Organic Electronics 27 (2015) 126e132

Contents lists available at ScienceDirect

Organic Electronics
journal homepage: www.elsevier.com/locate/orgel

Silicon-based electron-transport materials with high thermal stability

and triplet energy for efcient phosphorescent OLEDs
Seungjun Yi a, Jin-Hyoung Kim a, Woo-Ri Bae a, Jiwon Lee b, Won-Sik Han b, **,
Ho-Jin Son a, ***, Sang Ook Kang a, *
a
b

Department of Advanced Materials Chemistry, Korea University, Sejong 30019, South Korea
Department of Chemistry, Seoul Women's University, Seoul 01797, South Korea

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 23 June 2015
Received in revised form
4 August 2015
Accepted 30 August 2015
Available online 25 September 2015

A series of electron transporting materials was designed and used in organic light-emitting diodes
(OLEDs), exhibiting green phosphorescence. We used the tetrahedral structural motif of silicon atom,
which annulated with the 1,2-diphenyl-benzoimidazole (DBI) units in its periphery (1e4) and their
thermal, photophysical, and electrochemical properties were investigated. Among the series, the X-ray
crystal structure of compound 1 was obtained and investigated. Photophysical and electrochemical
properties showed that their LUMO levels can be slightly tuned as increasing number of DBI units and
enhancing the electron injection capability. Furthermore, thermal stability correlated well with an increase in the number of DBI units, showing a gradual increase in Tg values in the range of 100e141
C. The
electron-only devices (EOD) based on compounds 2 and 3 were fabricated; EOD device with compound 3
showed higher current densities at the same voltages, indicating higher electron transport (ET) capability
compared to compound 2. The electron mobilities (m) of compounds 2 and 3 were estimated as
1.93  105 cm2/V and 3.67  105 cm2/V at 1 MV/cm, respectively. We further investigated the excellent
ET property of compound 3 via the phosphorescent OLEDs in which the electron-transporting material
(ETM) was coupled with the green emitter, Ir(ppy)3. Finally, we compared it with the device based on
compound 2. The OLEDs device with compound 3 exhibited maximum current and external quantum
efciencies of 62.8 cd/A and 18.0%, respectively, with a small efciency roll-off at high current densities.
2015 Elsevier B.V. All rights reserved.

Keywords:
Electron transporting material
High thermal stability
High triplet

1. Introduction
Researchers have focused their attention on developing organic
light-emitting diodes (OLEDs) of high efciency. OLEDs have
numerous practical applications, because they can be used in at
panel displays and lighting [1e4]. OLEDs devices hold signicant
importance because they have high emission efciency and
appreciable lifetime.
To achieve high efciency, charge carrier balance has to be
optimized [5,6]. However, hole mobility is usually much higher
than electron mobility in OLEDs. As a result, the efciency of OLEDs
is drastically reduced [7e9]. Thus, several methodologies have been

* Corresponding author.
** Corresponding author.
*** Corresponding author.
E-mail addresses: wshan@swu.ac.kr (W.-S. Han), hjson@korea.ac.kr (H.-J. Son),
sangok@korea.ac.kr (S.O. Kang).
http://dx.doi.org/10.1016/j.orgel.2015.08.029
1566-1199/ 2015 Elsevier B.V. All rights reserved.

proposed in designing electron-transporting (ET) materials for

phosphorescent OLEDs (PhOLEDs); 1) utilization of electron decient heterocyclic compounds; 2) matching LUMO energy level
with emissive layer material is to be the next consideration; 3)
careful choice of linking unit is suggested avoiding unnecessary
triplet energy lowering by p-conjugation. Based on these guidelines, several potential phosphorescence electron transporting
materials have been developed such as oxadiazole [10,11], triazole
[12], and benzoimidazole [13e15]. These aza-aromatic compounds
showed high efciency in low voltage. Nevertheless, their use limits
the lifetime of the device.
Recently, we have reported silicon based blue phosphorescence
host materials [16]. Introducing silicon atom effectively breaks the
conjugation and attenuates the triplet wave function to enhance
device performance. Namely, silicon atom effectively distorts the
electronic p-type overlap in the surrounding electro-active organic
functional groups. Inspired by the successful exploit of silicon
centered compounds as blue phosphorescence host material, it was
rationally considered that the introduction of 1,2-diphenyl-

S. Yi et al. / Organic Electronics 27 (2015) 126e132

benzoimidazole (DBI) unit into the silicon atom may give a synergetic effect toward highly efcient and stable ET material. In this
report, as shown in Chart 1, we designed and prepared 1,2diphenyl-benzoimidazole (DBI) derivatives (1e4), which exuded
high thermal stability, wide energy gap, and high electron mobility.
The single-crystal molecular structure, thermal and photophysical
properties, and ET capability of these compounds were systematically investigated. Thereafter, green PhOLEDs using 2 and 3 as ET
materials were fabricated and their efciencies are reported.

