [V]
b
ss
IP [eV]
c
2
a
Compound
E
ox [V]
E
ox [V]
E
E
–
ox [V]
l
onset [nm]
l
onset [nm]
DBT
DBT-CM
cis-DBT-LM
trans-DBT-LM
1.64
1.31
1.25
1.10
1.81
1.34
1.26
1.12
1.58
1.25
1.22
1.05
5.98
5.65
5.62
5.45
334
350
375
385
–
1.50
1.60
1.58
434
445
396
a
b
onset
c
ss
Calculated from the potential at peak of oxidation.
E
E
ox ) eV. Measurements performed in a dilute CH
ox is the onset potentials for oxidation. Calculated using the empirical equation: IP ¼ꢀ(4.40
onset
d
solution (1 ꢂ 10ꢀ5 M). Measurements performed on vacuum-deposited film.
e
+
2
Cl
2
smaller changes. The carbon–sulfur bonds shorten by D ¼ 0.017
ꢁ
ꢁ
ꢁ
A, 0.031 A, and 0.030 A in DBT-CM, cis-DBT-LM, trans-DBT-
ꢁ
LM, respectively. A shortening of the single bond (D ¼ 0.018 A,
ꢁ ꢁ
.037 A, and 0.033 A in DBT-CM, cis-DBT-LM, trans-DBT-
0
LM, respectively) and lengthening of the double bond (D ¼ 0.015
ꢁ
A, 0.032 A, 0.029 A in DBT-CM, cis-DBT-LM, trans-DBT-LM,
ꢁ
ꢁ
respectively) in the vinylene bridge reduces considerably the BLA
ꢁ
ꢁ
ꢁ
in the oxidized species (BLA ¼ 0.080 A, 0.055 A, and 0.054 A in
DBT-CM, cis-DBT-LM, trans-DBT-LM, respectively). In DBT-
CM, the interplanar distance between the thiophene rings
ꢁ
increases slightly to 1.6 A, while the dihedral angle between the
ꢁ
DBT units in cis-DBT-LM closes substantially to 28.0 .
In contrast, there is minimal change to the carbon–sulfur bond
length on reduction of DBT-CM, while the bond-length change
ꢁ
in the linear analogs is reduced (D ¼ 0.015 A in both cis-DBT-
LM, trans-DBT-LM) versus oxidation. The bond-length changes
to the vinylene bridges, however, are substantial. As with
ꢁ
ꢁ
oxidation, the single bond shortens (D ¼ 0.022 A, 0.041 A, and
ꢁ
.037 A in DBT-CM, cis-DBT-LM, trans-DBT-LM, respec-
0
ꢁ
ꢁ
tively) and the double bond lengthens (D ¼ 0.025 A, 0.048 A,
ꢁ
.043 A in DBT-CM, cis-DBT-LM, trans-DBT-LM, respec-
0
tively). These geometric changes lead to a marked reduction in
ꢁ
ꢁ
BLA for the reduced species (BLA ¼ 0.066 A, 0.035 A, and 0.036
ꢁ
A in DBT-CM, cis-DBT-LM, trans-DBT-LM, respectively). In
DBT-CM, the interplanar distance between the thiophene rings
ꢁ
decreases slightly to 1.4 A, while the dihedral angle between the
ꢁ
DBT units in cis-DBT-LM closes slightly to 45.6 .
Based on the adiabatic potential energy surfaces for the
neutral and radical-ion states, the intramolecular reorganization
energies for the oxidation and reduction processes were deter-
mined, see Table S3. The intramolecular reorganization energy
combines the geometric relaxation energies of the electron-donor
and of the electron-acceptor upon a charge-transfer reaction. For
all three systems, the intramolecular reorganization energy for
Fig. 5 (a) Cyclic voltammogram calibrated with ferrocene/ferrocenium
+ 55
Fc/Fc ) redox system as an external potential marker and (b) differ-
(
ential pulse voltammetry of the cyclic and linear molecules in CH
solution at room temperature.
2 2
Cl
hole transport (l
CM, cis-DBT-LM, trans-DBT-LM, respectively) is smaller than
that for electron transport (l
t
(h) ¼ 0.151 eV, 0.320 eV, and 0.228 eV in DBT-
should be somewhat easier than either linear system (ꢀ0.25 eV
cis-DBT-LM and ꢀ0.32 eV trans-DBT-LM).
t
(e) ¼ 0.200 eV, 0.435 eV, and 0.265
eV in DBT-CM, cis-DBT-LM, trans-DBT-LM, respectively).
Importantly, the intramolecular reorganization energies for both
holes and electrons determined for the cyclic DBT-CM system
are smaller (a function of the smaller bond length changes across
the cyclic molecular structure both for oxidation and reduction)
than those for the linear systems, a result consistent with
The oxidation and reduction processes lead to sizeable
geometric changes for each molecule in the series, with the
changes (while similar in nature) being more pronounced in the
two linear structures. On oxidation to the radical-cation state,
the most significant bond-length changes occur for the carbon–
sulfur bond contained within the DBT moiety and the carbon–
carbon bonds that comprise the full vinylene bridge, while the
carbon–carbon bonds in the fused benzene rings undergo much
47
previous studies of macrocyclic systems. We note that the hole
reorganization energy for DBT-CM is only slightly larger than
that calculated for anthracene, 0.137 eV at the same level of
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 1313–1325 | 1317

View Article Online
4
8
theory. Between the linear configurations, the trans-configura-
tion has smaller intramolecular reorganization energies for both
charge carriers, a result that principally arises from the co-
planarity of the DBT moieties in trans-DBT-LM and the sizeable
changes to the dihedral angle between the DBT moieties in cis-
DBT-LM on oxidation and reduction.
oxidation. In the case of cis-DBT-LM, a number of low-energy
modes – in particular similar dihedral/ring breathing modes at
ꢀ
1
ꢀ1
19 cm and 418 cm – contribute along with a main high-energy
ꢀ
1
mode (1573 cm , a carbon–carbon stretch centered on the
ethylene bridge) to the rather large reorganization of the system
on reduction.
