Benzene substituted with bipyridine and terpyridine as electron- transporting materials for organic light-emitting devices

Article abstract of DOI:10.1039/c2jm16274a

New electron-transporting materials for organic light-emitting devices (OLEDs) based on trisubstituted benzene with both bipyridine and terpyridine, 1,3-bisbipyridyl-5-terpyridylbenzene (BBTB) and 1-bipyridyl-3,5- bisterpyridylbenzene (BTBB), were developed. Glass transition temperatures of BBTB and BTBB were 93 °C and 108 °C, respectively, and BTBB was completely amorphous with no melting point. Electron mobilities of BTBB exceeded the order of 10^-4 cm² V^-1 s^-1, while those of BBTB were very high and reached 10^-3 cm² V^-1 s^-1 at an electric field of approximately 500 kV cm^-2. These high mobilities contributed to a low voltage operation. For example, in the case of the conventional aluminum trisquinolinol (Alq)-based fluorescent OLED with BTBB, current densities of 3.5 mA cm^-2 and 100 mA cm^-2 were reached at voltages of 3.0 V and 4.5 V, respectively. In addition, ionization potentials of BBTB (6.33 eV) and BTBB (6.50 eV) were sufficiently large to confine holes in common emissive layers. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2jm16274a

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, 6765
www.rsc.org/materials
PAPER
Benzene substituted with bipyridine and terpyridine as electron-transporting
materials for organic light-emitting devices†
ab
Takayuki Yamamoto,a Hyeon-Gu Jeon,a Kouki Kase,c Shuichi Hayashi,c
*
Musubu Ichikawa,
Makoto Nagaokac and Norimasa Yokoyamac
Received 1st December 2011, Accepted 29th January 2012
DOI: 10.1039/c2jm16274a
New electron-transporting materials for organic light-emitting devices (OLEDs) based on trisubstituted
benzene with both bipyridine and terpyridine, 1,3-bisbipyridyl-5-terpyridylbenzene (BBTB) and
1-bipyridyl-3,5-bisterpyridylbenzene (BTBB), were developed. Glass transition temperatures of BBTB
and BTBB were 93 ꢀC and 108 ꢀC, respectively, and BTBB was completely amorphous with no melting
point. Electron mobilities of BTBB exceeded the order of 10ꢁ4 cm2 Vꢁ1 sꢁ1, while those of BBTB were very
high and reached 10ꢁ3 cm2 Vꢁ1 sꢁ1 at an electric field of approximately 500 kV cmꢁ2. These high mobilities
contributed to a low voltage operation. For example, in the case of the conventional aluminum
trisquinolinol (Alq)-based fluorescent OLED with BTBB, current densities of 3.5 mA cmꢁ2 and
100 mA cmꢁ2 were reached at voltages of 3.0 V and 4.5 V, respectively. In addition, ionization potentials
of BBTB (6.33 eV) and BTBB (6.50 eV) were sufficiently large to confine holes in common emissive layers.
higher Tg of 110 ꢀC, which is greater than the general criterion of
Introduction
thermal stability for practical use, but the electron mobility of the
material decreased to an order of 10ꢁ5 cm2 Vꢁ1 sꢁ1. Large bulk-
iness of a terpyridyl substituent might lead to a sufficient increase
in Tg, but it also leads to a sizable decrease in the electron
mobility. Here we demonstrate that asymmetrically substituting
benzene with bipyridine and terpyridine can be useful for over-
coming this trade-off, and we achieve a high mobility and Tg of
Organic light-emitting devices (OLEDs) are of potential interest
in applications such as flat televisions, mobile displays,1,2 lights,3
and optical communication light sources.4 OLEDs are usually
composed of functionally divided organic multilayers—for
example, hole-transporting (HT), emissive, and electron-trans-
porting (ET) layers.5,6 In the last decade, many kinds of
amorphous molecular semiconductor materials,7,8 HT
materials,9–16 and ET materials (ETMs)17–28 have been proposed.
Recently, it has been reported that pyridine-based materials
work well as ETMs with high electron mobility for OLEDs. For
example, Kido et al. reported that three compounds based on
triphenylbenzene with three pyridines were very good ETMs for
OLEDs, and in particular, the compound in which pyridine is
substituted at the 4 position showed extremely high electron
greater than 10ꢁ3 cm2 Vꢁ1 sꢁ1 and 100 C, respectively.
ꢀ
Experimental
Materials
Fig. 1 shows chemical structures of the newly developed mate-
rials, and Scheme 1 shows their synthetic routes; these materials
mobility above 10ꢁ3 cm2 Vꢁ1 sꢁ1 29
We also reported that
.
trisubstituted benzenes with bipyridine or terpyridine were good
ETMs, and tris(bipyridyl)benzene showed high electron mobility
above 10ꢁ4 cm2 Vꢁ1 sꢁ1 30
. However, the glass transition temper-
ature (Tg) of the materials was only 75 ꢀC and insufficient for
practical use. In contrast, tris(terpyridyl)benzene showed a much
aFunctional Polymer Science Course, Division of Chemistry and Materials,
Faculty of Textile Science and Technology, Shinshu University, 3-15-1
Tokida, Ueda, 386-8567, Japan
bPresto, Japan Science and Technology Agency (JST), 4-8-1 Honcho,
Kawaguchi, Saitama, 332-0012, Japan
cHodogaya Chemical Co., Ltd., 45 Miyukigaoka, Tsukuba, Ibaraki, 305-
0841, Japan
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c2jm16274a
Fig. 1 Chemical structures of BBTB and BTBB.
