Effective manipulation of the electronic effects and its influence on the emission of 5-substituted tris(8-quinolinolate) aluminum(III) complexes

Article abstract of DOI:10.1002/chem.200501403

The unique electron-transport and emissive properties of tris(8-quinolinolate) aluminum(III) (Alq₃) have resulted in extensive use of this material for small molecular organic light-emitting diode (OLED) fabrication. So far, efforts to prepare stable and easy-to-process red/green/blue (RGB)-emitting Alq₃ derivatives have met with only a limited success. In this paper, we describe how the electronic nature of various substituents, projected via an arylethynyl or aryl spacer to the position of the highest HOMO density (C5), may be used for effective emission tuning to obtain blue-, green-, and red-emitting materials. The synthetic strategy consists of four different pathways for the attachment of electron-donating and electron-withdrawing aryl or arylethynyl substituents to the 5-position of the quinolinolate ring. Successful tuning of the emission color covering the whole visible spectrum (λ = 450-800 nm) was achieved. In addition, the photophysical properties of the luminophores were found to correlate with the Hammett constant of the respective substituents, providing a powerful strategy with which to predict the optical properties of new materials. We also demonstrate that the electronic nature of the substituent affects the emission properties of the resulting complex through effective modification of the HOMO levels of the quinolinolate ligand.

Full text of DOI:10.1002/chem.200501403

FULL PAPER
DOI: 10.1002/chem.200501403
Effective Manipulation of the Electronic Effects and Its Influence on the
Emission of 5-Substituted Tris(8-quinolinolate) Aluminum(iii) Complexes
G
Victor A. Montes,[a] Radek Pohl,[a] Joseph Shinar,[b] and Pavel Anzenbacher, Jr.*[a]
Abstract: The unique electron-trans-
port and emissive properties of tris-
emission tuning to obtain blue-, green-,
and red-emitting materials. The syn-
thetic strategy consists of four different
pathways for the attachment of elec-
tron-donating and electron-withdraw-
ing aryl or arylethynyl substituents to
the 5-position of the quinolinolate ring.
Successful tuning of the emission color
covering the whole visible spectrum
(l=450–800 nm) was achieved. In ad-
dition, the photophysical properties of
the luminophores were found to corre-
late with the Hammett constant of the
(8-quinolinolate) aluminum(iii) (Alq3)
G
have resulted in extensive use of this
material for small molecular organic
light-emitting diode (OLED) fabrica-
tion. So far, efforts to prepare stable
and easy-to-process red/green/blue
(RGB)-emitting Alq3 derivatives have
met with only a limited success. In this
paper, we describe how the electronic
nature of various substituents, project-
ed via an arylethynyl or aryl spacer to
the position of the highest HOMO
density (C5), may be used for effective
respective substituents, providing
a
powerful strategy with which to predict
the optical properties of new materials.
We also demonstrate that the electron-
ic nature of the substituent affects the
emission properties of the resulting
complex through effective modification
of the HOMO levels of the quinolino-
late ligand.
Keywords: aluminum
·
electronic
organic light-emitting
quinolinolate semi-
structure
diodes
·
·
·
conductors
Organic light-emitting diodes (OLEDs) are a promising
technology for fabrication of full-color flat-panel displays,[1,2]
and some are already being commercialized.[3] Nevertheless,
a widespread use of OLED-based displays relies on wide
availability of stable materials emitting red/green/blue
(RGB) light of high color purity. Electroluminescent organ-
advantages are, in general, good emission-color purity, their
availability both as singlet[4] or triplet emitters,[5] and their
ease of deposition by means of thermal vacuum evapora-
tion. A major obstacle in the fabrication of SMOLED-based
full-color displays so far appears to be the limited availabili-
ty of long-lived complexes that cover the whole visible spec-
trum, while also possessing similar optoelectronic, semicon-
ductor, and charge-transport properties.[1,2]
A
of the so-called small-molecule OLEDs (SMOLEDs). Their
Success in the preparation of long-lasting SMOLED de-
A
[a] V. A. Montes, Dr. R. Pohl, Prof. P. Anzenbacher, Jr.
Department of Chemistry and Center for Photochemical Sciences
Bowling Green State University (BGSU)
Bowling Green, OH 43403 (USA)
proper alignment of the HOMO/LUMO levels with other
device components such as charge-injection and charge-
transport layers in order to balance the mobility of the carri-
ers and achieve efficient charge recombination in the emit-
ting layer.[6] In undoped structures, the materials must also
possess sufficiently large HOMO–LUMO gaps for the emis-
sion at the required wavelengths, high solid-state emission
quantum yields, and high emission-color stability.
Fax : (+1)419-372-9809
E-mail: pavel@bgnet.bgsu
[b] Prof. J. Shinar
Ames Laboratory-USDOE and
Department of Physics and Astronomy
Iowa State University
Ames, IA 50011 (USA)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author. It contains details
of the full synthesis and characterization of the precursors, ligands,
and complexes; numeric values used to construct the energy-gap law
correlations; and crystallographic data for 10c,d,j, and 11h.
aluminum(iii) (Alq3) has been used as an emitter and one of
A
the most stable electron-transporting materials and host for
saturated green and red colors currently used in SMO-
LEDs.[2,8] Although the past decade has seen great progress
Chem. Eur. J. 2006, 12, 4523 – 4535
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4523

in the preparation of hosts and emitters comparable to Alq3,
further improvement in both the efficiency and durability of
OLED materials is still highly desirable.[9] Hence, we decid-
ed to investigate Alq3-related materials in order to gain
better understanding of the processes that may allow ratio-
Perhaps the best example of a systematic effect on the
complex emission is represented by the introduction of a
(weakly electron-donating) methyl group into the quinolino-
late ligand. When a methyl group is attached to the elec-
tron-deficient pyridyl ring it causes a blueshift in the emis-
sion (lmax increases from 490 to 523 nm for positions 2 to 4,
respectively), whereas substitution on the electron-rich
phenoxide ring produces a redshift (lmax increases from 525
to 568 nm for positions 5 to 7, respectively).[12,16–17] However,
it should be noted that because of steric hindrance, the tris-
A
mentally interesting and industrially important material.
Both the semiconductor and emissive properties are large-
ly defined by the HOMO/LUMO levels of the quinolinolate
ligand and its lowest electronic p–p* transitions.[10,11]
A
number of studies aimed at tuning the HOMO/LUMO
levels and emission color in the Alq3-type materials were at-
tempted with a variable degree of success.[9] Physical and
theoretical studies providing an insight into the distribution
of the HOMO/LUMO densities in the quinolinolate ligand
were also performed. According to density functional theory
(DFT) calculations performed on mer-Alq3,[10b–c] the HOMO
orbitals are located mostly on the phenoxide side, and, in
particular, at C5 of the ligand, whereas the highest LUMO
density is found at the pyridine ring. It appears logical that
substitution with suitable functional groups at the phenoxide
ring would result in changes in the HOMO levels, and that
the LUMO levels would be tuned by substituents on the
pyridine ring.[12–13] However, the natural question as to
whether substitution in a HOMO node or LUMO node
would induce changes in both the HOMO and LUMO
levels remained unanswered.
Previous attempts to modify Alq3 by focusing mainly on
manipulating the HOMO–LUMO gap (i.e., the emission
energy) were motivated by the assumption that the substitu-
tion of electron-donating groups (EDGs) such as methyl at
the 2-, 3-, and 4-positions of the quinolinolate ligand would
induce a destabilizing effect on the LUMO, thus increasing
the LUMO energy and the HOMO–LUMO gap. By the
same token, substitution with electron-withdrawing groups
(EWGs) at the 5-, 6-, and 7-positions should deplete the
HOMO density, resulting in an increase in the energy of the
p–p* transition and an emission that is blueshifted relative
to the parent Alq3 (lmax =525 nm).[9,12] Although this
common-sense approach ought to provide a handle for
tuning the complex emission, the literature does not provide
evidence supporting this relationship between the electronic
nature of the substituent and the emissive properties of the
respective aluminum quinolinolate. In particular, the C5-
substituted complexes, the most widely represented group of
Alq3 derivatives, do not show this correlation. The only no-
table example of blueshifted fluorescence in an Alq3 deriva-
tive was obtained with the piperidyl amide of Alq3-5-sulfon-
ic acid with the emission maximum at l=480 nm.[14] Other
attempts at emission tuning failed to yield significant effects
or blue-emitting complexes. For example, the introduction
of a cyano group at the 5-position of the quinolinolate re-
sulted only in a 4 nm blueshift,[15] whereas the fluoro and
chloro[16] substituents yielded, against expectations, Alq3
complexes with slightly redshifted emissions of l=535 and
540 nm, respectively.
aluminum(iii) complex with 8-hydroxy-2-methylquinoline
G
(the bluest emitter) is not stable, so one of the three quinoli-
nolate ligands must be replaced by a smaller phenolato-type
ligand, thus creating the pentacoordinate complexes in-
stead.[1,18] Hence, one cannot be certain whether the dramat-
ic blueshifted emission (lꢀ490 nm) of the 8-hydroxy-2-
methylquinoline complexes is due to methyl-mediated elec-
tronic tuning or simply to the less sterically favorable geom-
etry of the 2-methylquinolinolate–aluminum(iii) bonding.
G
Thus, the convoluted effects observed in the materials
with the substituents attached directly to the quinolinolate
ligand may be due to the competing resonance and induc-
tive effects. In an effort to limit the direct mesomeric partic-
ipation of substituent p electrons, in order to better under-
stand the substituent effect on the energy and distribution
of the HOMO and LUMOs and its possible use in the emis-
sion-color tuning in Alq3 complexes, we decided to synthe-
size complexes 1a–k and 2a–n. The series comprise 5-substi-
tuted 8-quinoline ligands and various substituents connected
by an arylethynyl or aryl spacer (see Figure 1).
In this paper, we describe how the electronic nature of
various substituent groups, when projected through the se-
lected spacers, allows for systematic tuning of the HOMO/
LUMO levels and of the corresponding emission from the
aluminum(iii) quinolinolates.
G
Results and Discussion
Syntheses: The ligands and the corresponding AlIII com-
plexes were prepared following previously published meth-
ods.[19,20] The hydroxyl group in the starting material, 5-
bromo-8-hydroxyquinoline,[21] was protected by t-butyloxy-
carbonyl (Boc) or benzyl (Bn) protecting groups prior to
the Pd-catalyzed coupling reactions. The Boc group shows
excellent compatibility with the mild conditions for the So-
nogashira–Hagihara cross-coupling used to attach the ethyn-
yl-TMS (TMS=trimethylsilyl) and ethynylarenes to the qui-
nolinolate (Scheme 1). Because the electron-rich (EDG)
arenes do not undergo oxidative addition to Pd0 as readily
as electron-poor arenes, the EDG-substituted ethynylarenes
were prepared separately and coupled to 5-Br-8-O-Boc-qui-
noline 3 in 60–82% yield. For the introduction of EWG
arenes, quinoline 3 was first substituted with acetylene to
give 8-O-Boc-5-ethynylquinoline 4 (obtained in two steps
with 70% overall yield) followed by a cross-coupling reac-
tion with suitable EWG bromoarenes to give 5a,b,d–f (in
4524
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4523 – 4535

