Linear- and angular-shaped naphthodithiophenes: Selective synthesis, properties, and application to organic field-effect transistors

Article abstract of DOI:10.1021/ja110973m

A straightforward synthetic approach that exploits linear- and angular-shaped naphthodithiophenes (NDTs) being potential as new core structures for organic semiconductors is described. The newly established synthetic procedure involves two important steps

Full text of DOI:10.1021/ja110973m

ARTICLE
pubs.acs.org/JACS
Linear- and Angular-Shaped Naphthodithiophenes:
Selective Synthesis, Properties, and Application to Organic
Field-Effect Transistors
Shoji Shinamura,† Itaru Osaka,† Eigo Miyazaki,† Akiko Nakao,‡ Masakazu Yamagishi,§ Jun Takeya,§ and
||
Kazuo Takimiya*,†,
†Department of Applied Chemistry, Graduate School of Engineering, Hiroshima University, Higashi-Hiroshima 739-8527, Japan
‡High Energy Accelerator Research Organization (KEK), 1-1 Oho, Tsukuba, Ibaraki 305-0801 Japan
§ISIR, Osaka University, Ibaraki 567-0047, Japan
Institute for Advanced Materials Research, Hiroshima University, Higashi-Hiroshima 739-8530, Japan
S Supporting Information
b
ABSTRACT: A straightforward synthetic approach that exploits
linear- and angular-shaped naphthodithiophenes (NDTs) being
potential as new core structures for organic semiconductors is
described. The newly established synthetic procedure involves two
important steps; one is the chemoselective Sonogashira coupling
reaction on the trifluoromethanesulfonyloxy site over the bromine site
enabling selective formation of o-bromoethynylbenzene substructures
on the naphthalene core, and the other is a facile ring closing reaction
of fused-thiophene rings from the o-bromoethynylbenzene sub-
structures. As a result, three isomeric NDTs, naphtho[2,3-b:6,7-
b0]dithiophene, naphtho[2,3-b:7,6-b0]dithiophenes, and naphtho[2,1-b:6,5-b0]dithiophene, are selectively synthesized. Electro-
chemical and optical measurements of the parent NDTs indicated that the shape of the molecules plays an important role in
determining the electronic structure of the compounds; the linear-shaped NDTs formally isoelectronic with naphthacene have
lower oxidation potentials and more red-shifted absorption bands than those of the angular-shaped NDTs isoelectronic with
chrysene. On the contrary, the performance of the thin-film-based field-effect transistors (FETs) using the dioctyl or diphenyl
derivatives were much influenced by the symmetry of the molecules; centrosymmetric derivatives tend to give higher mobility (up to
1.5 cm2 V-1 s-1) than axisymmetric ones (∼0.06 cm2 V-1 s-1), implying that the intermolecular orbital overlap in the solid state is
influenced by the symmetry of the molecules. These results indicate that the present NDT cores, in particular the linear-shaped,
centrosymmetric naphtho[2,3-b:6,7-b0]dithiophene, are promising building blocks for the development of organic semiconducting
materials.
’ INTRODUCTION
theoretical investigation predicts that they are promising cores
for organic semiconductors.10
Thiophene-fused polycyclic aromatic compounds (hete-
roarenes) have been attracting a great deal of interest in view
of application to organic semiconducting materials for electronic
device applications, such as organic field-effect transistors
(OFETs)1-4 and organic photovoltaics (OPVs).5-7 In particu-
lar, linearly fused heteroarenes, benzodithiophene (BDT),1 and
anthradithiophene (ADT)2 (Figure 1), an isoelectronic analogue
of anthracene and pentacene, respectively, represent the
usefulness of this class of compounds to, for example, solution-
or vapor processable p- and n-channel organic semicon-
ductors,1-3 building block for polymers,4 and synthetic inter-
mediate for further extended heteroarenes.8 In contrast to
widely studied BDT and ADT derivatives whose efficient
synthetic methods are well-documented,9 linearly fused
naphthodithiophenes (NDTs, 1 and 2, Figure 1) isoelectronic
with naphthacene have been yet to be synthesized, although the
The isomeric NDTs with angular-fused structures (3 and 4,
Figure 1) were recently isolated by Tobe et al. from a
reaction mixture of the flash vacuum pyrolysis (FVP) of
dithienyldienynes.11a On the other hand, we have devised a
selective and scalable synthesis of 3 and applied the core to
organic semiconducting materials.12 In due course, the NDT
framework of 3 was found to be useful not only as a core for
molecular semiconductors12 but also as a building unit for
polymer semiconductors.13 In fact, both the 3-based molecular
and polymer semiconductors are applicable to OFETs, and they
showed field-effect mobility higher than 0.1 cm2 V-1 s-1. These
fruitful results on 3 as well as the superiority of BDT and ADT as
Received: December 7, 2010
Published: March 09, 2011
r
2011 American Chemical Society
5024
dx.doi.org/10.1021/ja110973m J. Am. Chem. Soc. 2011, 133, 5024–5035
|

Journal of the American Chemical Society
ARTICLE
Scheme 1. Reported Synthesis of 3
Figure 1. Molecular structures of BDT, ADT, and naphthodithiophene
(NDT) isomers. 1: naphtho[2,3-b:6,7-b0]dithiophenes, 2: naphtho[2,3-
b:7,6-b0]dithiophenes, 3: naphtho[1,2-b:5,6-b0]dithiophenes, 4: naph-
tho[2,1-b:6,5-b0]dithiophenes.
1,3,5,7-tetrabromo-2,6-dihydroxynaphthalene14 and 1,3,6-tri-
bromo-2,7-dihydroxynaphthalene,15 respectively (Scheme 2).
Being different from 7 with the chloro groups, the chemoselec-
tive Sonogashira coupling reactions of 10 and 13 at the trifluor-
omethanesulfonyloxy sites over the bromine sites seemed to be
very difficult. After many attempted reactions under various
conditions, however, we finally found that the reactions in polar
solvents such as THF or DMF gave the desired products, 11 and
14, with good selectivity in most cases (Scheme 2). The final ring
closing reaction proceeded smoothly to give the parent, di-n-
octyl, and diphenyl derivatives (1a-c and 2a-c) in good yields.
The chemoselective Sonogashira coupling at the trifluoro-
methanesulfonyloxy sites over the bromine sites on the naphtha-
lene derivatives also allowed synthesizing 4 from 1,5-dihy-
droxynaphthalene. It has been reported that bromination of
1,5-dihydroxynaphthalene gives 2,6-dibromo-1,5-dihydroxynap-
hthalene (15) selectively,16 whereas chlorination takes places
exclusively at 4- and 8-positions.17 Thus, 15 was converted into
the precursors (17) having the o-ethynylbromobenzene sub-
structures via the trifluoromethanesulfonyloxy intermediate
(16), and the final ring closing reaction with sodium sulfide gave
4a-c in good yields. All the NDT derivatives (1, 2, and 4) were
fully characterized with the spectroscopic and combustion ele-
mental analyses (see Experimental section).
Physicochemical Properties. Cyclic voltammograms of the
parent compounds, 1a-4a, are shown in Figure 2a and their
oxidation potentials are summarized in Table 1. Both 1a and 2a
show quasi-reversible oxidation peaks at þ0.94 V (onset, vs Ag/
AgCl), indicating that 1a and 2a have almost the same HOMO
energy levels. The good reversibility of their voltammograms
implies that their oxidized species are electrochemically stable,
and in fact, they did not polymerized on the working electrode
even on repeated cycles (Figure S1, Supporting Information).
These electrochemical behaviors of 1a and 2a sharply contrast to
that of isomeric 3a, which shows an irreversible oxidation peak at
much higher oxidation potential (þ1.38 V vs Ag/AgCl) and an
electrochemical polymerization on the working electrode on the
repeated cycles.9 On the other hand, 4a shows a rather similar
electrochemical behavior to that of 3a: relatively high oxidation
potential (þ1.33 V vs Ag/AgCl) to those of 1a and 2a and
electrochemical polymerization on the electrode (Figure S1,
Supporting Information). These results clearly indicate that the
direction of thiophene rings fused on the naphthalene core (i.e.,
anti vs syn at the 2,3/6,7-fused position or 1,2-b vs 2,1-b
annulation at the 1,2/5,6-fused position) does not affect to the
electronic structures significantly, whereas the position of thio-
phene rings (i.e., 1,2/5,6-fused vs 2,3/6,7-fused) does (Table 1).
building blocks for electronic materials prompted us to study the
linear-shaped NDTs, namely 1 and 2.
