Universal electron transporting layers via mixing two homostructure molecules with different polarities for organic light-emitting diodes

Article abstract of DOI:10.1016/j.orgel.2021.106220

In general, electron transport layer (ETL) in organic light-emtting diodes (OLEDs) consists of single component of electron transporting material (ETM) or a mixture with n-dopant such as 8-hydroxyquinolinolato-lithium (Liq). However, there exists a limit to controlling a wide range of carrier density in OLEDs according to the required characteristics of the devices due to electrically insulating property of Liq. Here, we suggest a universal strategy to construct an efficient ETL. We synthesized two ETMs, diphenyl-[4-(10-phenyl-anthracene-9-yl)-phenyl]-amine (An-Ph) and phneyl-[4-(10-phenyl-anthracene-9-yl)-phenyl]-pyridin-3-yl-amine (An-Py) that have the same core structures with different polarities in functional groups. The electrical characteristics of electron-only-devices (EODs) were investigated by space charge limited current (SCLC) modeling and impedance spectroscopy analysis. Interestingly, the homostructure type ETL composed of An-Ph and An-Py showed not only superior electron transporting capability, but also the possibility of controlling electron injection and transporting in a wide range compared to the heterostructure type ETL of An-Ph and Liq. Compared to the An-Ph-only EOD, the electron mobility in 75% An-Py-mixed homostructure EOD increased by almost 4 orders of magnitude. Such dramatic variation of electron mobility was achieved thanks to the molecular design strategy to separate charge injection and charge transport regions within a molecule, which consequently induced the giant surface potential (GSP) effect between the ETL/cathode interface. As a result, the external quantum efficiency (EQE) of blue fluorescent and phosphorescent OLEDs with the homostructure ETLs was enhanced by 28.6% and 34%, respectively, compared to that of each control device without manipulating outcoupling effects.

Full text of DOI:10.1016/j.orgel.2021.106220

Organic Electronics 96 (2021) 106220
Contents lists available at ScienceDirect
Organic Electronics
journal homepage: www.elsevier.com/locate/orgel
Universal electron transporting layers via mixing two homostructure
molecules with different polarities for organic light-emitting diodes
,
*
Ki Ju Kima, Hakjun Leea, Kyo Min Hwanga, Bubae Parka, Hyoung Yun Ohb
,
,c,**
Young Kwan Kima, Taekyung Kima
a Department of Information Display, Hongik University, Seoul, 121-791, South Korea
b LORDIN, 614 Dongtangiheung-ro, Hwaseong-si, Gyeonggi-do, 18469, South Korea
c Department of Materials Science and Engineering, Hongik University, Sejong, 30016, South Korea
A R T I C L E I N F O
A B S T R A C T
Keywords:
In general, electron transport layer (ETL) in organic light-emtting diodes (OLEDs) consists of single component of
electron transporting material (ETM) or a mixture with n-dopant such as 8-hydroxyquinolinolato-lithium (Liq).
However, there exists a limit to controlling a wide range of carrier density in OLEDs according to the required
characteristics of the devices due to electrically insulating property of Liq. Here, we suggest a universal strategy
to construct an efficient ETL. We synthesized two ETMs, diphenyl-[4-(10-phenyl-anthracene-9-yl)-phenyl]-amine
(An-Ph) and phneyl-[4-(10-phenyl-anthracene-9-yl)-phenyl]-pyridin-3-yl-amine (An-Py) that have the same core
structures with different polarities in functional groups. The electrical characteristics of electron-only-devices
(EODs) were investigated by space charge limited current (SCLC) modeling and impedance spectroscopy anal-
ysis. Interestingly, the homostructure type ETL composed of An-Ph and An-Py showed not only superior electron
transporting capability, but also the possibility of controlling electron injection and transporting in a wide range
compared to the heterostructure type ETL of An-Ph and Liq. Compared to the An-Ph-only EOD, the electron
mobility in 75% An-Py-mixed homostructure EOD increased by almost 4 orders of magnitude. Such dramatic
variation of electron mobility was achieved thanks to the molecular design strategy to separate charge injection
and charge transport regions within a molecule, which consequently induced the giant surface potential (GSP)
effect between the ETL/cathode interface. As a result, the external quantum efficiency (EQE) of blue fluorescent
and phosphorescent OLEDs with the homostructure ETLs was enhanced by 28.6% and 34%, respectively,
compared to that of each control device without manipulating outcoupling effects.
OLED
Electron transporting layer
Giant surface potential
1. Introduction
charge transporting materials [1–3]. However, it takes a lot of time and
cost to develop new charge transport materials according to device ar-
Organic light-emitting diodes (OLEDs) consist of an emission layer,
charge injection, and transport layers between two electrodes. Both ef-
ficiency and lifetime of OLED devices strongly depend on the balance of
electron and hole which are injected from the electrodes and transported
into the emission layer (EML). Therefore it is very important to control
the balance of charge injection for a variety of chromophores and host
materials for EMLs. In order to maximize the device performances out of
red, green, and blue EMLs simultaneously, not only various EML mate-
rials are investigated, but also charge transporting materials are devel-
oped. Thus, the research on OLEDs has been mainly conducted to
improve and optimize the charge balance in the EML by developing new
chitectures, EML materials, and the required device characteristics. As
widely known, the same hole transporting layer (HTL) has been used
since the OLED panels were commercially available for cellphones.