Organic Light-Emitting Diodes
FULL PAPER
moieties attached to the ethyn-
A
protected TFA salts were ob-
tained. This is due to the addi-
tion of TFA to the triple bond
followed by hydrolysis as
shown in Scheme 2. In the case
of the dimethylamino-deriva-
tive 6k, the resulting addition–
hydrolysis product (25–30%)
was isolated and its structure
determined by using NMR and
mass spectroscopies. The criti-
cal structural features in the
NMR spectra of 7k are charac-
teristic
signals
for
C=O
(13C NMR: d=194.5 ppm) and
CH2
4.73 ppm).
(1H NMR:
d=
ꢁ
ꢁ
The presence of a strong
base in the Suzuki–Miyaura
coupling reaction (such as
aqueous K2CO3) precluded
using the hydrolytically unsta-
ble Boc group as protection,
and the stable benzyl group
was used to protect the 8-hy-
droxy group instead. The syn-
thesis described in Scheme 3
employs an approach similar
to the one described in
Scheme 1. Here too, two dif-
ferent pathways were devised
to attach the electron-rich
EDG arenes by using previ-
ously prepared arene boronic
acids in cross-coupling reac-
tions with 8-benzyloxy-5-bro-
moquinoline 8, while the intro-
duction of electron-poor EWG
arenes proceeded with higher
yields when boronate ester 9
was cross-coupled to an EWG-
substituted haloarene. The pi-
nacolato boronate ester 9 was
prepared by Pd0-catalyzed bo-
Figure 1. Structure of Alq3 complexes 1a–k and 2a–n with electron-rich and electron-poor aromatic moieties.
A
Scheme 1. Reaction conditions: a) di-tert-butyl dicarbonate, DMAP, hexane, RT, 5 h; b) [Pd
(5%), DIPEA, THF, 608C, 24 h; c) KF (2 equiv), MeOH, RT, 3 h; d) piperidine (3 equiv), CH2Cl2, 5 min.
(PPh3)4] (5%), CuI
ternatively, the aryl–aryl cross-
coupling may be achieved
through a Stille reaction, in
62–97% yield). The piperidine-promoted removal of the
Boc group proceeds quantitatively, furnishing the free li-
gands 6a–k.
The removal of the Boc group may also be accomplished
by treatment of the Boc-protected intermediate with tri-
fluoroacetate (TFA) resulting in crystalline TFA salts of the
deprotected ligands. Unfortunately, with electron-rich aryl
which intermediate 8 is reacted with a trialkylstannyl arene.
This method, however, requires stannylarene intermediates,
which are less readily available compared with a large
number of commercially available boronic acids. Deprotec-
tion of the benzyl group was achieved by transfer hydroge-
nation using 1,4-cyclohexadiene as a hydrogen source.[22]
Chem. Eur. J. 2006, 12, 4523 – 4535
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4525

P. Anzenbacher, Jr., et al.
Alq3 (e=7”103 mꢁ1 cmꢁ1) was
also observed for the extended
chromophores in both series (1
and 2).
A simple visual examination
of
the
photoluminescence
upon excitation of sample solu-
tions in dichloromethane with
black light (365 nm) reveals
successful color tuning, mediat-
ed by the arylethynyl and aryl
Scheme 2. Deprotection of electron-rich ligands 5 by TFA resulted in formation of carbonyl-containing impuri-
ties of general structure 7. Values for characteristic signals from the NMR spectra of 7k are shown (see text
for more details).
electronic spacers (Figure 2).
Despite the functional-group-
mediated systematic tuning in
the acetylene series, it appears
that the through-the-bridge
conjugation is perhaps too ef-
fective, resulting in relatively
electron-rich species. We pre-
sume this to be the reason that
precluded obtaining a signifi-
cant blueshifted emission in
the acetylene series. Neverthe-
less, the clearly observable
substituent-mediated emission
tuning observed for the first
time in 5-substituted Alq3 de-
rivatives was encouraging.
We believed that the com-
pounds of the aryl series (2a–
n) with the electronic effectors
attached directly to the quino-
Scheme 3. Reaction conditions: a) Bn-Cl, K2CO3, CH3CN, reflux, 16 h; b) [Pd
ACHTREUNG
uene, TBACl, 908C, 24 h; c) [Pd
iPrOH, reflux, 3 h.
ACHTREUNG
linolate fluorophore (i.e., with-
The final AlIII complexes 1a–k and 2a–n were obtained
by reacting the deprotected ligands (free base or TFA salt)
with AlCl3·6H2O in ethanol, followed by neutralization with
triethylamine. The resulting complexes can exist as a mix-
ture of two geometrical isomers, fac and mer, in the solid
state.[23] Nevertheless, the mer-Alq3 isomer is predicted to be
more stable[10,24] and is the only species observed in solution
due to rapid interconversion from the fac form.[25] Indeed,
out the acetylene bridge) would show amplified emission
tuning. This hypothesis was confirmed by fluorescence spec-
tra recorded for both the ethynylarene 1a–k and aryl series
2a–n compounds (Figure 2). In both series, the emission
profiles cover a significant portion of the visible spectrum.
The emission maxima for 1a–k span l=80 nm (520–
600 nm), whereas for the series 2a–n they span l=120 nm
(490–612 nm).
The photophysical data obtained from spectroscopic
measurements of Alq3 complexes 1a–k and 2a–n are sum-
marized in Table 1. In both series, we observed a remarkable
correlation between the photophysical properties and the
electronic nature of the attached substituents. One can see
that both the absorption and the emission shift systematical-
ly from blue to green, yellow, and red depending on the
electronic nature of the modulator group. This attests to a
fine balance of electronic communication between the qui-
nolinolate fluorophores and the appended moieties, while
suppressing direct mesomeric participation of substituent p
electrons of the aryl substituent.
1
the H NMR and 19F NMR spectra recorded for complexes
1a–k and 2a–n support the presence of only one type of
isomer in solution, with a magnetic environment consistent
with the mer geometry (C3v symmetry).[26]
Spectroscopic properties: The complexes 1a–k and 2a–n
were studied by using UV-visible and fluorescence spectros-
copies and the resulting data are summarized in Table 1. For
both series, the p–p* absorption maxima of the complexes
shifted to lower energies relative to the transition of the
parent Alq3 (lambasx =388 nm), presumably the result of an ex-
tension in the conjugation of the quinoline chromophore.
As one could expect, the attachment of an aryl group onto a
triple bond results in a higher degree of conjugation (lower
p–p* transition energies) for the complexes 1a–k relative to
the aryl series 2a–n. A higher oscillator strength relative to
X-ray crystallographic structure determination for the li-
gands suggests that there is a higher degree of coplanarity
for the arylethynyl derivatives than for the aryl compounds
(see Figure 3). By the same token, the ortho interactions be-
4526
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4523 – 4535