To further assess the nature of reorganization processes, the
potential energy surfaces of the neutral and radical-ion states
were expanded in a power series of the normal-mode coordinates
to weigh the contribution of each vibrational mode to the indi-
Optical properties
The UV-vis spectra of DBT, DBT-CM, cis-DBT-LM, and trans-
DBT-LM in dilute CH Cl solution are displayed in Fig. 7a, with
vidual relaxation processes. The relaxation energy, l , can be
r
2
2
expressed in terms of the harmonic approximation as
each spectrum showing a number of features. DBT has three
main absorption bands with the strongest absorption at 236 nm,
and two weaker bands around 280 nm and 320 nm. Formation of
the cyclic dimer DBT-CM leads to slightly red-shifted, broader
absorption bands versus the DBT monomer, though one is able
to distinguish three bands in the 220–350 nm region. The cis-
isomer shows a very similar absorption profile, except that the
bands are much broader and the relative intensity of the weaker
bands at 300 nm and 320 nm increases significantly. With regard
to the trans-isomer, the absorption peaks at longer wavelength
(between 300 nm and 350 nm) continue to broaden, red-shift, and
intensify versus the cis-isomer.
X
X
-
hu
l
r
¼
l
i
¼
i
S
i
(1)
(2)
k
i
2
i
-
l
i
¼
DQ ; S
i
¼ l
i
=hu
2
i
where S denotes the Huang-Rhys factor (electron-vibration
coupling constant); DQ is the displacement along the normal
i
mode i between the equilibrium position of the ground and
radical-ion states; k and u are the force constants and vibra-
i
tional frequencies; and the summations run over all vibrational
9,49–53
modes.
For all three systems, the intramolecular reorgani-
(h) ¼ 0.151 eV, 0.359 eV, and
.229 eV in DBT-CM, cis-DBT-LM, trans-DBT-LM, respec-
Fig. 7b–d shows a comparison of the (normalized) solution
UV-vis spectra with those of vacuum-deposited thin films. The
maximum absorbance of DBT-CM in thin films occurs at 265 nm,
which is significantly red-shifted (20 nm, 0.38 eV), suggesting the
possible formation of J-type aggregates in the solid state that lead
zation energy for hole transport (l
t
0
tively) and electron transport (l
t
(e) ¼ 0.199 eV, 0.596 eV, and
0
.274 eV in DBT-CM, cis-DBT-LM, trans-DBT-LM, respec-
tively) is consistent with the picture derived by the analyses of the
adiabatic potential energy surfaces (Table S3).
54
to strong intermolecular interactions. The existence of such strong
interactions between adjacent molecules in the film is potentially
favorable for achieving good OFET performance. The cis- and
trans-isomers display similar trends to the cyclic compound in the
solid state with the absorption peaks bathochromically shifted by
35 nm (0.65 eV) and 49 nm (0.48 eV), respectively.
For holes, the intramolecular reorganization energy arises
from a series of high-energy vibrations in the range of 1000 z
ꢀ
1
1
600 cm and low-energy vibrations in the range of 100 z 600
ꢀ1
cm , with one vibrational mode, in general, playing a predomi-
nate role in each region (Fig. 6). The principal high-energy
Excited-state vertical transition energies (and wavelengths),
oscillator strengths, and excited-state electronic configurations as
determined with TDDFT at the B3LYP/6-31G** level of theory
are given in Table S4. The TDDFT excitation energies for the
first electronically-allowed transitions (4.01 eV, 3.71 eV, and 3.56
eV for DBT-CM, cis-DBT-LM, trans-DBT-LM, respectively)
show similar trends to the excitation energy onsets in solution
(3.54 eV, 3.30 eV, and 3.22 eV for DBT-CM, cis-DBT-LM, trans-
DBT-LM, respectively), though the TDDFT results (that are
obtained for isolated molecules) are blue-shifted with respect to
experiment. For DBT-CM, a number of low-lying excited states
have little-to-no transition dipole moment. This includes the first
transition, the HOMO (B )/LUMO (A ) electronic excitation,
ꢀ1
vibrational mode for the radical-cation states (1614 cm for
ꢀ1
ꢀ1
DBT-CM; 1604 cm for cis-DBT-LM; and 1577 cm for trans-
DBT-LM) can best be described as a carbon–carbon stretch
coupled with a carbon–hydrogen rock that is centered on the
bridging ethylene bond(s), with some additional carbon–carbon
stretch behavior within the DBT moieties. This result is consis-
tent with the substantial BLA changes on oxidation. The low-
ꢀ
1
ꢀ1
ꢀ1
energy modes (412 cm for DBT-CM; 18 cm , 302 cm , 396
ꢀ1
ꢀ1
ꢀ1
cm and 432 cm for cis-DBT-LM; and 151 cm for trans-
DBT-LM) can primarily be described as breathing modes
(
coupled to other carbon–carbon, carbon–sulfur, and carbon–
hydrogen motions) centered on the ethylene bridge(s) that move
the independent DBT structures in concert with each other–
a motion that is more pronounced in cis-DBT-LM and consistent
with the large change in the dihedral angle about the vinylene
bridge on oxidation.
u
u
which is symmetry forbidden due to the C2h symmetry of the
molecular structure. The first allowed transition occurs at 4.01
eV (310 nm), whose electronic configuration primarily includes
electronic excitations from HOMO-1 / LUMO (50%) and
HOMO/LUMO + 3 (24%). On the other hand, the trans-DBT-
LM (3.56 eV and 3.78 eV) and cis-DBT-LM (3.71 eV and 3.81
eV) conformers have two, low-lying excited states that are close
to each other energetically. In each case, the lowest-lying state is
principally described as a HOMO / LUMO electronic excita-
tion, while the second transition is HOMO/LUMO + 1 in the
case of the cis-isomer and HOMO/LUMO + 2 for the trans-
isomer; even though trans-DBT-LM has the same C2h
A similar situation arises for electrons, though the high-energy
modes contribute more significantly to the larger relaxation
ꢀ1
processes. For DBT-CM (1611 cm ) and trans-DBT-LM (1594
ꢀ1
cm ), the ethylene-centered carbon–carbon stretch is the main
mode of note; the contribution of this mode to the relaxation
process nearly doubles in the radical-anion state versus that
observed for the radical-cation, and is consistent with the larger
bond-length changes in the vinylene linkage on reduction versus
1
318 | J. Mater. Chem., 2012, 22, 1313–1325
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Fig. 6 Normal-mode contributions to the total intramolecular reorganization energy for holes (left) and electrons (right) as determined at the B3LYP/
6-31G(d,p) level of theory.
