Highly improved electroluminescence from a series of novel Eu111 complexes with functional single-coordinate phosphine oxide ligands: Tuning the intramolecular energy transfer, morphology, and carrier injection ability of the complexes

Article abstract of DOI:10.1002/chem.200700678

The functional single-coordinate phosphine oxide ligands (4-diphenylaminophenyl)diphenylphosphine oxide (TAPO), (4-naphthalen-1-yl- phenylaminophenyl)diphenylphosphine oxide (NaDAPO), and 9-[4- (diphenylphosphinoyl)phenyl]-9H-carbazole (CPPO). as the direct combinations of hole-transporting moieties, and electron-transporting triphenylphosphine oxide (TPPO) were designed and synthesized (amines or carbazole), together with their Eu^III complexes [Eu(tapo)₂(tta)₃] (1). [Eu(nadapo):(tta)₃] (2), and [Eu(cppo)₂(tta)₃] (3; TTA: 2-thenoyltrifluoroacetonate). The investigation indicated that by taking advantage of the modification inertia of the phosphine oxide ligands, the direct introduction of the hole-transport groups as chromophore made TAPO, NaDAPO. and CPPO obtain the most compact structure and mezzo Si and T ₁ energy levels, which improved the intramolecular energy transfer in their Eu^III complexes. The amorphous phase of 1-3 proved the weak intermolecular interaction, which resulted in extraordinarily low self-quenching of the complexes. The excellent double-carrier transport ability of the ligands was studied with Gaussian calculations, and the bipolar structure of TAPO and CPPO was proved. The great improvement of the double-carrier transport ability of 1-3 was shown by cyclic voltammetry. Their HOMO and LUMO energy levels of around 5.3 and 3.0 eV, respectively, are the best results for Eu^III complexes reported so far. A single-layer organic light-emitting diode of 2 had the impressive brightness of 59 cd m ^-2 which, to the best of our knowledge, is the highest reported so far. Both of the four-layer devices based on pure 1 and 2 had a maximum brightness of more than 1000 cdm^-2, turn-on voltages lower than 5 V, maximum external quantum yields of more than 3% and excellent spectral stability.

Full text of DOI:10.1002/chem.200700678

FULL PAPER
DOI: 10.1002/chem.200700678
Highly Improved Electroluminescence from a Series of Novel EuIII
Complexes with Functional Single-Coordinate Phosphine Oxide Ligands:
Tuning the Intramolecular Energy Transfer, Morphology, and Carrier
Injection Ability ofthe Complexes
Hui Xu,[b] Kun Yin,[c] and Wei Huang*[a]
Abstract: The functional single-coordi-
nate phosphine oxide ligands (4-diphe-
nylaminophenyl)diphenylphosphine
oxide (TAPO), (4-naphthalen-1-yl-phe-
nylaminophenyl)diphenylphosphine
oxide (NaDAPO), and 9-[4-(diphenyl-
phosphinoyl)phenyl]-9H-carbazole
(CPPO), as the direct combinations of
hole-transporting moieties, and elec-
tron-transporting triphenylphosphine
oxide (TPPO) were designed and syn-
thesized (amines or carbazole), togeth-
er with their EuIII complexes [Eu-
troduction of the hole-transport groups
as chromophore made TAPO,
CPPO was proved. The great improve-
ment of the double-carrier transport
ability of 1–3 was shown by cyclic vol-
tammetry. Their HOMO and LUMO
energy levels of around 5.3 and 3.0 eV,
respectively, are the best results for
EuIII complexes reported so far. A
NaDAPO, and CPPO obtain the most
compact structure and mezzo S1 and
T1 energy levels, which improved the
intramolecular energy transfer in their
EuIII complexes. The amorphous phase
of 1–3 proved the weak intermolecular
interaction, which resulted in extraordi-
narily low self-quenching of the com-
plexes. The excellent double-carrier
transport ability of the ligands was
studied with Gaussian calculations, and
the bipolar structure of TAPO and
single-layer
organic
light-emitting
diode of 2 had the impressive bright-
ness of 59 cdmꢀ2 which, to the best of
our knowledge, is the highest reported
so far. Both of the four-layer devices
based on pure 1 and 2 had a maximum
brightness of more than 1000 cdmꢀ2,
turn-on voltages lower than 5 V, maxi-
mum external quantum yields of more
than 3% and excellent spectral stabili-
ty.
A
N
G
N
A
ACHTREUNG
thenoyltrifluoroacetonate). The investi-
gation indicated that by taking advant-
age of the modification inertia of the
phosphine oxide ligands, the direct in-
Keywords: electroluminescence
europium hole transport
ligand design · P ligands
·
·
·
Introduction
For decades, lanthanide complexes have been studied as
emitters in organic light-emitting diodes (OLEDs) with the
advantages of a nearly monochromic characteristic emission,
chemical environmental stability, and approaching 100%
theoretical internal device quantum efficiency.[1–8] However,
for many reasons the electroluminescence (EL) perfor-
mance of these lanthanide materials is still much lower than
what people expect. The unbalanced hole and electron in-
jection and transport should be one of the most important
factors. In most lanthanide complexes, the materials tend to
transport electrons rather than holes.[1] Therefore, much
work was done to improve the hole-transport ability of the
lanthanide complexes through modification with hole-trans-
port moieties. Hole-transport moieties, such as N-ethylcar-
[a] Prof. W. Huang
Institute of Advanced Materials (IAM)
Jiangsu Key Laboratory of Organic Electronics
and Flat-Panel Displays
Nanjing University of Posts and Telecommunications (NUPT)
66 XinMoFan Road, Nanjing 21003 (China)
Fax : (+86)25-8349-2333
E-mail: wei-huang@njupt.edu.cn
[b] Dr. H. Xu
School of Chemistry and Materials
Heilongjiang University (HLJU)
74 Xuefu Road, Harbin 150080 (China)
[c] K. Yin
Institute of Advanced Materials (IAM)
Fudan University
220 Handan Road, Shanghai 200433 (China)
Chem. Eur. J. 2007, 13, 10281 – 10293
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10281