127

and the crude product was extracted twice using 100 mL of

dichloromethane. Finally, the crude product was washed with brine.
The combined organic layer was dried over MgSO4 and subjected to
ltration. The ltrates were dried under reduced pressure, and the
residue was subjected to column chromatography using ethyl acetate/n-hexane (v:v 1:10) as an eluent. Compound 1 was obtained
as a white solid with yield of 78%. We could not perform train sublimation of compound 1 due to its poor thermal stability. 1H
NMR(CDCl3): d 7.89 (dt, J 7.8 Hz, 0.9 Hz,1H, ArH), 7.48e7.56 (m, 5H,

Me

Si
N

Me

N
Me

Me

Me

Si
N

Me

Si

Si

Chart 1. Chemical structures of electron transport materials 1e4.

2. Experimental
2.1. General information
All the experimental procedures were carried out under a dry
nitrogen or argon atmosphere using standard Schlenk techniques.
Tetrahydrofuran (THF) was distilled freshly over sodium benzophenone. The 1H and 13C NMR spectra were recorded on a Varian
Mercury 300 spectrometer, which was operated at 300.1 and
75.4 MHz, respectively. 1H and 13C NMR chemical shifts were
measured in CDCl3 and referenced to relative peaks of CHCl3
(7.26 ppm for 1H NMR) and CDCl3 (77.16 ppm for 13C NMR). The
elemental analyses were conducted using a Carlo Erba Instruments
CHNSO EA 1108 analyzer. High resolution tandem mass spectrometry was performed using Jeol LTD JMS-HX 110/110A at the
Korean Basic Science Institute. Cyclic voltammetry (CV) was performed in an electrolytic solution prepared using 1 mM of electroactive compounds and 0.1 M tetrabutyl ammonium hexauorophosphate (NBu4PF6) at room temperature under an atmosphere of nitrogen. For this purpose, BAS 100B electrochemical
analyzer was used. Glassy carbon, platinum wire, and Ag/AgNO3
(0.1 M) were used as working, counter, and reference electrodes,
respectively. All the potentials were calibrated to the ferrocene/
ferrocenium (Fc/Fc) redox couple. n-BuLi (2.5 M solution in nhexane), trimethylchlorosilane, dimethyldichlorosilane, trichloromethylsilane, and tetrachlorosilane were purchased from
Aldrich and used in this experiment without further purication. 2(4-bromophenyl)-1-phenyl-1H-benzo[d] imidazole was synthesized according to the protocol mentioned in literature [17].
2.2. Synthesis
Compound 1 The solution of n-BuLi in 2.5 M n-hexane (2.64 mL,
6.6 mmol) was added dropwise to a solution of 2-(4-bromophenyl)1-phenyl-1H-benzo[d] imidazole (2.09 g, 6 mmol) in dry THF
(30 mL) at 78
C. At temperature of 78
C, the resulting reaction
mixture was stirred continuously. Thereafter, we slowly added trimethylchlorosilane (0.76 mL, 6 mmol) to the mixture. Then, the
reaction temperature was allowed to warm at room temperature
and stirred overnight. The reaction was quenched by adding water,