However, the electron transporting layer (ETL) has been continuously
and newly developed for every generation of cellphones with
newly-developed EMLs. In general, ETL is composed of an electron
transporting material (ETM) and an additive material in commercial
OLED products. Thus, for the evolution of OLED technologies, both new
ETMs and the optimized mixing ratio of the additive material should be
continuously developed. As additive materials in ETLs, n-dopant mate-
rials such as alkali metals or air-stable precursor molecules [4–8]. In the
* Corresponding author.
** Corresponding author. Department of Information Display, Hongik University, Seoul, 121-791, South Korea.
E-mail addresses: ohygo@lordin.net (H.Y. Oh), taekyung.kim@hongik.ac.kr (T. Kim).
https://doi.org/10.1016/j.orgel.2021.106220
Received 23 February 2021; Received in revised form 21 May 2021; Accepted 25 May 2021
Available online 7 June 2021
1566-1199/© 2021 Elsevier B.V. All rights reserved.

K.J. Kim et al.
Organic Electronics 96 (2021) 106220
OLED industry, 8-hydroxyquinolinolato-lithium (Liq) has been widely
used in the ETL for elongated device lifetime. However several issues
still remain with Liq. First, the charge transporting characteristics
strongly depend on the co-deposited ETMs. Second, there is a limit to
control electrical mobility and current density in a high electric field
depending on the content of Liq due to the electrically insulating nature
and variation of intermolecular packing property [9,10]. Finally, Liq is
known to form a trimer or tetramer structure under thermal stress,
affecting an undesirable effect on the charge balance in EML due to its
small size [11,12].
Here, we propose a universal methodology to construct an Liq-free
ETL that can adjust the current density of OLEDs in a wide range by
mixing two ETMs with the same core and different polarities due to
functional groups. The molecular design strategy was to separate charge
injection and charge transport regions within a molecule. Diphenyl-[4-
(10-phenyl-anthracene-9-yl)-phenyl]-amine (An-Ph) and phneyl-[4-
(10-phenyl-anthracene-9-yl)-phenyl]-pyridin-3-yl-amine (An-Py) were
designed and synthesized. The fused aromatic anthracene was chosen as
core units where electrons are delocalized and the overlap of intermo-
lecular wavefunctions easily occurs. On the other hand, functional
groups have similar structures, but different polarities induced by
phenyl and heteroaromatic rings. Since the size of the entire molecules
are almost identical and the molecular weights are similar, the deposi-
tion temperature and thermal stability would not be significantly
different.
Fig. 1. (a) The calculated LUMO of An-Ph and An-Py (b) Illustration of electron
injection from Al atoms in cathode to non-bonding sp2 orbital in pyridine. Cloud
shape represents the coordination reaction between Al and nitrogen.
atom in cathode to non-bonding sp2 orbital in pyridine of An-Py.
Furthermore, molecular dipole moment,
μ (D) of An-Py and An-Ph was
calculated to be 2.317 D and 0.2809 D, respectively. Thus, it is expected
that the degree of PDM will be increased when An-Py is mixed with
An-Ph.
2.2. Synthesis and photophysical and electrochemical analysis
As shown in Scheme 1, An-Ph and An-Py were synthesized by two-
step and one-step procedure, respectively. At first, 9-(4-bromophenyl)-
10-phenylanthracene was prepared through Suzuki coupling reaction
with (10-phenyl-9-anthryl)boronic acid and 1-chloro-4-iodo-benzene.
After then, Buchwald-Hartwig reaction was carried out with diphenyl-
amine to obtain An-Ph. Pre-purchased 9-(4-chlorophenyl)-10-phenyl-
anthracene and N-phenyl-3-pyridinamine were used to execute the same
amination reaction to acquire An-Py. The final products were purified by
recrystallization and vacuum sublimation to remove undesirable im-
When permanent dipole moment (PDM) induced by the asymmetric
molecular structure is present in a deposited film, spontaneous orien-
tation polarization (SOP) can be formed in a particular direction, which
induces counter charges on the heterointerface of the organic layer
under the electric field. The magnitude of the SOP results in a thickness-
dependent internal potential gradient, called giant surface potential
(GSP) effect, which was reported to improve the charge injection per-
formance in organic photovoltaic devices [13–15]. According to
Ref. [16] the dipole moment of EML can be adjusted by mixing polar and
non-polar materials at different ratios [16], thus a polar An-Py and a
non-polar An-Ph were mixed in a wide range (0–75 wt%) to construct an
efficient ETL.
purities. To confirm the chemical structure of final products, 1H and 13
C
nuclear magnetic resonance and mass spectroscopy were carried out
(See Figure S2~ S4).
