Design of efficient thermally activated delayed fluorescence materials for pure blue organic light emitting diodes

Article abstract of DOI:10.1021/ja306538w

Efficient thermally activated delayed fluorescence (TADF) has been characterized for a carbazole/sulfone derivative in both solutions and doped films. A pure blue organic light emitting diode (OLED) based on this compound demonstrates a very high external quantum efficiency (EQE) of nearly 10% at low current density. Because TADF only occurs in a bipolar system where donor and acceptor centered ³ππ* states are close to or higher than the triplet intramolecular charge transfer (³CT) state, control of the π-conjugation length of both donor and acceptor is considered to be as important as breaking the π-conjugation between them in blue TADF material design.

Full text of DOI:10.1021/ja306538w

Communication
pubs.acs.org/JACS
Design of Efficient Thermally Activated Delayed Fluorescence
Materials for Pure Blue Organic Light Emitting Diodes
Qisheng Zhang,† Jie Li,† Katsuyuki Shizu,† Shuping Huang,‡ Shuzo Hirata,† Hiroshi Miyazaki,†
and Chihaya Adachi*,†
†Center for Organic Photonics and Electronics Research (OPERA) and ‡Institute for Materials Chemistry and Engineering, Kyushu
University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
S
* Supporting Information
molecules have recently been designed and used successfully to
improve the light emitting capability of OLEDs by our group.6
ABSTRACT: Efficient thermally activated delayed fluo-
rescence (TADF) has been characterized for a carbazole/
sulfone derivative in both solutions and doped films. A
pure blue organic light emitting diode (OLED) based on
this compound demonstrates a very high external quantum
efficiency (EQE) of nearly 10% at low current density.
Because TADF only occurs in a bipolar system where
Thus, pure organic TADF materials take on the advantages of
chemical and thermal stability as compared to those
luminescent Cu(I) complexes, allowing them to be designed,
synthesized, and applied more easily in OLEDs.
In this paper, we report on one promising type of pure blue
TADF material and its design principle. The formulas of the
compounds discussed here are depicted in Figure 1. Reaction of
3
donor and acceptor centered ππ* states are close to or
higher than the triplet intramolecular charge transfer
(3CT) state, control of the π-conjugation length of both
donor and acceptor is considered to be as important as
breaking the π-conjugation between them in blue TADF
material design.
rganic light emitting diodes (OLEDs) containing Ir(III),
OPt(III), or Os(II) based phosphors can harvest both
singlet (25%) and triplet (75%) excitons through heavy atom
enhanced intersystem crossing (ISC), and therefore approach
an external electroluminescence (EL) quantum efficiency
(EQE) that is 4 times higher than that of conventional
fluorescent OLEDs.1 However, the high cost of these
phosphors hinders the prospective application of these devices
in many fields, including mobile displays, large-screen displays,
and general lighting. The development of low cost substitutes
for these noble metal-based phosphors has attracted increasing
attention over the past decade. In 2006, Wang’s group
demonstrated green EL devices based on a Cu(I) complex
that could achieve an EQE as high as 16%, indicating that both
the triplet and singlet excitons generate light in these devices.2
Subsequently, highly efficient EL from Cu(I) complexes was
also confirmed in some other Cu(I) systems.3 Because the
radiative decay rate of a triplet state is less favorable for the first
row transition metal complexes, the emission from these Cu(I)
complexes was believed to arise from an upper state which is in
thermal equilibrium with the lowest triplet state and involves
some singlet characters. In principle, when the energy
difference (ΔEST) between the lowest singlet state (S1) and
the lowest triplet state (T1) is small, reverse ISC can take place
even in pure aromatic organic compounds containing no heavy
metals.4 A small overlap between the highest occupied and
lowest unoccupied molecular orbitals (HOMO and LUMO) is
considered to be an important factor in realization of the small
ΔEST.1b,5 On the basis of this concept, several efficient green
emitting thermally activated delayed fluorescence (TADF)
Figure 1. Molecular structures of compounds 1−3.
diphenylamine, bis(4-tert-butylphenyl)amine or 3,6-di-tert-
butylcarbazole, with sodium hydride and bis(4-fluorophenyl)-
sulfone produces good yields of the compounds (SI section 3).