J. Mater. Chem., 2012, 22, 6765–6773 | 6765
This journal is ª The Royal Society of Chemistry 2012

View Article Online
were synthesized step-by-step by the Suzuki–Miyaura coupling
reaction with bipyridine and terpyridine bromides. Syntheses of
6-bromo[2,20]bipyridine and 6-bromo[2,20:60,200]terpyridine are
described in detail elsewhere.30 BBTB and BTBB were purified by
column chromatography and temperature-gradient sublimation
in a flow stream of pure Ar gas. Chemical structures were
confirmed by 1H NMR, 13C NMR, and mass spectroscopy. Other
OLED materials were obtained from industrial companies as
sublimation grades and used without further purification.
cyclohexane–toluene mixture as the eluent yielded the first
intermediate (1,3-dibromo-5-terpyridylbenzene, 2.45 g; 54.6%
yield; 98.4% purity by high-performance liquid chromatography
(HPLC)). Then, the obtained intermediate (7.98 g, 17.1 mmol)
was treated with bis(pinacolato)diboron (Pin2B2; 10.41 g, 41.0
mmol), potassium acetate (10.1 g, 103 mmol), and [1,10-bis-
(diphenylphosphino)ferrocene]dichloropalladium(II) (PdCl2(dppf);
0.83 g, 1.03 mmol) as the catalyst in dioxane (240 mL) at 80 ꢀC.
After 8 h, the reaction mixture was cooled to room temperature,
and then poured into a 3000 mL beaker filled with 1500 mL pure
water. After stirring for 40 min, a crude black powder (10.4 g)
was obtained. Adsorption purification with diol-bonded silica
(48 g) was carried out using a toluene (200 mL) solution of the
obtained crude powder. The solution was filtered to remove the
silica gel, distilled to remove toluene, washed with 150 mL
hexane, and dried at 60 ꢀC under vacuum for 1 h to give the
Synthesis
Synthesis of BBTB. First, (3,5-dibromophenyl)boronic acid
(2.69 g, 9.61 mmol) was treated with bromoterpyridine (3.0 g,
9.61 mmol) by the Suzuki–Miyaura coupling reaction using
K2CO3 (3.98 g, 28.83 mmol) with pure water (14.4 mL) as
the base, tetrakis(triphenylphosphine)palladium(0) (0.835 g,
0.72 mmol) as the catalyst, and a 4 : 1 toluene–ethanol mixture
(150 mL) as the solvent. The reaction mixture was refluxed with
stirring for 22 h and then cooled to room temperature. Pure
water was added and the mixture was stirred for 30 min. The
mixture was then extracted using toluene, dried with MgSO4,
filtered to remove precipitates, and distilled to remove toluene
giving a crude powder (5.64 g). Purification by silica-gel
(NH silica, 150 g) column chromatography using a 2 : 1
second
intermediate
(1,3-bis(pinacolatoboryl)-5-terpyr-
idylbenzene, 7.87 g; 82.0% yield; 90.4% purity by HPLC). The
purified second intermediate (7.50 g, 13.4 mmol) was treated with
bromobipyridine (6.96 g, 29.5 mmol) using K2CO3 (5.56 g, 40.2
mmol) with pure water (21 mL) as the base, tetrakis-
(triphenylphosphine)palladium(0) (0.77 g, 0.67 mmol) as the
catalyst, and a 4 : 1 toluene–ethanol mixture (10.4 g) as the
solvent. The reaction mixture was refluxed with stirring for 24 h
and then cooled to room temperature. Pure water was added and
the mixture was stirred for 30 min. Then, the mixture was
extracted using toluene, dried with MgSO4, filtered to remove
precipitates, and distilled to remove toluene giving a crude
powder (10.2 g). Adsorption purification by silica-gel column
chromatography (NH silica; 40 g) was carried out using a toluene
solution (150 mL) of the obtained crude powder, filtered to
remove the silica, and distilled to remove toluene. The obtained
ꢀ
solid was washed in hexane (200 mL) and then dried at 60 C
under vacuum for 2 h. Finally, purification by silica-gel (NH
silica, 240 g) column chromatography using a chloroform–
hexane mixture as the eluent, the ratio of which was changed
from 1 : 10 to 1 : 0 step by step, yielded BBTB (7.69 g; 97% purity
by HPLC). The powder was washed with 2-propyl ether
ꢀ
(200 mL), filtered, and dried at 60 C under vacuum for 2 h to
give pure BBTB (5.99 g; 70.0% yield; 98.6% purity by HPLC). 1H
NMR (CDCl3) dH (600 MHz, CDCl3): 9.03 (1H, d, J ¼ 1.4 Hz),
9.02 (2H, d, J ¼ 2.1 Hz), 8.79 (2H, d, J ¼ 7.6 Hz), 8.76 (1H, d, J ¼
7.6 Hz), 8.72 (3H, t, J ¼ 4.8 Hz), 8.67 (2H, d, J ¼ 7.6 Hz), 8.64
(2H, dd, J ¼ 7.6, 1.4 Hz), 8.49 (2H, dd, J ¼ 7.6, 1.4 Hz), 8.46 (1H,
dd, J ¼ 7.6, 1.4 Hz), 8.02–7.95 (8H, m), 7.88–7.85 (3H, m), 7.36–
7.33 (3H, m). dC (150 MHz, CDCl3): 156.37, 156.29, 156.21,
156.16, 155.87, 155.81, 155.54, 155.30, 149.15, 149.11, 140.33,
140.31, 137.82, 137.78, 137.69, 136.85, 136.83, 126.20, 126.18,
123.79, 123.74, 121.35, 121.25, 121.19, 121.03, 120.61, 120.59,
119.68, 119.61. Mass spectrum: 618.2406 and isotopes (100%)
([M + H]+; 618.24061894 for M: C41H28N7) after sublimation
purification.