In addition, we divided fabricated devices into two groups, homo-
structure (An-Ph mixed with An-Py) and heterostructure (An-Ph mixed
with Liq) type and investigated their electrical characteristics with
respect to the mixing ratio through space charge-limited current (SCLC)
theory and impedance spectroscopy. Finally, up to 33.7% and 34.4% of
efficiency enhancement with homostructure ETLs were confirmed in
blue fluorescent and phosphorescent OLEDs, respectively.
2.3. Physical properties
Fig. 2 shows UV–vis absorption and photoluminescence spectra of
An-Ph and An-Py in dilute solution. The vibrational bands at 357–397
nm wavelength originated from 0 to 4 to 0–1 transition of anthracene
moiety were measured in both An-Ph and An-Py [21]. The emission
energies of An-Py and An-Ph were estimated to be 2.65 and 2.54 eV from
peak wavelength, respectively. The electrochemical properties of An-Ph
and An-Py were measured by cyclic voltammetry (CV). ELUMO was ob-
tained from the onset of reduction potential (See Figure S5). ELUMO was
obtained from the formalism of ELUMO = EHOMO + Eg, where Eg is the
onset energy of the absorption spectra. Therefore, the HOMO/LUMO
levels of An-Ph and An-Py were estimated to be ꢀ 5.34/-2.41 and
ꢀ 5.46/-2.40 eV, respectively. The HOMO and LUMO levels of An-Py
were slightly lower than those of An-Ph due to electron-withdrawing
pyridine. The thermal properties of the molecules were measured and
their results are shown in Figure S6 and S7. The thermal decomposition
temperature (Td) of An-Ph and An-Py was found to be 367 and 329 ◦C,
respectively. The glass transition temperature (Tg) of An-Ph and An-Py
was 101.7 and 100.2 ◦C, respectively. The high Td ensures a good sur-
face morphology and high thermal stability for vacuum deposition in
OLED fabrication.
2. Results and discussion
2.1. Material design and calculation of electronic properties
Density functional theory (DFT) and time-dependent DFT (TD-DFT)
calculations were carried out for ground state molecules. The geometries
in the gas phase were optimized by the DFT method using the B3LYP
exchange-correlation function with the 6-31G* basis set in the Gaussian
09 program package [17]. Fig. 1(a) and Figure S1 show the electronic
populations of the lowest unoccupied molecular orbital (LUMO) state
and the highest occupied molecular orbital (HOMO) state of An-Ph and
An-Py, respectively. Regarding the HOMO states, the electron density of
An-Ph was localized on triphenyl amine, while those of An-Py on
anthracene. However, the electron density of the LUMO states in both
molecules was localized on the same anthracene core. Thus, it is ex-
pected that an electron pathway can be easily formed due to the overlap
of wavefunctions between two molecules when An-Py and An-Ph are
mixed. It has been reported that the interaction between nitrogen in
heteroaromatic rings of molecule and cathode metal, called coordina-
tion reaction reduces the electron injection barrier at ETL/metal surface
[18–20]. Fig. 1(b) illustrates an electron injection process from metal
2.4. Analysis of electron-only devices
The electron transporting characteristics of An-Ph and An-Py were
confirmed with electron-only devices (EODs). An-Py or Liq was mixed
with An-Ph in the EODs. The detailed device structure of the EODs is
2

K.J. Kim et al.
Organic Electronics 96 (2021) 106220
Scheme 1. Synthetic procedures of An-Ph and An-Py.
current injection defined by the transition voltage from Ohmic to trap
charge-limited (TCLC) region was also improved as shown in Figure S9
(a). The variation of the transition voltage as a function of An-Py mixing
ratio can be attributed to the increased dipole moment in the ETL called
giant surface potential (GSP) effect. The GSP caused by doping polar
molecules generates a polarization of ETL. Consequently, the polariza-
tion of ETL induces counter-charge in the interface between the ETL and
cathode, lowering the electron injection barrier [22]. The GSP effect is
proportional to the size of the dipole moment in the organic layer and
relies on layer thickness as well as on the doping concentration of the
polar molecule [15]. To confirm the GSP effect in homostructure ETL,
the capacitance measurements were carried out for the EODs at 1 kHz
frequency as shown in Figure S9(b). When the thickness of the ETL
increased from 40 nm to 60 nm, the geometrical capacitance decreased
since the capacitance is reciprocal to thickness. However, the voltage for
carrier injection (Vin) was reduced from 2.24 V to 2.12 V, which clearly
proves the GSP is present in the homostructure ETL. The charge injection
mechanism of An-Ph and An-Py is schematically displayed in Figure S10.
Since the GSP induces an internal electric field, the slope of energy levels
are different in An-Ph and An-Py ETLs. Therefore, the flat band voltage
for electron injection in An-Py is lower than that of An-Ph. On the other
hand, the current density of Liq-mixed EODs exhibited significantly
different behavior from the An-Py-mixed EODs. Similar to 25%
An-Py-mixed EOD, the current density of 25% Liq-mixed EOD increased.
However, the concentration of Liq increased from 25% to 50%, the
current density surprisingly decreased. Moreover, the 75% Liq-mixed
EOD showed not only the significantly reduced current density, but
also the unstable current injection and transport behavior. This unusual
electrical behavior can be understood in terms of the poor electron
transporting property of Liq [9,23].