bazolyl, were introduced into the b-diketone ligand.[9] Fur-
thermore, since the most popular neutral ligands, such as
1,10-phenanthroline (phen), triphenylphosphine oxide
(TPPO) and 2,2’-bipyridine (bpy), also only exhibit very
good electron-transport ability,[10,11] the hole-transporting tri-
phenylamine and carbazole groups were introduced into
phen or bpy derivatives to obtain a double-carrier transport
ability.[12–15] Thus, the EL performance of these complexes
was much improved.
tion of the phosphine oxide ligand did not influence its coor-
dination ability.
In this work, we designed and prepared (4-diphenylami-
nophenyl)diphenylphosphine oxide (TAPO), (4-naphthalen-
1-yl-phenylaminophenyl)diphenylphosphine
oxide
(NaDAPO) and 9-[4-(diphenylphosphinoyl)phenyl]-9H-car-
bazole (CPPO) as neutral ligands. The tertiary complexes
[Eu
A
(tta)3] (1), [Eu
G
N
A
ACHTREUNG
The effect of these modifications was inspiring. However,
because the lone electron pair of the N atom in phen and
bpy is involved in the whole conjugated system, the intro-
duction of the modifying groups would influence the distri-
bution of the electron cloud of the whole nitrogen heterocy-
clic ring and further influence the coordination ability of the
ligands. To avoid reducing the coordination ability of these
phen or bpy derivatives, buffer groups were introduced be-
tween the hole-transport moieties and the coordination moi-
eties. However, the introduction of the buffer groups made
the synthetic routines much more complicated. Notably, the
EL efficiencies of the OLEDs based on these complexes
were not as high as expected. The intramolecular energy-
transfer process in EuIII complexes follows the Dexter mech-
anism,[16,17] and thus the introduction of the buffer groups in-
creased the volume of these ligands, lengthened the distance
between the ligands and EuIII ions, reduced the overlap of
their electron clouds and as a result decreased the efficiency
of the intramolecular energy transfer. So it is attractive and
significant to design and synthesize highly efficient electro-
luminescent EuIII complexes with multifunctional neutral li-
gands with good double-carrier transport ability and com-
pact structure.
phosphine oxide ligands were also prepared (see Scheme 1).
By taking advantage of the modification inertia of the phos-
phine oxide ligands, these functional ligands were designed
as the direct combinations of hole-transporting moieties (di-
phenylamine, naphthalenyl phenylamine and carbazole),
and electron-transporting TPPO to obtain double-carrier
transport ability without reducing the intramolecular energy
transfer. The compact structure of these ligands is helpful in
making the intramolecular energy transfer more direct and
efficient.
Compared with [Eu
N
N
much better PL performance. The improvement mechanism
of intramolecular energy transfer by TAPO, NaDAPO, and
CPPO was suggested by measuring their excited energy
levels. We studied the frontier orbital energy levels of the
functional ligands and complexes 1–3 by Gaussian calcula-
tions and cyclic voltammetry (CV), respectively. It was
shown that complexes 1–3 obtain excellent double-carrier
transport ability with the lowest unoccupied molecular orbi-
tal (LUMO) energy levels around 3.0 eV and the highest oc-
cupied molecular orbital (HOMO) energy levels around
5.3 eV. To the best of our knowledge, this is the best result
for the frontier orbital energy levels of electroluminescent
EuIII complexes.
Furthermore, the complexes unexpectedly exhibited an
amorphous state and extraordinarily low self-quenching.
Stable red emissions from the single-layer and four-layer
OLEDs based on pure complexes 1–3 were demonstrated.
The maximum brightness of the single-layer device based on
2 was 59 cdmꢀ2. We believe that this is the best performance
from single-layer devices based on pure EuIII complexes.
Both of the four-layer devices based on 1 and 2 had a maxi-
mum brightness of more than 1000 cdmꢀ2, turn-on voltages
lower than 5 V, external quantum yields of more than 3%,
and excellent spectral stability.
Tppo has been used as the neutral ligand in electrolumi-
nescent TbIII complexes. The best electroluminescent TbIII
complex [Tb
N
N
4-(2-ethylbutyryl)-5-pyrazolone) reported so far used TPPO
as the neutral ligand.[18] Compared with phen derivatives,
TPPO exhibits good electron-transport ability and strong co-
ordination ability. However, when using TPPO in EuIII com-
plexes, their EL performance, such as brightness and effi-
ciency of the corresponding OLED devices, was very limit-
ed.[19] The most important factor is that the first singlet ex-
cited energy level (S1) of TPPO is much higher than that of
anionic ligands, such as 2-thenoyltrifluoroacetonate (TTA).
The energy gap of about 1.5 eV between them is too large
to decrease the energy-transfer efficiency from TPPO to
TTA. Also, the electron-transporting TPPO could not afford
balanceable carrier injection and transport, which is another
important reason for the low EL efficiency.
Results and Discussion
Design and synthesis ofthe ufnctional single-coordinate
phosphine oxide ligands: On account of the volume effect
mentioned above, we thought that designing double-carrier-
transport neutral ligands with the most compact structure
would be very significant for comparison. In our former
work, the phosphine oxide ligands exhibited very strong
modification inertia. This advantage gave us a routine to
design double-carrier-transport neutral ligands based on
phosphine oxide compounds with the simplest and most
Recently, we reported the novel EuIII complex [Eu-
A
N
bis[2-(diphenylphosphino)phenyl] ether oxide (dpepo).[20]
As a result of the larger conjugate area of dpepo, the energy
gap between S1 levels of dpepo and TTA was reduced to
0.8 eV. The photoluminescence (PL) and EL performance of
[Eu
A
G
N
ACHTREUNG
10282
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10281 – 10293

Electroluminescence of Eu Complexes
FULL PAPER
Scheme 1. Synthetic route to TAPO, NaDAPO, and CPPO and their complexes [Eu
HTTA: 2-thenoyltrifluoroacetone.
A
(tta)3] (1), [Eu
U
(tta)3] (2), and [Eu
N
N
compact structure, which are the direct combinations of
hole-transport moieties and TPPO moieties. These function-
al single-coordinate phosphine oxide ligands have the most
compact structure, as TPPO here serves not only as the co-
ordinating group but also as the electron-transport group
and no buffer group was used between the hole-transport
groups and coordinating TPPO. Their compact structure can
reduce the distance between the ligands and EuIII ions,
which would increase the efficiency of the Dexter energy
transfer in the corresponding EuIII complexes. During the
process of EL the relatively short distance between the li-
gands and EuIII ions would make the ions more easily and
efficiently excited through the process of charge-transfer ex-
citation (Scheme 2). Simultaneously, because of their bipolar
structure, hole and electron can be readily injected into
these functional phosphine oxide ligands and their corre-
sponding complexes. Compared with TPPO and its EuIII
complex, the hole-transport ability of the ligands and their
Chem. Eur. J. 2007, 13, 10281 – 10293
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10283

W. Huang et al.
peak at 292.5 nm mainly originate from CPPO. In the solid,
all of the complexes have a wide absorbance peak from 300
to 400 nm. In the PL spectra (excitation at 380 nm), besides
the four main peaks at 579, 593, 611, and 653 nm, which are
the characteristic emissions from EuIII corresponding to
5D0–7Fj (j=0–3) transitions, wide and much weaker peaks in
the short-wavelength range are recognized as the emissions
from TAPO, NaDAPO, or CPPO. Further studies of the PL
spectra of 1 in different solvents (excitation at 360 nm)
showed that with increasing polarity of the solvents, the in-
creasing emission intensity of TAPO became stronger and
the emissions from EuIII were weakened markedly
(Figure 2). This finding proved that the solvent effect should
be one of the most important factors for this inefficient in-
tramolecular energy transfer, which induces the visible emis-
sions from TAPO, NaDAPO, and CPPO. The emission of
TAPO was red-shifted about 10 nm in ethanol solution,
which was induced by the protonic solvent effect on the p–
p* transition of TAPO.
As single-coordinate ligands, the structures of complexes
1–3 are not rigid enough such that the small solvent mole-
cules could enter the complexes, increase the distance be-
tween the phosphine oxide ligands and the EuIII centre, and
further reduce the Dexter energy-transfer efficiency. Due to
their coordination ability, THF and ethanol amplified the
solvent effect by entering the complexes as solvent ligands.
However, the intensity of the short-wavelength emissions
was much smaller than that of the emission from EuIII when
excited by an appropriate wavelength in CH2Cl2, which
means that only a small fraction of the energy of these neu-
tral ligands is given out as their own emissions rather than
transfers to EuIII. Furthermore, in the solid state the emis-
sions from these phosphine oxide ligands nearly disappear
and the main emissions of complexes 1–3 are at 615 nm
(Figure 1).
Scheme 2. Effect of the functional single-coordinate phosphine oxide li-
gands in their EuIII complexes.
complexes is much improved, and a balance of the hole and
electron injection and transport in the complexes can be ex-
pected.
These phosphine oxide ligands can be synthesized by a
simple three-step process. Amines or carbazole first under-
went the Ullmann reaction with 4-bromoiodobenzene to
yield the corresponding bromides. The bromides were then
successively reacted with n-butyllithium, chlorodiphenyl-
phosphine, and hydrogen peroxide to form the correspond-
ing phosphine oxide ligands. These functional phosphine
oxide ligands were characterized by NMR spectroscopy,
mass spectrometry (MS), and elemental analysis. Their EuIII
complexes were synthesized according to a well-established
procedure.[26]
Optical properties: The UV/Vis absorbance and PL spectra
of complexes 1–3 in dilute CH2Cl2 solutions (1”
10ꢀ5 molLꢀ1) and as the solids were measured (Figure 1). In
solution, complexes 1 and 2 have similar UV/Vis character-
istics with one main peak around 330 nm, originating from
the neutral ligands TAPO and NaDAPO, and another
shoulder peak at shorter wavelength contributed by the
anionic ligand TTA and the neutral ligands, respectively;
the main absorbance peak of 3 at 338.5 nm and a sharp
The relative PL quantum yields of complexes 1–3 were
measured by using [Ru
A
a standard (see
Table 1). Compared with [Eu
A
ACHTREUNG
Figure 1. UV and PL spectra of 1–3 in CH2Cl2 solution and in the solid
state (insert). c: 1, a: 2, g: 3.
Figure 2. PL spectra of 1 in different solvents. c: in CH2Cl2, a: in
THF, g: in C2H5OH.
10284
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10281 – 10293