Organic Light-Emitting Diodes
FULL PAPER
Table 1. Photophysical properties[a] recorded for Alq3 and complexes 1a–
k and 2a–n in oxygen-saturated dichloromethane at 228C.
Complex
Alq3
l
(e)[b]
lF [nm]
FF
tF [ns]
abs
max
[c]
388 (7.0”103)
526
0.171
15.38
C5-ethynylarene series
1a
1b
1c
1d
1e
1 f
1g
1h
1i
407 (3.5”104)
520
538
541
545
559
573
561
561
569
573
600
0.317
0.228
0.235
0.203
0.037
0.003
0.088
0.092
0.047
0.036
0.009
11.85
8.67
8.95
6.61
2.14
–
3.56
3.30
1.86
1.28
0.89
410 (2.7”104)
414 (2.4”104)
414 (8.4”104)
423 (1.0”104)
429 (4.7”104)
421 (2.2”104)
421 (4.1”104)
421 (1.8”104)
385 (6.8”104)[d]
425 (3.9”104)
1j
1k
C5-aryl series
2a
2b
2c
2d
2e
2 f
2g
2h
2i
2j
2k
2l
2m
2n
abs
390 (2.7”104)
410 (2.7”104)
397 (2.0”104)
388 (1.1”104)
397 (7.5”103)
398 (1.4”104)
394 (1.4”104)
402 (1.0”104)
402 (1.0”104)
490
501
513
516
530
534
537
541
545
0.533
0.511
0.536
0.453
0.301
0.298
0.234
0.201
0.100
0.105
0.098
0.057
0.049
0.008
29.50
23.01
15.61
20.31
16.57
14.53
12.76
11.13
9.72
10.10
6.53
4.73
4.33
1.49
389 (1.7”104)[d]
385 (2.2”104)[d]
410 (4.7”104)
404 (9.8”104)
422 (1.0”104)
538
551
564
564
612
[a] l =absorption maximum, lF =fluorescence maximum, FF =fluores-
max
cence quantum yield, tF =fluorescence lifetime. [b] Units are [nm]
(mꢁ1 cmꢁ1). [c] Determined by using quinine sulfate (in 0.05m H2SO4) as a
standard. [d] Absorption maxima are dominated by the pyrene/anthra-
cene chromophores.
tween the aryl groups and quinoline reduce the degree of
conjugation between the two moieties, which reduces the ef-
ficacy of electronic communication. Specifically, compound
5i, in which the pyrene moiety is separated from quinoline
by ethynylene, displays a 308 dihedral angle between the
planes of quinoline and pyrene, whereas the angles between
quinoline and the 2-pyrimidyl (10c), chlorophenyl (11h),
pentafluorophenyl (10d), and 9-anthryl (10j) groups are 28,
50, 65, and 788, respectively.
Close inspection of the data in Table 1 also shows that the
fluorescence quantum yield and lifetime correlate with the
electronic nature of the ligands. For the complexes bearing
electron-poor substituents, the quantum yield and fluores-
cence lifetime were increased, whereas the electron-rich
moieties resulted in a decrease in both the emission quan-
tum yield and lifetime of the resulting complex with respect
to a neutral substituent such as a phenyl group. In accord-
ance with the energy-gap law,[27,28] analysis of the data con-
firmed an exponential dependence of the nonradiative-
decay rate constant (knr) on the energy gap between the sin-
glet and ground states, suggesting a uniform nature of the
photophysical behavior in each series of complexes (see the
Supporting Information for details). It is noteworthy that
compounds 1e and 1 f do not follow the trend in the aryl-
Figure 2. Emission of complexes 1a–k and 2a–n upon illumination with
black light and the corresponding fluorescence spectra of complexes.
tial localization of the LUMO density on the nitro and pyri-
dinium oxide moieties and alteration of the nature of the ex-
cited state take place.[13]
Furthermore, the fluorescence quantum yields and life-
times exhibited excellent Hammett correlations with the
constants sp.[29,30] This shows that the electronic effects are
projected through the conjugated spacers to the quinolino-
late emitter, thus providing an effective tool not only for
systematic tuning of the photophysical properties, but also
ethynyl series; this can be explained by considering that par-
Chem. Eur. J. 2006, 12, 4523 – 4535
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4527

P. Anzenbacher, Jr., et al.
tion process took place at voltages ranging from 0.27 to
1.32 V (Table 2). Also, in the derivatives with the most blue-
shifted emission (2a–d), the oxidation process became diffi-
cult, an observation in accordance with the expected behav-
Table 2. Redox potentials, determined from cyclic voltammetry experi-
ments,[a] of Alq3 and derivatives 1a–k and 2a–n, and the calculated
HOMO–LUMO gaps.
Complex
Alq3
Eox [V]
Ered [V]
HOMO–LUMO [eV]
2.95
1.02
ꢁ1.93
C5-ethynylarene series
1a
1b
1c
1d
1e
1 f
1g
1h
1i
1.12
0.96
0.73
0.75
0.66
0.81
0.64
0.62
ꢁ1.96
ꢁ1.93
ꢁ1.94
ꢁ1.89
ꢁ1.90
ꢁ1.64
ꢁ1.89
ꢁ1.93
3.08
2.89
2.67
2.64
2.56
2.45
2.53
2.55
2.35
2.35
2.20
Figure 3. Single-crystal X-ray structures of compounds 5i, 10c, 11h, 10d,
and 10j showing the difference in dihedral angle (a) between 5i with an
acetylene bridge and the compounds with an aryl–aryl bridge.
0.50, 1.10[b]
0.52, 1.18[b]
0.27, 0.52[b]
ꢁ1.85
ꢁ1.83
1j
1k
ꢁ1.93
for predicting the properties of the complexes based on the
Hammett correlation approach (Figure 4).
C5-aryl series
ꢁ1.93
2a
2b
2c
2d
2e
2 f
2g
2h
2i
2j
2k
2l
2m
2n
nd[c]
1.32
1.29
1.21
1.16
1.14
1.09
1.09
nd[c]
3.18
3.11
3.12
3.07
3.07
2.97
3.03
2.98
nd[d]
2.95
2.47
2.47
2.30
ꢁ1.86
Electrochemical studies—estimation of HOMO/LUMO
levels: Solution electrochemistry methods are known to pro-
vide insight into electron-transfer reactions in the condensed
phase.[31a] The cyclic voltammogram of Alq3 in solution is
characterized by irreversible single reduction and oxidation
processes. Experimental estimation of the HOMO–LUMO
energy gap is in good agreement[32,33] with theoretical mod-
els.[10d] In order to achieve better understanding of the effect
of the appended EWG/EDG moieties, we performed cyclic
voltammetry (CV) measurements for complexes 1a–k and
2a–n. The main motivation, however, was not only to assess
the energy gap of the materials, but to estimate the actual
position of the HOMO and LUMO levels.
ꢁ1.82
ꢁ1.91
ꢁ1.91
ꢁ1.93
ꢁ1.88
ꢁ1.94
1.08
ꢁ1.90
0.95[d]
1.03[b] (br)
0.53, 1.04[b]
0.53, 0.97[b]
0.36, 0.53[b]
ꢁ1.74[d]
ꢁ1.92
ꢁ1.94
ꢁ1.94
ꢁ1.94
[a] Measurements were performed in deoxygenated solutions of freshly
distilled CH2Cl2 at room temperature. Potentials are reported against the
ferrocene/ferrocenium redox couple (estimated error 50 mV). [b] Ac-
cording to literature reports, these additional peaks can be assigned to
oxidation processes from pyrene, anisol, and dimethylaniline moieties.[31]
[c] No detectable oxidation peak was observed. [d] Redox behavior at-
tributed to the anthracene moiety. nd=not detected.[31a]
In both series 1a–k and 2a–n, we observed an electro-
chemical behavior similar to Alq3.[32] Interestingly, the re-
duction process for the majority of the materials was ob-
served to remain constant circa ꢁ1.9 V, whereas the oxida-
ior of the more electron-defi-
cient systems.
Because the actual HOMO
and LUMO energy levels can
be calculated from redox po-
tentials by calibrating the scale
of
a reference to the zero
vacuum level (Fermi level or
absolute scale),[31a,34] we could
estimate the frontier orbital
energies for 1a–k and 2a–n
(see the Supporting Informa-
tion for details). In both series
the LUMO was found to be
largely unaffected throughout
the series, and only the energy
level of the HOMO was being
~
Figure 4. Hammett correlation of the quantum yield (left) and lifetime (right) for complexes 1d,g,j,k ( ) and
*
2 f–i,l,n ( ).
4528
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4523 – 4535

Organic Light-Emitting Diodes
FULL PAPER
effectively manipulated. The data are shown in the graph in
Figure 5.
These results represent unambiguous proof that the elec-
tronic properties of the moieties attached to the quinolino-
Figure 5. HOMO/LUMO energy levels calculated from the redox poten-
tials of 1a–e,g–k and 2b–i,k–n.
~
&
Figure 6. HOMO–LUMO gap ( ) and HOMO level ( ) for 1a–k and
2a–n as a function of the fluorescence emission energy.
late ligand through aryl and arylethynyl modulators are di-
rectly correlated with the position of the HOMO level of
the complexes, supporting our initial notion for tuning the
emission of Alq3 derivatives through position C5, which has
the highest HOMO density. Although it was expected that
this approach may also lead to changes in the LUMO levels,
surprisingly, the LUMO levels were unperturbed within the
experimental error. Preserving the LUMO level in a more
or less constant position is very important for homogeneous
device manufacturing because these derivatives behave as
electron-transport layers.
For derivatives that could not be thermally evaporated,
the OLEDs were fabricated by means of spin coating: the
ITO substrates were spun with poly(3,4-ethylenedioxythio-
phene)/poly(styrene sulfonate) (PEDOT/PSS) at 4000 rpm
and were annealed at 1508C for 60 min. The emitters were
dissolved in a poly(9-vinylcarbazole) matrix (10 mgmLꢁ1 in
toluene) and spin coated at 2000 rpm onto the PEDOT/PSS.
A 30 nm-thick layer of 2,9-dimethyl-4,7-diphenyl-1,10-phen-
anthroline (BCP) was then vacuum deposited onto the emit-
The HOMO–LUMO gaps calculated for 1a–k and 2a–n
from electrochemical experiments show excellent correla-
tion with the emission energies obtained from the photolu-
minescence (PL) spectra (see Figure 6). These data provide
unequivocal confirmation not only of our structural design,
but also of the photophysical nature of the tuning process.
ter layer, followed by the CsF and Al cathode.
Although we were originally concerned with the possible
effect of the acetylene moiety on the stability of the actual
devices, the OLEDs made of acetylene-substituted acenes[36]
encouraged us to incorporate some of the acetylene-bearing
materials in the OLEDs. Unfortunately, the thermal and
electric stability of 1a–k precluded their vacuum deposition,
and the performance of the spin-coated devices was likewise
unsatisfactory. We believe this to be due to the instability of
the bridge triple bond. On the other hand, the performance
of OLEDs fabricated by using complexes 2a–n confirmed
that these complexes are, indeed, electroluminophores, and,
more importantly, that the electroluminescence (EL) spectra
essentially reproduced the emission-tuning pattern observed
in solution (Figure 7). Thus, the presence of electron-poor
OLED fabrication: The OLED fabrication was performed
by using vacuum deposition and the typical architecture
used for Alq3.[35] The OLEDs were fabricated by using
indium tin oxide (ITO)-coated glass substrates (150–200 nm;
anode), copper phthalocyanine (5 nm), N,N’-diphenyl-N,N’-
bis(1-naphthylphenyl)-1,1’-biphenyl-4,4’-diamine (60 nm), an
Alq3-type emitter (50 nm), a CsF (1 nm) buffer layer, and
an Al cathode (150 nm).
Chem. Eur. J. 2006, 12, 4523 – 4535
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4529