Fig. 2. UV absorption (dot) and PL spectra (solid line) of An-Ph and An-Py in
DCM at room temperature.
described in the experimental section. Fig. 3(a) and (b) show the current
density (J)-voltage (V) curves of the An-Ph-based EODs as a function of
An-Py or Liq mixing ratio. Among the homostructure EODs, the device
without An-Py mixed showed the lowest current density. However, as
the mixing ratio of An-Py increased, the current density drastically
increased. Figure S8 shows the J-V characteristics of An-Py-only and An-
Ph-only EODs with and without EIL. The difference in electron injection
and transport properties is distinct in the EODs without EILs. The elec-
trons from Al cathode can be injected even without the help of EIL.
Recently, coordination reaction between nitrogen-containing heterocy-
cles with diffused Al atoms from cathode can reduce the electron in-
jection barrier and facilitate the electron injection from cathode [20].
Even with Liq inserted as an EIL, An-Py showed much superior electron
injection and transport properties to An-Ph. According to Ref. [20], the
role of Liq as EIL is to simply reduce the workfunction of cathode due to
its low charge neutrality level. These results strongly indicate that enven
though LUMO levels of An-Ph and An-Py are almost identical, the
electron injection from cathode to An-Ph does not occur efficiently since
no coordination reaction between An-Ph and Al cathode is present. The
electrons from cathode are easily and efficiently transferred to
non-bonding sp2 orbital in pyridine of An-Py and transported through
anthracene cores as illustrated in Fig. 1(b). In addition to the enhance-
ment of current density as a function of the An-Py mixing ratio, the
For further analysis of charge transporting properties, SCLC model
was applied. SCLC model is a versatile method to obtain field-dependent
mobility of organic layer. The field-dependent mobility can be expressed
as Eqn. (1) taking into account the Pool-Frenkel effect [24–26], where
μ0 is zero-field mobility, β is field activation of mobility and E is an
electric field.
(
)
√̅̅̅
μ
=
μ0 exp β
E
(1)
A monopolar SCLC conforming to Eqn. (1) proposed by Murgatroyd
can be approximated as follows [27].
(
)
E2
√̅̅̅
9
8
J = εε μ0 exp 0.89β
E
(2)
0 L
3

K.J. Kim et al.
Organic Electronics 96 (2021) 106220
Fig. 3. The current density (J)- voltage (V) and space charge limited current–electric fields characteristics of (a) and (c) An-Py- and (b) and (d) Liq-mixed EODs. The
black lines are numerical fitting curves to the space charge limited behavior by Eqn. (2).
The dielectric constant εε0 of organic bulk layer was assumed to be 3,
and we estimated μ0 and β from Eqn. (2). The fitting results are shown as
the black solid lines in Fig. 3(c) and (d) and the calculated values are
summarized in Table 1. The zero field mobility, μ0 corresponds to charge
hopping distance and molecular packing proclivities, and the electric
field activation β is related to the degree of energetic disorder of organic
film [24,28]. Thus, the greater the degree of μ0 and/or β in terms of both
morphology and density of state (DOS) is, the higher the charge mobility
is. In our experimental results, the degree of energetic disorder (β) was
not much different in the EODs. However, the zero field mobility of the
An-Py-mixed EODs was up to 50 times larger than that of the Liq-mixed
EODs at 75% mixing ratio. Therefore, as clearly shown in Table 1, the
large difference in current density and the electron mobility was caused
by the large difference in the zero field mobility rather than the electric
field activation. Figure S11 shows schematic diagrams of An-Py- and
Liq-mixed films under the electric field. In the homostrucutre ETL, the
electron transporting pathway is formed by the orbital overlap of
anthracene cores in An-Ph and An-Py molecules, resulting in a long
charge hopping distance as in Figure S11(a). On the contrary, even
though the electrons are injected from Liq molecules in EIL directly to
Liq or to An-Ph molecules in the ETL, the orbital overlap between An-Ph
molecules is significantly reduced as the mixing ratio of Liq increases as
shown in Figure S11(b). Thus, the hopping distance with Liq mixed was
significantly reduced.
In order to further analyze the charge injection and transporting
properties in the ETLs, impedance spectroscopy (IS) measurements were
carried out. The IS a widely-used non-destructive method to analyze
charge dynamics of organic photovoltaic devices and OLEDs. An OLED
can be represented with an equivalent circuit consisting of series and
parallel resistances and capacitors as Eqn. (3) [29–32].
R
wR2C
Z = Re(Z) + iIm(Z) = RL +
ꢀ i
(3)
1 + w2R2C2
1 + w2R2C2
RL is the contact resistance, R is the resistance of organic layer, w is
the angular frequency in the measurement, C is the capacitance. When
voltage is applied, two types of current flow are present. One is the
displacement current that flows through the capacitor C of the organic
layer and the other is the conductive current through the resistance R
[33]. Since the organic layers are sandwiched between anode and
cathode, the dielectric capacitance is present in the OLEDs. However,
very little displacement current flows in low voltage regime as a result of
the space charge and low carrier mobility of the organic materials. As
the voltage increases, the conductive current begins to flow through the
resistance of the organic material at the maximum point of Re(Z) when
the reactance and the resistance become the same as described in Eqn.