In this paper, we first report a newly established selective
syntheses of 1 and 2. In the meantime, it turned out that the
synthetic method developed for 1 and 2 is applicable to the
selective synthesis of the angular-shaped NDT, 4, which enables
us to access all four NDT isomers (Figure 1), being promising
cores for organic semiconductors, and thereby to compare their
properties. We thus describe the comparison of electronic
structures of the four parent NDTs (1a-4a) in detail. Then,
the thin-film FET devices based on the dioctyl- (1b, 2b, and 4b)
and diphenyl derivatives (1c, 2c, and 4c) are reported to
evaluate the usefulness of the NDT cores for organic semicon-
ductors. Among them, vapor-processed 1c-based devices showed
good FET characteristics with mobilities as high as 1.5 cm2
V-1 s-1 and Ion/Ioff of 107, which is almost comparable with the
isomeric 2,7-diphenyl[1]benzothieno[3,2-b][1]benzothiophene
(DPh-BTBT)-based OFETs (∼2.0 cm2 V-1 s-1).3d Finally, to
get insight into the structure-properties relationship of the
isomeric heteroarenes, detailed comparison between 1c and
DPh-BTBT were carried out in terms of the intermolecular
orbital couplings in the solid state, molecular reorganization
energy, and ionization potentials of the evaporated thin films.
’ RESULTS AND DISCUSSION
Synthesis. The reported selective synthesis of 3 is depicted in
Scheme 1,12 in which the thiophene annulation reaction from the
precursor (8) with two o-chloroethynylbenzene substructures9c
was employed as the key step for the synthesis. For the selective
preparation of 8, the higher reactivity of the trifluoromethane-
sulfonyloxy site than that of the chloro site in the Sonogashira
coupling is crucial, and thus, 1,5-dichloro-2,6-bis(trifluoro-
methanesulfonyloxy)naphthalene (7), readily converted from
2,6-dihydroxynaphthalene (6), is the actual key intermediate
for the synthesis of 3.
For the synthesis of 1 and 2, however, it is practically
impossible to synthesize the counterparts of 7 possessing two
chloro and trifluoromethanesulfonyloxy groups at the appropri-
ate positions on the naphthalene core. Instead of the chloro
intermediate, the bromo compounds (10 and 13, Scheme 2)
were accessible via 3,7-dibromo-2,6-dihydroxynaphthalene (9)
and 3,6-dibromo-2,7-dihydroxynaphthalene (12), respectively,
via selective debromination reaction mediated by tin metal from
5025
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
Scheme 2. Synthesis of 1, 2, and 4a
a TMS groups on the precursors (11a, 14a, and 17a) were removed spontaneously during the final ring closing reaction to give the parent NDTs.
Figure 2. (a) Cyclic voltammograms and (b) UV-vis spectra of 1a, 2a, 3a, and 4a.
The difference in electronic structures can be qualitatively
understood by considering the isoelectronic hydrocarbon-based
fused aromatic systems. Apparently, 1a and 2a are isoelectronic
with naphthacene with four benzene rings fused in a linear
manner, whereas 3a and 4a are with chrysene with a kinked
molecular structure.18 The HOMO energy levels of acenes such
as anthracene and naphthacene elevate significantly with increas-
ing the number of the fused rings. However, the elevation of the
HOMO energy level of the phenes such as phenanthrene
and chrysene is not significant with increasing the number of
the rings.18,19 It is thus rational to consider that 3a and 4a
isoelectronic with chrysene have higher oxidation potentials than
those of 1a and 2a. These qualitative considerations are well
reproduced by the theoretical calculations using the DFT
methods,20 which clearly show the similarity/difference of the
energy levels of the frontier orbitals as well as the molecular
electronic structures among 1a-4a (Figure 3).
Absorption spectra of 1a and 2a in the solution are also similar
to each other (Figure 2b), reflecting that these two isomers have
basically the same electronic structures. Because of their acene-
like structures, their absorption edges reach up to 410 nm, which
is strikingly different from those of 3 and 4 (absorption edge:
∼350 nm) with the phene-like electronic structures.
It is already reported that photochemical- and oxidative-
stability of large oligoacenes, such as naphthacene and pentacene
are not very good.21 In fact, absorption spectra of naphthacene in
air-saturated solution under illumination of ambient light
showed rapid degradation in several hours (Figure S2, Support-
ing Information). On the contrary, the absorption spectra of 1a
and 2a under the identical conditions showed no apparent
change even after several days, indicating that 1a and 2a are
quite stable compounds, despite the fact that they are isoelec-
tronic with naphthacene. These results show that the linear-fused
heteroarenes (1 and 2) are promising building blocks with
5026
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
Table 1. Electronic Properties of the Parent NDTs (1a-4a)
factor that affects to the field-effect mobility in the thin film
transistors.
compound
Eonset/Va
HOMO/eVb
λedge/nm
Eg/eVd
Similar dependence of the mobilities on the symmetry of the
core structure is observed for the diphenyl derivatives. The
mobility of 1c-based devices is among the highest (1.5 cm2 V-1
s-1) with a large Ion/Ioff ratio of 107, whereas 2c-based ones
showed lowest mobility (0.06 cm2 V-1 s-1) among the present
diphneyl NDT derivatives including previously reported 3c-
based ones (0.3 cm2 V-1 s-1).12
Even though the semiconducting molecules are isoelectronic,
the obvious differences between 1b/c- and 2b/c-based devices
are quite striking and should be related to the difference of the
symmetry, which may affect the intermolecular orbital overlap in
the solid state (vide infra). On the other hand, 4c-based FETs
showed high mobility (0.80 cm2 V-1 s-1), which is ca. three
times higher than that of the isoelectronic 3c-based one.12
Although the reasons for the difference between the FET
performances of 3c- and 4c-based devices are not very clear,
we expect that the extent of the intermolecular orbital overlap
through sulfur-involving nonbonded contacts in the solid state
can be related; since the NDT core of 4 has two sulfur atoms in
the molecular peripherally that face outwardly, whereas the sulfur
atoms in the core of 3 reside in the bay region of the molecule, the
former should be more favorable to the effective intermolecular
interaction than the latter through no-bonded contacts involving
sulfur atoms (vide infra).
1a
þ0.94
þ0.94
þ1.38
þ1.33
þ0.85
þ1.46
-5.3
-5.3
-5.8
-5.7
-5.2
-5.9
405
410
320c
350
490
330
3.0
3.1
3.9
3.5
2.5
3.8
2a
3a
4a
Naphthacene
Chrysene
a V vs Ag/AgCl. All the potentials were calibrated with the Fc/Fcþ (E1/2
= þ0.43 V measured under identical conditions). b Estimated with a
following equation: EHOMO (eV) = -4.4 - Eonset. c Forbidden triplet-
triplet bands were not considered (see ref 12). d Calculated from λedge
.
Another point that should be addressed here is that, although
the FET devices based on the centrocymmetric diphenyl deri-
vatives (1c and 4c) showed good performances comparable
with that of the BTBT counterpart (DPh-BTBT, μ ≈ 2.0 cm2
V-1 s-1),3d the FET mobilities of the octyl derivatives (1b and
4b) -based devices are fairly low relative to those based on C8-
BTBT (μ ≈ 2.9 cm2 V-1 s-1).3k To clarify these different
tendencies depending on the substituent on the NDT cores, we
then carried out structural investigations by means of XRD
measurements on the evaporated thin films.
Figure 3. Calculated HOMOs and LUMOs of 1a-4a with the TD-
DFT B3LYP/6-31(d) level. The calculated electronic transition energies
are 3.17 (1a), 3.20 (2a), 3.82 (3a), and 3.90 eV (4a).
Thin Films of Dioctyl (1b-4b) and Diphenyl Derivatives
(1c-4c). AFM images of evaporated thin films (Figure 5a-f)
show distinct surface morphology depending on the substituents
on the NDT cores. The thin films of the octyl derivatives
(1b-4b) have a rather flat surface without clear grain boundary,
whereas the surface of the thin films of diphenyl derivatives
(1c-4c) consists of granulose crystallites. The size of the grains
depends on Tsub for all the case (1c-4c) as observed for many
related organic semiconductors; for 1c and 2c, the grain are ∼0.1
μm in size for the thin films deposited at room temperature
relatively low oxidation potentials and good stability for the
development of new organic semiconductors.22
Thin Film FET Devices of Dioctyl (1b-4b) and Diphenyl
Derivatives (1c-4c). To evaluate the utility of 1, 2, and 4 as the
core structure for organic semiconductors, we fabricated FET
devices using vapor deposited thin films of the dioctyl- (1b, 2b,
and 4b) and diphenyl derivatives (1c, 2c, and 4c) with a bottom
gate, top-contact configuration on Si/SiO2 substrates. Depicted
in Figure 4 are the output- and transfer- characteristics of the
FETs, and their field-effect mobilities (μFET) extracted from the
saturation regime are summarized in Table 2.
As expected from the previous results on the related hetero-
arene-based OFET devices,3 both the dioctyl and diphenyl
derivatives showed typical p-channel FET responses. Moreover,
their FET characteristics were found to be largely dependent on
the core structures and substituents. In case of the dioctyl
derivatives, 1b and 4b with the centrosymmetic (C2h) cores
showed moderately high mobility (∼0.3 cm2 V-1 s-1), whereas
2b with the axisymmetric (C2v) core showed lower mobility
(∼0.015 cm2 V-1 s-1) than those of 1b and 4b by 1 order of
magnitude. Although the electronic structures of the NDT cores
are determined by the molecular shapes, that is, the linear-shaped
1a and 2a are very similar to each other, but quite different
from that of the angular-shaped 4a, the transport properties in
the FET devices are not the case. Rather than the molecular
shape, the symmetry of the molecule seems to be an important
(section 8, Supporting Information), whereas ∼0.2 μm at Tsub
=
100 °C (Figure 5d and e). Apparently, the larger grains con-
tribute to enhance μFET both for 1c and 2c (Table 2). On the
other hand, when 4c was deposited at high Tsubs, larger but
noncontinuous crystallites were observed (section 8, Supporting
Information), which is consistent with no field-effect response in
actual devices.