Electroluminescence of Eu Complexes
FULL PAPER
Table 1. The relative PL quantum yield (QY) of the complexes.
neutral ligands are higher than
that of TTA (Figure 5). Alter-
natively, compared with TPPO
the introduction of the amines
and carbazole as chromophore
groups into phosphine oxide
compounds remarkably reduces
Complex
Absorption peak [nm]
PL emission peaks [nm]
CH2Cl2 Film
Relative PL QY [%]
26.7
CH2Cl2
Film
[Eu
1
2
A
G
–
580, 617, 680
–
352
364
341
408, 579, 593, 611, 653 579, 592, 615 39.4
414, 579, 593, 611, 653 579, 592, 615 36.1
383, 579, 593, 611, 653 579, 593, 615 42.5
266, 335
293, 339
3
relative PL quantum yields of 1–3 are improved much more.
It is shown that the introduction of chromophore groups,
such as amines and carbazole, can improve the PL proper-
ties of the corresponding complexes. As steric effects induce
a decrease of the intramolecular energy-transfer efficiency,
the compact neutral ligands in 1–3 can limit this effect.
Intramolecular energy transfer in the complexes: The ab-
sorption and PL spectra of the ligands (TAPO, NaDAPO,
CPPO, and TTA) were measured (Figure 3). There are over-
laps between the emission peaks of these neutral ligands
and the absorbance peak of TTA, which shows the possibili-
ty of energy transfer from the ligands to TTA.
Figure 4. Smoothed phosphorescence spectra of the GdIII complexes con-
taining TAPO, NaDAPO, or CPPO. c: curve of [Gd
N
G
a: curve of [Gd(NO3)3(Nadapo)2], d: curve of [Gd(NO3)3
A
G
A
ACHTREUNG
Table 2. The singlet- and triplet-state energy levels of ligands.
Ligand
S1 energy level [eV]
T1 energy level [eV]
TAPO
3.50
3.32
3.54
4.512.35
3.12
2.90
2.92
2.84
NaDAPO
CPPO
TPPO[22]
TTA[20]
2.35
Figure 3. UV and PL spectra of the ligands TAPO, NaDAPO, CPPO, and
TTA in CH2Cl2 solution. c: curves of TAPO, a: curves of
*
NaDAPO, d: curves of CPPO, : curves of TTA.
The first triplet excited energy levels (T1) of TAPO,
NaDAPO, and CPPO were obtained by measuring the phos-
phorescence spectra of the gadolinium complexes based on
these neutral ligands corresponding to their peak emission
wavelengths (2.90 (428), 2.92 (425), and 2.84 eV (437 nm);
Figure 4). The first singlet excited energy levels (S1) of
TAPO, NaDAPO, and CPPO are estimated by referencing
their absorbance edges, which are 3.50 (354.5), 3.32 (373.5),
and 3.54 eV (350 nm). For comparison, the singlet- and trip-
let-state energy levels are summarized in Table 2 and illus-
trated in Figure 5.
The energy absorbed by the neutral ligands can be trans-
ferred to the S1 level of TTA, as all of the S1 levels of these
Figure 5. Schematic energy-level diagram and the energy-transfer pro-
cess: S1, first excited singlet state; T1, first excited triplet state.
Chem. Eur. J. 2007, 13, 10281 – 10293
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10285

W. Huang et al.
their S1 levels. The degrees of reduction of the S1 levels of
TAPO and CPPO are equivalent, and because of the longer
conjugated length of naphthalenyl moieties NaDAPO has
the lowest S1 level. In [Eu
N
N
from the S1 level of TPPO to that of TTA is inefficient, as
their energy gap is as large as 1.4 eV. Clearly, by modifica-
tion with chromophore groups, the S1 levels of TAPO,
NaDAPO, and CPPO are reduced to fit that of TTA. In par-
ticular, for TAPO and CPPO the mezzo energy gaps of
about 0.4 eV between their S1 levels and that of TTA make
the energy transfer between them more efficient. Notably,
contrary to the reduction of their S1 levels, the T1 levels of
TAPO, NaDAPO, and CPPO rise to positions between the
S1 and T1 energy levels of TTA, and the energy gaps be-
tween the T1 levels of these neutral ligands and that of TTA
are about 0.5 eV. Without modification, TPPO has an equiv-
alent T1 energy level with TTA, which makes the positive
energy transfer from the T1 level of TPPO to that of TTA
Figure 6. TGA curves of 1–3. c: curve of 1, a: curve of 2, d:
curve of 3.
much more difficult in [Eu
N
N
of their higher T1 energy level the positive energy transfer
from T1 levels of TAPO, NaDAPO, and CPPO to that of
TTA becomes preponderant, and the reverse energy transfer
from TTA to neutral ligands is restrained. Consequently,
most of the triplet excited energy of these neutral ligands
can be efficiently transferred to 5D0 of EuIII through the
T1 level of TTA. In brief, modification with chromophore
groups reduces the S1 levels of TAPO, NaDAPO, and CPPO
to fit that of TTA much more, which improves the intramo-
lecular singlet energy transfer in the complexes and increas-
es the S1 energy-transfer efficiency. Simultaneously, because
TAPO, NaDAPO, and CPPO have mezzo T1 energy levels
between the S1 and T1 energy levels of TTA, positive triplet
energy transfer from these ligands to TTA becomes the
main process. The triplet energy of the neutral ligands can
be transferred to central ions more efficiently. These find-
ings mean that the introduction of suitable chromophore
groups can support phosphine oxide ligands with mezzo S1
and T1 energy levels and improve the intramolecular energy
transfer in their EuIII complexes.
OLEDs during operation, which is significant for improving
the lifetime of the devices. It is known that for electrolumi-
nescent EuIII complexes, the triplet–triplet state (T–T)
energy quenching highly decreases their EL efficiency. Thus,
most of the electroluminescent EuIII complexes were used as
dopants in the OLED devices. However, the energy-transfer
efficiency between host materials and EuIII complexes was
still not sufficient, and the phase separation during long-
term operation would greatly decrease the performance of
the devices. The amorphous state of 1–3 shows that the in-
termolecular interaction in these complexes is much weaker
and the weaker interaction could decrease the T–T energy
quenching. Therefore, the pure EuIII complexes 1–3 can be
used directly as emitting layers to support excellent EL per-
formance.
Theoretical calculations: After modification with the hole-
transport groups, the change in the carrier injection and
transport ability of TAPO, NaDAPO, and CPPO was what
we were interested in. It is known that the carrier injection
and transport ability is relative to the frontier orbitals (the
LUMO and HOMO) of the compounds. The geometry opti-
mization and frontier orbital population analysis of TPPO,
TAPO, NaDAPO, and CPPO were performed at the
B3LYP/6-31G* level. All calculations were carried out by
using the Gaussian 03 program package.
The electron clouds of the HOMO and LUMO energy
levels of TPPO are mainly localized at the three phenyl
groups (Table 3). However, in TAPO, NaDAPO, and CPPO
the HOMO electron clouds are localized at the hole-trans-
porting moieties (such as triphenylamine, naphthalenyl di-
phenylamine, and N-phenylcarbazole) and the LUMO elec-
tron clouds are mainly localized at the TPPO moieties
except in NaDAPO, the LUMO electron cloud of which is
mainly localized at the naphthalenyl moiety with a larger
conjugated area. Compared with TPPO, the HOMO energy
levels of TAPO, NaDAPO, and CPPO rise remarkably, the
Thermal properties: Thermogravimetric analysis (TGA)
showed that the temperatures of the thermal decomposition
(Td) of complexes 1–3 are higher than 3008C (3348C for 1,
3128C for 2, and 3218C for 3; Figure 6), which makes device
fabrication by vacuum evaporation more feasible.
Notably, no melting point of complexes 1–3 was observed
by cyclic differential scanning calorimetry (DSC) analysis
(in every heating/cooling programme five cycles were
scanned for each complex). This means that complexes 1–3
are amorphous, or at least it is difficult for them to crystal-
lize. The formation of the amorphous states of these com-
plexes is due to their single-coordinate phosphine oxide li-
gands, whose bigger volumes and degrees of freedom in-
crease the volume of the complexes and decrease the inter-
molecular interaction. The stable amorphous states of com-
plexes 1–3 are very important for increasing the
morphological stability of the emitting-layer films in their
10286
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10281 – 10293