ArH), 7.43e7.54 (m, 2H, ArH), 7.31e7.36 (m, 3H, ArH), 7.25 (dd,
J 5.4 Hz, J 0.6 Hz, 2H, ArH), 0.25 (s, 9H, SieCH3). 13C NMR (CDCl3):
d152.4, 143.0, 142.3, 137.4, 137.1, 133.2, 130.1, 129.9, 128.6, 128.4,
127.5, 123.3, 123.0, 119.8, 110.4, 1.2. According to electrospray
ionization mass spectrometry (ESI-MS) technique, the calculated for
C22H22N2Si was 342.1552. Found: 342.1549. Calcd for C22H22N2Si: C,
77.15; H, 6.47; N, 8.18. Found: C, 77.12; H, 6.49; N, 8.17.
Compound 2 We used a procedure analogous to the one used in
the preparation of compound 1. In this case, dichlorodimethylsilane
was used as the starting material. After the reaction, the solvent
was removed and the residue was subjected to column chromatography (ethyl acetate/n-hexane, 1:5). Compound 2 was produced
as a white solid; the yield of this reaction was 72%. Before conducting device fabrication, 52% yield of pure compound 2 was obtained by train sublimation. 1H NMR (CDCl3) d 7.88 (dt, J 7.8 Hz,
0.9 Hz, 2H, ArH), 7.49e7.55 (m, 10H, ArH), 7.39e7.41 (m, 4H, ArH),
7.31e7.34 (m, 6H, ArH), 7.25e7.27 (m, 4H, ArH), 0.51 (s, 6H, SieCH3).
13
C NMR (CDCl3): d152.2, 143.0, 139.5, 137.4, 137.0, 134.1, 130.6,
129.9, 128.7, 128.5, 127.5, 123.4, 123.0, 119.9, 110.5, 2.6. According
to ESIeMS technique, calculated for C40H32N4Si was 596.2396.
Found: 596.2391. Calcd for C40H32N4Si: C, 80.50; H, 5.40; N, 9.39.
Found: C, 80.52; H, 5.41; N, 9.37.
Compound 3 We used a procedure analogous to the one used in
the preparation of compound 1 In this case, trichloromethylsilane
was used as the starting material. After the reaction, the solvent
was removed and the residue was subjected to column chromatography (ethyl acetate/n-hexane, 1:4), and compound 3 was obtained as a white solid. The yield of this reaction was 54%. For device
fabrication, compound 3 was further puried by train sublimation.
The yield of this reaction was 47% .1H NMR (CDCl3) d 7.88 (d,
J 7.8 Hz, 3H, ArH), 7.47e7.55 (m, 15H, ArH), 7.31e7.37 (m, 15H,
ArH), 7.22e7.27 (m, 6H, ArH), 0.78 (s, 3H, SieCH3). 13C NMR (CDCl3):
d152.0, 143.0, 137.4, 137.0, 136.9, 135.1, 131.0, 130.0, 128.8, 128.6,
127.4, 123.5, 119.9, 110.5, 3.8. The ESI-MS calculated for C58H42N6Si
was 850.3240. Found: 850.3244. Calcd for C58H42N6Si: C, 81.85; H,
4.97; N, 9.87. Found: C, 81.83; H, 4.99; N, 9.88.
Compound 4 The procedure used for synthesizing compound 4
was analogous to the one used in the preparation of compound 1. In
this case, tetrachlorosilane was used as the starting material. After
completion of the reaction, the solvent was recovered, and the

128

S. Yi et al. / Organic Electronics 27 (2015) 126e132

residue was subjected to column chromatography (ethyl acetate/nhexane, 1:1). Thus, we obtained pure compound 4 in the form of a
white solid. The yield of this reaction was 33%. However, we could
not perform train sublimation of compound 4 might be due to the
high molecular weight. 1H NMR (CDCl3) d 7.85e7.89 (m, 4H, ArH),
7.49e7.57 (m, 10H, ArH), 7.37e7.43 (m, 18H, ArH), 7.32e7.38 (m, 10H,
ArH), 7.20e7.28 (m, 10H, ArH). Owing to the low solubility of compound 4, 13C NMR could not be performed. The ESI-MS calculated for
C76H52N8Si was 1104.4084. Found: 1104.4088. Calcd for C76H52N8Si:
C, 82.58; H, 4.74; N, 10.14. Found: C, 82.61; H, 4.77; N, 10.12.

deposit organic and Al layers. The OLED devices were packaged in a

glove box lled with N2 of ultra-high purity; the concentrations of
oxygen and water were maintained at approximately 0.5 ppm in this
box. A glass lid was attached to the substrate through an epoxy seal,
which was applied around the perimeter of this box. The currentevoltage characteristics of OLEDs were measured using a source
measurement unit (Keithley 2635B). The electroluminescence
spectra, luminance, and CIE coordinates were measured using a
spectroradiometer (Konica Minolta CS-2000). Assuming Lambertian
emission, we calculated the external quantum efciency (EQE) from
the luminance, current density, and electroluminescence spectrum.

2.3. Crystal structure determination

Single crystal of compound 1 was mounted on a diffractometer.
To perform preliminary examination and data collection, we used a
Bruker SMART CCD detector system single-crystal X-ray diffractometer that was equipped with a sealed-tube X-ray source
(50 kV  30 mA); monochromatic Mo Ka radiation (l 0.71073 )
corresponding to graphite was obtained from the X-ray source. At
this preliminary stage, the unit cell constants were determined
from a set of 45 narrow-frame (0.3
in u) scans. A double-pass
method of scanning was used to exclude any noise. The collected
frames were integrated using an orientation matrix, which was
developed using narrow frame scans. The SMART software package
was used for data collection, while SAINT software package was
used for frame integration [18]. Final cell constants were determined by conducting global renement of the xyz centroids of the
reections, which were harvested from the entire data set. The
structural solution and renement were carried out using SHELXTL
PLUS software package [19].
2.4. Absorption and emission spectra
The absorption and photoluminescence spectra were recorded
on an SHIMADZU UV-3101PC UV/VIS/NIR scanning spectrophotometer and a VARIAN Cary Eclipse uorescence spectrophotometer, respectively.
2.5. Thermal properties
Thermal properties were measured using differential scanning
calorimetry (Pyris Diamond DSC; Perkin Elmer). A heating rate of
10
C/min was used for rst melting the compound followed by a
rapid cooling rate of 40
C/min to room temperature.