Synthesis of BTBB. First, (3,5-dibromophenyl)boronic acid
(12.1 g, 43.3 mmol) was treated with bromobipyridine (10.2 g,
43.3 mmol) using K2CO3 (18.0 g, 130 mmol) with pure water
(130 mL) as the base, tetrakis(triphenylphosphine)palladium(0)
(2.50 g, 2.19 mmol) as the catalyst, and a 4 : 1 toluene–ethanol
mixture (338 mL) as the solvent. The reaction mixture was
Scheme 1 Synthetic route to BBTB and BTBB.
6766 | J. Mater. Chem., 2012, 22, 6765–6773
This journal is ª The Royal Society of Chemistry 2012

View Article Online
refluxed with stirring for 68 h and then cooled to room temper-
ature. Saturated sodium chloride aqueous solution was added
and the mixture was stirred for 30 min. The resulting mixture was
extracted using toluene, dried with MgSO4, filtered to remove
precipitates, and distilled to remove the toluene giving a grayish-
white crude powder. The crude powder was washed in refluxing
methanol (600 mL) for 1 h. The mixture was cooled to room
temperature and filtered. The first intermediate (1,3-dibromo-5-
bipyridylbenzene, 8.12 g; 48.4% yield; 76.8% purity by HPLC)
was obtained. The obtained intermediate (8.12 g, 20.9 mmol) was
treated with Pin2B2 (12.75 g, 50.2 mmol), potassium acetate
(12.3 g, 125.4 mmol), and PdCl2(dppf) 1.02 g, 1.25 mmol) as the
then the substrate was immediately loaded into a high-vacuum
chamber (base pressure below ꢂ1 ꢃ 10ꢁ4 Pa) where organic
layers, 0.5 nm thick LiF, and 200 nm thick aluminum cathode
layers were deposited by thermal evaporation. Deposition rates
ꢁ1
ꢁ1
ꢁ1
ꢀ
were 0.6 A s for organic materials, 0.1 A s for LiF, and 2 A s
ꢀ
ꢀ
for Al. The structure of the devices fabricated in this study is
shown in Fig. 2. The current density–applied voltage–luminance
(J–V–L) characteristics of the OLEDs were measured by
a
commercial OLED-characteristics measurement system
(Precise Gauge EL1003) and a source meter (Keithley 2400). EL
spectra were recorded simultaneously by EL1003 during J–V–L
measurements.
catalyst in dioxane (240 mL) at 80 C. After 5.5 h, the reaction
1H and 13C NMR spectra were recorded on a Fourier
transform NMR spectrometer (JEOL JNM-ECA600). Mass
spectroscopic analysis was carried out on a matrix-assisted laser
desorption ionization TOF mass spectrometer (Bruker Daltonics
Autoflex) with no internal standard. Thermal analyses were
performed on a differential scanning calorimeter (Seiko Instru-
ments DSC-6200) at a heating rate of 10 ꢀC minꢁ1 under N2 gas.
UV and visible absorption spectra were recorded on a spectro-
photometer (Shimadzu UV-3150); photoluminescence (fluores-
cence and phosphorescence) spectra were recorded on
ꢀ
mixture was cooled to room temperature and then poured into
a 3000 mL beaker filled with 1500 mL pure water. After stirring
for 2 h, the mixture was filtered to give a crude powder (11.6 g).
Adsorption purification with diol-bonded silica (80 g) was
carried out using a toluene solution (200 mL) of the obtained
crude powder. The solution was filtered for removing silica,
distilled to remove toluene, and then recrystallized from a 1 : 1
toluene–hexane mixture to give the second intermediate
(1,3-bis(pinacolatoboryl)-5-bipyridylbenzene, 5.08 g; 50.3%
yield; 94.5% purity by HPLC). The purified second intermediate
(4.98 g, 10.3 mmol) was treated with bromoterpyridine (7.07 g,
22.7 mmol) using K2CO3 (4.27 g, 30.9 mmol) with pure water
(16 mL) as the base, tetrakis(triphenylphosphine)palladium(0)
(0.60 g, 0.52 mmol) as the catalyst, and a 4 : 1 toluene–ethanol
mixture (50 mL) as the solvent. The reaction mixture was
refluxed with stirring for 72 h and then cooled to room temper-
ature. Note that 0.6 g catalyst was added to the reaction mixture
at each elapsed time of 5, 21, 27, and 68 h. After filtering off white
precipitates, which appear while cooling the mixture in ice,
saturated sodium chloride aqueous solution was added to the
filtered solution and the mixture was extracted using toluene. The
toluene solution was dried with MgSO4 and filtered to remove
MgSO4. The solution was distilled to remove toluene and dried
under vacuum to give crude BTBB (14 g). Purification by silica-
gel (NH silica, 300 g) column chromatography using toluene as
the eluent yielded BTBB (6.02 g; 97.5% purity by HPLC). BBTB
was washed with refluxing methanol (100 mL), filtered, and dried
at 80 ꢀC under vacuum for 1 h to give pure BTBB (2.02 g; 20.0%
yield; 98.4% purity by HPLC). dH (600 MHz, CDCl3): 1H NMR
(CDCl3) d: 9.03 (2H, d, J ¼ 1.4 Hz), 9.00 (1H, t, J ¼ 1.7 Hz), 8.79
(1H, d, J ¼ 6.9 Hz), 8.76 (2H, d, J ¼ 7.6 Hz), 8.74–8.72 (3H, m),
8.68–8.65 (2H, m), 8.49 (1H, dd, J ¼ 7.6, 1.4 Hz), 8.46 (2H, dd,
J ¼ 7.6, 1.4 Hz), 8.03–7.96 (7H, m), 7.89–7.86 (3H, m), 7.36–7.34
(3H, m). dC (150 MHz, CDCl3): 156.39, 156.3, 156.19, 156.13,
155.86, 155.80, 155.55, 155.30, 149.15, 149.12, 140.29, 140.27,
137.83, 137.79, 137.70, 136.86, 128.23, 126.17, 125.23, 123.81,
123.75, 121.37, 121.26, 121.21, 121.05, 120.60, 120.58, 119.69,
119.63. Mass spectrum: 695.2672 and isotopes (100%) ([M + H]+;
695.26716804 for M: C46H31N8) after sublimation purification.