P. Anzenbacher, Jr., et al.
the green- and yellow-emitting devices based on 2g,h and
2l. OLEDs based on complexes 2j and 2k carrying 9-an-
A
file with two components. In these cases, one of the EL
maxima resembles the observed fluorescence spectra while
the second lower-energy transition suggests exciplex emis-
sion generated between anthracene or pyrene moieties from
the Alq3 derivatives and the triarylamine (a-NPD) from the
hole-transporting layer (see Figure 8). This hypothesis was
Figure 7. The OLEDs fabricated by using 2a–g,i–l,n and the correspond-
ing EL spectra confirm effective blue-to-red emission tuning in the solid
state.
substituents resulted in blueshifted EL maxima whereas the
presence of electron-rich groups resulted in redshifted EL.
Overall, the emission obtained from OLEDs fabricated
from 2a–n covered most of the visible region with the peak
emission of the most blueshifted complex at l=479 nm (2a)
and the most redshifted derivative centered at l=616 nm
(2n). To the best of our knowledge, this represents the most
blue- and redshifted emission obtained from Alq3 host mate-
rials reported to date. Interestingly, the blueshifted emitting
materials 2a,d,e displayed slightly blueshifted EL maxima
compared with the corresponding PL patterns (see Table 3).
A similar but less pronounced effect was also observed for
Figure 8. Electroluminescence dependence on voltage for the OLED
based on 2k.
indirectly supported by co-spin coating the respective com-
plexes with a-NPD and recording solid-state PL spectra,
which also showed similar bimodal patterns. In the absence
of a-NPD, the films from 2j and 2k showed only one-com-
ponent emission in agreement with their recorded solution
spectra. In the OLEDs, the relative peak intensity of the ex-
ciplex maxima decreases with increased driving voltage. This
is explained by moving the charge-recombination zone from
the emitter–HTL (HTL=hole-transporting layer) interface
with increasing voltage. Efforts toward utilizing this feature
in the design of OLEDs emitting white light are currently
being investigated.
Table 3. HOMO levels, photophysical data, and electroluminescence
(EL) recorded for Alq3 and complexes 2a–n.
[a]
[b]
[c]
Complex
EHOMO
lPL
lEL
FPL
Cd/Amax
hmax
[eV]
[nm]
[nm]
Alq3
2a
2b
2c
2d
2e
2 f
2g
2h
2i
2j
2k
2l
2m
2n
5.85
nd[d]
6.13
6.09
6.01
5.96
5.95
5.90
5.90
5.88
5.75[e]
5.84
5.33
5.33
5.16
526
490
501
513
516
530
534
537
541
545
538
551
564
564
612
520
479
505
528
500
522
541
531
0.171
0.533
0.511
0.536
0.453
0.301
0.298
0.234
0.201
0.100
0.105
0.098
0.057
0.049
0.008
2.57
1.62
0.31
0.34
0.69
1.45
1.24
2.37
0.04
0.35
0.07
1.01
0.58
nd[d]
0.10
0.83
0.90
0.12
0.12
0.29
0.42
0.38
0.78
0.01
0.19
0.03
0.30
0.18
nd[d]
0.06
The luminance and efficiency of the OLEDs based on
2a–n were generally lower than that of homogeneous Alq3
devices (see Table 3). The slightly inferior performance for
devices constructed with our materials can be attributed to a
number of reasons. In particular, the misalignment of the
HOMO levels between the employed hole-transport layer
531
551
(for a-NPD EHOMO =5.40 eV, for PEDOT/PSS EHOMO
=
542/672
555/660
562
4.90 eV)[14b,37] and the materials 2a–f with EWG (EHOMO
=
5.95–6.13 eV for 2a–f) would hinder hole injection into the
emitter material, shifting the recombination zone towards
the interface with the hole-transport layer thus lowering the
EL efficiency. Likewise, in electron-rich complexes, low
emission quantum yields would be a direct result of the
energy-gap law dependence. Also, an exciplex formation
leading to new redshifted emission bands also reduces the
nd[d]
616
[a] The peak EL wavelength did not change significantly with applied
voltage. [b] Maximum ratio between the forward-directed luminance of
the device and the measured current density passing through the device.
[c] Maximum external efficiency of the device. [d] Not detected.
[e] Redox behavior attributed to anthracene moiety.[31a]
4530
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4523 – 4535

Organic Light-Emitting Diodes
FULL PAPER
efficiency of the devices. In spite of the fact that the OLED
architecture and components were not performance opti-
mized for the new materials, two derivatives, 2a and 2 f,
show a performance comparable to Alq3 that is particularly
significant for the blue 2a-based OLED due to the scarcity
of stable efficient blue emitters and hosts. Furthermore, the
EL performance of these materials could be enhanced by in-
troducing a different hole-transporting material with suit-
Experimental Section
General: Commercially available solvents and reagents were used as re-
ceived from the chemical suppliers. Reactions that required anhydrous
conditions were carried out under an inert atmosphere of argon in oven-
dried glassware. Tetrahydrofuran (THF) was distilled from a K-Na alloy
under argon, and dichloromethane (CH2Cl2) was distilled from CaH2
under argon. All reactions were monitored by using Whatman K6F silica
gel 60 Š analytical TLC plates with UV detection (l=254 and 365 nm).
Silica gel (60 Š, 32–63 mm) from EMD Science was used for column
chromatography. Melting points (uncorrected) were measured by using
Thomas Hoover capillary melting point apparatus. 1H, 13C ATP (APT=
attached proton test), and 19F NMR spectra were recorded by using a
Varian Unity 400 spectrometer with a working frequency of 400.0 MHz.
Chemical shifts were referenced to the residual resonance signal of the
deuterated solvent. Low-resolution mass spectra were recorded on a Shi-
madzu GC–MS QP5050A instrument equipped with a direct probe (ioni-
zation 70 eV). High-resolution mass spectra (HRMS) were measured on
a magnetic sector mass spectrometer using electron impact (EI) or elec-
tron spray ionization (ESI). Elemental analyses were measured by Atlan-
tic Microlab, Inc. and ICT, Prague. Absorption spectra were recorded by
using a Hitachi U-3010 double-beam spectrophotometer, accurate to
ꢂ0.3 nm. Steady-state and time-resolved fluorescence measurements
were performed on a single-photon-counting spectrofluorimeter from Ed-
inburgh Analytical Instruments. The solutions of the complexes were
30 mm in dichloromethane. Quantum yields were determined by using
quinine sulfate (in 0.05m H2SO4) as a standard. X-ray data were collected
at room temperature on a Nonius Kappa CCD diffractometer using a
graphite monochromator with MoKa radiation (l=0.71073 Š).
A
holes and electrons in the emitter layer. Individual optimiza-
tion of the devices by using combinatorial matrix array tech-
niques[38] is expected to yield devices with significantly
higher performance.
Conclusion
The synthesis, optical properties, electrochemistry, electrolu-
minescence, and OLED performance of a series of 5-substi-
tuted tris(8-quinolinolate) aluminum(iii) complexes bearing
A
ED/EW groups connected through aryl or arylethynyl moi-
eties were described. Four pathways based on Sonogashira–
Hagihara and Suzuki–Miyaura methodologies were devised
to carry out the preparation of materials of diverse electron-
ic nature in a simple yet effective fashion. From the spectro-
scopic measurements, it was shown that the substituent elec-
tronic effects, when projected to the quinolinolate chromo-
phore via the selected conjugated spacers, provide an effec-
tive tool for a remarkable and systematic tuning of the emis-
sion over l=350 nm (450–800 nm). Additionally, the corre-
lation between the electronic nature of the substituents and
the emissive properties of the complexes suggests that our
strategy may be used even to predict the properties of new
Alq3-based electroluminescent materials.
The HOMO–LUMO gaps estimated for 1a–k and 2a–n
from electrochemical studies are in agreement with the
emission energies obtained from the photoluminescence
spectra. More importantly, the calculated levels for the fron-
tier orbitals obtained demonstrated that the HOMO level
was effectively manipulated, while the LUMO level was
found to be largely unaffected in both series of derivatives.
In conjunction, these results confirm the hypothesis that the
electronic properties of the moieties attached to the quinoli-
nolate ligand through aryl and arylethynyl modulators are
directly correlated with the position of the HOMO level of
the complexes.
CCDC 287976–287979 contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif.
Synthesis
5-Bromo-8-tert-butoxycarbonyloxyquinoline (3): 4-Dimethylaminopyri-
dine (DMAP; 136 mg, 1.12 mmol) was added at room temperature to a
stirred suspension of 5-bromoquinolin-8-ol (5 g, 22.32 mmol) and di-tert-
butyl dicarbonate (4.87 g, 22.32 mmol) in hexane (500 mL). The mixture
was stirred for 5 h at room temperature, filtered by using a paper filter,
and the filtrate was concentrated under vacuum to provide a crude prod-
uct as a yellowish oil. The product was recrystallized from hexanes to
yield white crystals (6.27 g, 87%). M.p. 86–888C; 1H NMR (CDCl3): d=
1.60 (s, 9H), 7.42 (d, J=8.1 Hz, 1H), 7.55 (dd, J=4.1, 8.5 Hz, 1H), 7.81
(d, J=8.1 Hz, 1H), 8.53 (dd, J=1.5, 8.5 Hz, 1H), 8.96 ppm (dd, 1H, J=
1.5, 4.1 Hz); 13C APT NMR (CDCl3): d=27.7 (CH3), 84.1 (C), 118.7 (C),
121.6 (CH), 122.8 (CH), 128.6 (C), 129.8 (CH), 135.9 (CH), 141.7 (C),
147.1 (C), 150.8 (CH), 151.6 ppm (C); EI-MS (direct insertion probe
(DIP), 70 eV): m/z (%): 225 (100) [M+ꢁBoc]; elemental analysis calcd
(%) for C14H14BrNO3 (324.17): C 51.87, H 4.35, N 4.32, Br 24.65; found:
C 51.62, H 4.57, N 4.15, Br 24.60.
8-tert-Butoxycarbonyloxy-5-ethynylquinoline (4): A solution of 2 (3.0 g,
9.25 mmol) and diisopropylethylamine (DIPEA; 30 mL) in THF
(300 mL) was purged with argon for 10 min at room temperature. Trime-
thylsilylacetylene (6.54 mL, 46.27 mmol) was added at room temperature
under an argon atmosphere followed by addition of [Pd(PPh3)4] (530 mg,
U
Finally, OLED fabrication using materials 2a–n, compris-
ing an aryl group as an electronic spacer, was successfully
achieved and the resulting devices displayed strong electro-
luminescence. Their spectra essentially reproduced the emis-
sion-tuning pattern observed in solution suggesting that this
approach may be used to generate potentially useful elec-
tron-transporting and emissive OLED materials that could
act either as energy hosts for guest dyes or as red/green/blue
(RGB) emitters. This work has given insight into the design
criteria for the development of potentially useful Alq3-based
materials and OLEDs.
0.46 mmol) and CuI (88 mg, 0.46 mmol). The mixture was stirred for 24 h
in a sealed tube under an argon atmosphere at 608C. The mixture was
cooled to room temperature, poured into diethyl ether (300 mL), and fil-
tered through a silica gel pad (10 g) by using diethyl ether (100 mL) as
the eluent. The filtrate was concentrated under vacuum and the residue
was redissolved in methanol (300 mL). KF (717 mg, 12.24 mmol) was
added to the solution and the mixture was stirred for 3 h at room temper-
ature. The reaction mixture was filtered using a paper filter, and water
(300 mL) was added. Methanol was evaporated under vacuum at room
temperature and the residue was extracted with dichloromethane
(300 mL). The dichloromethane layer was separated, dried over anhy-
drous sodium sulfate, filtered, and concentrated under vacuum. The resi-
due was purified by using column chromatography on silica gel (mobile
Chem. Eur. J. 2006, 12, 4523 – 4535
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4531