symmetry representation as DBT-CM, the HOMO (A
u
)/
) transition is electronically allowed and therefore
which the material is deposited. Therefore, thin films of the cyclic
molecule and the linear isomers deposited at different substrate
temperatures were investigated by atomic force microscopy
LUMO (B
observed.
g
(
AFM) and X-ray diffraction (XRD). Thin films of DBT-CM
and cis-/trans-DBT-LM with a thickness of 50 nm were fabri-
cated by vacuum-deposition on OTS-treated SiO /Si substrates.
Film morphology and characteristics
2
OFET device performance strongly depends on the morphology
and crystallinity of the semiconductor thin films, both of which
can be largely influenced by the temperature of the substrate on
As shown in the AFM images (Fig. 8), the thin-film morphology
of DBT-CM undergoes obvious changes with increased substrate
ꢁ
temperature. When the substrate is held at 15 C, relatively
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 1313–1325 | 1319

View Article Online
ꢀ
5
Fig. 7 UV-vis absorption spectra of (a) all the three target compounds and dibenzothiophene in dilute CH
2
Cl
2
solution (1 ꢂ 10 M); (b) DBT-CM, (c)
cis-DBT-LM, and (d) trans-DBT-LM in dilute CH Cl solution (Sol.) and in the solid state (Film). Thin films with about 50 nm thickness were fabricated
2
2
by a vacuum deposition method on quartz substrates at room temperature.
densely-packed, two-dimensional belt-like crystalline grains
form with an average length of several micrometres. The good
network interconnection between the crystallites might account
for the larger charge-carrier mobilities observed for films
deposited at this temperature (vide infra). With the substrate
crystallinity in the film. However, the film shows less continuity
and larger grain boundaries at the same time, which results in
a decrease in OFET performance (vide infra). Raising the
ꢁ
substrate temperature to 100 C leads to a further reduction in
the peak intensities.
ꢁ
temperature held at 60 C, the belt-like grains grow in size,
The AFM and XRD analyses of the thin films of the two linear
isomers at room temperature are displayed in Fig. 9. For the cis-
isomer, the AFM image reveals small grains and a discontinuous
film. The film of the trans-isomer shows a similar morphology
with much smaller grains; however, several strong reflection
peaks are observed in the XRD analysis, indicative of the
formation of a more polycrystalline film. The results are
consistent with the fact that the trans-isomer shows slightly better
performance in OFET devices than the cis-isomer, though the
charge-carrier mobilities for both systems are one order of
magnitude lower than that of the cyclic macromolecule (vide
infra).
leading to increased film discontinuity and intergrain boundaries.
ꢁ
When the substrate temperature is brought to 100 C, both the
grain sizes and intergrain gaps increase, a result that is expected
to have a negative effect on OFET performance.
The crystallinity and molecular orientation of the thin films
ꢁ
were also analyzed by XRD (Fig. 8). At 15 C, the first peak
ꢁ
ꢁ
appears at 8.8 with a d-spacing of 10.0 A, corresponding to the
thickness of a molecular monolayer. The molecular orientation
on the substrate can be estimated by comparing the monolayer
thickness with the actual molecular length; according to the
ꢁ
molecular height (10.4 A) derived from single crystal data, it is
estimated that the molecules exhibit end-to-end packing on the
substrate, resulting in an almost perpendicular arrangement of
cyclic molecules on the dielectric surface. A relatively weak peak
OFET measurements and charge mobility
ꢁ
ꢁ
at 17.6 with a d-spacing of 5.0 A is the second-order reflection
OFET devices were fabricated and measured in a top-contact
configuration, where the drain and source electrodes are depos-
ited on top of the active layer. Thin films of 50 nm thickness were
ꢁ
peak, whereas a much weaker peak at 11.7 is caused by a tilted
arrangement, rather than multiple orders of the dominant peak.
ꢁ
When the substrate temperature is increased to 60 C, the peak in
ꢁ
2
evaporated on OTS-modified SiO /Si substrates and gold elec-
the 11–12 region disappears and the intensity of the two
trodes were deposited subsequently by vacuum evaporation on
the active layers through a shadow mask. The cyclic molecule
remaining major peaks increases, indicating an increased
1
320 | J. Mater. Chem., 2012, 22, 1313–1325
This journal is ª The Royal Society of Chemistry 2012

View Article Online
and linear isomers each exhibit p-type semiconductor charac-
teristics in OFET devices. The OFET devices for the three
compounds were tested at room temperature in air. The mobil-
ities in the saturation regime from the gate sweep were deter-
mined using eqn (3):
2
ID ¼ (mFETC W/2L) (VG ꢀ VT)
(3)
i
where ID denotes the drain-source current in the saturated
regime; (mFET, the field-effect transistor (FET) mobility; C , the
i
capacitance of the gate dielectric layer; W and L, the channel
G T
width and length; V , the gate voltage; and V , the threshold
voltage. OFET characteristics for the three compounds, with
varying substrate deposition temperatures, are summarized in
Table 2.
OFET performance is found to be closely related to the
substrate temperature. In general, DBT-CM exhibits better
overall device performance versus the two linear isomers under
the same experimental conditions. For DBT-CM, the best OFET
performance is obtained at room temperature with a charge-
2
ꢀ1 ꢀ1
carrier mobility as high as 0.026 cm V
s and an on/off ratio of
5
0 . The output and transfer characteristics of a typical OFET
1
ꢁ
device for DBT-CM deposited at 15 C are provided in Fig. 10.