a
spectrofluorometer (Horiba FluoroMax4P). Ionization
potential (Ip) was determined by a photoelectron-emission-yield
spectrometer (Riken Keiki AC-3). Optical band gaps were
determined by the spectral onset of each UV-visible absorption
spectrum; Ea was then estimated by subtracting the band gap
energy from Ip. Spectroscopic measurements were conducted
with thin films prepared by thermal evaporation on quartz
substrates. We also determined densities of the thin films
prepared on quartz substrates by thermal evaporation using
X-ray reflectometry with a Rigaku SmartLab.
Electron mobility was measured by conventional TOF tech-
niques. The excitation light was a 500 ps duration optical pulse
from a N2 gas laser (l ¼ 337 nm, Lasertechnik Berlin MNL 200).
Test samples were prepared by thermal evaporation in vacuum
and encapsulated with fresh desiccant under a highly inert N2
atmosphere at the O2 concentration of <20 ppm. A 100 nm thick
fullerene C60 layer was used as the charge-generation layer for
optical excitation.31 The structures of the TOF samples are
shown in Fig. 2.
Computational chemistry was conducted using commercial
software (Wavefunction Spartan 10W) on a personal computer
with a quad-core processor. Density functional theory was used
for determining optimized geometries and their energies with
a B3LYP hybrid functional. The basis sets 6-31G* and
6-311+G** were used to calculate the optimized geometries and
their energies, respectively.
Results and discussion
Thermal and electronic properties
Fig. 3 shows differential scanning calorimetry (DSC) curves of
the two ETMs. The materials form amorphous glassy thin films,
as glass transition behaviors were observed in the second slow
heating after melting followed by quick coolingꢀ. That is, we
Device fabrication and measurements
All OLEDs were fabricated on 150 nm thick layers of indium-tin
oxide (ITO) that were commercially precoated onto glass
substrates with a sheet resistance of 14 U sqꢁ1. The solvent-
cleaned ITO surface was treated with O2 plasma for 5 min and
ꢀ
found baseline shifts at 93 C for BBTB and 108 C for BTBB.
BTBB, which has two terpyridines, exhibited a higher glass
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 6765–6773 | 6767

View Article Online
Fig. 2 Structures of devices and chemicals used in this study.
fact that BTBB is completely amorphous. The Tgs of these
materials increased in comparison with that oꢀf tris(bipyridyl)
benzene, which, as we reported before, was 75 C. An obvious
reason for the increase in Tg is the increase in the molecular
weight by adding one or two pyridyl groups. Moreover, the Tg of
BTBB was almost the same as that of tris(terpyridyl)benzene
(110 ꢀC). This indicates that another factor is responsible for the
increase in Tg.
Fig. 4a shows the most probable geometry of BTBB
determined by computational chemistry. The geometry can be
understood as that in which a pyridyl group is deleted from the
optimized geometry of tris(terpyridine)-substituted benzene with
C3 symmetry.30 However, the geometry can be easily converted
to others, as shown in Fig. 4b, by rotating bi- or ter-pyridine
wings around the single bond connecting the wing with the
central phenyl ring. Usually, the energy barrier for such a rota-
tion is only a few times higher than the thermal energy of room
temperature (2.51 kJ molꢁ1). Fig. 4b shows the energy difference
between the total electron energies of the most stable geometry
and the other three geometries. The maximum energy difference
is about 2 kJ molꢁ1. This means that the four geometries should
coexist in a solid. This mixing probably also results in the
material being amorphous, similar to the symmetrical tris(bi- or
ter-pyridyl)benzenes reported before.30 However, four geome-
tries contribute to the amorphous nature of BBTB and BTBB,
while two geometries contribute in the case of the tris-substituted
analogues.30 This difference between the numbers of contributing
geometries is probably one of the reasons why BTBB shows
a comparable Tg to tris(terpyridine)-substituted benzene, which
Fig. 3 DSC curves of BBTB and BTBB in the first (black line) and
second (red for BBTB and blue for BTBB) heating. DSC curves in first
cooling are also shown. Dotted lines show vertically 10 times magnified
curves around the temperature range of each Tg. Heating rates are 10 and
30 ꢀC minꢁ1 for heating and cooling, respectively.
transition temperature (Tg) of 108 ꢀC than BBTB, which has one
terpyridine. This is probably because of its larger molecular
weight (MW) and the bulkiness of terpyridine. In addition,
BTBB is completely amorphous. That is, the glass transition
occurred and no melting point appeared in the first heating.