Electroluminescence of Eu Complexes
FULL PAPER
Table 3. The calculated geometry, contours, and energy levels of the frontier orbitals.
Ligands
Geometry
LUMO energy orbital
HOMO energy orbital
LUMO HOMO
[eV]
[eV]
TPPO
ꢀ0.844
ꢀ6.721
TAPO
ꢀ0.816
ꢀ5.143
NaDAPO
ꢀ1.224
ꢀ5.224
CPPO
ꢀ1.007
ꢀ5.442
LUMO energy levels of NaDAPO and CPPO fall substan-
tially, and the LUMO energy level of TAPO is equivalent to
that of TPPO. Both TAPO and CPPO exhibit the character-
istic bipolar structure. It is clear that at the same time as im-
proving their hole-injection ability by introducing hole-
transport groups, TAPO, NaDAPO, and CPPO also have ex-
cellent electron-transport ability. With this double-carrier
transport ability, these single-coordinate phosphine oxide li-
gands can greatly improve the relevant performance of their
EuIII complexes, possibly balance the carrier injection and
transport in the complexes, and increase the EL efficiency
of the OLEDs.
phosphine oxide ligands, a balanceable electron and hole in-
jection and transport in complexes 1–3 was expected.
The electrochemical properties of complexes 1–3 were in-
vestigated by CV at room temperature in acetonitrile mea-
sured against an Ag/Ag+ (0.1m in acetonitrile) electrode
(0.34 V versus saturated calomel electrode (SCE); Figure 7)
and the data are listed in Table 4 for comparison. All of
complexes 1–3 displayed a single, reversible, one-electron
oxidation wave at 0.75, 0.78 or 1.32 V. Their onset potentials
were 0.5, 0.551and 0.948 V, respectively. According to the
equation reported by de Leeuw et al.,[23] EHOMO
=
(EoOnxsyet!SCE +4.4 eV) and ELUMO =(ERonesdet!SCE +4.4 eV). So the
EHOMO values of complexes 1–3 are about ꢀ5.24, ꢀ5.29, and
ꢀ5.69 eV, respectively, which approach the HOMO energy
level (ꢀ5.2 eV) of N,N-bis(a-naphthylphenyl)-4,4’-biphenyl-
diamine (NPB) or N,N’-diphenyl-N,N’-bis(3-methylphenyl)-
1,1’-biphenyl-4,4’-diamine (TPD) (the materials usually used
as hole-transport layer). It is clear that the HOMOs of com-
Electrochemical properties: The redox behavior of com-
plexes 1–3 is of much interest as it can give information
about the HOMOs and LUMOs of the complexes, which
are considered to determine the carrier injection and trans-
port ability. As they contain double-carrier-transporting
Chem. Eur. J. 2007, 13, 10281 – 10293
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10287

W. Huang et al.
sponding carrier-transport layers (NPB and BCP), com-
plexes 1–3 would have an excellent double-carrier injection
and transport ability, which is very important for carrier-
transport balance in emitting layers of OLEDs. It is proved
that by taking advantage of the unique structure of phos-
phine oxide ligands, while at the same time preserving the
good electron-transport ability of TPPO moieties, the intro-
duction of hole-transport moieties in phosphine oxide li-
gands can greatly improve the hole-transport ability of the
corresponding EuIII complexes.
EL performance of OLEDs: Encouraged by the results of
CV and other analyses, we studied the EL properties of
complexes 1–3. As mentioned above, the amorphous state
of complexes 1–3 shown by DSC decreases the T–T quench-
ing, which means that these complexes can be used as an
emitting layer directly. Furthermore, their LUMOs and
HOMOs that fit to the corresponding carrier-transport
layers (NPB and BCP) support enough carrier-transport
ability without using any host materials. Two types of
OLED structure were designed and fabricated: single-layer
devices A–C (based on complexes 1–3, respectively) with
the structure indium–tin oxide (ITO)/EuIII complex (60 nm)/
Mg0.9Ag0.1 (200 nm)/Ag (80 nm); and four-layer devices D–F
(based on complexes 1–3, respectively) with structure ITO/
NPB (30 nm)/EuIII complex (40 nm)/BCP (30 nm)/tris(8-hy-
Figure 7. Cyclic voltammograms of 1–3. c: curve of 1, d: curve of 2,
g: curve of 3.
Table 4. Electrochemical properties of the complexes.
Complex
VoltageOonxsyet [V]
(EHOMO [eV])
VoltageRonesdet [V]
(ELUMO [eV])
A
ACHTREUNG
[Eu
1
2
A
(tta)3][24]
– (ꢀ6.3)
– (ꢀ3.1)
0.500 (ꢀ5.24)
0.551( ꢀ5.29)
0.948 (ꢀ5.69)
ꢀ1.741 (ꢀ3.00)
ꢀ1.660 (ꢀ3.08)
ꢀ1.713 (ꢀ3.03)
3
droxyquinoline)aluminum
(Alq3;
30 nm)/Mg0.9Ag0.1
(200 nm)/Ag (80 nm). All of the devices gave the pure char-
plexes 1–3 have been increased high enough to improve
their hole injection and transport ability by introducing
hole-transport groups in the neutral ligands. The results
show that the hole injection and transport ability of these
complexes is as excellent as that of common hole-transport-
ing materials. On the other hand, all the complexes 1–3 had
two reversible, one-electron reduction waves (ꢀ2.02 and
ꢀ2.35 V for 1, ꢀ2.00 and ꢀ2.35 V for 2, and ꢀ2.25, and
ꢀ2.56 V for 3) with onset potentials of ꢀ1.741, ꢀ1.660, and
ꢀ1.713 V, respectively. According to the equation mentioned
above, the ELUMO values of the complexes are ꢀ3.00, ꢀ3.08,
and ꢀ3.03 eV, which also approach the LUMO energy level
of 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline (BCP; the
material usually used as the electron-transporting hole-
blocking layer).
acteristic emission of the EuIII ion at 616 nm corresponding
5
to D0–7F2, which demonstrated the recombination of carri-
ers in the emitting layers.
The brightness–current density–voltage curves of the
single-layer devices are shown in Figure 8. The turn-on volt-
age of device A was 8.2 V at 1cdm ꢀ2, and its maximum
brightness of 35 cdmꢀ2 was achieved at 19.0 V with the cur-
rent density of 161.2 mAcmꢀ2. Device B had a turn-on volt-
age at 8.2 V. The maximum brightness of device B was as
high as 59 cdmꢀ2 at 18.2 V with a current density of
To the best of our knowledge, among TTA-containing
EuIII complexes, the lowest LUMO and the highest HOMO
energy levels reported so far are located at ꢀ3.1and
ꢀ6.3 eV for [Eu
C
N
G
N
for example, increases as much as 0.96 eV. This great im-
provement in hole injection and transport ability of com-
plexes 1–3 is mainly caused by the functional phosphine
oxide ligands modified by the hole-transport moieties. Si-
multaneously, with the contribution from the rest of the
phosphine oxide moieties of the ligands, the complexes, the
ELUMO values of which are equivalent to that of [Eu
(tta)3], still have a very good electron injection and trans-
port ability. With LUMOs and HOMOs fitting to the corre-
(phen)-
Figure 8. Brightness–current density–voltage curves of the single-layer
ACHTREUNG
!
&
*
devices A–C. Luminescence: : device A, : device B, : device C; cur-
!
*
rent density: : device A, &: device B, : device C.
10288
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10281 – 10293