The results are consistent with the AFM images that show
formation of a continuous polycrystalline film with relatively
densely-packed, two-dimensional belt-like crystalline grains.
ꢁ
With the substrate temperature at 60 C, the mobility decreases
by an order of magnitude; this can be attributed to the observed
film discontinuity and the formation of large intergrain gaps,
even though XRD indicates an increase in the crystallinity. When
Fig. 8 AFM images and XRD patterns of 50-nm-thick thin films of
2
compound DBT-CM vacuum deposited on OTS treated SiO /Si
ꢁ
substrates at different temperatures: (a) and (b) Tsub at 15 C; (c) and (d)
ꢁ ꢁ
ꢁ
the substrate temperature increases further to 100 C, the
Tsub at 60 C; (e) and (f) Tsub at 100 C. The scale of each AFM image is 5
mobility of DBT-CM decreases, which is possibly due to the
much larger intergrain boundaries.
mm ꢂ 5 mm.
The performance of the two linear isomers is significantly
worse under the same device fabrication and measurement
conditions, consistent with previous results for triphenylamine
31,34
and carbazole dimers with cyclic and linear architectures.
ꢁ
When the substrate temperature increases to 60 C, the grain
sizes and boundaries of the thin films of cis- and trans-isomers
become considerably larger (see Fig. S3); consequently, these
highly discontinuous films show no field-effect properties.
Based on the crystal structures of DBT-CM, trans-DBT-LM,
and cis-DBT-LM, the electronic band structures (Fig. 11) were
calculated along a few high-symmetry directions in reciprocal
space to provide a description of the hole-transport properties.
The intermolecular electronic couplings (transfer integrals) – indi-
cators of the extent of wave-function overlap between pairs of
molecules in the crystal – were also evaluated for holes using
Table 2 OFET characteristics of devices based on the three compounds
ꢁ
fabricated under different substrate temperatures ( C)
ꢁ
2
m [cm V
ꢀ1 ꢀ1
Compound
T
sub [ C]
s
]
Ion/Ioff
Vth[V]
ꢀ
2
3
3
3
5
DBT-CM
15
2.6 ꢂ 10
4.8 ꢂ 10
2.7 ꢂ 10
1.5 ꢂ 10
–
10
10
10
10
–
ꢀ43.9
ꢀ51.1
ꢀ53.9
ꢀ51.7
–
ꢀ
ꢀ
ꢀ
6
5
6
6
1
0
00
cis-DBT-LM
15
60
20
Fig. 9 AFM images and XRD patterns of 50-nm-thick thin films
vacuum deposited on OTS treated SiO
2
/Si substrates at room tempera-
ꢀ3
5
trans-DBT-LM
–6 .6 ꢂ 10
10
–
–ꢀ55.4
tures: (a) and (b) for cis-DBT-LM; (c) and (d) for trans-DBT-LM. The
scale of each AFM image is 5 mm ꢂ 5 mm.
6
0
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 1313–1325 | 1321

View Article Online
molecule). The effective mass for hole transport along the GZ direction
(b-axis) is 3.0 m . The computed HOMO–HOMO (70 meV) transfer
0
integral between molecules denoted 2 and 3 in the DBT-CM unit cell
(stacked along the b-axis, see Fig. 12) are fairly significant, as the
prototypical organic semiconductor pentacene has a calculated trans-
fer integral for holes of approximately 100 meV. We note that the
highest VB width is only slightly smaller than four times the transfer
integral, showing good correlation in the context of the tight-binding
approximation. The band dispersion within the a–c plane (along the
a*- and c*-directions) is very flat (not shown), a result in agreement
with the minimal electronic couplings calculated for molecule pairs 1–2
and 1–3. Even though there exist short S/S contact distances between
neighboring columns (that are smaller than twice the van der Waals
radius of S atoms), the electronic coupling is calculated to be very small,
indicating that interstack charge-transfer pathways through S/S
contacts are modest at best in this system.
The band structure of cis-DBT-LM shows that the CB
minimum is located at (k ¼ 0, 0.3375, 0) and the VB maximum is at
the G point, indicating that cis-DBT-LM is also an indirect band-
gap semiconductor. The smallest effective mass for hole transport
Fig. 10 OFET characteristics for devices based on DBT-CM OTS-
modified SiO /Si substrate at room temperature: (a) output curve at
2
different gate voltages; (b) transfer curve in the saturated regime at
a constant source–drain voltage of ꢀ100 V and square root of the
absolute value of the current as a function of the gate voltage.
5
2
the fragment orbital approach at the B3LYP/6-31G** level of theory,
see Fig. 12 and Table 3. For DBT-CM, the conduction band (CB)
minimum is found at the G-point, while the valence band (VB)
maximum is close to the Z-point with a k-vector of (0, 0.4559, 0); these
results indicate the presence of an indirect band gap. Large dispersions
for the highest VB (235 meV) are observed along the GZ direction,
which corresponds to a direction associated with the b-axis in the
crystal structure (the primary stacking direction of the DBT-CM
Fig. 12 Select electronic coupling pathways as determined from the
crystal structure.
Fig. 11 Electronic band structures for the three compounds determined with the PW91 functional and ultra-soft pseudopotentials. For all the three
crystal structures, GZ direction is along the b-axis of the crystal structure, GY is along the direction perpendicular to the c-axis within the a–c-plane of the
crystal structure.
1
322 | J. Mater. Chem., 2012, 22, 1313–1325
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Table 3 Hole (HOMO-HOMO) electronic coupling (Hij, meV) between
adjacent molecules (see labeling in Fig. 12) for the three compounds as
determined at the B3LYP/6-31G(d.p) level of theory
and CH Cl were dried and distilled over CaH . Conventional
2
2
2
Schlenk techniques were adopted and reactions were carried out
under a nitrogen atmosphere unless otherwise stated. 2,8-
Dibromodibenzothiophene (1), 2,8-diformyldibenzothiophene
H
12
H
13
H
23
(
2), 2-bromodibenzothiophene (3), 2-formyldibenzothiophene
4), 2-hydroxymethyldibenzothiophene (5) and (dibenzothio-
DBT-CM
cis-DBT-LM
trans-DBT-LM
4
15
26
4
–
5
70
31
–
(
phene-2-methyl)triphenylphosphonium bromide (6) were
40–44
synthesized according to literature procedures.