P. Anzenbacher, Jr., et al.
phase hexanes/acetone 9:1) to provide a white powder (1.72 g, 69%).
The analytical sample was recrystallized from acetone/hexanes to provide
white crystals. M.p. 116–1188C; 1H NMR (CDCl3): d=1.59 (s, 9H), 3.47
(s, 1H), 7.48 (d, J=7.9 Hz, 1H), 7.52 (dd, J=4.2, 8.5 Hz, 1H), 7.76 (d,
J=7.9 Hz, 1H), 8.63 (dd, J=1.7, 8.5 Hz, 1H), 8.96 ppm (dd, J=1.7,
4.2 Hz, 1H); 13C APT NMR (CDCl3): d=27.6 (CH3), 80.0 (CH), 82.6
(C), 84.0 (C), 118.0 (C), 120.5 (CH), 122.5 (CH), 130.0 (C), 131.3 (CH),
134.5 (CH), 140.9 (C), 148.2 (C), 150.8 (CH), 151.6 ppm (C); EI-MS
(DIP, 70 eV): m/z (%): 169 (100) [M+ꢁBoc]; elemental analysis calcd
(%) for C16H15NO3 (269.30): C 71.36, H 5.61, N 5.20; found: C 71.44, H
5.43, N 5.16.
using filtration and the filter cake was washed with acetone and dried to
provide a product.
8-Hydroxy-5-[2-(4,6-dimethoxy-1,3,5-triazinyl)ethynyl] quinoline (6a):
1
Yellowish powder (79%); m.p. 228–2308C; H NMR (CDCl3): d=4.11 (s,
6H), 7.18 (d, J=8.0 Hz, 1H), 7.60 (dd, J=4.3, 8.5 Hz, 1H), 7.93 (d, J=
8.0 Hz, 1H), 8.75 (dd, J=1.5, 8.5 Hz, 1H), 8.85 ppm (dd, J=1.5, 4.3 Hz,
1H); 13C APT NMR (CDCl3): d=55.5 (CH3), 88.9 (C), 90.8 (C), 108.4
(C), 110.1 (CH), 123.2 (CH), 129.7 (C), 134.9 (CH), 135.4 (CH), 137.8
(C), 148.7 (CH), 155.1 (C), 163.0 (C), 172.4 ppm (C); EI-MS (DIP,
70 eV): m/z (%): 308 (100) [M+]; elemental analysis calcd (%) for
C16H12N4O3 (308.29): C 62.33, H 3.92, N 18.17; found: C 62.61, H 3.96, N
18.16.
General method A
General method for the acidic deprotection of a Boc derivative—prepa-
ration of 7d,g,i–k: Trifluoroacetic acid (0.42 mL, 616 mg, 5.40 mmol) was
added to a stirred solution of a Boc derivative (1.08 mmol) in dry CH2Cl2
(40 mL) under an argon atmosphere at room temperature. The mixture
was stirred for 24 h at room temperature under an argon atmosphere.
The solvent was evaporated under vacuum and the residue was recrystal-
lized from acetone.
Sonogashira–Hagihara coupling for preparation of 5a,b,d–f: A solution
of 4 (400 mg, 1.49 mmol), haloarene (1.35 mmol), and DIEPA (4 mL) in
THF (40 mL) was purged with argon for 10 min at room temperature. A
mixture of solid [Pd(PPh3)4] (78 mg, 0.068 mmol) and CuI (13 mg,
T
0.068 mmol) was added into the solution under an argon atmosphere and
the mixture was stirred at 608C for 24 h. The reaction mixture was
cooled to room temperature, diluted with diethyl ether (40 mL), and fil-
tered through a silica gel pad (5 g) by using diethyl ether (40 mL) as the
eluent. The filtrate was concentrated under vacuum and the residue was
purified by using column chromatography on silica gel.
5-(4-Cyanophenylethynyl)-8-hydroxyquinolinium trifluoroacetate (7d):
Yellow powder (99%); m.p. 181–1828C; 1H NMR ([D6]DMSO): d=7.21
(d, J=8.1 Hz, 1H), 7.82 (dd, J=4.3, 8.4 Hz, 1H), 7.85–7.96 (m, 5H), 8.86
(dd, J=1.5, 8.4 Hz, 1H), 9.00 ppm (dd, J=1.5, 4.3 Hz, 1H); 13C APT
NMR ([D6]DMSO): d=90.7 (C), 91.7 (C), 108.9 (C), 110.7 (C), 112.4
(CH), 118.5 (C), 123.3 (CH), 127.4 (C), 129.1 (C), 132.0 (CH), 132.6
(CH), 133.5 (CH), 136.1 (CH), 136.6 (C), 148.3 (CH), 154.5 (C),
158.2 ppm (C; CF3COOꢁ, JCF =36.6 Hz); EI-MS (DIP, 70 eV): m/z (%):
270 (100) [M+ꢁCF3COOH]; elemental analysis calcd (%) for
8-tert-Butoxycarbonyloxy-5-[2-(4,6-dimethoxy-1,3,5-triazinyl) ethynyl]quin-
A
haloarene. Mobile phase for column chromatography: hexanes/acetone
4:1. Yield: 503 mg (83%). An analytical sample was recrystallized from
acetone/hexanes to provide off-white crystals. M.p. 141–1428C; 1H NMR
(CDCl3): d=1.60 (s, 9H), 4.12 (s, 6H), 7.55 (d, J=7.9 Hz, 1H), 7.58 (dd,
J=4.3, 8.5 Hz, 1H), 7.97 (d, J=7.9 Hz, 1H), 8.77 (dd, J=1.7, 8.5 Hz,
1H), 8.99 ppm (dd, J=1.7, 4.3 Hz, 1H); 13C APT NMR (CDCl3): d=27.7
(CH3), 55.6 (CH3), 84.2 (C), 87.0 (C), 91.8 (C), 116.3 (C), 120.6 (CH),
122.9 (CH), 130.4 (C), 133.0 (CH), 134.5 (CH), 141.0 (C), 149.7 (C),
151.1 (CH), 151.4 (C), 162.7 (C), 172.5 ppm (C); EI-MS (DIP, 70 eV): m/
C20H11F3N2O3·3= CH2Cl2 (448.01): C 55.64, H 2.79, N 6.25; found: C 55.82,
4
H 2.53, N 5.94.
8-(Benzyloxy)-5-(4,4,5,5-tetramethyl[1,3,2]dioxaborolan-2-yl)quinoline
H
(9): A flame-dried screw-cap flask was loaded with 8-(benzyloxy)-5-bro-
moquinoline (4.713 g, 15.00 mmol), dry Et3N (15.00 mL), and dry THF
(75.00 mL). The solution was purged with argon for 10 min, then 4,4,5,5-