Out-of-plane XRD patterns of the evaporated thin films
(Figure 6a-f) show that these materials afford crystalline thin films
with the lamella-like layered structure on the Si/SiO2 substrate. The
calculated interlayer spacing (d-spacing), molecular length obtained
from the optimized molecular structure computed with the MO-
PAC-PM3 calculations, and the estimated molecular inclination
from the substrate normal of 1b-4b and 1c-4c are listed in
Table 3, together with their isomeric BTBT counterparts, C8-
BTBT and DPh-BTBT.
5027
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
Figure 4. Output and Transfer characteristics of FETs fabricated on OTS-treated Si/SiO2 substrate (Tsub = rt unless otherwise mentioned): (a) 1b,
(b) 2b, (c) 4b, (d) 1c (Tsub = 100 °C), (e) 2c (Tsub = 100 °C), and (f) 4c.
Figure 5. AFM images (2 ꢀ 2 μm) of evaporated thin films deposited on OTS-treated Si/SiO2 substrate: (a) 1b, (b) 2b, (c) 4b, (d) 1c (Tsub = 100 °C),
(e) 2c (Tsub = 100 °C), and (f) 4c.
The dioctyl derivatives of NDTs (1b, 2b, 4b) tend to show
larger inclination (35-40°) than that of C8-BTBT (30°), which
is consistent with the observed lower mobilities of the formers
(∼0.3 cm2 V-1 s-1) than that of the latter (∼2.9 cm2 V-1 s-1).23
On the other hand, despite the fact that the mobilities of the
vapor-processed devices are quite different between 1b and 2b by
1 order of magnitude, very similar thin film structures as well as
molecular orientations are proposed for these isoelectronic NDT
5028
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
derivatives by their XRD patterns. Theses contradictory results
can be interpreted by the distinct intermolecular interaction
caused by their different symmetry. In 1b and 2b, the relative
positions of sulfur atoms with a large atomic radius, which
influence the continuous intermolecular orbital overlap en-
hanced by sulfur-induced intermolecular contacts, are different;
the centrosymmetric 1b can afford delocalized intermolecular
interaction over a wide range of molecular array through such
intermolecular interactions, whereas axisymmetric 2b containing
the sulfur atoms at the same side of the molecular periphery tends
to afford much localized intermolecular interaction between
neighboring two molecules. This speculation is supported by
their absorption spectra in the thin film state (Figure 7a); the
centrosymmetric 1b thin film shows a much-pronounced red
shift than that for the 2b thin film.
Different from the octyl derivatives, the molecules of the
diphenyl derivatives tend to stand upright in the thin film state
regardless of the central heteroarene core. Among these, very
similar molecular ordering structure with almost the same
interlayer distances (d = ca. 20 Å) are suggested for the 1c and
2c thin films as in the case for previously discussed 1b and 2b.
Considering the large difference of the FET mobilities between
these two, it is speculated that the difference in the intermole-
cular interaction caused by the molecular symmetry again plays
crucial role for the transport properties in the thin film state. In
this case also, a distinct red-shifted absorption was observed for
the 1c thin film (Figure 7b), indicating the different extents of the
intermolecular interaction for 1c and 2c in the thin film state.
Single Crystal X-Ray Structural Analysis of 1c and 4c.
Single crystals of 1c and 4c with sufficient quality for X-ray
structural analysis were grown in careful recrystallization,
although single crystals of other compounds suitable for the
analysis were not obtained. Their packing structures are depicted
in Figures 7 and 8, respectively.
The simulated powder pattern from the bulk 1c crystal (Figure
S3, Supporting Information) well reproduces the XRD pattern of
the evaporated thin film of 1c (Figure 6d), suggesting that the
both phases are the same. The packing structure of 1c well
represents the typical organic semiconductor’s one, where the
planar 1c molecules pack efficiently in the layer-by-layer struc-
ture with the molecular long axis direction along the crystal-
lographic c-axis (Figure 8a). The 1c layer in the crystallographic
ab plane consists of a herringbone packing structure reminiscent
of high performance organic semiconductors such as pentacene
(Figure 8b). No strong intermolecular sulfur-sulfur contacts
shorter than the sum of van der Waals radii (∼3.6 Å) is observed
Table 2. FET Characteristicsa of 1b-, 2b-, 4b-, 1c-, 2c-, 3c-,
and 4c-Based OFETs Fabricated on OTS-Treated Si/SiO2
Substrate
d
compound
Tsubb/°C
μFETccm2 V-1 s-1
Ion/Ioff
Vth/V
1be
2be
4be,f
1c
rt
rt
0.20
0.015
0.30
0.50
0.80
1.5
106
106
105
105
107
107
106
106
107
-19
-21
-10
-5
rt
rt
60
100
rt
-5
-5
2c
0.030
0.050
0.060
0.30
0.80
-16
-21
-22
-32
-20
60
100
100
rt
3cf,g
4ce
105
106∼107
a FET devices were fabricated on octyltrichlorosilane (OTS)-treated
substrates with W/L = 30 unless otherwise stated. b Substrate tempera-
ture during vapor deposition of the semiconductors. The mobility values
are typical ones from more than ten devices tested. c extracted from the
saturation regime. d Ion and Ioff were the source-drain current measured
at Vg = -60 V and Vg = 0 V, respectively. e Vapor deposition at high Tsub
f
gave no continuous thin films. Data from the devices fabricated on
hexamethyldisilazane (HMDS)-treated Si/SiO2 substrate. g See ref 12.
Figure 6. Out-of-plane XRDs of the evaporated thin films deposited on OTS-treated Si/SiO2 substrate: (a) 1b, (b) 2b, (c) 4b, (d) 1c, (e) 2c, and (f) 4c.
5029
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
Table 3. Interlayer Spacings (d-Spacings) of Evaporated Thin Films on the Si/SiO2 Substrate of 1b, 2b, 4b, 1c, 2c, and 4c
ARTICLE
compound
2θ/°a
d-spacingsb/Å
molecular length (l)/Åc
inclination (θtilt) from the substrate normal/° d
1b
3.26
3.26
3.12
4.28
4.30
4.76
3.05
4.54
27.1
27.1
28.3
20.6
20.5
18.5
28.9
19.4
30.8
30.8
40
40
35
0
2b
4b
30.7
1c
20.2 (19.97)
20.0
2c
0
4c
19.0 (18.67)
30.8 (30.11)
19.7 (19.22)
0
C8-BTBTe
DPh-BTBTf
30
0
a (001) peaks. b Calculated interlayer spacings from the (001) reflections. c Obtained from optimized molecular geometry using MOPAC-PM3-MO
calculations. Values in parentheses are obtained from X-ray structural analyses. d Determined as θtilt = cos-1(d/l). e See also refs 3g and 3k. f See also
ref 3d.
thus speculate that such effective intermolecular interaction
between 4c molecules is also operative in the thin film phase
and thus can rationalize the higher mobility of 4c-based devices
than that of 3c.
Detailed Comparison of 1c and DPh-BTBT. Above discus-
sions based on the XRD studies gave qualitative understanding of
the structure-property relationships within the NDT-based
organic semiconductors and the rationale for the fact that 1c
gave the best FET performances among the present NDT-based
organic semiconductors. In fact, the mobility of 1c-based FETs as
high as 1.5 cm2 V-1 s-1 is comparable to those of the recently
developed superior organic semiconductors including DPh-
BTBT. Interestingly, the core structures of BTBT and NDT
Figure 7. Absorption spectra of thin films of (a) 1b and 2b and (b) 1c
and 2c.
are isomeric to each other with two thiophene and two benzene
rings fused in linear manners. Thus, it is very interesting to
compare these two isomeric high-performance organic semicon-
in the 1c layer, instead, there exists the molecular ribbon-like
networks with sulfur-sulfur (∼3.88 Å) and edge-to-face CH-π
(∼2.84 Å) contacts in the crystallographic (-1 1 0) and (1 1 0)
directions. We thus conclude that the pronounced red shift of 1c
thin film (Figure 7b) as well as the better mobility of 1c-based
FET than that of 2c should originate from the continuous
intermolecular interaction based on the centrosymmetric (C2h)
structure of 1c. Although the exact crystal structure of 2c is not
clear, the kind of intermolecular network structure could be very
difficult to occur for 2c with C2v symmetry.