(4).
Table 1
Summary of zero-field mobility, field activation of mobility, mobility at 1 MV/
cm for EODs.
1
R
2
R
2
Re(Z)max
=
= RL +
≈
(4)
Mixing
ratio
An-Ph: An-Py x %
An-Ph: Liq x %
ω
C
μ0 (cm
β (cm/
μ
(cm
μ0 (cm
β (cm/
μ (cm
In other words, Re(Z)max is the point at which the main current flow
is converted from displacement to conductive current, and the voltage of
Re(Z)max can be regarded as carrier injection voltage (Vin) from elec-
trode to organic layer. Fig. 4 shows Re(Z) of the EODs as a function of
voltage measured under 1 kHz, 50 mV AC bias conditions. First, in the
case of homostructure type EODs, Vin at 0, 25, 50, 75, and 100% of An-
Py mixing ratio gradually decreased to be 3.27, 2.48, 2.36, 2.30, and
2.24 V, respectively (see Fig. 4(a) inset). Note that the magnitude of Re
(Z)max decreases, which is also consistent with the increase of the current
density in SCLC region in Fig. 3(a). Therefore, it is shown that the charge
[2]/Vs)
V)
[2]/Vs)
[2]/Vs)
V)
[2]/Vs)
0%
4.23 ×
10ꢀ 10
2.49 ×
10ꢀ 9
0.141
2.80 ×
4.23 ×
10ꢀ 10
1.21 ×
10ꢀ 9
0.141
2.80 ×
10ꢀ 9
10ꢀ 9
25%
50%
75%
100%
0.185
0.188
0.192
0.194
7.95 ×
10ꢀ 7
0.184
0.171
0.168
–
7.04 ×
10ꢀ 7
1.18 ×
4.47 ×
8.17 ×
10ꢀ 10
4.16 ×
10ꢀ 9
2.29 ×
10ꢀ 8
10ꢀ 6
10ꢀ 7
2.62 ×
10ꢀ 8
1.18 ×
10ꢀ 5
3.09 ×
10ꢀ 7
3.29 ×
1.66 ×
–
–
10ꢀ 8
10ꢀ 5
4

K.J. Kim et al.
Organic Electronics 96 (2021) 106220
Fig. 4. The Re(Z) – Voltage characteristics of EODs.
injection barrier and the electrical conductivity can be controlled ac-
cording to the magnitude of the dipole moment of the ETL through
mixing homostructure molecules with different polarities. In the case of
the heterostructure EODs, as the mixing ratio of Liq increased from 25 to
50%, Vin surprisingly decreased while Re(Z)max increased. However,
neither Vin nor Re(Z)max was not clearly defined in the 75% Liq-mixed
EOD even though the role of Liq is to help charge injection for the ex-
istence of free Li with low workfunction between the cathode and ETL
[23]. This result strongly suggests that it is extremely difficult to control
the charge injection and the transport property at the same time by
mixing Liq in a wide range.
2.5. Performances of OLEDs
Since the hole mobility is generally much higher than the electron
mobility [34], the bottleneck of perfect charge balance and high effi-
ciency is to supply electrons into EML. To test the capability of the
electron injection and transporting by the homostructure ETL, we
Fig. 5. The device performances of FOLEDs (left) and PhOLEDs (right). (a) and (b) current density-voltage, (c) and (d) luminance-voltage, and (e) and (f) external
quantum efficiency versus luminance.
5

K.J. Kim et al.
Organic Electronics 96 (2021) 106220
fabricated a series of fluorescent OLEDs (FOLEDs) and phosphorescent
OLEDs (PhOLEDs). In addition, 50 wt% Liq-mixed heterostructure ETL
was also applied for comparison. The device structures are described in
Experimental section and energy level diagram of FOLEDs and PhO-
LEDs are represented in Figure S12.
EML. At last, The FOLED and PhOLED with 50 wt% Liq doped ETL
showed higher EQEmax (3.47%, 10.03%) than 0% device (2.82%,
9.72%), but the EQE roll-off were significantly severe than the other
devices. Hence, we confirmed that the homostructure ETL with An-Ph
and An-Py can be used universally for OLED, while Liq can not have
wide range of mixing ratio in ETL affecting device performance. The
electroluminescent spectra of FOLEDs and PhOLEDs are shown in Figure
S13.
The current density (J)-voltage (V)-luminance (L) of FOLEDs and
PhOLEDs are shown in Fig. 5(a)~(f). The turn-on voltages of 0% An-Py-
mixed FOLED and PhOLED were 4.36 V and 4.20 V at 1 cd/m2,
respectively due to the inefficient electron injection by non-polar An-Ph.