The absorption and emission spectra of these compounds in
low polar toluene are shown in Figure 2. At room temperature,
all three compounds exhibit broad and structureless emission
bands with a maximum at 402−419 nm, which can be ascribed
to the intramolecular charge-transfer (CT) transition because
of their dipolar nature. In contrast, their phosphorescence
spectra at 77 K are well resolved and show a characteristic
vibrational structure for triphenylamine or 9-phenylcarbazole
3
(see Figure S3), indicating that their T1 states are ππ* states
located mostly on the electron donating moieties. For the 3ππ*
state, the highest energy peak of its emission identifies its zero−
zero energy (E0−0). However, for a CT state without any
vibronic structure, estimation of E0−0 from the onset of its
broad emission band provides a more comparable result, as
demonstrated by some previous studies.3d,7 From Figure 2, the
Received: July 4, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja306538w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
smaller energy difference between its singlet and triplet CT
states.
The photoluminescence quantum yields (PLQY) of 1, 2, and
3 in oxygen-free toluene are 0.57, 0.65, and 0.69, respectively.
The transient photoluminescence (PL) decay characteristics of
these solutions show two-component decays, that is, a fast
decay with a lifetime of 3−5 ns and a slow decay with a lifetime
of 90−270 μs (see Table S1), which can be assigned to
fluorescence and TADF decay, respectively. The proportion of
the slow decay component decreases with an increase in the
ΔEST of these compounds, and almost disappears in the case of
1 (Figure 4). Similar to the case of our previous report,6a the
Figure 2. Absorption and emission spectra of compounds 1−3 in
toluene at 300 K and their phosphorescence spectra (delayed by 10
ms) in toluene at 77 K.
ΔEST values of compounds 1, 2, and 3 are calculated to be 0.54,
0.45, and 0.32 eV, respectively. This change in ΔEST can be
explained well in terms of the modification of the molecular
structures. The introduction of tert-butyl groups on the
diphenylamine unit enhances its electron donating ability, and
consequently lowers the CT energy and the ΔEST. On the
other hand, the replacement of a diphenylamine unit with a
1
carbazole unit slightly raises the CT state, and considerably
raises the 3ππ* state, resulting in a further decrease in its ΔEST.
Time dependent density functional theory (TD-DFT)
calculations can also predict the ΔEST changes of these three
compounds. Using the most popular B3LYP density functional,
the ΔEST values of 1, 2, and 3 were calculated to be 0.65, 0.60,
and 0.34 eV, respectively, based on their ground state
geometries when optimized by the DFT (B3LYP/6-31G*)
method. These results are well matched with the experimental
values. The electron densities of the HOMO and LUMO of
each of these three compounds are depicted in Figures 3 and
Figure 4. Emission decay of compounds 1−3 in toluene at 300 K.
Inset: Emission decay of 3 in oxygen-free hexane, chloroform, and
methanol at 300 K.
slow decay component cannot be observed in an air-
equilibrated solution as a consequence of oxygen quenching.
Here, the observed TADF emission is believed to occur
through a mechanism that involves a reverse internal
3
conversion (IC) from the donor centered ππ* state to the
3
3CT state and a successive reverse ISC from the CT state to
the 1CT state. The reversible energy transfer between the 3ππ*
3
and CT states can be regarded as a rapid intramolecular
electron exchange process, and has been demonstrated by some
transition metal CT complexes with an energy separation of
less than 0.20 eV between the two states.3d,8 Similar to the
reversible ISC process,6b the energy difference between
E
0−0(3CT) and E0−0(3ππ*) determines the reverse IC rate.