Electroluminescence of Eu Complexes
FULL PAPER
131.7 mAcmꢀ2. To the best of our knowledge, this is the
highest luminescence from single-layer devices based on
pure EuIII complexes reported so far.[9,12–14] Device C had a
higher turn-on voltage of 10 V. Its maximum brightness of
21cdm ꢀ2 was achieved at 17.8 V and 58.4 mAcmꢀ2. Com-
pared with [Eu
N
N
which had a brightness less than 1cdm ꢀ2, the performance
of single-layer devices A–C was significantly improved. No-
tably, compared with the highest single-layer device bright-
ness of 20 cdmꢀ2 reported by Huang and co-workers,[14] the
maximum brightness of devices A and B was remarkably in-
creased, especially for device B, the highest brightness of
which was nearly triple that of the largest value reported
formerly.
The efficiencies of the devices at 10 cdmꢀ2 were calculat-
ed and are listed in Table 5. The efficiencies of devices A
and B are much higher than that of device C. This trend
Figure 9. Brightness–current density–voltage curves of the four-layer de-
!
&
*
vices D–F. Luminescence: : device D, : device E, : device F; Current
!
*
density: : device D, &: device E, : device F.
Table 5. Efficiencies of devices A–F.
structure, their EL performance
mainly originated from the rele-
vant properties of the emitting
complexes. It is clear that turn-
on voltage is relative to the car-
rier injection and transport abil-
ity of the EuIII complexes as
emitting layers. The much
lower turn-on voltages of devi-
ces D and E compared with F
show the much easier carrier in-
jection and transport in 1 and 2.
Device Current efficiency [cdAꢀ1]/voltage [V] Power efficiency [lmWꢀ1]/voltage [V]
External quantum
yield [%]
(current density [mAcmꢀ2])
100 cdmꢀ2
(current density [mAcmꢀ2])
100 cdmꢀ2
Max.
Max.
Max. 100 cdmꢀ2
A
B
C
D
E
F
0.186/13 (5.33)[a]
0.217/13 (4.66)[a]
0.07/14.8 (14.1)[a]
5.07/4.4 (0.021)
5.88/5.0 (0.032)
3.25/8.4 (0.028)
–
–
–
0.045/13 (5.33)[a]
0.052/13 (4.66)[a]
0.015/14.8 (14.1)[a]
3.62/4.4 (0.021)
3.69/5.0 (0.032)
1.23/8.4 (0.028)
–
–
–
0.12[a]
0.09[a]
0.04[a]
3.2
–
–
–
1.31
1.21
0.90
2.08/10.4 (4.82)
1.92/10.0 (5.24)
1.44/14.2 (6.92)
0.63/10.4 (4.82)
0.60/10.0 (5.24)
0.32/14.2 (6.92)
3.71
2.08
[a] At a brightness of 10 cdmꢀ2
.
agrees with the variation of hole injection and transport
ability of complexes 1–3 (C based on 3, with the weakest
hole injection and transport ability, exhibited the lowest effi-
ciencies). It is demonstrated that through tuning the hole in-
jection and transport ability of the complexes by introducing
different hole-transport moieties, a balance of carrier trans-
port can be achieved which leads to the great improvement
of the EL performance of the complexes. The lower turn-on
voltages and higher brightness of devices A–C show that
complexes 1–3 have excellent and balanceable carrier-trans-
port ability by themselves and much weaker self-quenching.
Therefore, the four-layer devices were also fabricated by
using pure complexes 1–3 as emitting layers directly. The
carrier-transport layers were injected not only to enhance
the carrier transport in devices but also to improve the inter-
faces between electrodes and the emitting complexes.
The brightness–current density–voltage curves of the four-
layer devices D–F were measured (Figure 9). The turn-on
voltage of device D was as low as 4.4 V at 1cdm ꢀ2, and its
greatest brightness as high as 1195 cdmꢀ2 was achieved at
19.6 V with a current density of 313.26 mAcmꢀ2. Device E
had a turn-on voltage at 4.8 V. The maximum brightness of
device E was 1158 cdmꢀ2 at 18.0 V with the current density
of 340.03 mAcmꢀ2. Device F could be turned on at 8.0 V
with a maximum brightness of 852 cdmꢀ2 at 22.2 V with a
current density of 288.48 mAcmꢀ2. As devices D–F were
based on the pure EuIII complexes with the same device
As all three complexes have an equivalent electron injection
and transport ability (demonstrated by CV analysis), the
stronger hole injection and transport ability of 1 and 2 com-
pared with 3 should be the main factor inducing this differ-
ence. Furthermore, the much larger maximum brightness of
devices D and E compared with F shows that more holes
can be injected into 1 and 2 at lower operating voltages and
further recombined with electrons to form more excitons,
which would originate from the more balanceable carrier in-
jection and transport ability of 1 and 2 than 3.
The efficiencies of devices D–F were calculated and the
curves are shown in Figures 10 and 11. The data are listed in
Table 5 for comparison. D and E showed very high current
efficiencies of more than 5 cdAꢀ1 and an external quantum
yield of more than 3% at very small current densities. At
the practical luminescence of 100 cdmꢀ2, the current effi-
ciencies and external quantum yield of D and E are still
about 2 cdAꢀ1 and 1.2%, respectively. The brightness of
over 1000 cdmꢀ2 and the high efficiencies at 100 cdmꢀ2 of
the undoped devices indicate the very low self-quenching of
1 and 2. This means that the much weaker intermolecular
interaction and the amorphous state of 1 and 2 induced by
the larger and freer functional single-coordinate phosphine
oxide ligands can efficiently inhibit the self-quenching
effect. At current densities less than 10 mAcmꢀ2, D and E
had much higher efficiencies than F. This finding means that
Chem. Eur. J. 2007, 13, 10281 – 10293
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10289