However, once An-Py was mixed with An-Ph, the turn-on voltage was
significantly reduced down to 3.49 V and 3.88 V for 25% An-Py-mixed
FOLED and PhOLED, respectively. Such dramatic decreases can be
attributed to the efficient electron injection process from EIL to non-
bonding sp2 orbital in pyridine of An-Py and the GSP effect. Interest-
ingly, the turn-on voltages were not decreased significantly in the 50 and
75% An-Py-mixed devices. This result implies that the number of non-
bonding sp2 orbital sites and the intensity of the GSP were almost
saturated even in 25% An-Py-mixed ETL. However, compared to the
turn-on voltages, the driving voltages at 10 mA/cm2 of current density
gradually decreased as the An-Py mixing ratio increased, indicating that
the electron transporting property was sensitively affected by An-Py
mixing ratos. In 50% Liq-mixed FOLED and PhOLED, 10.82 V and
9.75 V of the driving voltages at 10 mA/cm2 were obtained, respectively
due to the electrically insulating property of Liq. The electrical trans-
porting properties of the FOLEDs and the PhOLEDs well corresponded to
the results of the EOD experiments.
3. Conclusion
In this study, we proposed a univseral method to construct ETL called
homostructure type that mixes two ETMs with the same core but with
different polarity substituents. The electrical characteristics of the newly
designed ETL were investigated through SCLC modeling and impedance
measurements. As a consequence, the electron injection and transport
capabilities were continuously improved depending on the mixing ratio
in homostructure ETL, which has superior electron transport charac-
teristics to the ETL with Liq. The mobility of homostructure type EOD
doped 50% is 4.47 × 10ꢀ 6 cm/V in 1 MV/cm electric field, which is 20
times higher than that of heterostructure type EOD with Liq at the same
condition. The GSP effect and continuously decreasing the electron in-
jection voltage of homostructure type ETL were confirmed by imped-
ance measurement, which shows the possibility of controlling the
electron density of the newly proposed ETL. These results are directly
related to the performances of blue fluorescent and phosphorescent
OLED following improvement of their device efficiency up to 33.7% and
34.4% respectively, as well as the driving characteristics. Consequently,
the ETL with two homostructure materials that can easily manipulate
electron transport characteristics is expected to be an effective alterna-
tive to conventional ETLs with Liq in display manufacturing.
As well known, the EQE of OLED consists of the following factors,
which are charge balance (γ), radiative exciton ratio (ηS/T), photo-
luminescence quantum yield (ΦPL), and outcoupling efficiency (ηOUT) as
shown in Eqn. (5).
EQE = γ × ηS/T × φPL
×
ηOUT
(5)
4. Experimental section
As the mixing ratio increased, the EQEmax also increased in the
FOLEDs. With an assumption of ηS/T PL, and ηOUT to be 0.25, 0.79 [35],
Φ
4.1. Physical measurements
and 0.2, respectively, the charge balances of 0, 25, 50, and 75%
An-Py-mixed FOLEDs were 0.71, 0.83, 0.91, and 0.95, In our FOLEDs,
the increase of electron transporting into EML proportionally increase
the charge balance and the device efficiency. On the other hand, the
EQEmax of the PhOLEDs shows a different feature from that of the
FOLEDs. The largest EQEmax was obtained from the PhOLED with 25%
An-Py-mixed ETL. mCP is a widely-used host with hole transporting
characteristics, which implies the host in the PhOLEDs is generally in an
electron-deficient state. Thus, as the electron transporting increased by
varying the mixing ratio of An-Py, the efficiency of the PhOLED was
expected to increase accordingly. However, as the An-Py mixing ratio
increased over 25%, the long-living triplet excitons were densely
populated in the EML, leading to undesirable triplet-triplet annihilation
and/or triplet-polaron quenching processes. In both the FOLED and the
PhOLED with 50% Liq-mixed ETL, the EQEmax increased from 2.82% to
9.72%–3.47% and 10.03%, respectively as in An-Py-mixed devices.
However, the roll-off characteristics of the Liq-mixd devices were
distinctively different those of the An-Py-mixed devices. As the device
performances were summarized in Table S1, when An-Py was mixed in
the ETL, the EQEs at 10 mA/cm2 maintained over 90% and 70% of the
EQEmax of the FOLEDs and the PhOLEDs, respectively. The new concept
of ETL by utilizing homostructure molecules with different polarities in
this work clearly demonstrated a strategy to construct a universal ETL
which operates in a wide range of current density and simultaneously
exhibits superb roll-off characteristics and efficiency in fluorescent and
phosphorescent devices.
To obtain HOMO and LUMO of An-Ph and An-Py, cyclic voltammetry
(CV) was conducted in nitrogen-purged dichloromethane with tetrabu-
tylammonium hexafluorophosphate (TBAPF6) (0.1 M) as the supporting
electrolyte at room temperature by electrochemical analyzer
(PGSTAT101, E-Chem Technology). We used ferrocenium-ferrocene
(Fc+/Fc) as standard material to calculate HOMO/LUMO of organic
materials. The UV–vis absorption results and PL spectra were measured
with dichloromethane as solvent (10ꢀ 5 M) using a UV–Visible spec-
troscopy (Agilent 8453) and fluorescence spectrometer (LS55, Perki-
nElmer precisely). Thermogravimetric analysis (TGA) and differential
scanning calorimetry (DSC) was carried out under 100 ◦C/min condition
in nitrogen atmosphere by Seiko Exstar 6000 instrument.