0
is 2.1 m , along a path nearly aligned with the a-axis. The effective
(
Z, Z)-cyclo[2](dibenzo[b,d]thiophen-2, 8-yl-ethylene) (DBT-
CM)
mass for holes is smaller than in the DBT-CM crystal, a result that
is attributed to the strong intramolecular coupling between the
two DBT fragments within the linear isomer. However, the
transfer integrals for holes are small (15 meV–30 meV), principally
due to the ‘V-shape’ of cis-DBT-LM that severely limits the
intermolecular electronic coupling within the crystal. For trans-
DBT-LM, the band dispersion for the two top valence bands is
virtually flat along the selected high-symmetry directions. The VB
maximum is found at the Y-point (k ¼ 0.5, 0, 0), where the highest
VB shows no dispersion; the intermolecular electronic coupling
for hole transport is also determined to be small (26 meV).
Zn powder (5.43 g, 83 mmol) was suspended in 1,4-dioxane
(
2 4
300 mL) under N . A solution of TiCl (9.2 mL, 83 mmol) in
1
,4-dioxane (50 mL) was added carefully to the suspension with
ꢁ
stirring and then the suspension was heated to reflux at 130 C for 1
0
h. A solution of 2 (1 g, 4.2 mmol) and pyridine (5.4 mL, 66 mmol) in
1
,4-dioxane (150 mL) was added dropwise to the reaction mixture.
The mixture was stirred and heated at reflux for 10 h, and after
cooling to room temperature, saturated aqueous NaHCO (200
3
mL) was added and the reaction mixture was stirred for an addi-
tional0.5h. Thereactionmixturewasfilteredandmostofthefiltrate
was removed under reduced pressure. The residual mixture was
extracted with dichloromethane (3 ꢂ 100 mL) and the combined
organic extracts were dried over anhydrous sodium sulfate, filtered,
and concentrated under reduced pressure. The residue was purified
by flash chromatography on silica gel using petroleum ether as the
eluent to afford the crude product. Recrystallization from
Conclusion
Three dibenzothiophene dimers have been synthesized by
McMurry coupling and Wittig reactions, including one cyclic
molecule (DBT-CM) and two linear compounds (cis- and trans-
DBM-LM). The structures were thoroughly characterized by
means of NMR, HRMS, elemental analysis, electrochemistry,
absorption spectroscopy, quantum-chemical analysis, and
employed as the active materials in organic TFTs. The cyclic
architecture affects significantly the molecular stacking pattern
in the crystals, as well as the molecular and crystal electronic
properties.
ꢁ
dichloromethane gave a yellow crystal (0.28g, 16%). M.P. 310 C;
1
H NMR (400 MHz, DMSO) d 9.33 (s, 4H), 7.83–7.81 (d, J ¼ 8.3
13
Hz, 4H), 7.33–7.31 (d, J ¼ 8.3 Hz, 4H), 7.25 (s, 4H); C NMR (100
3
MHz, CDCl ) d 138.7, 134.7, 132.6, 132.2, 127.7, 123.1, 122.2; EI-
+
MS: m/z ¼ 416 [M] ; HRMS (GCT-MS): calculated for C28
H S :
16 2
+
416.0693; found: 416.0698 [M] ; elemental analysis calculated (%)
Several prominent reasons related to the cyclic architecture
appear to contribute to the large mobility differences between the
cyclic and linear compounds. First, the cyclic structure signifi-
cantly improves the molecular ordering within the crystal, leading
to larger band dispersions (and intermolecular electronic
couplings) for hole transport. The better packing arrangements
also lead to the formation of relatively densely-packed, two-
dimensional belt-like crystalline grains with fine continuity. Even
though smaller grains were obtained for the linear compounds, the
denser packing motif of the cyclic structure leads to better charge-
carrier mobility through the film. Moreover, the cyclic structure
leads to a marked reduction in intramolecular reorganization
energy for hole transport versus the linear isomers, a feature that
aids in the improved hole-transport characteristics. Each of these
characteristics leads to OFET devices based on DBT-CM having
for C28H S : C 80.73, H 3.87; found: C 81.01, H 3.73.
16 2
(Z)-1,2-bis(dibenzo[b,d]thiophen-2-yl)ethene (cis-DBT-LM)
To a solution of 2-formyldibenzothiophene (0.106 g, 0.5 mmol)
and (dibenzothiophene-2-methyl)triphenylphosphonium bromide
t
0.27 g, 0.5 mmol) in THF/MeOH (6 mL, 2 : 1), BuOK (0.112 g,
(
1
mmol) was added slowly. The resulting mixture was stirred at
room temperature for 6 h. Water (10 mL) was added and the
solution was extracted with chloroform (3 ꢂ 15 mL). The
combined extracts were dried with Na SO , and purified by
2
4
chromatography using petroleum as the eluent on a short column
ꢁ
1
to give a white powder (0.03 g, 39%). M.p. 185 C; H NMR (400
MHz, CDCl
) d 8.12 (s, 2H), 7.97 (d, J ¼ 7.6 Hz, 2H), 7.84 (d, J ¼
.7 Hz, 2H), 7.67 (d, J ¼ 8.3 Hz, 2H), 7.45–7.37 (m, 6H), 6.85 (s,
2H); C NMR (100 MHz, CDCl ) d 139.9, 138.5, 135.9, 135.5,
33.8, 130.3, 127.8, 126.9, 124.6, 123.0, 122.6, 122.1, 121.7; EI-
3
7
13
2
ꢀ1 ꢀ1
3
the largest charge-carrier mobility of 0.026 cm V s , a one-
order-of-magnitude improvement over either linear analog.