Furthermore, no recrystallization occurred in the cooling process
with a relatively slow descent rate of ꢁ10 ꢀC minꢁ1, and the glass
transition behavior could be observed again. All this supports the
6768 | J. Mater. Chem., 2012, 22, 6765–6773
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Fig. 4 (a) Optimized geometries of BTBB. (b) Difference of total electron energy (DE) of the four most stable geometries for BBTB and BTBB, of which
geometries can be mutually converted to each other by rotating the single bonds connecting the coplanar bipyridine or terpyridine wings with the central
phenyl ring. The structures in panel (b) represent the geometries of BTBB with the 2nd, 3rd, and 4th lowest energies from left to right. Note that the
dihedral angles between the central benzene and the other bipyridine and terpyridine were not zero but almost the same (23–25ꢀ).
has a larger MW than BTBB. This is because crystallizations
must require that molecules are in the same conformation. In
addition, BTBB has two terpyridine wings, which requires
a larger free volume for rotation. In addition, the structures in
Fig. 4b indicates that one terpyridine must rotate to change the
3rd and 4th lowest geometries to the optimized one. We think
that this probably accounts for the perfectly amorphous nature
of BTBB.
BTBB has a higher Ip than BBTB indicates that the additional
pyridyl group makes the highest occupied molecular orbital
(HOMO) deeper without affecting Eg. This tendency is also
observed in the former symmetrical tris(bi- or ter-pyridine)-
substituted benzenes;30 changing bipyridine with terpyridine
increased Ip with no change in Eg. On the other hand, the
resulting electron affinities (Eas) of BBTB and BTBB, estimated
by subtracting Eg from Ip for each, were 2.78 eV and 2.91 eV,
respectively. The higher Ea of BTBB suggests that its lowest
unoccupied molecular orbital (LUMO) is deeper than that of
Table 1 also shows the densities of BBTB and BTBB measured
in their thin-film states. The densities of BBTB and BTBB
are both equal to 1.27 g cmꢁ3, which is larger than those of
tris(bipyridine)- and tris(terpyridine)-substituted benzene
(1.21 g cmꢁ3), which were also measured in this study. These high
densities may also contribute to the increase in Tg according to
the free-volume theory,32 the standard theory for glass transition.
Briefly, this is because a larger density means a smaller free-
volume fraction (ratio of free volume to total volume) if solids
are composed of similar elements.
Fig. 5 shows the UV-visible absorption and fluorescence
spectra of BBTB and BTBB in their thin-film states. The figure
also shows their phosphorescence spectra at 77 K. Both materials
show optical absorption onsets at a wavelength of about 350 nm
and UV fluorescence peaks at 370 nm. Band gaps (Egs) of the
materials were estimated from the onsets of the optical absorp-
tion spectra and are summarized in Table 1. The values of Egs
were as high as about 3.6 eV, and the values of ionization
potentials (Ips), which were determined by photoemission-elec-
tron-yield spectroscopy, varied from 6.3 to 6.5 eV. The fact that
Fig. 5 UV-visible absorption, photoluminescence (PL), and phospho-
rescence (at 77 K) spectra of BBTB and BTBB in their thin-film states.
T
The two arrows are the indications of the peaks used for determining Eg
.
Table 1 Basic properties of BBTB and BTBB. Tm, Tg, and Td were measured in the solid state. The other properties were measured in the thin-film
statea
Material
Tm (ꢀC)
Tg (ꢀC)
Density (g cmꢁ3
)
Ip (eV)
Ea (eV)
Eg (eV)
EgT (eV)
BBTB
BTBB
226
n.d.
93
108
1.27
1.27
6.33
6.50
2.78
2.91
3.55
3.59
2.53
2.53
a
n.d: not detected.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 6765–6773 | 6769

View Article Online
BBTB. This can lead to better electron injection, resulting in the
potential for reducing the operation voltage. The modestly large
Ips of BBTB and BTBB probably produce an effect on hole-
blocking. Fig. 5 also shows their phosphorescence spectra. The
phosphorescence spectra are also relatively similar. Estimated
triplet band gaps (EgTs) for BBTB and BTBB are shown in Table
same as that of the reference device with Alq as the ETM. BBTB
and BTBB potentially possess hole-blocking properties because
of their large Ips, but no differences among the devices were
detected, because very slow hole mobility34 in Alq of about
10ꢁ8 cm2 Vꢁ1 sꢁ1 can practically prevent hole-leaking to the
cathode. As shown in the inset of the figure, the electrolumi-
nescence (EL) spectrum of the BTBB device indicates that
recombination of both charge carriers (electrons and holes)
should occur in the Alq layer. In addition, in the UV-blue region,
there was no EL assigned to other materials, even in the
magnified spectrum, which also supports the fact that no charge-
carrier leakage occurred.
1, where we assumed that each shoulder marked with the arrows
T
in the figure corresponded to each Eg
.
OLED performance
Fig. 6a shows J–V curves of the fluorescent devices (test OLED-1
shown in Fig. 2) with BBTB and BTBB. The devices with the new
materials showed a lower operation voltage in comparison with
the reference device using Alq as the ETM. The voltage reduc-
tions reached 1.5 V–2.0 V. In particular, the device with BTBB
exhibited a very low operation voltage; a current density of 3.5
mA cmꢁ2 corresponded to a voltage of 3.0 V. In addition, current
densities of the device with the new ETMs rapidly increased with
a small boost of applied voltage, and the BTBB device reached
J ¼ 100 mA cmꢁ2 at a mere 4.5 V. These low operation voltages
and the rapid increase in current densities with small over-
potentials strongly imply that BBTB and BTBB probably possess
both good electron-injection properties from the cathode and
fast electron mobilities (electron mobility will be discussed later).
In addition, BTBB had a higher Ea than BBTB, which probably
caused it to have a lower operation voltage owing to better
electron injection from the cathode. Simple estimates of the
electron-injection barriers at the interfaces for BBTB/Al and
BTBB/Al are 0.67 eV and 0.80 eV, respectively. Here we assumed
the work function of the Al cathode combined with the
LiF-injection layer was equal to that of pure aluminum (3.58 eV),
and even this probably led to overestimations. These barriers are
slightly higher in comparison with those estimated in the
same manner for other ETMs reported with good electron-
injection properties from the LiF/Al cathode.33 We think that
bipyridine and/or terpyridine can help smooth electron injections
owing to their chelating abilities for metal atoms, which can lead
to complex formations at the interface, resulting in midgap
states.