4.2. OLED fabrication
Indium tin oxide (ITO) coated glass with sheet resistance of 30 Ω/sq
and thickness of 150 nm were cleaned in ultrasonic baths (acetone,
methyl alcohol, distilled water, and ethyl alcohol) for 15 min each.
Before depositing organic layers, pre-cleaned ITO-coated glass sub-
strates were dried in a convection oven at 120 ◦C for 10 min. After then,
the ITO/glass substrates were treated with O2 plasma at 2 × 10ꢀ 2 Torr
and 125 W condition for 2 min. Organic layers were deposited by
thermal evaporator with shadow mask under high vacuum (5 × 10ꢀ 7
Torr) condition. Aluminum (Al) cathode was deposited at a rate of 10 Å/
s. The deposition rates were controlled with a quartz crystal monitor. In
encapsulation process, glass cap and UV-epoxy resin were used to pre-
vent degradation from ambient condition. The getter was attached
under the glass cap to absorb the residue moisture in the encapsulated
device, which has an active layer of 3 × 3 mm2.
The charge balance effect by homostructure ETL in PhOLEDs also can
be noticed through current efficiency (See Table S1). There is significant
increase from 19.25 cd/A in 0% device to 25.11 cd/A in 25% device and
the current efficiency starts to decrease in highly doped devices. It is due
to charge unbalance resulting from excessive injection of electrons into
The EODs were fabricated as the following structure: ITO/TmPyPB
6

K.J. Kim et al.
Organic Electronics 96 (2021) 106220
(5 nm)/An-Ph:An-Py (or Liq) x wt.% (40 nm)/Liq (2 nm)/Al (100 nm),
where x = 0, 25, 50, 75, 100 for homostructure EOD and 0, 25, 50 for
heterostructure EOD.
References
[1] C.J.T. Abhishek, P. Kulkarni, Babel Amit, Samson A. Jenekhe, Chem. Mater. 16
(2004) 4556–4573.
The detailed device stack of emitting devices are ITO (150 nm)/2-
TNATA (30 nm)/NPB (10 nm)/MADN:Pyrene-CN 2 wt% (20 nm)/
TmPyPB (5 nm)/ETL (40 nm)/Liq (2 nm)/Al (100 nm) for FOLEDs and
ITO (150 nm)/2-TNATA (40 nm)/NPB (10 nm)/mCP (10 nm)/mCP:
FIrpic 8 wt% (20 nm)/TmPyPB (5 nm)/ETL (40 nm)/Liq (2 nm)/Al (100
nm) for PhOLEDs, where ETL is composed of An-Ph:An-Py(or Liq) x wt.
%, where x = 0, 25, 50, 75 for homostructure ETL and 50 for hetero-
structure ETL. In the device structure, 2-TNATA is 4,4′,4′′-Tris [2-
naphthyl (phenyl)amino]triphenylamine, NPB is N,N′-Di (1-naphthyl)-
N,N′-diphenyl-(1,1′-biphenyl)-4,4′-diamine, TmPyPB is 1,3,5-Tri (m-
[2] S. Kwon, K.-R. Wee, A.-L. Kim, S.O. Kang, J. Phys. Chem. Lett. 1 (2009) 295–299.
[3] H.O. Shigeke Naka, Hiroyoshi Onnagawa, Yoshihisa Yamaguchi, Tetsuo Tsutsui,
Synth. Met 111 (2000) 3.
[4] J.-H. Lee, M.-H. Wu, C.-C. Chao, H.-L. Chen, M.-K. Leung, Chem. Phys. Lett. 416
(2005) 234–237.
[5] Z. Bin, L. Duan, Y. Qiu, ACS Appl. Mater. Interfaces 7 (2015) 6444–6450.
[6] Y. Chen, X. Wei, Z. Li, Y. Liu, J. Liu, R. Wang, P. Wang, Y. Yamada-Takamura,
Y. Wang, J. Mat. Chem. C 5 (2017) 8400–8407.
[7] F.-S. Juang, A. Chittawanij, L.-A. Hong, Y.-S. Tsai, K.-K. Huang, IEICE Trans.
Electron. E98.C (2015) 66–72.
[8] D.P. Tsang, T. Matsushima, C. Adachi, Sci. Rep. 6 (2016) 22463.
[9] H.M. Kim, J.H. Seo, W.K. Han, K.H. Lee, H.J. Kang, S.S. Yoon, Y.K. Kim, Mol. Cryst.
Liq. Cryst. 538 (2011) 53–60.
pyridin-3-ylphenyl)benzene.
2-Methyl-9,10-bis(naphthalen-2-yl)
[10] S.H. Choi, S.J. Lee, K.Y. Lee, H.S. Yang, K.-i. Han, K.H. Kim, S.D. Kim, H.D. Bae, Y.
H. Tak, Jpn. J. Appl. Phys. (2009) 48.
anthracene (MADN) and 1,3-Bis(N-carbazolyl)benzene (mCP) were
used as host. (1,6-Bis(N-phenyl–CN–phenylamino)-pyrenes (Pyrene-CN)
and Bis [2-(4,6-difluorophenyl)pyridinato-C2, N]iridium (III) (FIrpic)
were used for sky blue fluorescent and phosphorescent dopant
respectively.