1
+
MS: m/z ¼ 392 [M] ; HRMS (GCT-MS): calculated for
+
: 392.0693; found: 392.0700 [M] .
C
26
H
16
S
2
Experimental section
Synthesis
(E)-1,2-bis(dibenzo[b,d]thiophen-2-yl)ethene (trans-DBT-LM)
All reagents and solvents were purchased from commercial
suppliers and used directly as received unless otherwise stated.
Anhydrous diethyl ether, THF and 1,4-dioxane were distilled
over Na/benzophenone. Pyridine was distilled over NaOH. DMF
To a solution of 2-formyldibenzothiophene (0.76 g, 3.59 mmol)
and diethyl dibenzothiophene-2-methylphosphonate (1.13 g,
t
3.37 mmol) in THF (100 mL), BuOK (0.73 g, 6.47 mmol) was
added slowly. The resulting suspension was stirred at room
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 1313–1325 | 1323

View Article Online
temperature overnight, followed by filtration to afford a light
yellow powder that was then washed with water several times and
further purified by recrystallization in toluene to provide white
of 8817 reflections, 1550 were unique; GOF ¼ 1.037; 136
parameters; R ¼ 0.0352, wR ¼ 0.0905 (for all reflections).
1
2
ꢁ
1
floccules (0.50 g, 38%). M.P. 281 C; H NMR (400 MHz, CDCl
3
)
Crystallographic data for cis-DCM-LM. 0.10 ꢂ 0.10 ꢂ 0.02
3
mm ; monoclinic; P 2
ꢁ
d 8.32(s, 2H), 8.24(d, J ¼ 8.1Hz, 2H), 7.87(d, J ¼ 8.1Hz, 4H), 7.72
1
/c; Z ¼ 2; a ¼ 18.924(4) A, b ¼ 6.2925(13)
ꢁ
ꢁ
ꢁ
3
ꢀ3
(
d, J ¼ 8.3 Hz, 2H), 7.51–7.49 (m, 4H), 7.42 (s, 2H); EI-MS: m/z ¼
A, c ¼ 7.9164(16) A; V ¼ 932.7(3) A ; rcalculated ¼ 1.402 g cm ; of
7994 reflections, 1775 were unique; GOF ¼ 1.159; 136 parame-
+
92 [M] ; HRMS (GCT-MS): calculated for C26
3
H
16
S
2
: 392.0693;
+
found: 392.0699 [M] ; elemental analysis calculated (%) for
: C 79.55, H 4.11; found: C 79.67, H 4.35.
ters; R
1
¼ 0.0879, wR ¼ 0.1423 (for all reflections).
2
26 16 2
C H S
Crystallographic data for trans-DCM-LM. 0.31 ꢂ 0.10 ꢂ 0.10
3
ꢁ
ꢁ
mm ; monoclinic; P 2 /c; Z ¼ 2; a ¼ 6.792(15) A, b ¼ 10.677(2) A,
1
ꢁ
ꢁ
3
ꢀ3
Measurements and characterization
c ¼ 12.729(3) A; V ¼ 914.8(4) A ; r
¼ 1.425 g cm ; of
calculated
1
1960 reflections, 2007 were unique; GOF ¼ 1.234; 136 param-
€
Melting points were measured on a BUCHI Melting Point B-500.
eters; R ¼ 0.0660, wR ¼ 0.1176 (for all reflections).
1 13
H NMR and C NMR spectra were performed on Bruker
1
2
Advance 400 MHz spectrometers. MS (EI) measurements were
performed on SHIMADZU GCMSQP2010 or UK GCT-
Micromass spectrometers. HRMS (GCT-MS) measurements
were performed on a Bruker APEX II FTICRMS spectrometer.
Elemental analyses were measured on a Carlo Erba 1106
elemental analyzer.
FET device fabrication and measurement
OFETs based on vacuum-deposited thin films were fabricated in
a top-contact device configuration. A heavily doped n-type Si
2
wafer with a 500 nm-thick SiO layer and a capacitance of 7.5 nF
ꢀ2
cm was used as the substrate. The gate dielectric was treated
with octadecyltrichlorosilane (OTS) by a vapor deposition
method. All three materials were deposited on substrates by
UV-vis spectra were recorded on a JASCO V-570 spectrom-
eter. Cyclic voltammetry (CV) measurements were carried out on
a CHI660C analyzer in a conventional three-electrode cell setup
with a platinum rod as the working electrode, a platinum wire as
the counter electrode, Ag/AgCl as the reference electrode and
ꢀ
4
thermal evaporation under a pressure of 4–6 ꢂ 10 Pa at
ꢁ
ꢀ1
ꢁ
a deposition rate gradually increased from 0.1 A s to 0.5 A s
ꢀ1
+
for the first 10 nm and then maintained until the thickness of the
film was 50 nm. The deposition rate and film thickness were
monitored by ULVAC CRTM-6000. Subsequently, 20 nm thick
gold source and drain electrodes were deposited on the films via
a shadow mask. The channel length and width were 0.11 mm and
calibrated with ferrocene/ferrocenium (Fc/Fc ) as an external
55
potential marker in anhydrous CH Cl solution containing 0.1
2
2
n
M Bu NPF as a supporting electrolyte at a scan rate of 100
4
6
ꢀ1
mVs under a nitrogen atmosphere at room temperature.
Differential pulse voltammetry (DPV) was also performed on
a CHI660C analyzer in the same conventional three-electrode
cell setup aforementioned (without any potential marker) at
a pulse width of 0.05 s in a pulse period of 0.2 s with a sensitivity
5
.30 mm, respectively. The FET characteristics were measured at
room temperature in air using Keithley 4200 SCS.
ꢀ5
of 1 ꢂ 10 under a nitrogen atmosphere at room temperature.