Fig. 7a shows J–V curves of phosphorescent OLEDs (test
OLED-2 shown in Fig. 2) with BBTB and BTBB. Here
1,1-bis((di-4-tolylamino)phenyl)cyclohexane (TAPC) was used
as the HT material for reducing the operation voltage and pre-
venting triplet energy leakage from the emissive layer to the HT
layer because of the material’s high hole mobility35 and large Eg
(2.87 eV).36 The reference device used 1,3,5-tris(1-phenyl-1H-
T
benzimidazol-2-yl)benzene (TPBi) as the ETM, which exhibits
T
high electron mobility37 and larger Eg (2.58 eV)38 than Irppy
(2.42 eV)39. The phosphorescent OLED with BTBB exhibited
a much lower voltage operation than the reference one with
TPBi. The reported LUMO of TPBi is ꢁ2.9 eV,40 and thus, the
electron-injection barriers at the interfaces, BTBB/LiF/Al and
TPBi/LiF/Al, are probably almost the same because Ea of BTBB
is 2.91 eV. Therefore, the large difference between the operation
voltages for both devices suggests that the electron mobility of
BTBB should be much higher than that of TPBi. Note that
TAPC has an extremely high hole mobility35 of about
10ꢁ2 cm2 Vꢁ1 sꢁ1 despite the fact that TAPC is an amorphous
organic semiconductor. Thus, the charge balance in the devices,
including the BBTB and BTBB devices, is probably biased
towards being hole-rich, and hence, the recombination zone of
holes and electrons would be located in a part of the emissive
layer near the interface between the emissive layer and the BBTB,
BTBB, or TPBi layer. Given that the luminous efficiency is
almost the same for all devices, in this situation this indicates that
the new ETMs do not work as energy leakage paths for triplet
T
excitons of Irppy, although the Eg of BTBB is neither smaller
Fig. 6b shows luminous efficiency–current density (h–J)
characteristics of the devices with BBTB and BTBB. The
luminous efficiency of the devices with the new ETMs were the
nor much larger than that of Irppy.
The luminous efficiency of the BTBB device at a current
density of 0.11 mA cmꢁ2 was 47.4 cd Aꢁ1, and this efficiency
Fig. 6 (a) Current density–voltage (J–V) and (b) luminous efficiency–current density (h–J) curves of devices with BBTB and BTBB. The inset in panel
(b) shows the EL spectrum of the BTBB device at a current density of 80 mA cmꢁ2. Dashed curve in the inset presents the spectrum magnified by 10 times.
6770 | J. Mater. Chem., 2012, 22, 6765–6773
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Fig. 7 (a) Current density–voltage (J–V), (b) luminous efficiency–current density (h–J) and (inset) luminous power efficiency–voltage curves of
phosphorescent OLED with BBTB and BTBB.
corresponded to an external quantum efficiency (EQE) of 11.9%
under the assumption of perfect Lambertian light distribution.
This EQE is common for the conventional Irppy-based phos-
phorescent OLEDs with a single emissive layer and a hole-
blocking layer such as bathocuproine (BCP),30 and thus, BTBB
should have hole-blocking ability. These results suggest that
BTBB can be used as an ETM with hole-blocking ability for
phosphorescent OLEDs. In addition, the inset of Fig. 7b shows
that the BTBB device exhibited a higher luminous power
efficiency than other devices, which is because of its lower voltage
operation, while the EQE and luminous efficiency of the BTBB
device are the same as those for other devices, as shown in
Fig. 7b.
Fig. 8 Transient photocurrent curves of TOF measurement of BTBB at
applied voltages of 430 and 550 V. Red straight lines in the inset are for
guides to determine the transit time.
Electron mobility
Finally, we would like to briefly discuss electron mobility in the
new materials. Fig. 8 shows the transient electron photocurrent
waveforms, while the inset presents this data using a double
logarithmic scale. These waveforms were obtained by time-of-
flight (TOF) measurement using the BTBB film whose thickness
is 5.17 mm. Although each photocurrent waveform was not
clearly non-dispersive, we can clearly see that the transit time (tT)
of the photogenerated-charge carriers—electrons in this case—
decreased with increasing applied voltages because the photo-
currents decayed faster in such a case. In accordance with the
conventional method to determine tT for dispersive trans-
portations, we drew two auxiliary straight lines, as shown in the
figure, and determined tT at the crossing point of the auxiliary
lines. Note that the slopes of the two auxiliary lines for each
photocurrent are the same and the slopes are larger than ꢁ1 for
before tT and smaller than ꢁ1 for after tT, respectively, which
also support the argument that the observed photocurrents are
due to drift motions in the film. If we consider the electron
mobility m as
orders of magnitude higher than for Alq (2.0 ꢃ 10ꢁ6 cm2 Vꢁ1 sꢁ1)41
and 3 times higher than for TPBi (9 ꢃ 10ꢁ5 cm2 Vꢁ1 sꢁ1)37 at
approximately the same electric field. This high mobility probably
accounts for the better J–V curves for OLEDs with BTBB in
comparison with those of the OLEDs with the corresponding
reference. Note that all the transient photocurrent waveforms for
both materials are shown in Fig. S1 and S2.†
Fig. 9 shows the electron mobility as a function of the square
root of the applied electric field for several materials, including
BBTB and BTBB. Both BBTB and BTBB show mobilities that are
about 3 orders of magnitude higher than for Alq. In particular, the
electron mobility of BBTB is greater than 10ꢁ3 cm2 Vꢁ1 sꢁ1 at an
electric field of greater than or equal to 490 kV cmꢁ1. The higher
electron mobility of BBTB should contribute to reducing the
operation voltage below that of the BTBB devices, but in fact,
just the opposite occurs. We think that the smooth electron
injection at the interface between BTBB and the cathode owing
to the relatively large Ea of BTBB (compared with BBTB) can
probably account for the reason why the BTBB devices exhibit
better J–V characteristics than the BBTB devices in which the
electron mobility is two times higher. However, unfortunately,
we are unsure at present why BBTB and BTBB exhibited higher
mobilities than their symmetric analogues, tris(bipyridyl)benzene
Dd
m ¼
;
(1)
tTV
where D is the distance between the two electrodes (5.27 mm), d is
the thickness of the prepared BTBB layer, and V is the applied
voltage, we can determine, for example, m as 2.8 ꢃ 10ꢁ4 cm2 Vꢁ1 sꢁ1
for an electric field E ¼ V/D ¼ 816 kV cmꢁ1. This mobility is 2
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 6765–6773 | 6771

View Article Online
probably because of the comparably higher densities of BBTB
and BTBB thin films. These high mobilities contributed to a low
voltage operation; for example, in the case of the Alq-based
device with BTBB, current densities of 3.5 mA cmꢁ2 and 100 mA
cmꢁ2 were reached at voltages of 3.0 V and 4.5 V, respectively.