Different from 1c, where the structures both in the thin film
and the bulk single crystal are likely the same, the single crystal
structure of 4c in the bulk crystal is apparently different from that
in the thin film, judged from the comparison of the XRD pattern
of the thin film (Figure 6f) and the simulated powder pattern
(Figure 9c). Thus, it is not appropriate to correlate directly
between the crystal structure in the bulk single crystal phase and
the transport properties in the thin film state. Nevertheless,
careful inspection of the packing structure in the bulk single
crystal of 4c would be beneficial to understand its structural
features in the solid state. In the actual crystal structure, the 4c
molecules form stacking structures along the crystallographic
b-axis direction with an interplanar separation of ca. 3.3 Å
(Figure 9a). In addition, the molecular stacks interact to one
another via sulfur-sulfur contacts (3.71 Å) with the edge-to-face
manner (Figure 9b). The latter interactive structure via sulfur
atoms should be characteristic of 4c with two sulfur atoms
pointing outward, since the isoelectronic 3c with the sulfur
atoms in the molecular bay region did not show such interactive
short S-S contacts (Figure S4, Supporting Information).12 We
ductors in detail, which will give us an insight into the electronic
structures both at the solid state as well as the molecular levels.
First, to qualitatively understand the electronic structures in
the solid state, we carried out calculations of transfer integrals (t)
between HOMOs in the semiconducting layer elucidated by the
single crystal X-ray analysis using the Amsterdam Density
Functional (ADF) program package.24,25 Calculated transfer
integrals are shown in Figure 9. For 1c, both the transfer integrals
along the stacking direction designated as ta (19 meV) and those
in the side-by-side direction designated as tp and tq (61 and 58
meV, respectively) are quite large, indicating that the two-
dimensional strong intermolecular orbital coupling is realized
as observed for high performance organic semiconductors
(Figure 10a).26,27 Similarly, large transfer integrals both in the
stacking (ta: 34 meV) and transverse directions (tp: 65 meV) are
also calculated for DPh-BTBT (Figure 10b). These results are
quite reasonable and consistent with their similar high mobilities.
However, detailed comparison of these two implies that the
difference of the core structure may cause the subtle change in
the solid-state electronic structure. When compared the transfer
integrals in the stacking (ta) and the transverse directions (tp and
tq), ta in the 1c structure is ca. one-third of tp or tq, indicating that
moderate anisotropy for 1c could be likely. In case of the DPh-
BTBT structure, ta:tp ratio is 1:2, implying the smaller anisotropy
for DPh-BTBT than that for 1c. Note that the BTBT core
structure has two sulfur atoms in the center of the molecule,
facilitating direct sulfur-sulfur intermolecular contacts. In fact,
nonbonded intermolecular sulfur-sulfur distance in DPh-BTBT
crystal is 3.51 Å, which is much shorter than the sum of the van
deer Waals radii of sulfur atom (3.60 Å), in the molecular stacks
5030
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
Figure 8. Packing structure of 1c representing the typical molecular lamella structure (a) with a herringbone packing motif (b). Face-to-edge molecular
array with nonbonded weak sulfur-sulfur contacts (dotted line: 3.86-3.88 Å) and edge-to-face CH-π contacts (solid line: 2.83-2.84 Å) also exist (c).
Figure 9. Packing structure of 4c: (a) stacking structure along the b-axis direction, (b) b-axis projection (dotted line: 3.71 Å), and (c) simulated powder
pattern from the bulk single crystal of 4c (blue solid line), which is apparently different from that of the evaporated thin film (red dotted line, Figure 6f).
along the crystallographic a-axis direction, which certainly con-
tributes to the larger ta than that in the 1c. In contrast, no such
short sulfur-sulfur contact exists in the 1c crystal structure,
thereby resulting in small ta. Thus, the large ta and thereby small
anisotropy in the solid-state electronic structure of DPh-BTBT
can be boiled down to the molecular structure level.28
Another important parameter that may potentially affect the
transport properties is molecular reorganization energy (λ).29 In the
present p-channel organic semiconductors’ cases, λ for hole (λh)
should be in concern, and λhs for the cores, that is, 1a and the parent
BTBT, are calculated to be 105 and 226 meV, respectively.10,19 The
smaller λh for 1a than that of BTBT can be qualitatively understood
Figure 10. Calculated transfer integrals of the semiconducting layer of
(a) 1c and (b) DPh-BTBT.
by the consideration that the 1a has an acene-like electronic
structure that is an ideal structure for minimizing λ.10,30 From the
viewpoint of λh, the core structure of 1a should be advantageous for
efficient carrier transport. However, in case of the diphenyl deriva-
tives, the difference in λh is much smaller than the parent (λh; 1c:
170 meV, DPh-BTBT: 233 meV), implying that the effect from the
difference of λh may be rather marginal.
Finally, we shall mention their molecular electronic structures
that determine their HOMO energy levels. As already discussed,
the NDT core (1a) is isoelectronic with naphthacene, whereas
the BTBT core is with chrysene.18 Since 1c is hardly soluble to
common organic solvents, it is not possible to estimate its
5031
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
Figure 11. FET characteristics of (a) DPh-BTBT- and (b) 1c-based devices in the linear (Vd = -5 V) and saturated regimes (Vd = -60 V) (insets:
output characteristics at the low Vd region).
HOMO energy level with electrochemical methods. Instead, we
measured the ionization potentials (IP) of the evaporated thin
films of 1c and DPh-BTBT with photoemission yield spectros-
copy in air. The IP of DPh-BTBT (5.6 eV) is relatively high to
ordinary organic semiconductors such as pentacene (5.0 eV).3d
In contrast, the IP of 1c is determined to be 5.1 eV, which is very
close to that of pentacene. It has been pointed out that large IP
(i.e., deep HOMO energy level) often causes large contact
resistance between the gold source electrode (work function:
∼5.0 eV) and the semiconducting layer that gives rise to poor
FET characteristics in the linear regime,31 although it brings
enhanced air-stability.32 In fact, the FET characteristics of DPh-
BTBT-based devices in the linear regime (Vd = -5 V) are much
influenced by the large contact resistance; the FET mobility in
the linear regime (μlin: ∼0.1 cm2 V-1 s-1) is less than one tenth
of that in the saturated regime (μsat: ∼1.5 cm2 V-1 s-1), and
nonlinear behavior at the low Vd region was observed
(Figure 11a). In contrast, such effects from the contact resistance
are not siginificant for the 1c-based devices (Figure 11b); μlin:
1.0 cm2 V-1 s-1 and almost linear increase of Id at the low Vd
region. This can be ascribed to smaller IP of 1c than that of DPh-
BTBT. In addition to the small contact resistance, 1c-based
devices showed fairly good air stability; even after storage under
ambient lab conditions during several months (Figure S5,
Supporting Information), the devices showed basically no de-
gradation at all. As a result, 1c is a potential organic semicon-
ductor that can give high mobility FETs with good air stability
and low contact resistance.
to determine the molecular electronic structure of the resulting
heteroarenes.
In contrast to the molecular properties, the device character-
istics are fairly influenced by the symmetry of the molecules, and
centrosymmetric 1 and 4 gave high performance OFETs with
mobilities higher than 0.1 cm2 V-1 s-1. Among the present
NDT-based organic semiconductors, the diphenyl derivative of
the linear-shaped, centrosymmetric one (1c) gave the best
devices showing mobilities of up to 1.5 cm2 V-1 s-1 with Ion/
Ioff of 107. The high mobility of 1c-based transistor is well
interpreted by the elucidation of the solid-state electronic
structure based on the molecular arrangement. Not only the
high mobility, its good performances were preserved for more
than one month under ambient conditions, showing the air-
stability of the OFET devices as well as the material itself.
Moreover, facile carrier injection from the gold electrode into
the semiconducting layer and thus fairly good FET character-
istics in the linear regime were observed for the 1c-based devices,
reflecting the lower IP of the semiconductor (5.1 eV) than that of
isomeric DPh-BTBT (5.6 eV) showing comparable FET mobi-
lity to that of 1c. These are important experimental evidence that
indicates the linear-shaped, centrosymmetric NDT (1) is a highly
promising core structure for developing new organic semicon-
ducting materials that simultaneously assume high stability and
good carrier injection/transport properties. For these reasons,
further material developments based on 1, both for small
molecules and polymeric materials, by making use of the
chemically reactive R positions of the thiophene moieties, are
now actively investigated in our group.
’ CONCLUSION
’ EXPERIMENTAL SECTION
In summary, we have successfully established selective and
straightforward synthetic routes to linear-and angular-shaped
NDTs (1, 2, and 4). Physicochemical evaluation of the parent
NDTs revealed that linear-shaped 1a and 2a have lower oxida-
tion potentials than those of angular-shaped 3a and 4a, reflecting
that 1a and 2a are formally isoelectronic with naphthacene,
whereas 3a and 4a are with chrysene. These results indicate that
the direction of thiophene rings fused on the naphthalene core
(i.e., anti vs syn at the 2,3/6,7-fused position or 1,2-b vs 2,1-b
annulation at the 1,2/5,6-fused position) does not largely
affect to the electronic structures, whereas the position of
thiophene rings (i.e., 1,2/5,6-fused vs 2,3/6,7-fused) is crucial
Synthesis. General. All chemicals and solvents are of reagent grade
unless otherwise indicated. THF was purified with a standard distillation
procedure prior to use. 1,3,5,7-Tetrabromo-2,6-dihydroxynaphtha-
lene,14 3,6-dibromo-2,7-dihydroxynaphthalene,15 2,6-dibromo-1,5-
dihydroxynaphthalene16 were synthesized as reported. Melting points
were uncorrected. All reactions were carried out under nitrogen atmo-
sphere. Nuclear magnetic resonance spectra were obtained in deuterated
chloroform (CDCl3) with TMS as internal reference unless otherwise
stated; chemical shifts (δ) are reported in parts per million. IR spectra
were recorded using a KBr pellet for solid samples. EI-MS spectra were
obtained using an electron impact ionization procedure (70 eV).