3. Results and discussion

3.1. Synthesis and characterization
Initially, 2-(4-bromophenyl)-1-phenyl-1H-benzo[d]imidazole
was prepared in quantitative yield from 4-bromobenzaldehyde and
N-phenylbenzene-1,2-diamine as the starting materials [17].
Compounds 1e4, which were the nal products, were synthesized
by conducting a direct substitution reaction between (4-(1-phenyl1H-benzo[d]imidazol-2-yl)phenyl)lithium and MenSiClm (n 0, 1,
2, 3, and m 4n) according to the protocol in the literature [16].
The nal products were puried by silica-gel column chromatography, and compounds 2 and 3 were further puried by vacuum
train sublimation. Compounds 1 and 4 were not able to further
purication by train sublimation. The structures and chemical purities of nal products were adequately veried by 1H NMR, 13C
NMR, mass spectrometry, and elemental analysis, respectively (see
Experimental Section).
3.2. Crystal structure
Fig. 1 displays the molecular structure of a single crystal for
compound 1, which was readily obtained by slow evaporation of its
dissolution in dichloromethane/n-hexane. Table S1eS3 summarized structural parameters, bond lengths, and angles detected by
X-ray. Compound 1 was crystallized in the monoclinic crystal system using P21/c space group at R1 0.0555. The silicon atom had a
tetrahedral geometry, which was dened by the three methyl
groups [1.850(3) (averaged)] and one ipso-carbon atom in DBI
[1.885(2) ]. These values are in good agreement with 1.865 ,
which corresponds with the typical tetrahedral silanes in the International Tables for X-ray Crystallography, Vol. III [22]. The

2.6. DFT calculations

The derivatives were theoretically calculated using the
Gaussian09 package [20]. The ground-state geometries of compounds 1e3 were optimized using the B3LYP density functional
theory (DFT) and 6/31G (d,p) basis set. Then, time-dependent DFT
(TDDFT) calculations were performed using the same functional
theory and basis set, which were previously used to estimate the
energies and oscillator strengths of derivatives. The contours of
electron density were plotted using Chem3D ver.10 [21].
2.7. Device fabrication and characterization
The OLED devices were fabricated on glass substrates, which
were pre-coated with a 150 nm indium tin oxide (ITO) layer having
sheet resistance of 10 U/square. The ITO glass was pre-cleaned by a
conventional solvent cleaning method. Using oxygen plasma treatment, the ITO surface was cleaned again immediately before
depositing the hole injection layer. At a background pressure of
~3  107 Torr, thermal evaporation was performed to sequentially

Fig. 1. ORTEP (III) [23] structure of compound 1 showing thermal ellipsoids; the
probability level of this structure is 30%. Hydrogen atoms were omitted for the purpose
of clarity.

S. Yi et al. / Organic Electronics 27 (2015) 126e132

The thermal stability of compounds 1e4 were investigated using the techniques of thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). The DSC experiments were
carried out at a temperature ranging from 30 to 430
C under an
atmosphere of nitrogen. In subsequent heating cycles, we observed
endothermic step transitions of weak intensity, corresponding to
compounds 2 and 3 in the DSC curve; the glass phase transition
temperature (Tg) of the compounds was determined from these
curves. For compounds 1 and 4, we could not observe endothermic
step transitions of well-dened nature. The thermal properties of
the derivatives, 2 and 3, increased in proportion to the number of
substituted DBI units at the silicon central atom, as shown in Fig. S1.
In the consecutive heating and cooling cycles, the glass phase
transition temperature, Tg, remained constant in the DSC thermogram. This indicates that the amorphous glass state of these compounds was highly stable. The glass-transition temperature of
compounds 2 and 3 are 100 and 141
C, respectively. Compounds 2
and 3 have excellent thermal stability, which could be ascribed to
the increase in their molecular size. In particular, the compound 3
showed higher thermal stability compared with carbon-based DBI
compound, TPBi (2-[3,5-bis(1-phenylbenzimidazol-2-yl)phenyl]-1phenylbenzimidazole, Tg 124
C) [24], and this higher Tg value of
compound 3 would gave advantages for the high morphologic
stability of the amorphous phase in a deposited lm, which is a
prerequisite for applications in OLEDs.