However, we observed another mobility–Tg trade-off relation-
ship in the new materials. Further investigations are required to
promote the material science of amorphous organic
semiconductors. Ips of BBTB and BTBB are sufficiently deep to
confine holes in emissive layers.
Acknowledgements
The authors would like to thank Riken Keiki and Rigaku for
ionization potential and density measurements. This work was
supported by Program for Fostering Regional Innovation in
Nagano, granted by MEXT, Japan.
Fig. 9 Electron mobilities of BBTB and BTBB as a function of the
square root of the electric field. Data for Alq were referred from the
literature by Murata (see text).41
References
(1.4 ꢃ 10ꢁ4 cm2 Vꢁ1 sꢁ1) and tris(terpyridyl)benzene (3.2 ꢃ
10ꢁ5 cm2 Vꢁ1 sꢁ1) in an electric field of 680 kV cmꢁ1 30
The
.
1 J.-h. Choi, K.-H. Kim, S.-J. Choi and H. H. Lee, Nanotechnology,
2006, 17, 2246.
densities of the BBTB and BTBB thin films are higher than those
of their symmetric analogues, which we believe may also
contribute to their higher electron mobilities, similarly to the case
of Tg; a higher density leads to a shorter distance between
hopping sites. We regret that we do not fully understand why
BBTB and BTBB, asymmetric derivatives of trisubstituted
benzene with both bi- and ter-pyridines, exhibit higher densities
than their symmetric analogues. This is still a question that
goes beyond the frontiers of molecular material science to the
best of our knowledge. Note that it was difficult to determine tT
or mobility for BTBB under applied electric fields below
680 kV cmꢁ1 (370 V). On the other hand, tT for BBTB were
successfully determined in smaller electric fields. This difference
might imply that there is larger number of electron traps in the
BTBB film compared with the BBTB film. Further detailed
studies about the point are needed.
2 Y. Nakajima, T. Takei, T. Tsuzuki, M. Suzuki, H. Fukagawa,
T. Yamamoto and S. Tokito, J. Soc. Inf. Disp., 2009, 17, 629–634.
3 K. T. Kamtekar, A. P. Monkman and M. R. Bryce, Adv. Mater.,
2010, 22, 572–582.
4 Y. Ohmori, Laser Photonics Rev., 2010, 4, 300–310.
5 C. W. Tang and S. A. VanSlyke, Appl. Phys. Lett., 1987, 51, 913–915.
6 C. Adachi, T. Tsutsui and S. Saito, Appl. Phys. Lett., 1990, 57, 531–
533.
7 Y. Shirota, J. Mater. Chem., 2000, 10, 1–25.
8 U. Mitschke and P. Baeuerle, J. Mater. Chem., 2000, 10, 1471–1507.
9 S. A. VanSlyke, C. H. Chen and C. W. Tang, Appl. Phys. Lett., 1996,
69, 2160–2162.
10 M. Thelakkat and H.-W. Schmidt, Adv. Mater., 1998, 10, 219–223.
11 K. Katsuma and Y. Shirota, Adv. Mater., 1998, 10, 223–226.
12 H. Tanaka, S. Tokito, Y. Taga and A. Okada, Chem. Commun., 1996,
2175–2178.
13 M. Ichikawa, K. Hibino, N. Yokoyama, T. Miki, T. Koyama and
Y. Taniguchi, Synth. Met., 2006, 156, 1383–1389.
14 Y.-L. Liao, W.-Y. Hung, T.-H. Hou, C.-Y. Lin and K.-T. Wong,
Chem. Mater., 2007, 19, 6350–6357.
15 C.-Y. Lin, Y.-M. Chen, H.-F. Chen, F.-C. Fang, Y.-C. Lin,
W.-Y. Hung, K.-T. Wong, R. C. Kwong and S. C. Xia, Org.
Electron., 2009, 10, 181–188.
16 Z. Jiang, Z. Liu, C. Yang, C. Zhong, J. Qin, G. Yu and Y. Liu, Adv.
Funct. Mater., 2009, 19, 3987–3995.