W. Huang et al.
favorably with these results. The functional single-coordi-
nate phosphine oxide ligands support the excellent EL prop-
erties of EuIII complexes with a very simple structure, which
should originate from the structure characteristic of phos-
phine oxide compounds and their special coordination
mode.
The EL spectra of devices D–F show the characteristic
emissions from EuIII with the main peak at 616 nm
(Figure 12). Devices D and E exhibited excellent spectral
stability during the increase in operation voltage. However
in the EL spectrum of F, the emission from Alq3 at about
520 nm became more and more visible during voltage in-
crease, which was also reported in previous work.[5,14,25] Ac-
cording to the energy-level schemes in Figure 12, it is clear
that in devices D and E, because of the higher HOMO
energy levels of 1 and 2, the hole cannot jump across energy
gaps as big as ꢁ1.5 eV to enter into Alq3 even at breakdown
voltages, so no emission from Alq3 was observed. However,
in F the lower HOMO energy level of 3 nearly halves the
energy gap between hole-transporting NPB and hole-block-
ing BCP. The hole can make use of the HOMO energy level
of 3 to cross the layer of BCP and enter into Alq3 at higher
voltages. This result proves that increasing the HOMO
energy level of the EuIII complexes can efficiently improve
the EL efficiency by limiting holes in emitting layers, and
enhances the spectral stability of the complexes.
Figure 10. Current efficiency–current density curves of the four-layer de-
vices D–F.
Conclusion
Three functional single-coordinate phosphine oxide ligands
with the most compact structure and their EuIII complexes
were designed and synthesized. Our investigations show that
the introduction of hole-transport groups in single coordi-
nate phosphine oxide ligands significantly improves the
properties of their EuIII complexes. The introduction of
chromophore groups makes TAPO, NaDAPO, and CPPO
obtain mezzo S1 and T1 energy levels, which fit those of
TTA much more. Simultaneously, the ligands were designed
with a structure that is the simplest and most compact possi-
ble in order to shorten the distance between ligands and
EuIII ions. The match of the excited energy levels and the
compact structure of the complexes improve the intramolec-
ular energy-transfer efficiency. On the other hand, the modi-
fication increases the volumes of the complexes, which re-
duces the intermolecular interaction and results in the amor-
phous phase and the extraordinarily low self-quenching of
the complexes. The LUMO and HOMO energy levels of
complexes 1–3 were tuned to fit the corresponding carrier-
transport layer. The single-layer and four-layer OLEDs
based on pure EuIII complexes 1–3 had excellent EL perfor-
mance, in which the single-layer device of 2 had the highest
brightness of 59 cdmꢀ2 reported so far and both of the four-
layer devices based on 1 and 2 had a maximum brightness
of more than 1000 cdmꢀ2, turn-on voltages lower than 5 V,
external quantum yields of more than 3%, and excellent
spectral stability. Compounds 1 and 2 are among the most
Figure 11. Power efficiency–current density curves of the four-layer devi-
ces D–F.
when the carrier density in the device is limited, the capabil-
ity of capturing carriers of the complexes directly influences
both the brightness and the efficiency of the devices, since
the self-quenching effect is not evident at low exciton densi-
ty. However, when the current density became more than
ꢀ2
1 0 mAcm , along with the current density increase the effi-
ciencies of D–F decreased very rapidly. Notably, the current
efficiency of E was reduced faster than those of D and F
(Figure 10). The p–p stacking trend of the naphthalenyl
group with a longer conjugated length in NaDAPO would
enhance the intermolecular interaction in 2, which should
be the main factor that induces the worse self-quenching in
E. Nevertheless, because of the stronger carrier injection
and transport ability of 1 and 2, D and E had lower opera-
tion voltages at the same current density compared with F,
which made the power efficiencies of D and E constantly
higher than that of F.
Table 6 shows recent representative work on electrolumi-
nescent functional EuIII complexes. The EL performance of
the single-layer and four-layer devices based on 2 compared
10290
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10281 – 10293

Electroluminescence of Eu Complexes
FULL PAPER
Table 6. Summary of the EL performance of the EuIII complexes with functionalized ligands.
efficient EuIII complexes. Our
results demonstrate the further
application of single-coordinate
phosphine oxide ligands in
high-performance electrolumi-
nescent EuIII complexes by ap-
propriate chemical modifica-
tions.
Ligand
Turn-on voltage
Current efficiency Power efficiency
Ref.
[13]
(maximum brightness/voltage [cdAꢀ1
[V, cdmꢀ2/V])
]
[lm Wꢀ1
]
Single-Layer
Device
Multi-layer
Device
8 (19/13.5)
6 (1305/16)
5 (1500/16)
–
–
1.44
Experimental Section
7 (20/16)
2.7
[14]
[8]
Materials and instrumentation: All the
reagents and solvents used for the syn-
thesis of 1–3 were purchased from Al-
drich and Acros. Alq3, BCP, and NPB
used for electroluminescent device
fabrication were purchased from Lum-
thec Corporation. All the reagents
were used without further purification.
1H NMR spectra were recorded by
using a Varian Mercury Plus 400NB
spectrometer relative to tetramethylsi-
lane (TMS) as internal standard. Mo-
lecular masses were determined by a
Shimadzu laser desorption/ionization
–
5.7 (957/19.1) 4.14
4.8 (1152/18) 5.88
2.28
this
work
8.2 (59/18.2)
3.69
Figure 12. EL spectra and energy-level schemes of devices D–F.
Chem. Eur. J. 2007, 13, 10281 – 10293
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10291