1.0
0.8
0.6
0.4

1 absorption
2 absorption
3 absorption
4 absorption

350

1 emission
2 emission
3 emission
4 emission

200

300
250

150
100

0.2

50

0.0
250

300

350

400

450

500

550

0
600

Emission Intensity (a.u)

3.3. Thermal analysis

dened phosphorescence spectra, which were obtained after dissolving them in 2-methyl-tetrahydrofuran (2-MeTHF) at 77 K. In
the phosphorescence spectra, the vibronic sub-band with the
highest energy represented triplet energies (ET) of 14 were
observed; 2.71, 2.69, 2.66, and 2.65 eV for 1, 2, 3, and 4, respectively.
The triplet energies of 14 are higher than carbon-based compound, TPBi (ET 2.63 eV) [24], which indicates that silicon atom
effectively breaks the conjugation and attenuates the triplet wave
function. Furthermore, the triplet energy levels of compounds were
sufciently higher than that of Ir(ppy)3, a phosphorescent dopant
that emits green light. Therefore, we expect that these electrontransporting materials, 1e4, would effectively conne the triplet
excitons on the dopant and prevent the back energy transfer.

Absorption

imidazole ring and the phenyl ring substituted at 2-position was

tilted by 41.09(8)
. On the other hand, the phenyl ring at 1-position
(C17eC18eC19eC20eC21eC22) was twisted further at 59.56(7)
.

129

Wavelength (nm)
Fig. 2. Steady-state absorption and emission spectra of compounds 1e4 in dichloromethane (DCM) solution.

3.4. Photophysical properties

24000

1
2
3
4

22000

Intensity (A. U.)

We measured the steady-state absorption and emission spectra

of compounds 1e4 in dichloromethane (DCM) solution; the
resulting spectral data are summarized in Table 1. As shown in
Fig. 2, all the compounds exhibited similar absorption and emission
peaks of different intensities depending on the number of DMI
units. The absorption peaks around 300 nm were mainly attributed
to the pp* transitions of the DBI unit, while the emission peak at
around 340 nm could be assigned unambiguously to the lowest
pp* transition of the DBI chromophore in the molecule. The absorption peaks of compounds 1 and 2 appeared at 301 and 302 nm,
respectively. In comparative terms, there was a slight red shift in
the absorption peaks of compounds 3 and 4, they appeared at 307
and 306 nm, respectively. As shown in Table 1, the emission peaks
of compounds 3 and 4 also appeared to have a slight red shift. This
red shift can be attributed to the increasing number of DBI units,
which act as substituents in compounds 3 and 4 (vide infra, DFT
calculations). As shown in Fig. 3, all the compounds exhibited well-

20000
18000
16000
14000
12000
10000
8000

450

500

550

600

Wavelength(nm)
Fig. 3. Emission spectra of compounds 1, 2, 3, and 4 in 2-methyl-tetrahydrofuran (2MeTHF) solution at 77 K.

Table 1
Photophysical and electrochemical properties of 14.

1
2
3
4
a
b
c
d
e

Absorption lmax. (nm)a (,  103 (M1 cm1))

Emission lmax. (nm)aat 293 K

Emission lmax. (nm)b at 77 K

Ered (V)c

HOMO (eV)d

LUMO (eV)e

Eg (eV)

301
302
307
306

358
359
361
363

457,
461,
466,
468,

1.85
1.84
1.88
1.88

6.56
6.57
6.50
6.49

2.89
2.90
2.86
2.86

3.67
3.67
3.64
3.63

(105.4)
(149.8)
(235.6)
(260.8)

490,
493,
495,
501,

523
528
523
531

Measured in DCM.
Measured in 2-MeTHF.
Measured in THF.
The HOMO levels were determined using the following equation: EHOMO(eV) ELUMOEg.
The LUMO levels were determined using the following equation: ELUMO(eV) e(Ered(V vs Fc/Fc) 4.74 V).

S. Yi et al. / Organic Electronics 27 (2015) 126e132

3.6. Electron-only devices (EOD)

To compare the electron-transport properties of TPBi, compounds 2 and 3 we fabricated electron-only devices (EODs) containing TPBi and each ET materials. Fig. 5a shows the current
densityevoltage (JeV) characteristics of these EODs with the
structures of ITO (150 nm)/Al (50 nm)/TPBi or 2 or 3 (50 nm)/Liq
(1 nm)/Al (100 nm). Under the same voltages, the EOD containing
compound 3 showed signicantly higher current densities than the
EODs containing TPBi and compound 2 while the EOD containing

(a)
2

Current density (mA/cm )