Summary
We have developed new ETMs for OLEDs based on benzene-
substituted bipyridine and terpyridine (BBTB and BTBB)
groups. From the viewpoint of the trade-off relationship between
mobility and Tg appearing in our former trisubstituted benzenes
with bipyridine or terpyridine, the new materials could achieve
better properties. Tgs of BBTB and BTBB were 93 ꢀC and 108 ꢀC,
respectively, of which the Tg of BTBB was almost comparable to
that of tris(terpyridyl)benzene, although tris(terpyridyl)benzene
has a larger molecular weight. This is due to the additional
pyridyl group in BTBB. The substitution with both bi- and ter-
pyridines leads to conformational variation, and the four
conformers have similar energies with the maximum difference
less than the thermal energy of room temperature. This confor-
mational variation contributes to the formation of the glassy
state. In particular, we found BTBB to be completely amorphous
with no melting point. Electron mobilities of BTBB exceeded the
order of 10ꢁ4 cm2 Vꢁ1 sꢁ1, while those of BBTB were very high
and reached 10ꢁ3 cm2 Vꢁ1 sꢁ1. The reason these mobilities are
higher than those of the former trisubstituted derivatives is
17 C. Adachi, T. Tsutsui and S. Saito, Appl. Phys. Lett., 1989, 55, 1489.
18 Y. Hamada, C. Adachi, T. Tsutsui and S. Saito, Jpn. J. Appl. Phys.,
1992, 31, 1812–1816.
19 K. Tamao, M. Uchida, T. Izumizawa, K. Furukawa and
S. Yamaguchi, J. Am. Chem. Soc., 1996, 118, 11974–11975.
20 M. Uchida, T. Izumizawa, T. Nakano, S. Yamaguchi, K. Tamao and
K. Furukawa, Chem. Mater., 2001, 13, 2680–2683.
21 J. Bettenhausen and P. Strohriegl, Adv. Mater., 1996, 8, 507–810.
22 S. B. Heidenhain, Y. Sakamoto, T. Suzuki, A. Miura, H. Fujikawa,
T. Mori, S. Tokito and Y. Taga, J. Am. Chem. Soc., 2000, 122,
10240–10241.
23 G. Hughes and M. R. Bryce, J. Mater. Chem., 2005, 15, 94–107.
24 M. Ichikawa, T. Kawaguchi, K. Kobayashi, T. Miki, K. Furukawa,
T. Koyama and Y. Taniguchi, J. Mater. Chem., 2006, 16, 221–225.
25 H. Sasabe, E. Gonmori, T. Chiba, Y.-J. Li, D. Tanaka, S.-J. Su,
T. Takeda, Y.-J. Pu, K.-i. Nakayama and J. Kido, Chem. Mater.,
2008, 20, 5951–5953.
26 Y.-J. Pu, M. Yoshizaki, T. Akiniwa, K.-i. Nakayama and J. Kido,
Org. Electron., 2009, 10, 877–882.
27 H. Tsuji, K. Sato, Y. Sato and E. Nakamura, J. Mater. Chem., 2009,
19, 3364–3366.
28 L. Na, W. Pengfei, L. Shiu-Lun, L. Weimin, L. Chun-Sing, L. Shuit-
Tong and L. Zengtao, Adv. Mater., 2010, 22, 527–530.
6772 | J. Mater. Chem., 2012, 22, 6765–6773
This journal is ª The Royal Society of Chemistry 2012

View Article Online
€
35 P. M. Borsenberger, L. Pautmeier, R. Richert and H. Bassler, J.
Chem. Phys., 1991, 94, 8276.
29 S.-J. Su, T. Chiba, T. Takeda and J. Kido, Adv. Mater., 2008, 20,
2125–2130.
30 M. Ichikawa, K. Wakabayashi, S. Hayashi, N. Yokoyama,
T. Koyama and Y. Taniguchi, Org. Electron., 2010, 11, 1966–
1973.
31 M. Ichikawa, N. Hiramatsu, N. Yokoyama, T. Miki, S. Narita,
T. Koyama and Y. Taniguchi, Phys. Status Solidi RRL, 2007, 1,
R37–R39.
36 K. Goushi, R. Kwong, J. J. Brown, H. Sasabe and C. Adachi, J. Appl.
Phys., 2004, 95, 7798–7802.
37 H. Wen-Yi, K. Tung-Huei, L. Yu-Ting, W. Chung-Chih, H. Tsung-
Hsi, C. Teng-Chih, W. Ken-Tsung and I. W. Chih, Appl. Phys.
Lett., 2006, 88, 064102.
38 H. Fukagawa, K. Watanabe and S. Tokito, Org. Electron., 2009, 10,
798–802.
39 K. Goushi, Y. Kawamura, H. Sasabe and C. Adachi, Jpn. J. Appl.
Phys., 2004, 43, L937–L939.
40 X.-M. Yu, H.-S. Kwok, W.-Y. Wong and G.-J. Zhou, Chem. Mater.,
2006, 18, 5097–5103.
41 H. Murata, G. G. Malliaras, M. Uchida, Y. Shen and Z. H. Kafafi,
Chem. Phys. Lett., 2001, 339, 161–166.
32 L. H. Sperling Introduction to Physical Polymer Science 4th ed.; John
Wiley & Sons: Hoboken, 2006.
33 T. Matsushima, M. Takamori, Y. Miyashita, Y. Honma, T. Tanaka,
H. Aihara and H. Murata, Org. Electron., 2010, 11, 16–22.
34 R. G. Kepler, P. M. Beeson, S. J. Jacobs, R. A. Anderson,
M. B. Sinclair, V. S. Valencia and P. A. Cahill, Appl. Phys. Lett.,
1995, 66, 3618–3620.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 6765–6773 | 6773