Computational methodology
Thermogravimetric analysis (TGA) was performed on a Shi-
ꢀ1
ꢁ
madzu DTG 60 instrument at a heating rate of 10 C min under
a nitrogen atmosphere with runs recorded from room tempera-
Analyses of the neutral ground-state and radical-ion structures
for the cyclic and linear molecules were carried out using density
57–59
functional theory (DFT). The B3LYP functional
ꢁ
ture to 550 C.
was used in
XRD measurements of thin films were performed in reflection
mode at 40 kV and 200 mA with Cu-Ka radiation using a 2 kW
Rigaku X-ray diffractometer. AFM images of the thin films were
obtained on a Nanoscope IIIa AFM (Digital Instruments)
operating in tapping mode.
conjunction with the 6-31G(d,p) basis set. Time-dependent DFT
(TDDFT) calculations were employed to analyze the vertical
transition energies, oscillator strengths, and electronic configu-
rations for low-lying singlet excited states. All calculations were
60
performed with Gaussian 09 (Revision A.02), and molecular
61
orbital density plots were generated with ArgusLab 4.0.1.
Band-structure calculations at the DFT level were carried out
62
using the Vienna Ab Initio Simulation Package (VASP) with
Single-crystal structure determination
ultra-soft pseudopotentials and the GGA exchange and corre-
X-ray crystallographic data were collected with a Bruker Smart
CCD diffractometer through using graphite-monochromated
63
lation functional of Perdew and Wang (PW91). A plane wave
ꢀ6
cut-off of 400 eV and a total energy convergence of 10 eV for
the self-consistent iterations were applied in the calculations.
ꢁ
Mo-Ka radiation (l ¼ 0.71073 A). The data were collected at 113
K and 173 K for the cyclic and two linear isomers, respectively.
The structures were resolved by the direct method and refined by
2
full-matrix least-squares on F using the SHELXL-97 program.
56
Acknowledgements
All non-hydrogen atoms were refined anisotropically. Complete
crystallographic data for the three compounds are available.†
The authors are grateful to financial support from Solvay S. A.,
the National Natural Science Foundation of China, the State
Key Basic Research Program, and the Chinese Academy of
Sciences. The authors would like to thank Dr Stephen Barlow
and Dr Veaceslav Coropceanu for fruitful discussions.
Crystallographic data for DCM-CM. Crystal size: 0.18 ꢂ 0.12
3
ꢁ
ꢂ 0.10 mm ; monoclinic; P 2
1
/c; Z ¼ 2; a ¼ 12.339(3) A, b ¼ 4.748
ꢁ
ꢁ
ꢁ
3
ꢀ3
(
9) A, c ¼ 16.949(3) A; V ¼ 929.8(3) A ; rcalculated ¼ 1.483 g cm ;
1
324 | J. Mater. Chem., 2012, 22, 1313–1325
This journal is ª The Royal Society of Chemistry 2012

View Article Online
3
7 L. Tan, L. Zhang, X. Jiang, X. D. Yang, L. J. Wang, Z. Wang,
L. Q. Li, W. P. Hu, Z. G. Shuai, L. Li and D. B. Zhu, Adv. Funct.
Mater., 2009, 19, 272.
Notes and references
1
2
3
A. Tsumura, H. Koezuka and T. Ando, Appl. Phys. Lett., 1986, 49,
210.
C. D. Dimitrakopoulos and P. R. L. Malenfant, Adv. Mater., 2002,
4, 99.
T. W. Kelley, P. F. Baude, C. Gerlach, D. E. Ender, D. Muyres,
M. A. Haase, D. E. Vogel and S. D. Theiss, Chem. Mater., 2004,
1
38 M. Mas-Torrent, P. Hadley, S. T. Bromley, X. Ribas, J. Tarres,
M. Mas, E. Molins, J. Veciana and C. Rovira, J. Am. Chem. Soc.,
2004, 126, 8546.
39 J. H. Gao, L. Q. Li, Q. Meng, R. J. Li, H. Jiang, H. X. Li and
W. P. Hu, J. Mater. Chem., 2007, 17, 1421.
1
1
6, 4413.
40 K. Memminger, T. Oeser and T. J. J. Muller, Org. Lett., 2008, 10,
2797.
41 J. E. Mcmurry, Chem. Rev., 1989, 89, 1513.
42 Y. Nishide, H. Osuga, M. Saito, T. Aiba, Y. Inagaki, Y. Doge and
K. Tanaka, J. Org. Chem., 2007, 72, 9141.
4
5
Y. M. Sun, Y. Q. Liu and D. B. Zhu, J. Mater. Chem., 2005, 15, 53.
S. Allard, M. Forster, B. Souharce, H. Thiem and U. Scherf, Angew.
Chem., Int. Ed., 2008, 47, 4070.
M. Mas-Torrent and C. Rovira, Chem. Soc. Rev., 2008, 37, 827.
R. P. Ortiz, J. Casado, V. Hernandez, J. T. L. Navarrete, J. A. Letizia,
M. A. Ratner, A. Facchetti and T. J. Marks, Chem.–Eur. J., 2009, 15,
6
7
43 L. V. Dunkerton, B. C. Barot and A. Nigam, J. Heterocycl. Chem.,
1987, 24, 749.
5023.
S. R. Forrest, Chem. Rev., 1997, 97, 1793.
V. Coropceanu, J. Cornil, D. A. da Silva, Y. Olivier, R. Silbey and
J. L. Bredas, Chem. Rev., 2007, 107, 926.
0 A. R. Murphy and J. M. J. Frechet, Chem. Rev., 2007, 107, 1066.
1 J. Zaumseil and H. Sirringhaus, Chem. Rev., 2007, 107, 1296.
2 A. J. Baca, J. H. Ahn, Y. G. Sun, M. A. Meitl, E. Menard, H. S. Kim,
W. M. Choi, D. H. Kim, Y. Huang and J. A. Rogers, Angew. Chem.,
Int. Ed., 2008, 47, 5524.
44 S. Kumar, J. Chem. Soc. Perk. T., 2001, 1, 1018.
45 M. Thelakkat and H. W. Schmidt, Adv. Mater., 1998, 10, 219.
46 H. Meng, M. Bendikov, G. Mitchell, R. Helgeson, F. Wudl, Z. Bao,
T. Siegrist, C. Kloc and C. H. Chen, Adv. Mater., 2003, 15, 1090.
47 X. D. Yang, Q. K. Li and Z. G. Shuai, Nanotechnology, 2007, 18.
48 V. Coropceanu, M. Malagoli, D. A. da Silva, N. E. Gruhn, T. G. Bill
and J. L. Bredas, Phys. Rev. Lett., 2002, 89.