The photophysical properties of 3 were also measured in
various solvents with different polarities. In the polar solvents
chloroform and methanol, the emission maximum red-shifts to
446 and 469 nm (Figure S5a), respectively, and the lifetime of
the slow decay decreases dramatically to 56 and 83 μs (Figure 4
inset), respectively, compared to the values of 404 nm and 270
μs in toluene. Conversely, in the nonpolar solvent hexane, the
emission maximum blue-shifts (Figure S5a), while the slow
decay component disappears (Figure 4 inset). To explain this
change in emission decay, we consider that a CT state with a
large dipole moment change is more easily stabilized by a polar
solvent with respect to the ππ* state, leading to an approach to
Figure 3. The HOMO (lower image) and LUMO (upper image) of
compounds 1 (a) and 3 (b) calculated at the B3LYP/6-31G* level.
The calculated S1/T1 energies of 1 and 3 are 3.54/2.89 eV and 3.31/
2.97 eV, respectively, corresponding to the measured values of 3.30/
2.76 eV and 3.30/2.98 eV, respectively.
3
3
or inversion of the CT and ππ* states. The increased CT
characteristics of its phosphorescence spectra in chloroform
and methanol glasses at 77 K (Figure S5b) implies that the 3CT
state might be the lowest triplet state for 3 in these two solvents
at 300 K, when taking the growing solvation in a fluid solution
into account. Because of the lack of a reversible IC process, the
energy up-conversion from the T1 state to the S1 state becomes
S4. In comparison with the phenylamine derivatives, 1 and 2,
the carbazole derivative 3 shows a greater separation of the
HOMO and LUMO orbitals because of the relatively large
dihedral angle of 49° between the phenyl bridge and the donor
moiety, in contrast with that of ∼32° for 1 and 2. The lower
overlap between the HOMO and the LUMO of 3 indicates a
B
dx.doi.org/10.1021/ja306538w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
much faster in a polar solvent. In contrast, the difference
between E0−0(3CT) and E0−0(3ππ*) increases in nonpolar
hexane solvent, and as a result, the reverse IC process is
hindered and the total PLQY (0.52) is thus lower than that in
the three other solvents (∼0.70).
and S9). The Commission Internationale de L’Eclairage (CIE)
coordinates of the EL of the 3-based device are (0.15, 0.07),
which are very close to those of the National Television
Standards Committee (NTSC) standard blue of (0.14, 0.08).
Note that a pure blue fluorescent OLED generally has an EQE
of less than 3%,9 and pure blue emission with CIEy < 0.10 is still
a critical issue for phosphorescent transition-metal complexes.10
Here, Ir(fppz)2(dfbdp)10e was selected for comparison because
it is the most efficient blue phosphor, with a strong emission
band in the 400−450 nm region. Using the same configuration
described above, an Ir(fppz)2(dfbdp) based device exhibits a
similar maximum EQE of 8.8% with a reduced efficiency roll-off
when compared with that of the 3-based device (see Figure
S10). We believe that the design of an appropriate device
structure in a subsequent study can solve the roll-off problem.