5032
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
3,7-Dibromo-2,6-dihydroxynaphthalene (9). To a suspen-
sion of 1,3,5,7-tetrabromo-2,6-dihydroxynaphthalene14 (1.0 g, 2.1
mmol) in glacial acetic acid (20 mL) was added mossy tin (499 mg,
4.2 mmol) and the mixture was refluxed for 62 h. After cooling to room
temperature, water (20 mL) was added to the mixture. The resulting
precipitate was collected by filtration and washed with water to give 2,6-
dibromo-3,7-dihydroxynaphthalene (9, 530 mg, 79%) as a white solid.
Mp 243-245 °C; 1H NMR (400 MHz, CDCl3) δ 5.58 (s, 2H), 7.25 (s,
2H), 7.89 (s, 2H); 13C NMR (100 MHz, DMSO-d6) δ 109.1, 113.6,
128.8, 129.8, 149.9; EI-MS (70 eV) m/z = 316 (Mþ); IR (KBr) ν = 3265
(OH) cm-1; Anal. Calcd for C10H6Br2O2: C, 37.77; H, 1.90%. Found:
C, 37.77; H, 1.77%.
14.3, 19.8, 22.8, 28.7, 29.1, 29.3, 29.4, 32.0, 79.4, 96.9, 123.4, 124.8,
130.3, 131.4, 131.6; EI-MS (70 eV) m/z = 558 (Mþ); Anal. Calcd for
C30H38Br2: C,64.52; H, 6.86%. Found: C, 64.59; H, 6.87%.
2,6-Dibromo-3,7-bis(phenylethynyl)naphthalene (11c).
Eighty-two percent yield. Mp 213-214 °C; 1H NMR (400 MHz,
CDCl3) δ 7.39-7.41 (m, 6H), 7.62-7.64 (m, 4H) 7.97 (s, 2H), 8.07
(s, 2H); 13C NMR (100 MHz, CDCl3) δ 88.1, 95.3, 122.8, 123.4, 124.6,
128.6, 129.1, 130.7, 131.6, 132.0; EI-MS (70 eV) m/z = 486 (Mþ); Anal.
Calcd for C26H14Br2: C,64.23; H, 2.90%. Found: C, 64.11; H, 2.60%.
3,6-Dibromo-2,7-bis(trimethylsilylethynyl)naphthalene
(14a). Nineteen percent yield. Mp 168-170 °C; 1H NMR (270 MHz,
CDCl3) δ 0.30 (s, 18H), 7.90 (s, 2H), 7.95 (s, 2H); 13C NMR (100
MHz, CDCl3) δ -0.1, 100.9, 102.8, 123.8, 124.2, 130.0, 130.0, 133.0,
133.6; EI-MS (70 eV) m/z = 478 (Mþ); Anal. Calcd for C20H22Br2Si2:
C,50.22; H, 4.64%. Found: C, 49.93; H, 4.45%.
3,6-Dibromo-2,7-di(decyn-1-yl)naphthalene (14b). Eighty
percent yield. Mp 50-51 °C; 1H NMR (270 MHz, CDCl3) δ 0.89 (t, J =
6.8 Hz, 6H), 1.27-1.72 (m, 24H), 2.50 (t, J = 6.9 Hz, 4H) 7.81 (s, 2H),
7.93 (s, 2H); 13C NMR (100 MHz, CDCl3) δ 14.3, 19.8, 22.8, 28.7, 29.1,
29.3, 29.4, 32.0, 79.4, 96.5, 124.0, 124.5, 129.7, 130.3, 132.0, 132.8; EI-
MS (70 eV) m/z = 558 (Mþ); Anal. Calcd for C30H38Br2: C,64.52; H,
6.86%. Found: C, 64.49; H, 6.74%.
General Procedure for the Triflation of Dibromodihydrox-
ynaphthalene. To a suspension of dibromodihydroxynaphthalene
(3.0 g, 9.4 mmol), pyridine (4.5 mL, 56 mmol) in dichloromethane
(90 mL) was slowly added trifluoromethanesulfonic anhydride (3.3 mL,
21 mmol) at 0 °C. After the mixture was stirred for 4 h at room
temperature, water (10 mL) and hydrochloric acid (1 M, 10 mL) were
added. The resulting mixture was extracted with dichloromethane
(30 mL ꢀ 3), and combined organic layer was dried (MgSO4) and
concentrated in vacuo. The residue was purified by column chromatog-
raphy on silica gel eluted with dichloromethane (Rf = 0.95) to give
dibromobis(trifluoromethanesulfonyloxy)naphthalene as a white solid.
3,7-Dibromo-2,6-bis(trifluoromethanesulfonyloxy)naph-
thalene (10). Seventy-one percent yield. Mp 138-139 °C; 1H NMR
(270 MHz, CDCl3) δ 7.14 (s, 2H), 8.25 (s, 2H); 13C NMR (68.5 MHz,
CDCl3) δ 116.2, 118.8 (q, 2JC-F = 318.8 Hz, CF3), 120.0, 131.9, 133.9,
3,6-Dibromo-2,7-bis(phenyletynyl)naphthalene (14c). Sixty-
1
nine percent yield. Mp 164-165 °C; H NMR (270 MHz, CDCl3)
δ 7.38-7.42 (m, 6H), 7.62-7.65 (m, 4H), 8.01 (s, 2H), 8.03 (s, 2H); 13C
NMR (100 MHz, CDCl3) δ 88.0, 94.8, 122.8, 124.0, 124.2, 128.6, 129.0,
130.1, 130.3, 131.9, 132.4, 133.5; EI-MS (70 eV) m/z = 486 (Mþ); Anal.
Calcd for C26H14Br2: C,64.23; H, 2.90%. Found: C, 64.33; H, 2.67%.
145.8; EI-MS (70 eV) m/z
=
580 (Mþ); Anal. Calcd for
C12H4Br2F6O6S2: C, 24.76; H, 0.69%. Found: C, 24.80; H, 0.29%.
3,6-Dibromo-2,7-bis(trifluoromethanesulfonyloxy)naph-
2,6-Dibromo-1,5-bis(trimethylsilylethynyl)naphthalene
1
1
(17a). Thirty-four percent yield. Mp 201-202 °C; H NMR (270
thalene (13). Sixty percent yield. Mp 123-124 °C; H NMR (400
MHz, CDCl3) δ 7.86 (s, 2H), 8.19 (s, 2H); 13C NMR (100 MHz,
CDCl3) δ 116.5, 118.8 (q, 2JC-F = 319.3 Hz, CF3), 121.1, 131.3, 132.5,
132.8, 145.7; EI-MS (70 eV) m/z = 580 (Mþ); Anal. Calcd for
C12H4Br2F6O6S2: C, 24.76; H, 0.69%. Found: C, 25.03; H, 0.39%.
2,6-Dibromo-1,5-bis(trifluoromethanesulfonyloxy)naphthale-
ne (15). Fifty-eight percent yield. Mp 159-160 °C; 1H NMR (270 MHz,
CDCl3) δ 7.89 (d, J = 9.2 Hz 2H), 8.03 (d, J = 9.2 Hz 2H); 13C NMR (100
MHz, CDCl3) δ 116.8, 118.8 (q, 2JC-F = 319.4 Hz, CF3), 123.0, 128.8, 133.3,
142.7; EI-MS (70 eV) m/z = 580 (Mþ); Anal. Calcd for C12H4Br2F6O6S2: C,
24.76; H, 0.69%. Found: C, 25.03; H, 0.35%.
MHz, CDCl3) δ 0.35 (s, 18H), 7.71 (d, 2H, J = 8.8 Hz), 8.14 (d, 2H, J =
8.8 Hz); 13C NMR (100 MHz, CDCl3) δ 0.1, 100.8, 106.6, 123.0, 126.0,
128.1, 131.5, 133.1 EI-MS (70 eV) m/z = 478 (Mþ); Anal. Calcd for
C20H22Br2Si2: C,50.22; H, 4.64%. Found: C, 49.98; H, 4.49%.
2,6-Dibromo-1,5-di(decyn-1-yl)naphthalene (17b). Sixty-
1
one percent yield. Mp 70-71 °C; H NMR (270 MHz, CDCl3) δ
0.89 (t, 6H, J = 6.8 Hz), 1.21-1.76 (m, 24H), 2.62 (t, 4H, J = 7.0 Hz),
7.67 (d, 2H, J = 9.3 Hz), 8.11 (d, 2H, J = 9.3 Hz); 13C NMR (100 MHz,
CDCl3) δ 14.2, 20.1, 22.8, 28.8, 29.1, 29.3, 29.4, 32.0, 77.5, 102.0, 123.8,
125.1, 127.2, 131.1, 133.3; EI-MS (70 eV) m/z = 558 (Mþ); Anal. Calcd
for C30H38Br2: C,64.52; H, 6.86%. Found: C, 64.28; H, 6.84%.
2,6-Dibromo-1,5-bis(phenyletynyl)naphthalene (17c). Fifty-
General Procedure for the Palladium-Catalyzed Sonoga-
shira Coupling of Dibromobis(trifluoromethanesulfonylo-
xy)naphthalene with Terminal Alkynes. To a degassed solution
of dibromobis(trifluoromethanesulfonyloxy)naphthalene (582 mg, 1.0
mmol) in DMF (7 mL) and diisopropylamine (7 mL) was added
Pd(PPh3)2Cl2 (70 mg, 0.05 mmol, 10 mol %), CuI (38 mg, 0.1 mmol,
20 mol %) and terminal alkyne (2.0 mmol). After the mixture was stirred
for 11 h at room temperature, water (1 mL) and hydrochloric acid (1 M,
1 mL) were added. The resulting mixture was extracted with dichlor-
omethane (5 mL ꢀ 3), and the combined organic layer was dried
(MgSO4) and concentrated in vacuo. The residue was purified by
column chromatography on silica gel eluted with hexane to give
dibromodiethynylnaphthalene analogue as a white solid.
2,6-Dibromo-3,7-bis(trimethylsilylethynyl)naphthalene
(11a). Forty-one percent yield. Mp 196-197 °C; 1H NMR (270 MHz,
CDCl3) δ 0.30 (s, 18H), 7.87 (s, 2H), 7.97 (s, 2H); 13C NMR (100
MHz, CDCl3) δ -0.1, 101.4, 102.8, 123.2, 124.5, 130.6, 131.9, 132.2;
EI-MS (70 eV) m/z = 478 (Mþ); Anal. Calcd for C20H22Br2Si2:
C,50.22; H, 4.64%. Found: C, 50.43; H, 4.65%.
1
three percent yield. Mp 223-224 °C; H NMR (270 MHz, CDCl3) δ
7.42-7.44 (m, 6H), 7.69-7.72 (m, 4H), 7.79 (d, 2H, J = 8.9 Hz), 8.27 (d,
2H, J = 8.9 Hz); 13C NMR (125 MHz, CDCl3) δ 86.1, 100.1, 122.8, 123.2,
125.8, 127.8, 128.7, 129.2, 131.5, 131.9, 133.0; EI-MS (70 eV) m/z = 486
(Mþ); Anal. Calcd for C26H14Br2: C,64.23; H, 2.90%. Found: C, 64.05;
H, 2.56%.
General Procedure for the Cyclization of Dibromodiethy-
nylnaphthalene Analogues. A suspension of Na2S 9H2O (202
3
mg, 0.84 mmol) in NMP (6 mL) was stirred for 15 min at room
temperature. The mixture was added dibromodiethynylnaphthalene
analogue (0.2 mmol), and was heated at 185 °C. After the mixture
was stirred for 12 h at the same temperature, the resulting mixture was
added into saturated aqueous ammonium chloride (20 mL). The
resulting precipitate was collected by filtration and washed with water,
methanol and hexane. The residue was purified by vacuum sublimation
to give naphthodithiophene derivative as a pale yellow solid.
Naphtho[2,3-b:6,7-b0]dithiophene (1a). Eighty-five percent
1
2,6-Dibromo-3,7-di(decyn-1-yl)naphthalene (11b). Eighty-
seven percent yield. Mp 98-99 °C; 1H NMR (270 MHz, CDCl3) δ 0.89
(t, J = 7.0 Hz, 6H), 1.27-1.37 (m, 20H), 1.61-1.72 (m, 4H), 2.51 (t, J =
6.6 Hz, 4H) 7.79 (s, 2H), 7.95 (s, 2H); 13C NMR (100 MHz, CDCl3) δ
yield. Mp > 300 °C; H NMR (270 MHz, CDCl3) δ 7.43 (d, J = 5.8
Hz, 2H), 7.51 (d, J = 5.8 Hz, 2H), 8.41 (s, 2H), 8.52 (s, 2H); EI-MS (70
eV) m/z = 240 (Mþ); Anal. Calcd for C14H8S2: C,69.96; H, 3.35%.
Found: C, 69.76; H, 2.99%.
5033
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
2,7-Dioctylnaphtho[2,3-b:6,7-b0]dithiophene (1b). Seventy-
eight percent yield. Mp 269-271 °C; 1H NMR (400 MHz, CDCl3) δ
0.89 (t, J = 7.4 Hz, 6H), 1.28-1.50 (m, 20H), 1.75-1.83 (m, 4H), 2.92 (t,
J = 7.4 Hz, 4H), 7.06 (s, 2H), 8.16 (s, 2H), 8.32 (s, 2H); EI-MS (70 eV)
m/z = 464 (Mþ); Anal. Calcd for C30H40S2: C,77.53; H, 8.67%. Found:
C, 77.48; H, 8.46%.
synchrotron (λ = 1.00000 Å) at beamline BL-8B of Photon Factory
(PF), High Energy Accelerator Research Organization (KEK). Crystal-
lographic data for DPh-BTBT: Crystallographic data for DPh-BTBT:
C26H16S2 (392.53), colorless plate, 0.10 ꢀ 0.10 ꢀ 0.01 mm3, monoclinic,
space group, P21/c (#14), a = 6.334(2), b = 7.462(5), c = 19.516(2) Å, β
= 93.06(3)°, V = 921.1(7) Å3, Z = 2, R = 0.0843 for 1284 observed
reflections (I > 2σ(I)) and 152 variable parameters, wR2 = 0.2377 for
all data.
Fabrication and Evaluation of FET Devices. OFETs were
fabricated in a “top-contact” configuration on a heavily doped nþ-Si
(100) wafer with a 200 nm thermally grown SiO2 (Ci = 17.3 nF cm-2).
The substrate surfaces were treated with octyltrichlorosilane (OTS) or
hexamethyldisilazane (HMDS) as reported previously.3d A thin film
NDT derivatives as the active layer was vacuum-deposited on the Si/
SiO2 substrates maintained at various temperatures (Tsub) at a rate of 1 Å
2,7-Diphenylnaphtho[2,3-b:6,7-b0]dithiophene (1c). Eighty-
six percent yield. Mp > 300 °C; EI-MS (70 eV) m/z = 392 (Mþ); Anal.
Calcd for C26H16S2: C,79.55; H, 4.11%. Found: C, 79.55; H, 3.90%.
Naphtho[2,3-b:7,6-b0]dithiophene (2a). Sixty-eight percent
1
yield. Mp >300 °C; H NMR (270 MHz, CDCl3) δ 7.43 (d, J = 5.5
Hz, 2H), 7.50 (d, J = 5.5 Hz, 2H), 8.45 (s, 2H), 8.47 (s, 2H); EI-MS (70
eV) m/z = 240 (Mþ); Anal. Calcd for C14H8S2: C,69.96; H, 3.35%.
Found: C, 69.77; H, 3.15%.
2,7-Dioctylnaphtho[2,3-b:7,6-b0]dithiophene (2b). Ninety-
s
-1 under a pressure of ∼10-3 Pa. On top of the organic thin film, gold
1
five percent yield. Mp 240-241 °C; H NMR (400 MHz, CDCl3) δ
films (80 nm) as drain and source electrodes were deposited through a
shadow mask. For a typical device, the drain-source channel length (L)
and width (W) are 50 μm and 1.5 mm, respectively. Characteristics of
the OFET devices were measured at room temperature under ambient
conditions with a Keithley 4200 semiconducting parameter analyzer.
Field-effect mobility (μFET) was calculated in the saturation (Vd = -60 V)
or linear regime (Vd = -5 V) of the Id using the following equation,
0.88 (t, J = 7.0 Hz, 6H), 1.28-1.81 (m, 24H), 2.92 (t, J = 7.3 Hz, 4H),
7.05 (s, 2H), 8.21 (s, 2H), 8.26 (s, 2H); EI-MS (70 eV) m/z = 464
(Mþ); Anal. Calcd for C30H40S2: C,77.53; H, 8.67%. Found: C, 77.32;
H, 8.69%.
2,7-Diphenylnaphtho[2,3-b:7,6-b0]dithiophene (2c). Seventy-
one percent yield. Mp >300 °C; EI-MS (70 eV) m/z = 392 (Mþ); Anal.
Calcd for C26H16S2: C,79.55; H, 4.11%. Found: C, 79.32; H, 4.15%.
Naphtho[2,1-b:6,5-b0]dithiophene (4a). Sixty percent yield.
Mp 274-275 °C; 1H NMR (270 MHz, CDCl3) δ 7.43 (d, 2H, J = 5.4
Hz), 8.05 (d, 2H, J = 5.5 Hz), 8.05 (d, 2H, J = 8.9 Hz), 8.30 (d, 2H, J = 8.9
Hz); 13C NMR (125 MHz, CDCl3) δ 120.9, 121.2, 122.4, 126.4, 126.7,
136.8, 137.0; EI-MS (70 eV) m/z = 240 (Mþ); Anal. Calcd for C14H8S2:
C,69.96; H, 3.35%. Found: C, 70.10; H, 2.95%.
2
Saturation regime : Id ¼ CiμFETðW=2LÞðVg - VthÞ
Linear regime : Id ¼ VdCiμFETðW=LÞðVg - VthÞ
where Ci is the capacitance of the SiO2 insulator, and Vg and Vth are the
gate and threshold voltages, respectively. Current on/off ratio (Ion/Ioff)
was determined from the Id at Vg = 0 V (Ioff) and Vg = -60 V (Ion). The
μFET data reported are typical values from more than ten different devices.
2,7-Dioctylnaphtho[2,1-b:6,5-b0]dithiophene (4b). Sixty-
1
seven percent yield. Mp 154-155 °C; H NMR (270 MHz, CDCl3)
δ 0.89 (t, 6H, J = 6.8 Hz), 1.21-1.76 (m, 24H), 2.62 (t, 4H, J = 7.0 Hz),
7.67 (d, 2H, J = 9.3 Hz), 8.11 (d, 2H, J = 9.3 Hz); 13C NMR (125 MHz,
CDCl3) δ 14.3, 22.8, 29.4, 29.4, 29.5, 31.1, 31.7, 32.0, 119.2, 119.9,
120.7, 126.3, 135.8, 137.0, 147.1; EI-MS (70 eV) m/z = 464 (Mþ);
Anal. Calcd for C30H40S2: C,77.53; H, 8.67%. Found: C, 77.64;
H, 8.85%.
’ ASSOCIATED CONTENT
S
Supporting Information. Instrumentation for the char-
b
acterization of compounds and physicochemical evaluation,
repeated voltammograms of 1a-4a, absorption spectra of 1a
and naphthacene in air saturated solution, crystal structure of 3c,
simulated powder patterns from the bulk single crystal structures
of 1c and DPh-BTBT, shelf-life time test of 1c-based FET,
calculated transfer integrals (t) of 3c and 4c based on the bulk-
single crystal structures, results on TD-DFT calculations of 1a-
4a, AFM images of evaporated thin film of 1c, 2c (Tsub = rt), and
4c (Tsub = 100 °C), complete ref 20, NMR spectra of com-
pounds, crystallographic information file (CIF) for 1c, 4c and
DPh-BTBT. This material is available free of charge via the
Internet at http://pubs.acs.org.
2,7-Diphenylnaphtho[2,1-b:6,5-b0]dithiophene (4c). Fifty-
six percent yield. Mp >300 °C; 1H NMR (400 MHz, CDCl3) δ 7.39-
7.40 (m, 2H), 7.47-7.51 (m, 4H), 7.82-7.84 (m, 4H), 8.01 (d, 2H, J =
8.6 Hz), 7.71 (s, 2H), 8.05 (d, 2H, J = 8.6 Hz); EI-MS (70 eV) m/z = 392
(Mþ); Anal. Calcd for C26H16S2: C,79.55; H, 4.11%. Found: C, 79.31;
H, 3.73%.
Single Crystal X-Ray Analysis. Single crystals of 1c and 4c
suitable for X-ray structural analysis were obtained by careful recrystalli-
zation from chlorobenzene. The X-ray crystal structure analyses were
made on a Rigaku Mercury-CCD (Mo KR radiation, λ = 0.71069 Å,
graphite monochromator, T = 296 K, 2θmax = 55.0°). The structure was
solved by the direct methods.33 Non-hydrogen atoms were refined
anisotropically, and hydrogen atoms were included in the calculations
but not refined. All calculations were performed using the crystal-
lographic software package TeXsan 1.2.34
Crystallographic data for 1c: C26H16S2 (392.53), yellow plate, 0.60 ꢀ
0.50 ꢀ 0.10 mm3, monoclinic, space group, P21/n (#14), a = 5.922(10),
b = 7.60(1), c = 40.63(7) Å, β = 93.753(6)°, V = 1825(5) Å3, Z = 4, R =
0.0949 for 2229 observed reflections (I > 2σ(I)) and 253 variable
parameters, wR2 = 0.3005 for all data.
Crystallographic data for 4c: C26H16S2 (392.53), colorless plate, 1.0
ꢀ 0.20 ꢀ 0.05 mm3, monoclinic, space group, P21/c (#14), a = 9.517(1),
b = 5.8072(8), c = 16.988(2) Å, β = 90.954(6)°, V = 938.8(2) Å3, Z = 2, R
= 0.0657 for 1563 observed reflections (I > 2σ(I)) and 127 variable
parameters, wR2 = 0.2223 for all data.
’ AUTHOR INFORMATION
Corresponding Author
ktakimi@hiroshima-u.ac.jp
’ ACKNOWLEDGMENT
This work was partially supported by a Grant-in-Aid for
Scientific Research (No. 20350088) from the Ministry of Educa-
tion, Culture, Sports, Science and Technology, Japan and The
Strategic Promotion of Innovative Research and Development
from the Japan Science and Technology Agency. Combustion
elemental analyses were carried out at the Natural Science Center
for Basic Research and Development (N-BARD), Hiroshima
University.
The X-ray crystal structure analysis of DPh-BTBT was made on a
Rigaku DSC imaging plate system by using Si-monochromated
5034
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035

Journal of the American Chemical Society
ARTICLE
’ REFERENCES
(10) Coropceanu, V.; Kwon, O.; Wex, B.; Kaafarani, B. R.; Gruhn,
N. E.; Durivage, J. C.; Neckers, D. C.; Bredas, J.-L. Chem.—Eur. J. 2006,
12, 2073–2080.
(11) (a) Umeda, R.; Fukuda, H.; Miki, K.; Rahman, S. M. A.; Sonoda,
M.; Tobe, Y. C. R. Chimie 2009, 12 378-384. see also (b) Tilak, B. D.
Proc. Indian Acad. Sci. Sect. A 1951, 33A, 71–77.
(12) Shinamura, S.; Miyazaki, E.; Takimiya, K. J. Org. Chem. 2010,
75, 1228–1234.
(13) Osaka, I.; Abe, T.; Shinamura, S.; Miyazaki, E.; Takimiya, K.
J. Am. Chem. Soc. 2010, 132, 5000–5001.
(14) Mataka, S.; Takahashi, K.; Ikezaki, Y.; Hatta, T.; Tori-i, A.;
(1) (a) Laquindanum, J. G.; Katz, H. E.; Lovinger, A. J.; Dodabalapur,
A. Adv. Mater. 1997, 8, 36–39. (b) Takimiya, K.; Kunugi, Y.; Konda, Y.;
Niihara, N.; Otsubo, T. J. Am. Chem. Soc. 2004, 126, 5084–5085. (d)
Takimiya, K.; Kunugi, Y.; Ebata, H.; Otsubo, T. Chem. Lett. 2006,
35, 1200–1201. (e) Kashiki, T.; Miyazaki, E.; Takimiya, K. Chem. Lett.
2008, 37, 284–285. (f) Kashiki, T.; Miyazaki, E.; Takimiya, K. Chem. Lett.
2009, 38, 568–569.
(2) (a) Laquindanum, J. G.; Katz, H. E.; Lovinger, A. J. J. Am. Chem.
Soc. 1998, 120, 664–672. (b) Payne, M. M.; Parkin, S. R.; Anthony, J. E.;
Kuo, C.-C.; Jackson, T. N. J. Am. Chem. Soc. 2005, 127, 4986–4987.
(c) Dickey, K. C.; Anthony, J. E.; Loo, Y. L. Adv. Mater. 2006,
18, 1721–1726. (d) Chen, M.-C.; Kim, C.; Chen, S.-Y.; Chiang, Y.-J.;
Chung, M.-C.; Facchetti, A.; Marks, T. J. J. Mater. Chem. 2008,
18, 1029–1036.
Tashiro, M. Bull. Chem. Soc. Jpn. 1991, 64, 68–73.
(15) (a) Cooke, R. G.; Johnson, B. L.; Owen, W. R. Aust. J. Chem.
1960, 13, 256–260. (b) Kitamura, C.; Ohara, T.; Kawatsuki, N.; Yoneda,
A.; Kobayashi, T.; Naito, H.; Komatsu, T.; Kitamura, T. CrystEngComm
2007, 9, 644–647.
(3) (a) 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. (b) Sirringhaus, H.; Friend, R. H.;
Wang, C.; Leuninger, J.; Mullen, K. J. Mater. Chem. 1999, 9, 2095–2101.
(c) Anthony, J. E. Chem. Rev. 2006, 106, 5028–5048. (d) Takimiya, K.;
Ebata, H.; Sakamoto, K.; Izawa, T.; Otsubo, T.; Kunugi, Y. J. Am. Chem.
Soc. 2006, 128, 12604–12605. (e) Takimiya, K.; Kunugi, Y.; Konda, Y.;
Ebata, H.; Toyoshima, Y.; Otsubo, T. J. Am. Chem. Soc. 2006,
128, 3044–3050. (f) Yamamoto, T.; Takimiya, K. J. Am. Chem. Soc.
2007, 129, 2224–2225. (g) Ebata, H.; Izawa, T.; Miyazaki, E.; Takimiya,
K.; Ikeda, M.; Kuwabara, H.; Yui, T. J. Am. Chem. Soc. 2007,
129, 15732–15733. (h) Takimiya, K.; Kunugi, Y.; Otsubo, T. Chem.
Lett. 2007, 36, 578–583. (i) Gao, J.; Li, R.; Li, L.; Meng, Q.; Jiang, H.; Li,
H.; Hu, W. Adv. Mater. 2007, 19, 3008–3011. (j) Anthony, J. E. Angew.
Chem., Int. Ed. 2008, 47, 452–483. (k) Izawa, T.; Miyazaki, E.; Takimiya,
K. Adv. Mater. 2008, 20, 3388–3392. (l) Gao, P.; Beckmann, D.; Tsao,
H. N.; Feng, X.; Enkelmann, V.; Baumgarten, M.; Pisula, W.; M€ullen, K.
Adv. Mater. 2009, 21, 213–216.
(4) (a) Pan, H.; Li, Y.; Wu, Y.; Liu, P.; Ong, B. S.; Zhu, S.; Xu, G.
Chem. Mater. 2006, 18, 3237–3241. (b) Pan, H.; Li, Y.; Wu, Y.; Liu, P.;
Ong, B. S.; Zhu, S.; Xu, G. J. Am. Chem. Soc. 2007, 129, 4112–4113. (c)
Pan, H.; Wu, Y.; Li, Y.; Liu, P.; Ong, B. S.; Zhu, S.; Xu, G. Adv. Funct.
Mater. 2007, 17, 3574–3579.
(5) (a) Liang, Y.; Wu, Y.; Feng, D.; Tsai, S.-T.; Son, H.-J.; Li, G.; Yu,
L. J. Am. Chem. Soc. 2009, 131, 56–57. (b) Liang, Y.; Feng, D.; Wu, Y.;
Tsai, S.-T.; Li, G.; Ray, C.; Yu, L. J. Am. Chem. Soc. 2009,
131, 7792–7799. (c) Huo, L.; Hou, J.; Zhang, S.; Chen, H. Y.; Yang,
Y. Angew. Chem. Inter. Ed. 2010, 49, 1500–1503. (d) Liang, Y.; Xu, Z.;
Xia, J.; Tsai, S.-T.; Wu, Y.; Li, G.; Ray, C.; Yu, L. Adv. Mater. 2010,
22, E135–E138. (e) Piliego, C.; Holcombe, T. W.; Douglas, J. D.; Woo,
C. H.; Beaujuge, P. M.; Frꢀechet, J. M. J. Am. Chem. Soc. 2010,
132, 7595–7597. (f) Zhang, Y.; Hau, S. K.; Yip, H.-L.; Sun, Y.; Acton,
O.; Jen, A. K. Y. Chem. Mater. 2010, 22, 2696–2698. (g) Zou, Y.; Najari,
A.; Berrouard, P.; Beauprꢀe, S.; Aïch, B. R.; Tao, Y.; Leclerc, M. J., M.
J. Am. Chem. Soc. 2010, 132, 5330–5331.
(16) Wheeler, A. S.; Ergle, D. R. J. Am. Chem. Soc. 1930, 52,
4872–4880.
(17) Wheeler, A. S.; Mattox, W. J. J. Am. Chem. Soc. 1933, 55, 686–690.
(18) Takimiya, K.; Yamamoto, T.; Ebata, H.; Izawa, T. Sci. Tech. Adv.
Mater. 2007, 8, 273–276.
(19) (a) Suresh, C. H.; Gadre, S. R. J. Org. Chem. 1999,
64, 2505–2512. (b) Poater, J.; Visser, R.; Sola, M.; Bickelhaupt, F. M.
J. Org. Chem. 2007, 72, 1134–1142.
(20) MO calculations were carried out with the DFT/TD-DFT
method at the B3LYP/6-31g(d) level using Gaussian 03 program
package. Frisch, M. J. et al. Gaussian 03, revision C.02; Gaussian, Inc.:
Wallingford, CT, 2004.
(21) Maliakal, A.; Raghavachari, K.; Katz, H.; Chandross, E.; Siegrist,
T. Chem. Mater. 2004, 16, 4980–4986.
(22) (a) Tang, M. L.; Okamoto, T.; Bao, Z. J. Am. Chem. Soc. 2006,
128, 16002–16003. (b) Valiyev, F.; Hu, W.-S.; Chen, H.-Y.; Kuo, M.-Y.;
Chao, I.; Tao, Y.-T. Chem. Mater. 2007, 19, 3018–3026. (c) Tang, M. L.;
Mannsfeld, S. C. B.; Sun, Y.-S.; Becerril, H. t. A.; Bao, Z. J. Am. Chem. Soc.
2009, 131, 882–883.
(23) Takimiya, K.; Kunugi, Y.; Toyoshima, Y.; Otsubo, T. J. Am.
Chem. Soc. 2005, 127, 3605–3612.
(24) ADF2008.01. SCM Theoretical Chemistry; Vrije Universiteit:
Amsterdam, The Netherlands, http://www.scm.com.
(25) (a) Senthilkumar, K.; Grozema, F. C.; Bickelhaupt, F. M.;
Siebbeles, L. D. A. J. Chem. Phys. 2003, 119, 9809–9817. (b) Prins, P.;
Senthilkumar, K.; Grozema, F. C.; Jonkheijm, P.; Schenning, A. P. H. J.;
Meijer, E. W.; Siebbeles, L. D. A. J. Chem. Phys. B 2005, 109,
18267–18274.
(26) (a) Valiyev, F.; Hu, W.-S.; Chen, H.-Y.; Kuo, M.-Y.; Chao, I.;
Tao, Y.-T. Chem. Mater. 2007, 19, 3018–3026. (b) Kim, E.-G.;
Coropceanu, V.; Gruhn, N. E.; Sanchez-Carrera, R. S.; Snoeberger, R.;
Matzger, A. J.; Bredas, J.-L. J. Am. Chem. Soc. 2007, 129, 13072–13081.
(27) These transfer integrals are also larger than those calculated
for 3c (∼25 meV) and 4c (∼11 meV) using the bulk single crystal
structures. See Supporting Information.
(6) Lloyd, M. T.; Mayer, A. C.; Subramanian, S.; Mourey, D. A.;
Herman, D. J.; Bapat, A. V.; Anthony, J. E.; Malliaras, G. G. J. Am. Chem.
Soc. 2007, 129, 9144–9149.
(7) (a) M€uhlbacher, D.; Scharber, M.; Morana, M.; Zhu, Z.; Waller,
D.; Gaudiana, R.; Brabec, C. Adv. Mater. 2006, 18, 2884–2889. (b) Hou,
J.; Chen, H.-Y.; Zhang, S.; Li, G.; Yang, Y. J. Am. Chem. Soc. 2008,
130, 16144–16145.
(28) Sanchez-Carrera, R. S.; Atahan, S.; Schrier, J.; Aspuru-Guzik, A.
J. Phys. Chem. C 2010, 114, 2334–2340.
(29) (a) Bredas, J.-L.; Beljonne, D.; Coropceanu, V.; Cornil, J. Chem.
Rev. 2004, 104, 4971–5004. (b) Sakanoue, K.; Motoda, M.; Sugimoto,
M.; Sakaki, S. J. Chem. Phys. A 1999, 103, 5551–5556.
(30) λhs for 2a-4a are 100, 266, and 245 meV, respectively,
reflecting their electronic structures.
(8) (a) Wang, C.-H.; Hu, R.-R.; Liang, S.; Chen, J.-H.; Yang, Z.; Pei,
J. Tetrahedron Lett. 2005, 46, 8153–8157. (b) Ebata, H.; Miyazaki, E.;
Yamamoto, T.; Takimiya, K. Org. Lett. 2007, 9, 4499–4502. (c) Zhou, Y.;
Liu, W.-J.; Ma, Y.; Wang, H.; Qi, L.; Cao, Y.; Wang, J.; Pei, J. J. Am. Chem.
Soc. 2007, 129, 12386–12387.
(31) (a) Kano, M.; Minari, T.; Tsukagoshi, K. Appl. Phys. Lett. 2009,
94, 143304. (b) Kobayashi, N.; Sasaki, M.; Nomoto, K. Chem. Mater.
2009, 21, 552–556.
(32) Meng, H.; Bao, Z.; Lovinger, A. J.; Wang, B.-C.; Mujsce, A. M.
J. Am. Chem. Soc. 2001, 123, 9214–9215.
(9) (a) Beimling, P.; Koβmehl, G. Chem. Ber. 1986, 119, 3198–3203.
(b) Takimiya, K.; Konda, Y.; Ebata, H.; Niihara, N.; Otsubo, T. J. Org.
Chem. 2005, 70, 10569–10571. (c) Kashiki, T.; Shinamura, S.; Kohara,
M.; Miyazaki, E.; Takimiya, K.; Ikeda, M.; Kuwabara, H. Org. Lett. 2009,
11, 2473–2475.
(33) Sheldrick, G. M. SHELXL (SHELX97). Programs for the refine-
ment of crystal structures. University of Goettingen: Germany, 1997.
(34) teXsan: Single Crystal Structure Analysis Software, Version 1.19;
Molecular Structure Corporation and Rigaku Corporation: 2000.
5035
dx.doi.org/10.1021/ja110973m |J. Am. Chem. Soc. 2011, 133, 5024–5035