8
9
1
1
1
49 P. F. Barbara, T. J. Meyer and M. A. Ratner, J. Phys. Chem., 1996,
100, 13148.
1
3 C. D. Sheraw, L. Zhou, J. R. Huang, D. J. Gundlach, T. N. Jackson,
M. G. Kane, I. G. Hill, M. S. Hammond, J. Campi, B. K. Greening,
J. Francl and J. West, Appl. Phys. Lett., 2002, 80, 1088.
4 H. Klauk, M. Halik, U. Zschieschang, G. Schmid, W. Radlik and
W. Weber, J. Appl. Phys., 2002, 92, 5259.
5 T. W. Kelley, L. D. Boardman, T. D. Dunbar, D. V. Muyres,
M. J. Pellerite and T. Y. P. Smith, J. Phys. Chem. B, 2003, 107, 5877.
6 H. E. Katz, Chem. Mater., 2004, 16, 4748.
50 K. D. Demadis, C. M. Hartshorn and T. J. Meyer, Chem. Rev., 2001,
101, 2655.
51 J. L. Bredas, D. Beljonne, V. Coropceanu and J. Cornil, Chem. Rev.,
2004, 104, 4971.
52 E. F. Valeev, V. Coropceanu, D. A. da Silva, S. Salman and
J. L. Bredas, J. Am. Chem. Soc., 2006, 128, 9882.
53 C. Risko, V. Coropceanu, S. Barlow, V. Geskin, K. Schmidt,
N. E. Gruhn, S. R. Marder and J. L. Bredas, J. Phys. Chem. C,
2008, 112, 7959.
54 J. Bouchard, M. Belletete, G. Durocher and M. Leclerc,
Macromolecules, 2003, 36, 4624.
55 K. M. Omer, S. Y. Ku, K. T. Wong and A. J. Bard, J. Am. Chem.
Soc., 2009, 131, 10733.
56 G. M. Sheldrick, SHELXL-97, Program for the Solution and
Refinement of Crystal Structures, University of G o€ ttingen,
G o€ ttingen, Germany, 1997.
57 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
58 A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098.
59 A. D. Becke, J. Chem. Phys., 1993, 98, 1372.
60 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. M. Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin,
V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,
N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth,
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels,
O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox,
Gaussian, Incorporated, Wallingford, CT, 2009.
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
7 J. S. Moore, Acc. Chem. Res., 1997, 30, 402.
8 S. Hoger, J. Polym. Sci., Part A: Polym. Chem., 1999, 37, 2685.
9 M. M. Haley, J. J. Pak and S. C. Brand, Top. Curr. Chem., 201, 81.
0 C. Grave and A. D. Schluter, Eur. J. Org. Chem., 2002, 3075.
1 D. Zhao and J. S. Moore, Chem. Commun., 2003, 807.
2 Y. Yamaguchi and Z. Yoshida, Chem.–Eur. J., 2003, 9, 5430.
3 S. Hoger, Chem.–Eur. J., 2004, 10, 1320.
4 S. Hoger, Angew. Chem., Int. Ed., 2005, 44, 3806.
5 W. Zhang and J. S. Moore, Angew. Chem., Int. Ed., 2006, 45, 4416.
6 I. Hisaki, M. Sonoda and Y. Tobe, Eur. J. Org. Chem., 2006, 833.
7 A. Mishra, C. Q. Ma and P. Bauerle, Chem. Rev., 2009, 109, 1141.
8 P. Bauerle, Adv. Mater., 1992, 4, 102.
9 Y. Y. Noh, J. J. Kim, Y. Yoshida and K. Yase, Adv. Mater., 2003, 15,
699.
3
0 A. Dadvand, F. Cicoira, K. Y. Chernichenko, E. S. Balenkova,
R. M. Osuna, F. Rosei, V. G. Nenajdenko and D. F. Perepichka,
Chem. Commun., 2008, 5354.
3
3
3
1 H. Xu, G. Yu, W. Xu, Y. Xu, G. L. Cui, D. Q. Zhang, Y. Q. Liu and
D. B. Zhu, Langmuir, 2005, 21, 5391.
2 H. Xu, Y. Wang, G. Yu, W. Xu, Y. B. Song, D. Q. Zhang, Y. Q. Liu
and D. B. Zhu, Chem. Phys. Lett., 2005, 414, 369.
3 Y. B. Song, C. Di, X. D. Yang, S. P. Li, W. Xu, Y. Q. Liu,
L. M. Yang, Z. G. Shuai, D. Q. Zhang and D. B. Zhu, J. Am.
Chem. Soc., 2007, 129, 1470.
3
3
3
4 T. Y. Zhao, Z. M. Wei, Y. B. Song, W. Xu, W. P. Hu and D. B. Zhu,
J. Mater. Chem., 2007, 17, 4377.
5 Y. B. Song, C. A. Di, Z. M. Wei, T. Y. Zhao, W. Xu, Y. Q. Liu,
D. Q. Zhang and D. B. Zhu, Chem.–Eur. J., 2008, 14, 4731.
6 Y. M. Sun, L. Tan, S. D. Jiang, H. L. Qian, Z. H. Wang, D. W. Yan,
C. G. Di, Y. Wang, W. P. Wu, G. Yu, S. K. Yan, C. R. Wang,
W. P. Hu, Y. Q. Liu and D. B. Zhu, J. Am. Chem. Soc., 2007, 129,
61 M. A. Thompson, Planaria Software, LLC, Seattle, WA.
62 G. Kresse and J. Furthmuller, Comput. Mater. Sci., 1996, 6, 15.
63 Y. Wang and J. P. Perdew, Phys. Rev. B: Condens. Matter, 1991, 43,
8911.
1882.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 1313–1325 | 1325