W. Huang et al.
time-of-flight (LDI-TOF) mass spectrometer or a Shimadzu GCMS-
QP2010 Plus instrument equipped with a DB-5 column. Elemental analy-
ses were performed on a Vario EL III elemental analyzer. TGA and
DSC were performed on Shimadzu DTG-60A and DSC-60A thermal an-
alyzers under a nitrogen atmosphere at a heating rate of 108Cminꢀ1. Ab-
sorption and PL emission spectra of the target compounds were mea-
sured in dichloromethane with a Shimadzu UV-3150 spectrophotometer
and a Shimadzu RF-5301PC spectrophotometer, respectively. Phosphor-
escence spectra were measured in dichloromethane by using an Edin-
burgh FLS 920 fluorescence spectrophotometer at 77 K with cooling by
liquid nitrogen.
9-[4-(Diphenylphosphinoyl)phenyl]-9H-carbazole (CPPO): Cppo was
prepared from 9-(4-bromophenyl)-9H-carbazole in a yield of 53% as a
white powder. 1H NMR (CDCl3): d=8.12–8.17 (d, J=7.6 Hz, 2H), 7.86–
7.95 (t, J=10.0 Hz, 2H), 7.69–7.85 (m, 6H), 7.46–7.64 (m, 8H), 7.39–7.46
(t, J=7.6 Hz, 2H), 7.28–7.36 ppm (q, JH,H =10.0 Hz, JP, H =6.8 Hz, 2H);
LDI-TOF-MS: m/z (%): 443 (100) [M+]; elemental analysis calcd (%)
for C30H22NOP: C 81.25, H 5.00, N 3.16, O 3.61; found: C 81.65, H 4.88,
N 3.57, O 4.00.
General procedure ofcomplexation : The complexes were prepared ac-
cording to well-established protocols.[26] 2-Thenoyltrifluoroacetone
(HTTA; 672.7 mg, 3 mmol) was dissolved in ethanol, and NaOH
(120 mg, 3 mmol) in aqueous solution (2m) was added to remove the H+
in the TTA molecule. Then EuCl3·6H2O (370.1mg, 1mmol) in aqueous
solution was added dropwise, the mixture was stirred at 608C for 30 min,
TAPO (890.0 mg, 2 mmol; or 990.0 mg of NaDAPO, or 886.0 mg of
CPPO) in ethanol was added dropwise, and the mixture was stirred at
608C for 4 h to afford the complexes 1–3. Purification was accomplished
by precipitation from concentrated ethanol and water solution.
(4-Bromophenyl)diphenylamine:
Potassium
carbonate
(1.95 g,
14.1 mmol), copper sulfate pentahydrate (0.176 g, 0.705 mmol), 4-iodo-
bromobenzene (7.64 g, 26.0 mmol), and diphenylamine (2.40 g,
14.1 mmol) were added to a 100 mL two-necked round-bottomed flask.
The mixture was heated to 2208C for 15 h. Water (10 mL) was added to
stop the reaction, and then the mixture was extracted with dichlorome-
thane (3”30 mL). The organic layer was dried with MgSO4. The solvent
was removed in vacuo and the residue was purified by flash column chro-
matography by using petroleum ether as the eluent. Small white crystals;
yield: 2.92 g of (62%); 1H NMR (CDCl3, 400 MHz): d=7.32 (d, J=
8.0 Hz, 2H), 7.23–7.27 (m, 4H), 6.62 (t, J=8.0 Hz, 2H), 7.01–7.08 (m,
6H), 6.94 ppm (d, J=8.0 Hz, 2H); GCMS: m/z (%): 323 (100) [M+].
[Eu
A
G
ꢀ
ꢀ
400 MHz): d=7.56–7.77 (brs, 16H; Ph H), 7.22–7.36 (m, 14H; Ph H),
7.13–7.22 (m, 14H; Ph H), 7.05–7.12 (t, J=7.6 Hz, 4H; Ph H), 6.75–6.82
(d, 3H, J=4.8 Hz; Th H), 6.40–6.47 (t, 3H, J=4.4 Hz; Th H), 6.10–6.18
(d, 3H, J=2.8 Hz; Th H), 4.04 ppm (s, 3H; COCHCO); elemental anal-
ysis calcd (%) for C84H60EuF9N2O8P2S3: C 59.12, H 3.54, Eu 8.91, N 1.64,
O 7.50, S 5.64; found: C 59.01, H 3.685, Eu 9.14, N 1.408, O 7.14, S 5.80.
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
(4-Diphenylaminophenyl)diphenylphosphine oxide (TAPO): (4-Bromo-
phenyl)diphenylamine (324 mg, 1mmol) was dissolved in absolute ether
(10 mL) and lithium salts were formed with 1.6m nBuLi/hexane solution
(0.75 mL, 1.2 mmol) at 08C for 1h. Then diphenylphosphine chloride
(0.23 mL, 1.3 mmol) was added dropwise at 08C. The mixture was react-
ed at 08C for 1h and then at room temperature overnight. The phos-
phine compound (4-diphenylaminophenyl)diphenylphosphine was ob-
tained and purified by flash column chromatography by using dichloro-
[Eu
N
N
(CDCl3): d=7.90–7.98 (d, J=8.4 Hz, 4H), 7.86–7.90 (d, J=8.4 Hz, 4H),
7.77–7.83 (d, J=8.0 Hz, 4H), 7.52–7.73 (brs, 16H), 7.43–7.50 (m, 7H),
7.33–7.41 (m, 6H), 7.15–7.24 (m, 11H), 6.72–6.80 (d, J=4.4 Hz, 3H),
6.36–6.44 (t, J=4.0 Hz, 3H), 6.06–6.12 (d, J=2.4 Hz, 3H), 4.22 ppm (s,
3H); elemental analysis calcd (%) for C92H64EuF9N2O8P2S3: C 61.16, H
3.57, Eu 8.41, N 1.55, O 7.08, S 5.32; found: C 61.35, H 3.69, Eu 8.87, N
1.54, O 7.12, S 5.34.
1
methane as the eluent. Yellow powder; yield: 300 mg of (70%); H NMR
(CDCl3, 400 MHz): d=7.30–7.38 (m, 8H), 7.21–7.30 (m, 4H), 6.97–
7.20 ppm (m, 10H); LDI-TOF-MS: m/z (%): 429 (100) [M+]; elemental
analysis calcd (%) for C30H24NP: C 83.89, H 5.63, N 3.26; found: C 83.58,
H 5.51, N 3.09. The phosphine (429 mg, 1 mmol) was dissolved in 1,4-di-
oxane and sufficient dilute sulfuric acid was added to protect the
N atoms. Then the phosphine was oxidized by 30% hydrogen peroxide
(1.1 mmol) for 4 h to form the ligand TAPO. Straw-yellow powder; yield:
441mg (99%); 1H NMR (400 MHz, CDCl3): d=7.65–7.76 (m, 4H), 7.37–
7.58 (m, 9H), 7.24–7.35 (m, 4H), 7.07–7.18 (m, 5H), 6.97–7.06 ppm (q,
[Eu
N
N
(CDCl3): d=7.18–8.26 (m, 44H), 6.55–6.82 (brs, 3H), 6.12–6.33 (brs,
3H), 5.56–5.76 (brs, 3H), 4.12 ppm (s, 3H); elemental analysis calcd (%)
for C84H56EuF9N2O8P2S3: C 59.26, H 3.32, Eu 8.93, N 1.65, O 7.52, S 5.65;
found: C 59.45, H 3.34, Eu 9.01, N 1.79, O 7.86, S 5.52.
General procedure for the preparation of gadolinium complexes: For
triplet energy level measurement, gadolinium complexes comprising
TAPO, NaDAPO, or CPPO without TTA were also synthesized.[22] The
phosphine oxide ligand (1mmol) was dissolved in ethanol (15 mL). [Gd-
J
H-H =8.8, JP-H =2.4 Hz, 2H); LDI-TOF-MS: m/z (%): 445 (100) [M+];
elemental analysis calcd (%) for C30H24NOP: C 80.88, H 5.43, N 3.14, O
3.59; found: C 80.37, H 5.77, N 3.06, O 3.73.
A
G
dropwise with stirring and the mixture was heated to reflux for 2 h. The
gadolinium complexes were recrystallized from concentrated methanol/
water mixed solvent.
(4-Bromophenyl)naphthalen-1-yl-phenylamine: (4-Bromophenyl)naph-
thalen-1-yl-phenylamine was prepared from naphthalen-1-yl-phenylamine
by following the same synthetic procedure as that for (4-bromophenyl)di-
phenylamine. White powder; yield: 71%; 1H NMR (CDC13): d=7.95–
8.01(d, J=8.0 Hz, 1H), 7.86–7.93 (d, J=8.0 Hz, 1H), 7.76–7.83 (d, J=
8.4 Hz, 1H), 7.46–7.59 (m, 2H), 7.39–7.46 (t, J=7.4 Hz, 1H), 7.30–7.38
(m, 3H), 7.28–7.20 (t, J=8.0 Hz, 2H), 7.02–6.94 (m, 3H), 6.81–6.72 ppm
(d, J=8.8 Hz, 2H); GCMS: m/z (%): 373 (100) [M+].
[Gd
E
G
C60H48GdN5O11P2: C 58.39, H 3.92, N 5.67, O 14.26; found: C 58.53, H
3.88, N 5.57, O 14.63.
[Gd
G
G
C68H52GdN5O11P2: C 61.21, H 3.93, N 5.25, O 13.19; found: C 61.47, H
3.92, N 5.11, O 13.25.
(4-Naphthalen-1-yl-phenylaminophenyl)diphenylphosphine
oxide
[Gd
G
N
(NaDAPO): NaDAPO was prepared from (4-bromophenyl)naphthalen-
1-yl-phenylamine by following the same two-step synthetic procedure as
that for TAPO. Pale-yellow powder; yield 43%; 1H NMR (CDCl3): d=
7.85–7.93 (brs, 2H), 7.78–7.84 (d, J=8.0 Hz, 1H), 7.62–7.76 (m, 5H),
7.31–7.61 (m, 14H), 7.21–7.30 (m, 2H), 7.15–7.21 ppm (d, J=8.0 Hz,
2H); LDI-TOF-MS: m/z (%): 495 (100) [M+]; elemental analysis calcd
(%) for C34H26NOP: C 82.41, H 5.29, N 2.83, O 3.23; found: C 82.06, H
5.77, N 2.80, O 3.48.
C60H44GdN5O11P2: C 58.58, H 3.61, N 5.69, O 14.31; found: C 58.78, H
3.50, N 5.45, O 14.53.
Fabrication and testing ofOLEDs : Single-layer and four-layer OLEDs
were fabricated by vacuum deposition with the configurations ITO/EuIII
complex (60 nm)/Mg0.9Ag0.1 (200 nm)/Ag (80 nm) and ITO/NPB (30 nm)/
EuIII complex (40 nm)/BCP (30 nm)/Alq3 (30 nm)/Mg0.9Ag0.1 (200 nm)/Ag
(80 nm), in which NPB is the hole-transporting layer, BCP is the elec-
tron-transporting/hole-blocking layer, Alq3 is the electron-transporting
layer, and ITO and MgAg were used as the anode and cathode, respec-
tively.[27] Before loading into a deposition chamber, the ITO substrate
was cleaned with detergents and deionized water, dried in an oven at
1208C for 4 h, and treated with UV/ozone for 25 min. Devices were fabri-
cated by evaporating organic layers at a rate of 0.1–0.3 nmsꢀ1 onto the
ITO substrate sequentially at a pressure below 1”10 ꢀ6 mbar. A layer of
N-(4-Bromophenyl)carbazole: N-(4-Bromophenyl)carbazole was pre-
pared from carbazole by following the same synthetic procedure as that
for (4-bromophenyl)diphenylamine. White powder; Yield: 77%;
1H NMR (CDCl3): d=8.10–8.18 (d, J=8.0 Hz, 2H), 7.69–7.75 (d, J=
8.4 Hz, 2H), 7.34–7.49 (m, 6H), 7.26–7.32 ppm (t, J=6.8 Hz, 2H);
GCMS: m/z (%): 321(100) [ M+].
10292
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10281 – 10293

Electroluminescence of Eu Complexes
FULL PAPER
MgAg 200 nm in thickness was deposited on the Alq3 layer at a rate of
0.4 nmsꢀ1 to improve electron injection. Finally, an 80 nm-thick layer of
Ag was deposited at a rate of 0.6 nmsꢀ1 as the cathode. The emission
area of the devices was 0.12 cm2, as determined by the overlap area of
the anode and cathode. The EL spectra and current–voltage–luminance
characteristics were measured with a Spectrascan PR 650 photometer
and a computer-controlled dc power supply under ambient conditions.
[10] R. Reyes, E. N. Hering, M. Cremona, C. F. B. da Silva, H. F. Brito,
C. A. Achete, Thin Solid Films 2002, 420–421, 23–29.
[11] X.-C. Gao, H. Cao, C.-H. Huang, S. Umitani, G.-Q. Chen, P. Jiang,
Synth. Met. 1999, 99, 127–132.
[12] Z. Bian, D. Gao, K. Wang, L. Jin, C. Huang, Thin Solid Films 2004,
460, 237–241.
[13] M. Sun, H. Xin, K. Wang, Y. Zhang, L. Jin, C. Huang, Chem.
Commun. 2003, 702–703.
[14] H. Xin, F. Y. Li, M. Guan, C. H. Huang, M. Sun, K. Z. Wang, Y. A.
Zhang, L. P. Jin, J. Appl. Phys. 2003, 94, 4729–4731.
[15] H. Xin, M. Sun, K. Wang, Y. Zhang, L. Jin, C. Huang, Chem. Phys.
Lett. 2004, 388, 55–57.
Acknowledgements
[16] D. L. Dexter, J. Chem. Phys. 1953, 21, 836–850.
[17] G. A. Crosby, R. E. Whan, R. M. Alire, J. Chem. Phys. 1961, 34,
743–748.
[18] J. Wang, R. Wang, J. Yang, Z. Zheng, M. D. Carducci, T. Cayou, J.
Am. Chem. Soc. 2001, 123, 6179–6180.
This work was financially supported by the National Natural Science
Foundation of China under grants 60325412, 90406021, and 50428303.
Helpful discussions with Professor C.-H. Huang, Professor F.-Y. Li, Dr.
R.-F. Chen, Dr. S.-J. Liu and Dr. Q. Zhao are highly appreciated.
[19] Y. J. Fu, T. K. S. Wong, Y. K. Yan, X. Hu, J. AlloysCompd. 2003,
358, 235–244.
[1] J. Kido, Y. Okamoto, Chem. Rev. 2002, 102, 2357–2368.
[2] M. D. Mcgehee, T. Bergstedt, C. Zhang, A. P. Saab, M. B. OꢁRegan,
G. C. Bazan, V. I. Srdanov, A. J. Heeger, Adv. Mater. 1999, 11, 1349–
1354.
[20] H. Xu, L. Wang, X. Zhu, K. Yin, G. Zhong, X. Hou, W. Huang, J.
Phys. Chem. B 2006, 110, 3023–3029.
[21] X. C. Gao, H. Cao, G. Q. Yao, C. H. Huang, L. Huang, J. Rare Earth
1999, 17, 170–174.
[3] L. Huang, K.-Z. Wang, C.-H. Huang, F.-Y. Li, Y.-Y. Huang, J. Mater.
Chem. 2001, 11, 790–793.
[4] T. Ohmori, H. Kajii, T. Sawatani, H. Ueta, K. Yoshino, Thin Solid
Films 2001, 393, 407–411.
[22] H. Xin, M. Shi, X. C. Gao, Y. Y. Huang, Z. Gong, D. Nie, H. Cao,
Z. Bian, F. Li, C. Huang, J. Phys. Chem. B 2004, 108, 10796–10800.
[23] D. M. de Leeuw, M. M. J. Simenon, A. R. Brown, R. E. F. Einer-
hand, Synth. Met. 1997, 87, 53–59.
[5] P.-P. Sun, J.-P. Duan, H.-T. Shih, C.-H. Cheng, Appl. Phys. Lett.
2002, 81, 792–794.
[24] Y. Ohmori, H. Kajii, T. Sawatani, H. Ueta, K. Yoshino, Thin Solid
Films 2001, 393, 407–411.
[6] P.-P. Sun, J.-P. Duan, J.-J. Lih, C.-H. Cheng, Adv. Funct. Mater. 2003,
13, 683–691.
[25] J. Kido, H. Hayase, K. Hongawa, K. Nagai, K. Okuyama, Appl.
Phys. Lett. 1994, 65, 2124–2126.
[7] X.-H. Zhu, L.-H. Wang, J.-R. , W. Huang, J.-F. Fang, D.-G. Ma, J.
Mater. Chem. 2004, 14, 2732–2733.
[26] L. R. Melby, N. J. Rose, E. Abranmson, J. C. Caris, J. Am. Chem.
Soc. 1964, 86, 5117–5125.
[8] J. Yu, L. Zhou, H. Zhang, Y. Zheng, H. Li, R. Deng, Z. Peng, Z. Li,
Inorg. Chem. 2005, 44, 1611–1618.
[27] J. F. Fang, D. G. Ma, Appl. Phys. Lett. 2003, 83, 4041–4043.
[9] M. R. Robinson, M. B. OꢁReganc, G. C. Bazan, Chem. Commun.
2000, 1645–1646.
Received: May 4, 2007
Published online: October 4, 2007
Chem. Eur. J. 2007, 13, 10281 – 10293
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10293