tetramethyl-1,3,2-dioxaborolane (2.400 mL, 2.112 g, 16.50 mmol) and [Pd-
z
(%): 308 (100) [M+ꢁBoc]; elemental analysis calcd (%) for
C21H20N4O5 (408.41): C 61.76, H 4.94, N 13.72; found: C 61.80, H 5.08, N
13.67.
A
24 h at 1008C under an argon atmosphere, cooled to room temperature,
diluted with diethyl ether (100 mL), and filtered through a silica gel pad
by using diethyl ether as the eluent. The filtrate was concentrated under
vacuum, and the residue was purified by using column chromatography
on silica gel (mobile phase: hexanes/acetone 4:1) to provide the product
as a yellowish oil (5.17 g, 95%), which was recrystallized from acetone/
hexanes to provide white crystals. M.p. 123–1248C; 1H NMR (CDCl3):
d=1.38 (s, 12H), 5.47 (s, 2H), 7.01 (d, J=7.9 Hz, 1H), 7.27 (m, 1H), 7.34
(m, 2H), 7.47 (dd, J=4.1, 8.6 Hz, 1H), 7.50 (m, 2H), 7.99 (d, J=7.9 Hz,
1H), 8.96 (dd, J=1.8, 4.1 Hz, 1H), 9.1 ppm (dd, J=1.7, 8.6 Hz, 1H); 13C
APT NMR (CDCl3): d=24.9 (CH3), 70.5 (CH2), 83.6 (C), 109.1 (CH),
121.8 (CH), 127.0 (CH), 127.8 (CH), 128.6 (CH), 133.4 (C), 136.7 (CH),
137.0 (CH), 140.4 (C), 148.9 (CH), 156.9 ppm (C); EI-MS (DIP, 70 eV):
m/z (%): 361 (22) [M+], 284 (17) [M+ꢁPh], 255 (25) [M+ꢁBnO], 91
(100) [tropylium+].
General method B
Sonogashira–Hagihara coupling for the preparation of 5c,g–k: A solution
of 3 (1.0 g, 3.09 mmol), the acetylene derivative (3.09 mmol), and DIPEA
(10 mL) in THF (100 mL) was purged with argon for 10 min at room
temperature. A solid mixture of [Pd(PPh3)4] (180 mg, 0.155 mmol) and
G
CuI (30 mg, 0.155 mmol) was added under an argon atmosphere and the
mixture was stirred at 608C for 24 h. The reaction mixture was cooled to
room temperature, diluted with diethyl ether (100 mL), and filtered
through a silica gel pad (10 g) by using diethyl ether (50 mL) as the
eluent. The filtrate was concentrated under vacuum and the residue was
purified by using column chromatography on silica gel (mobile phase:
hexanes/acetone 9:1).
8-tert-Butoxycarbonyloxy-5-(4-pyridylethynyl)quinoline (5c): 4-Ethynyl-
pyridine hydrochloride was used as the starting acetylene derivative.
Yield: 520 mg (49%) of a white crystalline compound. An analytical
sample was recrystallized from dichloromethane/hexanes. M.p. 101–
General method C
Suzuki–Miyaura coupling for the preparation of 10a–e,j,k: Compound 4
(1.242 g, 3.44 mmol), haloarene (3.44 mmol), tetrabutylammonium chlo-
1
1028C; H NMR (CDCl3): d=7.47 (m, 2H), 7.54 (d, J=7.8 Hz, 1H), 7.56
(dd, J=4.1, 8.4 Hz, 1H), 7.82 (d, J=7.8 Hz, 1H), 8.63–8.67 (m, 3H),
8.99 ppm (dd, J=1.7, 4.1 Hz, 1H); 13C APT NMR (CDCl3): d=27.7
(CH3), 84.1 (C), 90.1 (C), 91.8 (C), 117.9 (C), 120.6 (CH), 122.6 (CH),
125.5 (CH), 129.7 (C), 130.9 (C), 131.2 (CH), 134.2 (CH), 141.1 (C),
148.6 (C), 149.9 (CH), 150.9 (CH), 151.6 ppm (C); EI-MS (DIP, 70 eV):
m/z (%): 246 (100) [M+ꢁBoc]; elemental analysis calcd (%) for
C21H18N2O3 (346.38): C 72.82, H 5.24, N 8.09; found: C 72.40, H 5.24, N
8.01.
A
(0.180 g, 0.16 mmol) were added to an air-free two-phase mixture of tol-
uene (25 mL) and an aqueous 1m K2CO3 solution (25 mL). The reaction
mixture was intensively stirred under an argon atmosphere at 908C for
24 h. The organic layer was separated and the aqueous phase was extract-
ed with toluene (3”30 mL). The organic layers were combined, washed
with water (2”50 mL), and dried with anhydrous Na2SO4. The solvent
was evaporated and the residue was dissolved in dichloromethane and fil-
tered over silica gel by using 25% methanol/dichloromethane as the
mobile phase.
General method for the basic deprotection of a Boc derivative—prepara-
tion of 6a–c,e,f,h,k: Piperidine (0.1 mL, 1 mmol) was added to a stirred
solution of Boc derivative (0.35 mmol) in dry CH2Cl2 (1 mL) at room
temperature and the mixture was stirred for 5 min at the same tempera-
ture. The mixture was concentrated under vacuum and the residue was
treated with acetone (5 mL). The resulting precipitate was isolated by
8-(Benzyloxy)-5-[2-(4,6-dimethoxy-1,3,5-triazinyl)] quinoline (10a): 2-
Chloro-4,6-dimethoxy-1,3,5-triazine was used as the starting aryl halide.
Mobile phase for chromatography: hexanes/acetone 7:3. Yield: 0.190 g
(51%) of white solid. The analytical sample was obtained by crystalliza-
4532
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4523 – 4535