120
Electron-only devices
EOD with TPBi
EOD with 2
EOD with 3

100
80
60
40
20
0

10 11 12

Voltage (V)

(b)
1E-11
TPBi
2
3

1E-12
2

To investigate the electrochemical properties of compounds

1e4, we performed cyclic voltammetry (CV) by employing a threeelectrode cell system: platinum disk electrode was used as the
working electrode, while a platinum wire and Ag/AgNO3 were used
as counter and reference electrodes, respectively. As shown in
Fig.
S2,
irreversible
reduction
waves
appeared
at
Epc 1.85, 1.84, 1.88, and 1.88 V (versus Fc/Fc) for compounds 1e4, respectively. These peaks could be ascribed to the DBI
group in its reduced form. Using redox potentials, the energy levels
of the highest occupied molecular orbital (HOMO) and the lowest
unoccupied molecular orbital (LUMO) in compounds 1e4 were
estimated as 6.56/2.89, 6.57/2.90, 6.50/2.86, and 6.49/
2.86 eV, respectively (See Table 1). With an increase in the
number of electron withdrawing DBI units, the energy levels of
LUMO were slightly lowered, indicating that the electron-injection
to emissive layer could be improved in the development of OLEDs.
In order to understand transition states, we carried out quantum
chemical calculations (DFT/B3LYP/6-31G(d,p)) using Gaussian09
package [20]. Unfortunately, we could not decipher an optimized
structure for compound 4 due to discrepancy in the calculations
that were carried out in repeated trials. However, using DFT
calculation, we found that the optimized geometry of compound 1
(vide infra) had excellent agreement with crystal structures (SI,
Table S4). Therefore, although we could not obtain the crystal
structures of compounds 2 and 3, we surmised their optimized
geometries using DFT calculation. Thereafter, we analyzed their
transition states using these optimized geometries. Spatial distributions for the frontier orbitals of 1, 2, and 3 are depicted in Fig. 4. In
compound 1, the HOMO was mainly localized in benzoimidazole
moiety and phenyl rings. Furthermore, LUMO was also distributed
in benzoimidazole moiety and phenyl rings; the phenyl group
attached to the 1-position slightly contributed to the distribution of
LUMO. In compound 2, the HOMO was mainly localized in benzoimidazole moiety and phenyl rings in one of the two DBI groups,
while the LUMO was localized in the other DBI group: the localization of LUMO was mainly attributed to benzoimidazole and
phenyl rings, the phenyl group attached to 1-position also slightly
contributed to electron localization in LUMO. As the HOMO and
LUMO energy levels were almost separated spatially, we propose
that the HOMOeLUMO excitation restricts the electron density
distribution to one side of compound 2. The HOMO of compound 3
was distributed on the two of DBI groups, while the LUMO was
delocalized over two DBI groups. Thus, as an increasing number of
DBI units are attached to the silicon atom, intramolecular energy
transfer would be feasible. Thus, DBI units serve as electrontransporting channels.
To determine the functional groups associated with the excited

states in the absorption spectra of compounds 1e3, we calculated

the singlet excited states with the help of TD-DFT calculations.
Thereafter, as shown in Fig. S3, we compared the spectra with the
experimental data. The results indicated that the simulated absorption spectra agreed well with the experimental spectra. Based
on the TD-DFT results of compounds 1e3, we infer that the absorption bands in the region of 260e350 nm could be predominantly attributed to the excitation of electrons from HOMO to
LUMO. Furthermore, few electrons must have undergone the
following transition: HOMO1 / LUMO.

1E-13

3.5. Electrochemical properties and theoretical calculations

J/E (A/V )

130

1E-14
1E-15
1E-16
500

600

700

800
1/2

900
1/2

1000 1100 1200

1/2

E (V /cm )

Fig. 4. Spatial distributions for the frontier orbitals of compounds 1, 2, and 3.

Fig. 5. (a) JeV characteristics of electron-only devices (EODs) for TPBi, compounds 2
and 3. Devices were fabricated with the following structures: ITO (150 nm)/Al (50 nm)/
2 or 3 (50 nm)/Liq (1 nm)/Al (100 nm). (b) The JeV characteristics for TPBi, compounds
2
and
3
were
tted pto
SCLC
with
eld
dependent
mobility.

J 9=80 r E2 =Lm0 exp0:89b E, where J is the current density; 0r is the dielectric constant; E is the electric eld; L is the lm thickness; m0 is the zero-eld carrier
mobility; and b is the Poole Frenkel factor. The values of m0 and b were calculated by
assuming r is equal to 3. The eld
p dependent mobility was calculated using Poole
Frenkel equation: m m0 expb E.