[11] M. Rajeswaran, W.J. Begley, L.P. Olson, S. Huo, Polyhedron 26 (2007) 3653–3660.
[12] K.H.K. Man Jae Park, Olga Egorova, Change Woo Lee, Kyung Lae Rho, Nari Ahn,
2019. SID 2019.
¨
[13] L. Jager, T.D. Schmidt, W. Brütting, AIP Adv. 6 (2016).
[14] K. Osada, K. Goushi, H. Kaji, C. Adachi, H. Ishii, Y. Noguchi, Org. Electron. 58
(2018) 313–317.
[15] Y. Noguchi, W. Brütting, H. Ishii, Jpn. J. Appl. Phys. 58 (2019).
¨
[16] T. Morgenstern, M. Schmid, A. Hofmann, M. Bierling, L. Jager, W. Brütting, ACS
4.3. OLED characterizations
Appl. Mater. Interfaces 10 (2018) 31541–31551.
[17] K. Nasu, T. Nakagawa, H. Nomura, C.-J. Lin, C.-H. Cheng, M.-R. Tseng, T. Yasuda,
C. Adachi, Chem. Commun. 49 (2013).
The J-V-L characteristics and EL performance were measured using a
source meter (model 2400, Keithley) and Chroma meter (CS-1000A)
instrument. Impedance analyzer (Alpha-AN, Novocontrol) were used to
obtain Re(Z). All measurements were conducted at room temperature.
[18] Z. Bin, G. Dong, P. Wei, Z. Liu, D. Zhang, R. Su, Y. Qiu, L. Duan, Nat. Commun. 10
(2019) 866.
[19] H. Fukagawa, K. Suzuki, H. Ito, K. Inagaki, T. Sasaki, T. Oono, M. Hasegawa,
K. Morii, T. Shimizu, Nat. Commun. 11 (2020) 3700.
[20] T. Sasaki, M. Hasegawa, K. Inagaki, H. Ito, K. Suzuki, T. Oono, K. Morii, T. Shimizu,
H. Fukagawa, Nat. Commun. 12 (2021) 2706.
[21] S. Kang, J.S. Huh, J.J. Kim, J. Park, J. Mater. Chem. C 8 (2020) 11168–11176.
¨
[22] S. Altazin, S. Züfle, E. Knapp, C. Kirsch, T.D. Schmidt, L. Jager, Y. Noguchi,
Declaration of competing interest
W. Brütting, B. Ruhstaller, Org. Electron. 39 (2016) 244–249.
[23] O.V.S. Zugan Liu, Nigel Male, Synth. Met 128 (2002) 4.
[24] T. Yasuda, Y. Yamaguchi, D.-C. Zou, T. Tsutsui, Jpn. J. Appl. Phys. 41 (2002)
5626–5629.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
[25] J.C. Blakesley, F.A. Castro, W. Kylberg, G.F.A. Dibb, C. Arantes, R. Valaski,
M. Cremona, J.S. Kim, J.-S. Kim, Org. Electron. 15 (2014) 1263–1272.
[26] J.R.S.G.G. Malliaras, P.J. Brock, c. Scott, Phys. Rev. B 58 (1998) 4.
[27] P.N. Murgatroyd, J. Phys. D Appl. Phys. 3 (1970) 6.
[28] J.J.R.R.H. Young, J. Phys. Chem. 99 (1995) 11.
[29] X.-W. Zhang, J.-W. Xu, H.-R. Xu, H. Wang, C.-L. Xie, B. Wei, X.-Y. Jiang, Z.-
L. Zhang, J. Phys. Appl. Phys. 46 (2013).
Acknowledgment
This research was supported by MOTIE (20012556), Basic Science
Research Program through the National Research Foundation of Korea
[30] J. Bisquert, L. Bertoluzzi, I. Mora-Sero, G. Garcia-Belmonte, J. Phys. Chem. C 118
(2014) 18983–18991.
(NRF)
funded
by
the
Ministry
of
Education
(No.
[31] T. Okachi, T. Nagase, T. Kobayashi, H. Naito, Thin Solid Films 517 (2008)
1331–1334.
2015R1A6A1A03031833) and 2021 Hongik University Innovation
Support program Fund.
[32] H. Lee, K.J. Kim, Y.J. Moon, Y.K. Kim, T. Kim, Org. Electron. (2020) 84.
[33] S. Lee, J. Hur, S. Lim, J. Soc. Inf. Disp. 18 (2010).
[34] I.W. Wu, P.-S. Wang, W.-H. Tseng, J.-H. Chang, C.-I. Wu, Org. Electron. 13 (2012)
13–17.
Appendix A. Supplementary data
[35] K.-R. Wee, H.-C. Ahn, H.-J. Son, W.-S. Han, J.-E. Kim, D.W. Cho, S.O. Kang, J. Org.
Chem. 74 (2009) 8472–8475.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.orgel.2021.106220.
7