Organic Light-Emitting Diodes
FULL PAPER
tion from an acetone/hexanes mixture to provide white crystals. M.p.
122–1238C; 1H NMR (CDCl3): d=4.13 (s, 6H), 5.53 (s, 2H), 7.11 (d, J=
8.4 Hz, 1H), 7.30 (m, 1H), 7.37 (m, 2H), 7.53 (m, 2H), 7.55 (dd, J=4.0,
8.9 Hz, 1H), 8.53 (d, J=8.4 Hz, 1H), 9.01 (dd, J=1.7, 4.0 Hz, 1H),
9.72 ppm (dd, J=1.7, 8.9 Hz, 1H); 13C APT NMR (CDCl3): d=55.2
(CH3), 70.8 (CH2), 108.6 (CH), 122.4 (CH), 123.9 (C), 127.0 (CH), 127.9
(CH), 128.2 (C), 128.6 (CH), 132.5 (CH), 134.8 (CH), 136.1 (C), 140.3
(C), 149.1 (CH), 157.7 (C), 172.3 (C), 175.5 ppm (C); EI-MS (DIP,
70 eV): m/z (%): 374 (21) [M+], 297 (14), 268 (17) [M+ꢁBnO], 91 (100)
[tropylium+].
7.24 (dd, J=1.5, 4.9 Hz, 1H), 7.30 (dd, J=4.9, 8.3 Hz, 1H), 7.40–7.50 (m,
7H), 7.54 (dd, J=4.9, 8.3 Hz, 1H), 7.75 (m, 3H), 8.73 (dd, J=1.5, 8.3 Hz,
1H), 8.77 (dd, J=1.5, 8.3 Hz, 1H), 8.81 (m, 2H), 8.85 ppm (dd, J=1.5,
4.9 Hz, 1H); 13C APT NMR (CDCl3): d=40.4 (CH3), 84.0 (C), 84.3 (C),
84.7 (C), 92.8 (C), 93.1 (C), 93.4 (C), 106.3 (C), 106.4 (C), 106.7 (C),
110.3 (C), 110.5 (C), 110.7 (C), 111.9 (CH), 112.7 (CH), 113.2 (CH),
113.8 (CH), 121.5 (CH), 122.0 (CH), 122.2 (CH), 129.7 (C), 129.9 (C),
130.0 (C), 132.5 (CH), 134.7 (CH), 134.9 (CH), 135.2 (CH), 138.8 (CH),
138.9 (CH), 139.15 (C), 139.24 (CH), 139.8 (C), 142.7 (CH), 144.9 (CH),
145.2 (CH), 149.9 (C), 149.98 (C), 150.04 (C), 158.7 (C), 159.0 (C),
159.2 ppm (C); HRMS (ESI): m/z calcd for C57H46AlN6O3 [M+H+]:
889.3447; found: 889.3487; elemental analysis calcd (%) for
C57H45AlN6O3·5H2O (979.08): C 69.93, H 5.66, N 8.58; found: C 70.17, H
5.02, N 8.38.
General method D
Suzuki–Miyaura coupling for the preparation of 10 f–h,l–n: 8-(Benzyl-
A
tive (3.40 mmol), tetrabutylammonium chloride (TBACl; 0.110 g,
0.37 mmol), and tetrakis(triphenylphosphine) palladium (0.180 g,
0.16 mmol) were added to an air-free two-phase mixture of toluene
(30 mL) and an aqueous 1m K2CO3 solution (30 mL). The resulting mix-
ture was intensively stirred under an argon atmosphere at 908C for 24 h.
The organic layer was separated and the aqueous phase was extracted
with toluene (3”30 mL). The organic layers were combined, washed with
water (2”50 mL), and dried with anhydrous Na2SO4. The solvent was
evaporated and the residue was dissolved in dichloromethane and filtered
over silica gel by using 25% methanol/dichloromethane as the mobile
phase. The crude product was purified by recrystallization or column
chromatography.
General method for the preparation of the aluminum complexes 2a–n: A
mixture of the ligand (0.15 mmol) and AlCl3·6H2O (12 mg, 0.05 mmol) in
degassed ethanol (5 mL) was left at reflux for 3 h. The mixture was
cooled to room temperature and neutralized by using triethylamine.
Water (10 mL) was added and the precipitated complex was filtered off.
The filter cake was washed thoroughly with water, ethanol, and diethyl
ether, and dried under vacuum to provide the desired product.
Aluminum
A
tris{5-[2-(4,6-dimethoxy-1,3,5-triazinyl)]-8-quinolinolate}
(2a): Yellow solid (28 mg, 95%); m.p. >2608C; 1H NMR (CDCl3): d=
4.11 (s, 6H), 4.12 (s, 6H), 4.13 (s, 6H), 7.15 (m, 2H), 7.19 (d, J=8.5 Hz,
1H), 7.29 (dd, J=1.5, 4.7 Hz, 1H), 7.38 (dd, J=4.7, 8.7 Hz, 1H), 7.55
(dd, J=4.7, 8.7 Hz, 1H), 7.63 (dd, J=4.7, 8.7 Hz, 1H), 8.83 (dd, J=1.5,
4.7 Hz, 1H), 8.86–8.94 (m, 3H), 8.96 (d, J=8.5 Hz, 1H), 10.17 (dd, J=
1.5, 8.7 Hz, 1H), 10.21 (dd, J=1.5, 8.7 Hz, 1H), 10.26 ppm (dd, J=1.5,
8.7 Hz, 1H); HRMS (ESI): m/z calcd for C42H33AlN12O9 [M+H+]:
877.2387; found: 877.2409.
8-(Benzyloxy)-5-(4-cyanophenyl)quinoline (10 f): 4-Cyanophenylboronic
acid was used as the starting material for the cross-coupling. The result-
ing compound was recrystallized from ethanol to provide white crystals
1
(0.536 g, 46%). M.p. 180–1818C; H NMR (CDCl3): d=5.51 (s, 2H), 7.09
(d, J=7.9 Hz, 1H), 7.32 (m, 2H), 7.38 (m, 2H), 7.43 (dd, J=4.3, 8.5 Hz,
1H), 7.51–7.56 (m, 4H), 7.76 (m, 2H), 8.11 (dd, J=1.8, 8.5 Hz, 1H),
9.03 ppm (dd, J=1.8, 4.3 Hz, 1H); 13C APT NMR (CDCl3): d=70.9
(CH2), 109.4 (CH), 111.2 (C), 118.8 (C), 122.0 (CH), 127.1 (CH), 127.2
(C), 127.6 (CH), 128.0 (CH), 128.3 (C), 128.7 (CH), 130.3 (C), 130.8
(CH), 132.3 (CH), 133.4 (CH), 136.7 (CH), 140.6 (C), 144.3 (C), 149.5
(CH), 149.7 (C), 153.1 (C); EI-MS (DIP, 70 eV): m/z (%): 336 (20) [M+],
259 (7) [MꢁC6H6+], 230 (11) [Mꢁbenzaldehyde+], 91 (100) [tropylium+
].
Calculation of the energy-gap law correlation data: Values for the emis-
sion energy (Eem) and the radiative (kr) and nonradiative (knr) rate decay
rate constants were calculated by using the Equations (1)–(3):
hc
lmax
ð1Þ
ð2Þ
ð3Þ
Eem ¼ hnmax
¼
ꢀ
kr ¼
tF
General method for the deprotection of the benzyl derivative by catalytic
transfer hydrogenation—preparation of the ligands 11a–n: 1,4-Cyclohex-
adiene (0.500 mL, 5.4 mmol) was added in one portion to a mixture of
the benzyl derivatives 10 (1.80 mmol) and 10% Pd/C catalyst (300 mg) in
degassed isopropanol (30 mL). The mixture was heated at 1108C for 3 h
in a screw-cap flask. The solution was cooled to room temperature and
filtered over filter paper to remove the catalyst. The solvent was evapo-
rated and the crude product was purified by means of recrystallization or
column chromatography.
1
tF
knr
¼
ꢁkr
Estimation of the HOMO/LUMO energy levels for the complexes: The
HOMO and LUMO energies were calculated by using a commonly ac-
cepted procedure.[34] According to this, the value for ferrocene (Fc) with
respect to the zero vacuum level is estimated as ꢁ4.8 eV, determined
from ꢁ4.6 eV for the standard electrode potential EA of a normal hydro-
gen electrode (NHE) on the zero vacuum level, and 0.2 V for Fc versus
NHE. Thus, the values for the HOMO and LUMO levels were obtained
through Equations (4) and (5) as follows:
8-Hydroxy-5-[2-(4,6-dimethoxy-1,3,5-triazinyl)]quinoline (11a): 8-(Ben-
A
N
the starting material. The resulting product was recrystallized from etha-
nol to give white crystals (0.84 g, 59%). M.p. 227–2298C; 1H NMR
(CDCl3): d=4.15 (s, 6H), 7.25 (d, J=8.3 Hz, 1H), 7.56 (dd, J=4.1,
8.9 Hz, 1H), 8.76 (d, J=8.3 Hz, 1H), 8.81 (dd, J=1.6, 4.1 Hz, 1H),
9.89 ppm (dd, J=1.6, 8.9 Hz, 1H); 13C APT NMR (CDCl3): d=55.1
(CH3), 109.2 (CH), 122.3 (C), 122.9 (CH), 127.8 (C), 134.0 (CH), 136.0
(CH), 138.4 (C), 147.6 (CH), 156.5 (C), 172.6 (C), 175.7 ppm (C); EI-MS
(DIP, 70 eV): m/z (%): 284 (100) [M+], 269 (50), 212 (33).
HOMO ¼ ꢁðEox þ 4:8Þ eV
LUMO ¼ ꢁðEred þ 4:8Þ eV
ð4Þ
ð5Þ
in which Eox and Ered are the redox potentials referenced against ferro-
cene.
General method for the preparation of complexes 1a–k: A 0.05m aque-
ous solution of AlCl3·6H2O (0.07 mol) was added to a boiling solution of
the ligand (free base or TFA) (0.210 mmol) in ethanol (15 mL). The mix-
ture was left at reflux for 1 h, neutralized with Et3N and cooled to room
temperature. The precipitated aluminum complex was collected by using
filtration and the filter cake was washed with water, ethanol, and diethyl
ether to provide the desired product. Complexes 1a–k can be recrystal-
lized from ethanol or dichloromethane/hexanes.
Construction of the devices: The OLEDs were fabricated on Applied
Films Corp ITO-coated glass substrates (150–200 nm thick; resistance, R
ꢀ20 W/square). The ITO-coated substrates, which served as the anode,
were degreased by a detergent and organic solvents and then UV-ozone
cleaned to increase the ITO work function. The organic layers, CsF elec-
tron injection buffer layer, and Al cathode were deposited by thermal
evaporation in a high-vacuum (ꢀ106 torr) chamber installed in an Ar-
filled glove-box (<1 ppm O2, H2O). The organic layers consisted of a
5 nm-thick copper phthalocyanine (CuPc) layer, a 60 nm-thick N,N’-di-
Aluminum
A
tris[5-(4-dimethylaminophenylethynyl)]-8-quinolinolate
1
(1k): Orange solid (84%); m.p. 245–2478C (decomp); H NMR (CDCl3):
d=2.98 (s, 12H; CH3), 2.99 (s, 6H; CH3), 6.67 (m, 6H), 7.07 (m, 3H),
phenyl-N,N’-bis(1-naphthylphenyl)-1,1’-biphenyl-4,4’-diamine
(a-NPD)
layer, and a 50 nm-thick layer of the Alq3 derivative. The cathode consist-
Chem. Eur. J. 2006, 12, 4523 – 4535
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4533

P. Anzenbacher, Jr., et al.
ed of a ꢀ1 nm-thick CsF layer and a ꢀ150 nm-thick Al layer. The depo-
sition rate of the organic layers was 0.05–0.1 nmsꢁ1, controlled by a cali-
brated quartz-crystal thickness monitor. The CsF layer was deposited at
ꢀ0.01 nmsecꢁ1. The Al was deposited at 0.5–0.7 nmsecꢁ1 through a mask
of 3.0 mm diameter holes. The Alq3 derivatives that could not be ther-
mally evaporated were dissolved in toluene and embedded in a poly(9-vi-
nylcarbazole) (PVK) (10 mgmLꢁ1 in toluene) matrix. The “1”1” ITO
substrates were then spin coated at 4000 rpm with poly(3,4-ethylenediox-
ythiophene) (PEDOT)/poly(styrene sulfonate) (PSS) and baked at 1508C
for 60 min. The Alq3 derivative/PVK blend was then spin coated at
2000 rpm onto the PEDOT/PSS. A 30 nm-thick layer of 2,9-dimethyl-4,7-
diphenyl-1,10-phenanthroline (BCP), a hole-blocking electron-transport-
ing layer, was then thermally evaporated onto the spin-coated Alq3 deriv-
ative-doped PVK. CsF and Al films were deposited as described above.
Metaland Rare Earth Compounds: Excited States, Transitions, Inter-
actions I (Ed.: H. Yersin), Springer, Berlin, 2001.
[12] L. S. Sapochak, A. Padmaperuma, N. Washton, F. Endrino, G. T.
Schmett, J. Marshall, D. Fogarty, P. E. Burrows, S. R. Forrest, J. Am.
Chem. Soc. 2001, 123, 6300–6307.
[13] M. Sugimoto, M. Anzai, K. Sakanoue, S. Sakaki, Appl. Phys. Lett.
2001, 79, 2348–2350.
[14] a) T. A. Hopkins, K. Meerholz, S. Shaheen, M. L. Anderson, A.
Schmidt, B. Kippelen, A. B. Padias, J. H. K. Hall, N. Peyghambarian,
N. R. Armstrong, Chem. Mater. 1996, 8, 344–351; b) J.-A. Cheng,
C. H. Chen, C. H. Liao, Chem. Mater. 2004, 16, 2862–2868.
[15] P. E. Burrows, Z. Shen, V. Bulovic, D. M. McCarty, S. R. Forrest,
J. A. Cronin, M. E. Thompson, J. Appl. Phys. 1996, 79, 7991–8006.
[16] M. Matsumura, T. Akai, Jpn. J. Appl. Phys. Part 1 1996, 35, 5357–
5360.
Characterization of the OLEDs: The current–voltage and electrolumines-
cence intensity–voltage curves were measured by using a Keithley 2000
digital multimeter (DMM) and Minolta L100 luminance meter. The EL
spectra were measured by using an OceanOptics Chem-2000 spectrome-
ter. The external maximum efficiency of the diodes was determined by
using Equation (6):
[17] S. Anderson, M. S. Weaver, A. J. Hudson, Synth. Met. 2000, 111–112,
459–463.
[18] a) P. S. Bryan, F. V. Lovecchio, S. A. VanSlyke (assignee: Eastman
Kodak Company), United States Patent and Trademark Office
(USPTO), US 5,141,671, 1992; b) S. A. VanSlyke (assignee: Eastman
Kodak Company), USPTO, US patent, US 5,151,629, 1992; c) L. S.
Sapochak, P. E. Burrows, D. Garbuzov, D. M. Ho, S. R. Forrest,
M. E. Thompson, J. Phys. Chem. 1996, 100, 17766–17771.
[19] R. Pohl, P. Anzenbacher, Jr., Org. Lett. 2003, 5, 2769–2772.
[20] R. Pohl, V. A. Montes, J. Shinar, P. Anzenbacher, Jr., J. Org. Chem.
2004, 69, 1723–1725.
R
pe
Kmhc J
L
FðlÞdl
h ¼
R
ð6Þ
FðlÞyðlÞdl
l
in which L is the measured forward-directed luminance [cdmꢁ2], J is the
current density [Amꢁ2], Km is the maximum luminous efficacy (namely,
680 LmWꢁ1), y(l) is the normalized photopic spectral response function,
F(l) is the electroluminescence emission spectrum for the device, and l
is the wavelength.
[21] R. Prasad, H. L. D. Coffer, Q. Fernando, H. Freiser, J. Org. Chem.
1965, 30, 1251–1251.
[22] Y. Nakano, D. Imai, Synthesis 1997, 1425–1428.
[23] a) M. Coelle, R. E. Dinnebier, W. Bruetting, Chem. Commun. 2002,
23, 2908–2909; b) M. Coelle, S. Forero-Lenger, J. Gmeiner, W.
Bruetting, Phys. Chem. Chem. Phys. 2003, 5, 2958; c) M. Coelle, W.
Bruetting, Phys. Status Solidi A 2004, 201, 1095–1115; d) M. Utz,
M. Nandagopal, M. Mathai, F. Papadimitrakopoulos, Appl. Phys.
Lett. 2003, 83, 4023–4025.
[24] A. Curioni, W. Andreoni, J. Am. Chem. Soc. 1999, 121, 8216–8220.
[25] M. Utz, C. Chen, M. Morton, F. Papadimitrakopoulos, J. Am. Chem.
Soc. 2003, 125, 1371–1375.
[26] B. C. Baker, D. T. Sawyer, Anal. Chem. 1968, 40, 1945–1951.
[27] N. J. Turro, Modern Molecular Photochemistry, Benjamin–Cummings
Publishing, Menlo Park, 1978.
[28] B. R. Henry, W. Siebrand in Organic Molecular Photophysics, Vol. 1
(Ed.: J. B. Birks), Wiley, London, 1973.
[29] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.
[30] M. B. Smith, J. March, Marchꢀs Advanced Organic Chemistry: Reac-
tions, Mechanisms, and Structure, 5th ed., Wiley-Interscience, New
York, 2001.
[31] a) A. Bard, R. Faulkner, Electrochemical Methods—Fundamentals
and Applications, 2nd ed., Wiley, New York, 2001; b) A. Benniston,
A. Harriman, D. Lawrie, A. Mayeux, K. Rafferty, O. Russell, Dalton
Trans. 2003, 4762–4769; c) M. Hissler, A. Harriman, A. Khatyr, R.
Ziessel, Chem. Eur. J. 1999, 5, 3366–3381; d) P. Roy, M. Saha, T.
Okajima, T. Ohsaka, Electroanalysis 2004, 16, 289–297; e) T. Mizo-
guchi, R. N. Adams, J. Am. Chem. Soc. 1962, 84, 2058–2061; f) E.
Bosch, J. Kochi, J. Org. Chem. 1994, 59, 5573–5586.
[32] J. D. Anderson, E. M. Mc Donald, P. A. Lee, M. L. Anderson, E. L.
Ritchie, H. K. Hall, T. Hopkins, E. A. Mash, J. Wang, A. Padias, S.
Thayumavan, S. Barlow, S. R. Marder, G. E. Jabbour, S. Shaheen, B.
Kippelen, N. Peyghambarian, R. M. Wightman, N. R. Armstrong, J.
Am. Chem. Soc. 1998, 120, 9646–9655.
Acknowledgements
Financial support from the NSF (NER Grant 0304320, DMR Grant
0306117 to P.A.), the Alfred P. Sloan Research Foundation to P.A., the
BGSU (Technology Innovations Enhancement grant to P.A.), and the Mc
Master Endowment for a Mc Master fellowship to V.M. is gratefully ac-
knowledged. Ames Laboratory is operated by Iowa State University for
the US Department of Energy under contract W-7405-Eng-82. The work
in Ames was supported by the Director for Energy Research, Office of
Basic Energy Sciences. We thank Manuel Palacios for his help with the
cover graphic.
[1] S. M. Kelly in Flat Panel Displays: Advanced Organic Materials
(series Ed.: J. A. Connor), The Royal Society of Chemistry, Cam-
bridge, 2000.
[2] J. Shinar, Organic Light-Emitting Devices—A Survey, Springer,
Berlin, 2003.
[3] A. B. Holmes, Nature 2003, 421, 800–801.
[4] L. S. Hung, C. H. Chen, Mater. Sci. Eng. R 2002, 39, 143–222.
[5] E. Holder, B. M. W. Langeveld, U. S. Schubert, Adv. Mater. 2005, 17,
1109–1121.
[6] S. Naka, H. Okada, H. Onagawa, Mol. Electron. Bioelectron. 2000,
11, 70–77.
[7] C. W. Tang, S. A. VanSlyke, Appl. Phys. Lett. 1987, 51, 913–915.
[8] V. Bulovic, M. A. Baldo, S. R. Forrest in Organic Electronic Materi-
als (Eds.: R. Farchioni, G. Grosso), Springer, Berlin, 2001, pp. 391–
439, and references therein.
[33] I. G. Hill, A. Kahn, Z. G. Soos, R. A. Pascal, Chem. Phys. Lett. 2000,
327, 181–188.
[34] J. Pommerehne, H. Vestweber, W. Guss, R. Mahrt, H. Bässler, M.
Porsch, J. Daub, Adv. Mater. 1995, 7, 551–554.
[35] H. Mu, H. Shen, D. Klotzkin, Solid-State Electron. 2004, 48, 2085–
2088.
[36] S. A. Odom, S. R. Parkin, J. E. Anthony, Org. Lett. 2003, 5, 4245–
4248.
[9] C. H. Chen, J. Shi, Coord. Chem. Rev. 1998, 171, 161–174.
[10] a) A. Curioni, M. Boero, W. Andreoni, Chem. Phys. Lett. 1998, 294,
263–271; b) M. Sugimoto, S. Sakaki, K. Sakanoue, M. D. Newton, J.
Appl. Phys. 2001, 90, 6092–6097; c) M. D. Halls, H. B. Schlegel,
Chem. Mater. 2001, 13, 2632–2640; d) B. C. Lin, C. P. Cheng, Z.-Q.
You, C.-P. Hsu, J. Am. Chem. Soc. 2005, 127, 66–67.
[11] “Luminescent Metal Complexes: Diversity of Excited States”:
A. K. H. Vogler in Topics in Current Chemistry, Vol. 241, Transition
4534
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4523 – 4535

Organic Light-Emitting Diodes
FULL PAPER
[37] A. Van Dijken, J. J. A. M. Bastiaansen, N. M. M. Kiggen, B. M. W.
Langeveld, C. Rothe, A. Monkman, I. Bach, P. Stçssel, K. Brunner,
J. Am. Chem. Soc. 2004, 126, 7718–7727.
[38] a) L. Zou, V. Savvateꢁev, J. Booher, C.-H. Kim, J. Shinar, Appl.
Phys. Lett. 2001, 79, 2282–2284; b) K. O. Cheon, J. Shinar, Appl.
Phys. Lett. 2003, 83, 2073–2075.
Received: November 11, 2005
Published online: April 18, 2006
Chem. Eur. J. 2006, 12, 4523 – 4535
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4535