S. Yi et al. / Organic Electronics 27 (2015) 126e132

compound 2 did lower current densities than the EOD containing

TPBi. This indicates that the electron mobility of compound 3 is
higher than those of TPBi and compound 2 because TPBi, compounds 2 and 3 have almost the same LUMO level [25]. Murgatroyd

131

presented the space charge limited current (SCLC) model, which is

characterized by eld dependent mobility [26,27]; the electron
mobilities (m) of TPBi, compounds 2 and 3, which were estimated by
this model, were found to be 2.46  105 cm2/V, 1.93  105 cm2/V

Fig. 6. Energy level diagram and chemical structures of the used materials. Device structure: ITO (150 nm)/HATCN (10 nm)/TAPC (90 nm)/CBP:Ir(ppy)3 8% (20 nm)/2 or 3 (60 nm)/
Liq (1 nm)/Al (150 nm).

Fig. 7. (a) JeVeL characteristics, (b) current efciencyecurrent density curves, (c) power efciencyecurrent density curves, and (d) EQE-current density curves of devices I and II.

132

S. Yi et al. / Organic Electronics 27 (2015) 126e132

Table 2
EL performances of the devices with 2 and 3 as ETMs.
Device

Device I
Device II

Turn on (V)

3.0
3.6

Power efciency (lm/W)

Current efciency (Cd/A)

EQE (%)

Max

1000 nits

Max

1000 nits

Max

1000 nits

CIE

61.8
57.7

28.0
19.0

62.8
63.7

58.0
52.5

18.0
18.1

16.6
14.9

(0.30,0.62)
(0.29,0.63)

Device congurations: Device I: ITO (150 nm)/HATCN (10 nm)/TAPC (90 nm)/CBP:Ir(ppy)3 8% (20 nm)/3 (60 nm)/Liq (1 nm)/Al (150 nm), device II: ITO (150 nm)/HATCN
(10 nm)/TAPC (90 nm)/CBP:Ir(ppy)3 8% (20 nm)/2 (60 nm)/Liq (1 nm)/Al (150 nm).

and 3.67  105 cm2/V at 1 MV/cm, respectively. While performing

these calculations, the relative dielectric constant was assumed to
be equal to 3. As compounds 2 and 3 have the same DBI unit, we
attributed the improved electron transport ability of compound 3
to an increasing number of DBI units, resulting in intramolecular
interactions between DBI units.

Development Program (10041556) funded by the Ministry of Trade,

industry & Energy (MI, Korea), Basic Science Research Program
through the National Research Foundation of Korea (NRF) funded
by the Ministry of Education (NRF-2014R1A6A1030732), and Basic
Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future
Planning (2014R1A1A1005982 and 2014R1A1A1007625).

3.7. Green PhOLEDs performances

Encouraged by the excellent charge transport property of
compound 3, we fabricated a phosphorescent OLED device using
compound 3 as an electron transport material (device I). We also
prepared an OLED device containing compound 2. Thereafter, we
compared these two devices. Fig. 6 illustrates the chemical structures of materials used for device fabrication. Using the iridium
complex fac-tris(2-phenylpyridine) iridium (Ir(ppy)3) as a green
emitter, we fabricated devices having a structure of ITO (150 nm)/
HATCN (10 nm)/TAPC (90 nm)/CBP:Ir(ppy)3 8% (20 nm)/2 or 3
(60 nm)/Liq (1 nm)/Al (150 nm). The current densityevoltageeluminance (JeVeL) characteristics, such as current efciency vs current density, power efciency vs current density, and
EQE vs current density are depicted in Fig. 7, and the performances
of devices are summarized in Table 2.
In accord with the JeV characteristics of EODs, the current
densities of device I are higher than those of device II at the same
voltage (Fig. 7a). The higher electron mobility of compound 3 improves the charge carrier balance in device I. Therefore, compared
to device II, device I has a lower turn-on voltage but higher efciency at practical brightness, and lower efciency roll-off. The
power efciency, current efciency, and EQE of device I were
28 lm W1, 58 cd A1, and 16.6% at 1000 nits, respectively.
4. Conclusions
A series of benzoimidazole-based electron transporting materials (ETL) were designed and employed in organic light-emitting
diodes (OLEDs), which exhibited green phosphorescence. In this
diode, we used the tetrahedral structural motif of silicon atom. By
increasing the number of DBI units, we gradually enhanced thermal
stability but maintained their photophysical and electrochemical
properties. The compound 3 showed high thermal stability at
141
C; it also exhibited a high triplet energy of 2.64 eV, and a high
electron mobility of 3.67  105 cm2/V at 1 MV/cm. Compounds 2
and 3 were used as an electron transporter for green phosphorescent devices I and II, respectively. Compared with device II, device I
showed much better performances. Compound 3 effectively
transports electrons into the emissive layer of devices, resulting in
high efciencies at an EQE of 18.0%, current efciency of 62.8 cd A1,
and power efciency of 61.8 lm W1. Moreover, device I shows low
efciency roll-off at high current density or luminescence.
Acknowledgment
This research was supported by the Industrial Core Technology

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.orgel.2015.08.029.