In summary, although a high degree of CT in the excitation
process leads to a small exchange energy between the singlet
By doping of these three compounds in a high triplet energy
host, bis(2-(diphenylphosphino)phenyl)ether oxide (DPEPO),
with a concentration of 10 wt %, the TADF emissions were also
observed in vacuum conditions. The proportion of the delayed
component in the total emission occurs in the order 3 > 2 > 1
(see Figure S7), which agrees with the behavior in toluene. The
emission maximum and the PLQY of 1, 2, and 3 in the DPEPO
films are 421 nm/0.60, 430 nm/0.66, and 423 nm/0.80,
respectively. Using these doped DPEPO films as the emitting
layers (EML), multilayer OLEDs were fabricated with a
configuration of ITO/α-NPD (30 nm)/TCTA (20 nm)/CzSi
(10 nm)/EML (20 nm)/DPEPO (10 nm)/TPBI (30 nm)/LiF
(1 nm)/Al (Figure S8), where α-NPD, TCTA, CzSi and TPBI
represent N,N′-diphenyl-N,N′-bis(1-naphthyl)-1,10-biphenyl-
4,4′-diamine, 4,4′,4″-tris(N-carbazolyl) triphenylamine, 9-(4-
tert-butylphenyl)-3,6-bis(triphenylsilyl)-9H-carbazole and 1,3,5-
tris(N-phenylbenzimidazol-2-yl)benzene, respectively. As
shown in Figure 5a, the maximum EQE of the device based
3
and triplet CT states, the ππ* state is the lowest triplet state
for an aromatic CT compound in most cases. The energy
interchange between its singlet and triplet states can be efficient
3
only when the ππ* state is close to or even higher than the
3CT state. Therefore, to attain efficient blue TADF emission,
both the π-conjugation length and the redox potential of the
donor and acceptor moieties should be taken into account,
along with the interruption of the conjugation between them.
On the basis of this principle, a simple and efficient pure blue
TADF material has been designed and successfully applied to
OLEDs. Its higher EL efficiency relative to fluorescent materials
with similar CIE coordinates indicates that TADF materials
have great potential for OLED applications, even in the pure
blue region where noble metal based phosphors do not work
well.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental details; photophysical and molecular orbital data;
comparison of device data. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
adachi@cstf.kyushu-u.ac.jp
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors wish to thank Prof. Masahiro Kotani for
stimulating discussions with regard to this work. This work
was supported by a Grant-in-Aid from the Funding Program for
World-Leading Innovative R&D on Science and Technology
(FIRST).
Figure 5. (a) The EQE-current density characteristics of the OLEDs
based on compounds 1−3; (b) the EL and PL spectra of 10 wt % 3
doped in a DPEPO layer. Inset: The CIE coordinates of the EL
spectrum of a 3-based device.
REFERENCES
■
(1) (a) Adachi, C.; Baldo, M. A.; Thompson, M. E.; Forrest, S. R. J.
Appl. Phys. 2001, 90, 5048. (b) Highly Efficient OLEDs with
Phosphorescent Materials; Yersin, H., Ed.; Wiley-VCH, Weinheim,
2008. (c) Xiao, L.; Chen, Z.; Qu, B.; Luo, J.; Kong, S.; Gong, Q.; Kido,
J. Adv. Mater. 2011, 23, 926.
(2) Zhang, Q.; Zhou, Q.; Cheng, Y.; Wang, L.; Ma, D.; Jing, X.;
Wang, F. Adv. Funct. Mater. 2006, 16, 1203.
(3) (a) Tsuboyama, A.; Kuge, K.; Furugori, M.; Okada, S.; Hoshino,
M.; Ueno, K. Inorg. Chem. 2007, 46, 1992. (b) Deaton, J. C.; Switatski,
S. C.; Kondakov, D. Y.; Young, R. H.; Pawlik, T. D.; Giesen, D. J.;
on 1 is 2.9%, while the values for the devices based on 2 and 3
increase dramatically to 5.6% and 9.9%, respectively, although
the EQE roll-off of all these devices is significant, probably
because of the unbalanced charge capture ability of these large
band gap emitters in the EML. The EL spectra of these devices
are almost identical to the PL spectra of their EMLs (Figures 5b
C
dx.doi.org/10.1021/ja306538w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Harkins, S. B.; Miller, A. J. M.; Mickenberg, S. F.; Peters, J. C. J. Am.
Chem. Soc. 2010, 132, 9499. (c) Hashimoto, M.; Igawa, S.; Yashima,
M.; Kawata, I.; Hoshino, M.; Osawa, M. J. Am. Chem. Soc. 2011, 133,
10348. (d) Zhang, Q.; Komino, T.; Huang, S.; Matsunami, S.; Goushi,
K.; Adachi, C. Adv. Funct. Mater. 2012, 22, 2327.