Conict of interest
The authors declare that they have no competing nancial
interest.

References
[1] C.W. Tang, S.A. VanSlyke, C.H. Chen, J. Appl. Phys. 65 (1989) 3610.
[2] M.A. Baldo, D.F. O'Brien, Y. You, A. Shoustikov, S. Sibley, M.E. Thompson,
S.R. Forrest, Nature 395 (1998) 151.
[3] S.J. Su, T. Chiba, T. Takeda, J. Kido, Adv. Mater. 20 (2008) 2125.
[4] F. So, J. Kido, P. Burrows, MRS Bull. 33 (2008) 663.
[5] T.D. Anthopoulos, P.J. Jonathan Markham, E.B. Namdas, Ifor D.W. Samuel, Appl.
Phys. Lett. 82 (2003) 4824.
[6] S.W. Culligan, A.C.-A. Chen, J.U. Wallace, K.P. Klubek, C.W. Tang, S.H. Chen,
Adv. Funct. Mater. 16 (2006) 1481.
[7] G.G. Malliaras, J.C. Scott, J. Appl. Phys. 83 (1998) 5399.
[8] S. Naka, H. Okada, H. Onnagawa, Y. Yamaguchi, T. Tsutsui, Synth. Met.
111e112 (2000) 331.
[9] B. Krummachera, M.K. Mathai, V. Choongb, S.A. Choulis, F. So, A. Winnacker,
Org. Electron. 7 (2006) 313.
[10] C. Adachi, T. Tsutsui, S. Saito, Appl. Phys. Lett. 56 (1990) 799.
[11] H. Tang, F. Li, J. Shinar, Appl. Phys. Lett. 71 (1997) 2560.
[12] M. Thelakkat, H.W. Schmidt, Polym. Adv. Technol. 9 (1998) 429.
[13] M.Y. Lai, C.H. Chen, W.S. Huang, J.T. Lin, T.H. Ke, L.Y. Chen, M.H. Tsai, C.C. Wu,
Angew. Chem. Int. Ed. 47 (2008) 581.
[14] Y. Li, M.K. Fung, Z. Xie, S.T. Lee, L.S. Hung, J. Shi, Adv. Mater. 14 (2002) 1317.
[15] C.H. Chen, W.S. Huang, M.Y. Lai, W.C. Tsao, J.T. Lin, Y.H. Wu, T.H. Ke, L.Y. Chen,
C.C. Wu, Adv. Funct. Mater. 19 (2009) 2661.
[16] W.S. Han, H.J. Son, K.R. Wee, K.T. Min, S. Kwon, I.H. Suh, S.H. Choi, D.H. Jung,
S.O. Kang, J. Phys. Chem. C 113 (2009) 19686.
[17] M. Debeaux, M.W. Thesen, D. Schneidenbach, H. Hopf, S. Janietz, H. Krger,
A. Wedel, W. Kowalsky, H.H. Johannes, Adv. Funct. Mater. 20 (2010) 399.
[18] SMART and SAINT, Bruker Analytical X-Ray Division, Madison, WI, 2002.
[19] G.M. Sheldrick, SHELXTL-PLUS Software Package, Bruker Analytical X-Ray
Division, Madison, WI, 2002.
[20] M.J. Frisch, et al., Gaussian 09, Revision D.01, Gaussian Inc., Wallingford, CT,
2013.
[21] Cambridge Soft Corporation, Cambridge, MA, USA, http://www.cambridgesoft.
com/.
[22] K. Lonsdale, International Tables for X-ray Crystallography vol. III, D. Reidel
Publishing Company, 1983, p. 275.
[23] http://www.chem.gla.ac.uk/~louis/software/ortep3/download.html.
[24] F. Wang, J. Hu, X. Cao, T. Yang, Y. Tao, L. Mei, X. Zhang, W. Huang, J. Mater.
Chem. C 3 (2015) 5533.
[25] W. Brtting, S. Berleb, A.G. Mckl, Org. Electron. 2 (2001) 1.
[26] P.N. Murgatroyd, J. Phys. D. Appl. Phys. 3 (1970) 151.
[27] T.Y. Asuda, Y. Yamaguchi, D.C. Zou, T. Tsutsui, Jpn. J. Appl. Phys. 41 (2002)
5626.