(4) Valeur, B. Molecular Fluorescence Principles and Applications;
Wiley-VCH: Weinheim, 2002; p 41.
(5) Klessinger, M.; Michl, J. Excited States and Photochemistry of
Organic Molecules; VCH: Weinheim, 1995.
(6) (a) Endo, A.; Sato, K.; Yoshimura, K.; Kai, T.; Kawada, A.;
Miyazaki, H.; Adachi, C. Appl. Phys. Lett. 2011, 98, 083302. (b) Goushi,
K.; Yoshida, K.; Sato, K.; Adachi, C. Nat. Photonics 2012, 6, 253.
(c) Nakagawa, T.; Ku, S.-Y.; Wong, K.-T.; Adachi, C. Chem. Commun.
2012, DOI: 10.1039/C2CC31468A.
(7) Tyson, D. S.; Castellano, F. N. J. Phys. Chem. A 1999, 103, 10955.
(8) (a) Casadonte, D. J., Jr.; McMillin, D. R. J. Am. Chem. Soc. 1987,
109, 331. (b) Ford, W. E.; Rodgers, M. A. J. J. Phys. Chem. 1992, 96,
2917. (c) Leydet, Y.; Bassani, D. M.; Jonusauskas, G.; McClenaghan,
N. D. J. Am. Chem. Soc. 2007, 129, 8688. (d) Yarnell, J. E.; Deaton, J.
C.; McCusker, C. E.; Castellano, F. N. Inorg. Chem. 2011, 50, 7820.
(9) (a) Liao, S.-H.; Shiu, J.-R.; Liu, S.-W.; Yeh, S.-J.; Chen, Y.-H.;
Chen, C.-T.; Chow, T. J.; Wu, C.-I J. Am. Chem. Soc. 2009, 131, 763.
(b) Yang, Y.; Cohn, P.; Dyer, A. L.; Eom, S.-H.; Reynolds, J. R.;
Castellano, R. K.; Xue, J. Chem. Mater. 2010, 22, 3580. (c) Zhang, y.;
Lai, S.-L.; Tong, Q.-X.; Lo, M.-F.; Ng, T.-W.; Chan, M.-Y.; Wen, Z.-C.;
He, J.; Jeff, K.-S.; Tang, X.-L.; Liu, W.-M.; Ko, C.-C.; Wang, P.-F.; Lee,
C.-S. Chem. Mater. 2012, 24, 61.
(10) (a) Yook, K. S.; Lee, J. Y. Adv. Mater. 2012, 24, 3145.
(b) Holmes, R. J.; Forrest, S. R.; Sajoto, T.; Tamayo, A.; Djurovich, P.
I.; Thompson, M. E.; Brooks, J.; Tung, Y.-J.; D’Andrade, B. W.;
Weaver, M. S.; Kwong, R. C.; Brown, J. J. Appl. Phys. Lett. 2005, 87,
243507. (c) Chang, C.-F.; Cheng, Y.-M.; Chi, Y.; Chiu, Y.-C.; Lin, C.-
C.; Lee, G.-H.; Chou, P.-T.; Chen, C.-C.; Chang, C.-H.; Wu, C.-C.
Angew. Chem., Int. Ed. 2008, 47, 4542. (d) Chiu, Y.-C.; Hung, J.-Y.;
Chi, Y.; Chen, C.-C.; Chang, C.-H.; Wu, C.-C.; Cheng, Y.-M.; Yu, Y.-
C.; Lee, G.-H.; Chou, P.-T. Adv. Mater. 2009, 21, 2221. (e) Hsieh, C.-
H.; Wu, F.-I.; Fan, C.-H.; Huang, M.-J.; Lu, K.-Y.; Chou, P.-Y.; Yang,
Y.-H. O.; Wu, S.-H.; Chen, I.-C.; Chou, S.-H.; Wong, K.-T.; Cheng,
C.-H. Chem.Eur. J. 2011, 17, 9180.
D
dx.doi.org/10.1021/ja306538w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX