An azomethin-zinc complex for organic electroluminescence: Crystal structure, thermal stability and optoelectronic properties

Article abstract of DOI:10.1016/j.ica.2005.08.018

An azomethin-zinc complex, bis[salicylidene(4-dimethylamino)aniline] zinc(II) (Zn(sada)₂) was synthesized and structurally characterized by single-crystal X-ray crystallography. Crystal data for Zn(C ₁₅H₁₅N₂O)₂ was determined as follows: space group, triclinic, P1?; a = 10.2791(9) ?, b = 16.5008(14) ?, c = 17.5984(15) ?, α = 114.830(2)°, β = 96.579(2)°, γ = 97.674(2)°, Z = 4. Through thermal analysis characterization and FT-IR spectra, this complex was proved to have good thermal stability. The vapor-deposited films exhibited uniform and environment-stable morphology. The light emission and charge transporting performance of Zn(sada)₂ in organic light emitting diodes (OLEDs) were investigated preliminarily, and the results indicated the superior electron transporting property of this complex. Compared with the typical bilayer device of N,N′-diphenyl-N,N′-bis(1-naphthyl)-benzidine (NPB)/tris-(8- hydroxyquinoline)aluminum (Alq₃), the device with Zn(sada) ₂ as the electron transporting layer exhibited a much lower turn-on voltage of 2.5 V (it is usually 3.5 V for an NPB/Alq₃ device).

Full text of DOI:10.1016/j.ica.2005.08.018

Inorganica Chimica Acta 358 (2005) 4451–4458
www.elsevier.com/locate/ica
An azomethin-zinc complex for organic electroluminescence:
Crystal structure, thermal stability and optoelectronic properties
*
Junfeng Xie, Juan Qiao , Liduo Wang, Jing Xie, Yong Qiu
Key Laboratory of Organic Optoelectronics and Molecular Engineering of Ministry of Education, Beijing 100084, China
Department of Chemistry, Tsinghua University, Beijing 100084, China
Received 30 March 2005; received in revised form 11 August 2005; accepted 11 August 2005
Available online 26 September 2005
Abstract
An azomethin-zinc complex, bis[salicylidene(4-dimethylamino)aniline]zinc(II) (Zn(sada)2) was synthesized and structurally character-
ꢀ
ized by single-crystal X-ray crystallography. Crystal data for Zn(C15H15N2O)2 was determined as follows: space group, triclinic, P1;
˚
˚
˚
a = 10.2791(9) A, b = 16.5008(14) A, c = 17.5984(15) A, a = 114.830(2)ꢀ, b = 96.579(2)ꢀ, c = 97.674(2)ꢀ, Z = 4. Through thermal anal-
ysis characterization and FT-IR spectra, this complex was proved to have good thermal stability. The vapor-deposited films exhibited
uniform and environment-stable morphology. The light emission and charge transporting performance of Zn(sada)2 in organic light
emitting diodes (OLEDs) were investigated preliminarily, and the results indicated the superior electron transporting property of this
complex. Compared with the typical bilayer device of N,N0-diphenyl-N,N0-bis(1-naphthyl)-benzidine (NPB)/tris-(8-hydroxyquino-
line)aluminum (Alq3), the device with Zn(sada)2 as the electron transporting layer exhibited a much lower turn-on voltage of 2.5 V
(it is usually 3.5 V for an NPB/Alq3 device).
ꢁ 2005 Elsevier B.V. All rights reserved.
Keywords: Zinc complexes; Organic electroluminescence; X-ray crystal structures; Thermal stability; Optoelectronic properties
1. Introduction
was demonstrated to be a useful emitter material or elec-
tron-transporting hosts in OLEDs [2].
Since the first vacuum-deposited OLEDs using tris-(8-
hydroxyquinoline) aluminum (Alq3) were reported by Tang
and VanSlyke [1], organic metal-chelate compounds have
drawn a great deal of attention. These materials offer many
attractive properties such as the dual function of electron-
transporting and light emission, higher environmental sta-
bility, and ease of sublimation. By varying the central metal
atom (Al, Ga, In, Be, Mg, Ca, Zn, Cu, etc.), a variety of
other metal complexes with 8-hydroxyquinoline have been
designed and synthesized for use in OLEDs. Amongst
these, the zinc complex bis-(8-hydroxyquinoline)zinc
(Znq2) was found to have strong yellow fluorescence and
It is well understood that the charge balance in OLEDs is
one of the most important factors necessary to achieve high
device efficiency. Alq3 is the most widely used electron-trans-
porting material in OLEDs. However, the electron mobility
of Alq3 is lower than the hole mobility of traditional hole
transporting materials (HTMs), N,N0-diphenyl-N,N0-bis(3-
methyl)-benzidine (TPD) and N,N0-diphenyl-N,N0-bis(1-
naphthyl)-benzidine (NPB). Zinc complexes have been used
in OLEDs for more than a decade [2,3], but the best electro-
luminescent performance of these materials as emitters is just
comparable with that of Alq3. However, in many instances,
the electron-transporting mobility of zinc complexes goes
beyond that of Alq3. So zinc complexes may be potential can-
didates to enhance the electron-transporting properties for
OLEDs.
*
Corresponding author. Tel.: +86 10 6277 3109; fax: +86 10 6279 5137.
E-mail address: qjuan@mail.tsinghua.edu.cn (J. Qiao).
0020-1693/$ - see front matter ꢁ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2005.08.018

4452
J. Xie et al. / Inorganica Chimica Acta 358 (2005) 4451–4458
Since 1993, when OLEDs with Znq2 were reported [2],
ture of Zn(sada)2, the thermal stability of the racemic
structure and its optoelectronic properties.
studies of zinc complexes as active materials for OLEDs have
focused on improving electron mobility and/or producing
a blue shift, relative to Znq2, in the emission wavelength
maximum. Zn2+ ion is the only oxidation state of zinc atom
and has no unoccupied valence electron orbits. Therefore,
the oxidation and reduction of zinc complexes are mainly
carried out in the ligands, i.e., luminescence and charge
transporting may be ascribed to intraligand electron transi-
tion. One simple way to tune the emission wavelength is to
manipulate substituents in the 8-hydroxyquinoline rings.
2-Position substituted bis(2-methyl-8-hydroxyquinoline)-
zinc (ZnMq2) [3], 5-position substituted bis(8-hydroxy-5-
piperidinylsulfonamidoquinolate)zinc (Zn(QS)2) [4] and
(ZnLn)2 Æ 2H2O (L = 5-amido-substituted-8-hydroxyquino-
linate ligand) [5], etc., have been reported one after the other.
These complexes showed more or less either blue-shifting or
red-shifting with reference to Znq2. The other way is to intro-
duce non-oxidate ligands, including Schiff base ligands
[2,6,7], polycyclic aromatic ligands [8–11] and polypyridyl-
amines [12–14]. In addition, metalloporphyrin [15] and metal
clusters [16,17] were also used to obtain fine EL materials.
Most of these zinc complexes have been reported as potential
OLED materials. However, few of them have been used for
practical OLEDs.
One of the aims of molecular design is to obtain device-
quality materials. Studying structure–property relation-
ships provides guidance in molecular design for improved
molecules. The molecular structure of anhydrous Znq2
was determined in 1985 [18], and this complex was intro-
duced in OLEDs in 1993, but it was not until 2002 that
the relationship between the structure of tetrameric Znq2
and its performance in OLEDs was clearly studied by
Sapochak et al. [19]. A detailed investigation of zinc(II)
2-(2-hydroxyphenyl)benzothiazolate (Zn(BTZ)2) was just
accomplished last year [20], although it was reported as
an excellent white EL material as early as 1996 [9].
2. Experimental
2.1. Synthesis of sada
A mixture of salicylaldehyde and N,N-dimethyl-ben-
zene-1,4-diamine (purchased from Aldrich Chemical Co.)
in a 1:1 molar ratio was heated and the following recrystal-
lization with ethanol gave an orange-red precipitate with a
yield of about 87%. MS (EI) [m/z] 240. 1H NMR
(400 MHz, CD3Cl), d = 2.99 (s, 6H), 6.80–6.70 (m, 2H),
6.91 (t, 1H, J = 8.6 Hz), 7.00 (d, 1H, J = 8.7 Hz), 7.27 (d,
2H, J = 9.0 Hz), 7.30 (ddd, 1H, J = 8.6, 7.0, 1.8 Hz), 7.34
(dd, 1H, J = 7.8, 1.8 Hz), 8.61 (s, 1H), 13.73 (s, 1H).
2.2. Synthesis of Zn(sada)2
First, a solution of ZnCl2 (0.6815 g, 5 mmol) in ethanol
(30 ml) was gradually added to a solution of sada (1.20 g,
5 mmol) and piperidine (1.0 ml, 10 mmol) in 120 ml etha-
nol. After the mixture was stirred for 0.5 h while being
heated at reflux, and stirred again for 24 h at room temper-
ature, a yellow precipitate was produced. The crude prod-
uct was collected by filtration and washed with ethanol and
finally dried under an infrared lamp. The material was fur-
ther purified by vacuum train sublimation before analysis
and the fabrication of device. Yield: 86%. MS (EI) [m/z]
542. Anal. Calc. for C30H30N4O2Zn: C, 66.24; H, 5.56; N,
10.30. Found: C, 66.29; H, 5.43; N, 10.50%. 1H NMR
(400 MHz, [D6]DMSO), d = 2.85 (s, 12H), 6.70–6.62 (m,
8H), 7.22 (dd, 4H, J = 7.0, 1.9 Hz), 7.30 (ddd, 2H,
J = 8.7, 7.0, 1.9 Hz), 7.50 (dd, 2H, J = 8.0, 1.8 Hz), 8.78
(s, 2H). IR (cmꢀ1): 2898, 1612, 1579, 1531, 1518, 1463,
1442, 1387, 1365, 1348, 1331, 1317, 1255, 1226, 1190,
1180, 1149, 1125, 1064, 1032, 947, 922, 862, 818, 757, 717.
Recently, our group has focused on metal complexes
based on Schiff base ligands, which have proven to be
highly efficient luminescent materials for OLEDs [21,22].
Azomethin-zinc complexes are typical metal complexes
based on Schiff base ligands and early studies were focused
on their structures and reactions [23–27]. As early as 1993,
they were reported as potential OLED materials by Hama-
da et al. [2]. However, there are few investigations on the
relationships between the molecular structure and molecu-
lar packing, the morphological properties in thin films and
the corresponding optoelectronic properties of these com-
plexes. In this study, we used an azomethine ligand bearing
an electron-donating substituent, salicylidene(4-dimethyla-
mino)aniline (sada), and synthesized the corresponding
azomethin-zinc complex bis[salicylidene(4-dimethylamino)-
aniline]zinc(II) (Zn(sada)2) (Fig. 1(a)). By using X-ray single
crystal diffraction, Zn(sada)2 was structurally characterized
as a racemic compound, and this complex has very good
thermal stability in both the single crystal and thin film
forms. Herein, we describe the molecular and crystal struc-
2.3. X-ray crystallography
A single crystal of Zn(sada)2 suitable for X-ray crystallog-
raphy was obtained by gradient-temperature vacuum subli-
mation. Zn(sada)2 was heated incrementally in zone 1 of a
two zone furnace from 220 to 270 ꢀC under a N2 atmosphere
of about 2 Pa for 48 h. The single crystal was collected at
245 ꢀC and had a typical size of 0.3 · 0.2 · 0.2 mm3. The
solid-state structure was further confirmed by single-crystal
X-ray diffraction analysis. Room temperature (294 1 ꢀC)
single-crystal X-ray experiments were performed on a
Bruker SMART APEX CCD diffractometer equipped
with graphite-monochromatized Mo Ka radiation (k =
˚
0.71073 A). Direct phase determination yielded the positions
of Zn, O, N and most of the C atoms, and the other C atoms
were located in successive-difference Fourier syntheses.
Hydrogen atoms were generated theoretically and rode on
their parent atoms in the final refinement. All non-hydrogen
atoms were subjected to anisotropic refinement. The final

J. Xie et al. / Inorganica Chimica Acta 358 (2005) 4451–4458
4453
Fig. 1. (a) Molecular formula of Zn(sada)2; (b) ORTEP drawing of molecule 1, with 35% probability ellipsoids, showing the atomic numbering scheme; (c)
ORTEP drawing of molecule 2, with 35% probability ellipsoids, showing the atomic numbering scheme.
full-matrix least-square refinement on F2 converged with
R1 = 0.0385 and wR2 = 0.0822 for 7715 observed reflections
[I P 2r(I)]. The final difference electron density map shows
no features. The structural solutions and refinements were
performed using the SHELXTL NT ver. 5.10 program package
(Bruker, 1997) [28].
sition temperature (Tg) and melting point (Tm) were
determined by differential scanning calorimetry (DSC) per-
formed on a TA thermal analysis instrument (DSC 2910
Modulated DSC). Cyclic voltammetry measurements were
conducted on a model CH 600 voltammetric analyzer with
a platinum plate as the working electrode, a silver wire as
the pseudo-reference electrode, a polished platinum wire
as the counter electrode, and ferrocene as an internal
[30], at a scan rate of 50 mV/s. The supporting electrolyte
was 0.1 mol/dm3 tetrabutylammonium tetrafluoroborate
in DMF. Prior to electrochemical measurements, the solu-
tion was deoxygenated with bubbling nitrogen for 15 min.
2.4. Equipment
1H NMR spectra were recorded on a Bruker ARX400
NMR spectrometer with tetramethylsilane as the internal
standard. Infrared spectra were recorded on a Nicolet
Magna-IR 750 (USA) FT-IR microscope system. Absorp-
tion spectra were recorded with a UV–Vis spectrophotom-
eter (HP8452A), and PL spectra were obtained with a
fluorospectrophotometer (HITACHI, F4500). Relative
PL quantum efficiencies (UPL) were determined from de-
gassed dimethyl formamide (DMF) solutions, and the con-
centrations of the samples were carefully adjusted so that
the optical densities at 390 nm (excitation wavelength) were
<0.1 absorption units. PL quantum yields were finally cal-
culated relative to the known value of Alq3 in DMF
(UPL = 0.116) [29] and normalized to Alq3. The glass tran-
2.5. Fabrication of EL devices
Devices were grown on glass slides precoated with in-
dium tin oxide (ITO) with 30 X/square. The substrates were
ultrasonically cleaned in detergent solution for ꢁ1 min, fol-
lowed by thorough rinsing in de-ionized water. They were
then rinsed in acetone and methanol, and then dried in pure
nitrogen gas. The organic films were deposited layer by
layer onto the ITO surface. After deposition of the organic
layers, a Mg:Ag (10:1 mass ratio) electrode was deposited

4454
J. Xie et al. / Inorganica Chimica Acta 358 (2005) 4451–4458
Table 2
onto the organic layer without breaking the vacuum. The
chamber pressure was below 5 · 10ꢀ4 Pa during the deposi-
tion of organic materials and metals. The active area of the
devices was 0.35 cm2. Devices were tested in air with ITO as
the positive electrode in the forward bias configuration
without further encapsulation. Current–voltage–luminance
(I–V–L) curves of OLEDs were measured using a Keithley
4200 semiconductor characterization system. The EL spec-
˚
Selected bond lengths (A) and angles (ꢀ)
Molecule 1
Zn(1)–O(1)
Zn(1)–O(2)
Zn(1)–N(1)
Zn(1)–N(3)
1.9164(19)
1.908(2)
2.020(2)
2.012(2)
O(1)–Zn(1)–O(2)
O(2)–Zn(1)–N(3)
O(1)–Zn(1)–N(3)
O(2)–Zn(1)–N(1)
O(1)–Zn(1)–N(1)
N(1)–Zn(1)–N(3)
115.88(9)
97.15(8)
113.87(9)
113.00(9)
97.71(8)
tra were measured with
spectrophotometer.
a Photo Research PR650
120.47(9)
3. Results and discussion
Molecule 2
Zn(1A)–O(1A)
Zn(1A)–O(2A)
Zn(1A)–N(1A)
Zn(1A)–N(3A)
1.907(2)
1.914(2)
2.007(2)
2.006(2)
3.1. Molecular structure
The coordination numbers of Zn2+ ion can vary from 2
to 6, and the change of the ligands may result in distinct
molecular structures. Recent reports involved dimeric [20],
trimeric [11] and tetrameric structures [19]. The molecular
structure of Zn(sada)2 was elucidated as a monomer by
X-ray crystallography. The crystal data and selected bond
lengths and angles are listed in Tables 1 and 2, respectively.
A perspective drawing of the structure is shown in Fig. 1.
There are two independent molecules, 1 and 2 in the
Zn(sada)2 single crystal. It was considered that this phenom-
enon comes from the restriction in solid state. As seen in
Table 2, the averages of Zn–O and Zn–N bond lengths are
O(1A)–Zn(1A)–O(2A)
O(2A)–Zn(1A)–N(3A)
O(1A)–Zn(1A)–N(3A)
O(2A)–Zn(1A)–N(1A)
O(1A)–Zn(1A)–N(1A)
N(1A)–Zn(1A)–N(3A)
109.28(9)
96.30(8)
123.28(9)
118.38(9)
96.74(9)
114.40(8)
C6–C7–N1 hexatomic ring. The coordinate bond lengths
and bond angles are similar to the mentioned azomethin-
zinc complexes [23–27]. The distance of the two aniline rings
in molecule 2 is closer than that in molecule 1, resulting in an
increase of the torsion of the C(8A)–N(1A) and C(23A)–
N(3A) bonds, i.e., the dihedral angle of the two aromatic
rings of the same ligand in molecule 2 is bigger than that
in molecule 1. In fact, the dihedral angles between phenol
and aniline groups of the same sada ligand in molecule 1
are 9.7ꢀ and 19.8ꢀ, which are almost planar, while those in
molecule 2 are 42.0ꢀ and 45.6ꢀ, respectively. The C2 symme-
try of the Zn(sada)2 molecule in the ideal state disappears in
single crystals because of the molecule stacking effect.
˚
about 1.91 and 2.01 A, respectively. The zinc ion is
four-coordinate to the Schiff base ligands and has distorted
tetrahedron geometry. In both molecules, the bond angle
(96.30–97.71ꢀ) formed by oxygen and nitrogen atoms of
the same ligand together with zinc atom (O–Zn–N) is much
less than 109.5ꢀ, because of the tension of the Zn1–O1–C1–
Table 1
Crystallographic data
Formula
C30H30N4O2Zn
Formula weight
Color
543.95
yellow prism
0.2 · 0.2 · 0.3
triclinic
Crystal size (mm3)
Crystal system
Space group
ꢀ
P1 ðNo. 2Þ
˚
a (A)
10.2791(9)
16.5008(14)
17.5984(15)
114.830(2)
96.579(2)
97.674(2)
2636.0(4)
4
˚
b (A)
˚
c (A)
a (ꢀ)
b (ꢀ)
c (ꢀ)
3
˚
V (A )
Z
Dcalc (g cmꢀ3
)
1.371
0.966
l (cmꢀ1
)
2hmax (ꢀ)
F(000)
50
1136
Reflections measured
Rint
Goodness-of-fit on F2
R1, wR2 [I P 2r(I)]a
R1, wR2 (all data)a
9216/7715
0.0148
1.023
0.0385, 0.0822
0.0468, 0.0846
P
P
P
P
2
2 1=2
a
R1 ¼ jjF oj ꢀ jF cjj= jF oj; wR2 ¼ ½ wðF 2o ꢀ F c2Þ = wðF o2Þ ꢂ
.
Fig. 2. Crystal packing view of Zn(sada)2 along the a direction.

J. Xie et al. / Inorganica Chimica Acta 358 (2005) 4451–4458
4455
3.2. Thermal stability
3.3. Absorption and photoluminescence
As seen in Fig. 2, the Zn(sada)2 crystal is a racemic
compound. Experimental evidence for the stability of this
racemic compound was obtained from DSC/TGA thermal
analysis. The complex only had a ꢁ2% weight loss at
317 ꢀC and finally decomposed at temperatures over
400 ꢀC. The melting point was obtained at the first DSC
heating of about 274 ꢀC. By rapid quenching of the melting
sample, the glass transition was observed at 112 ꢀC for the
second heating, and no recrystallization peak was found. In
addition, the DSC scans at different heating rates for
Zn(sada)2 powders were measured. The single endothermic
transition (DHfusion = 54.94 J/mol) peak observed at
274 ꢀC narrowed as the heating rate decreased from 20 to
2 ꢀC/min. There was no additional thermal transition
detected, indicating that no phase transition happened in
the heating process. This behavior is in contrast to Alq3,
which shows several enthalpic transitions assigned to differ-
ent polymorphic phases. Furthermore, there were no
detectable changes in the FT-IR spectra for powder sam-
ples obtained by vacuum train sublimation or deposited
from the acetone solution and vapor-deposited film sample
on the silicon substrate. These results indicate that the
Zn(sada)2 complex is a stable racemic compound.
The UV–Vis absorption and PL spectra of Zn(sada)2
and sada in DMF solutions are shown in Fig. 3. The
maximum absorption peak of sada is at 390 nm, and that
of Zn(sada)2 is bathochromically shifted to 412 nm. The
broad absorption band closely matches the absorption
spectrum of the protonated ligand precursor and thus
was assigned as a ligand-based p–p* transition. The maxi-
mum emission peak of sada is at 553 nm, and that of
Zn(sada)2 is at 542 nm which does not change at different
excitation wavelengths. It is obvious that the sada ligand
acts as absorber and emitter in this molecule. The rota-
tional and vibrational degrees of freedom due to the
dimethylamino substitutes could increase the probability
of non-radiative relaxation processes, resulting in low fluo-
rescence quantum efficiency of Zn(sada)2 (0.007 in DMF
solution). Compared with the solution sample, the absorp-
tion and emission of the film exhibited a small red shift of
only several nanometers. This suggests that molecules of
Zn(sada)2 in the solid film have a weak intermolecular
interaction and the bulky dimethylamino substitutes may
prevent easy aggregation.
3.4. Electrochemical properties
An atomic force microscope (AFM) was utilized to ob-
serve the surface morphology of the vapor-deposited films
on a silica substrate. The as-deposited film of Zn(sada)2
exhibited an entirely amorphous and uniform surface and
a negligible change after being stored for 5 days at room
temperature in air. This environment-stable amorphous
morphology may come from the non-planar molecular
structure caused by the dimethylamino substituents and
an increased number of conformers in this molecule which
may prevent easy packing of molecules and hence ready
crystallization [31].
Energy level is an important property for organic mate-
rials used in OLEDs. The charge injection (transport) bar-
riers, which are one of the key conditions that influence the
operating voltage and efficiency of an OLED, are mainly
due to energetic differences between the work function of
the electrodes and the HOMO or LUMO energy of organic
materials. Commonly, the corresponding energy level can
be easily obtained by cyclic voltammetry. The oxidation
and reduction potentials of Zn(sada)2 were determined in
DMF solution using ferrocene as an internal reference.
1.2
Zn(sada)2
Absorption
Emission
sada
1.0
Zn(sada)2 film
0.8
0.6
0.4
0.2
0.0
300
350
400
450
500
550
600
650
700
750
Wavelength (nm)
Fig. 3. UV–Vis absorption spectra and normalized PL spectra of Zn(sada)2 and sada.

4456
J. Xie et al. / Inorganica Chimica Acta 358 (2005) 4451–4458
There are two oxidation maxima (0.704 and 0.863 V) in the
oxidation process, because of continuous oxidation of the
two ligands. As the potential is swept back, a cation reduc-
tion peak appears at 0.643 V, which shows that the mole-
cule oxidation process is somewhat reversible. The
reduction peak of Zn(sada)2 is ꢀ1.871 V, but this process
is irreversible as observed from the cyclic voltammogram.
Thus the HOMO and LUMO energy were calculated as
ꢀ5.20 and ꢀ2.63 eV, respectively, with an Eg of 2.57 eV.
The optical band gap evaluated from the absorption spec-
trum is about 2.59 eV, which agrees with that obtained
from cyclic voltammetry.
is similar to the PL spectrum of the Zn(sada)2 film. As
we expected, inferior EL performance with a maximum
brightness of 1470 cd/m2 (at 11.8 V) and a luminous effi-
1.0
0.8
0.6
0.4
0.2
0.0
Alq3
Zn(sada)2
Alq3/Zn(sada)2
3.5. OLED studies
To investigate the emission and electron transporting
characteristics of Zn(sada)2 in OLEDs, a device using
Zn(sada)2 as the electron transporting/emissive layer with
a configuration of ITO/NPB (60 nm)/Zn(sada)2 (60 nm)/
Mg:Ag (device A) was fabricated, where NPB was used
as a hole transporting material. As shown in Fig. 4, this
device (A) emitted yellow light centered at 544 nm, which
400
500
600
700
800
Wavelength (nm)
Fig. 4. EL spectra at 10 V of the type A, B and C devices, using Zn(sada)2,
Alq3/Zn(sada)2, and Alq3 as electron-transporting layers, respectively.
Fig. 5. Luminance, current and voltage characteristics of the type A, B and C device. (a) Luminance–voltage characteristics; (b) current–voltage
characteristic; (c) luminance efficiency–current density characteristics.

J. Xie et al. / Inorganica Chimica Acta 358 (2005) 4451–4458
4457
4. Summary
In conclusion, we successfully designed and synthesized
an azomethin-zinc complex, Zn(sada)2, which was structur-
ally characterized by single-crystal X-ray crystallography.
Further thermal analyses proved it has good thermal stabil-
ity. The study of the vapor-deposited thin film gave strong
evidence of Zn(sada)2 possessing excellent film-forming
capability. By using Zn(sada)2 as an electron transporting
layer, three types of organic light emitting devices were fab-
ricated. Compared with the typical bilayer device of NPB/
Alq3, the device with Zn(sada)2 showed much lower turn-
on voltage of 2.5 V (compared to 3.5 V for NPB/Alq3 de-
vices), which indicates that Zn(sada)2 may serve as a good
electron transporting material in OLEDs.
By changing the substituents on the ligand, we have also
recently synthesized a series of azomethin-zinc complexes.
The preliminary studies demonstrated that the substituents
on the ligand can also finely tune the luminescent proper-
ties of the corresponding metal complexes. A detailed study
is in process.
Fig. 6. Energy level of the OLED of ITO/NPB/Alq3/Zn(sada)2/Mg:Ag/
Ag.
ciency of 0.5 cd/A (at 20 mA/cm2) was obtained. The lower
EL efficiency is consistent with the significantly lower PL
efficiency in solution, and may be even lower in the solid
state.
Because of the poor luminous property of Zn(sada)2, de-
vice A could not indicate anything about the electron trans-
porting characteristics for this complex. Two additional
types of devices were designed and fabricated: ITO/NPB
(60 nm)/Alq3 (60 nm)/Mg:Ag (device B), and ITO/NPB
(60 nm)/Alq3 (30 nm)/Zn(sada)2 (30 nm)/Mg:Ag (device
C). The EL spectra of these devices are shown in Fig. 4,
and are almost identical at different operating voltages
(from 4 to 12 V). Devices B and C both emitted strong
green light centered at 520 nm, originating from the intrin-
sic emission of Alq3, with higher luminescent efficiencies of
2.5 and 3.0 cd/A (at 20 mA/cm2), respectively (Fig. 5(c)).
As seen in Fig. 5(b), the current of device C with Zn(sada)2
as the electron transport layer is much higher than that of
device B with Alq3 at the same voltage. Furthermore, the
turn-on voltages of devices B and C were about 3.5 and
2.5 V, respectively. (Herein, the turn-on voltage is defined
as the voltage required to give a brightness of 1 cd/m2.)
It is well-known that charge injection and transport are
the limiting factors in determining current density and
operating voltage. In both devices, Alq3 served as an emit-
ter. The only difference between devices B and C is the elec-
tron transporting layer: Alq3 for B and Zn(sada)2 for C.
From the energy level diagram (Fig. 6) it is clear that
Zn(sada)2 has the higher electron injection barrier with re-
gard to Alq3. Thus, the possibility of enhanced injection
efficiency for device C could be ruled out, and the decrease
of the turn-on voltage in device C could be ascribed to the
better electron transporting property of the Zn complex.
Actually, the introduction of the electron-donating dimeth-
ylamino substituents led to a relatively high HOMO energy
level of Zn(sada)2, which may not have been enough to
block the holes within the Alq3 emitting layer. Thus, the
redundant current coming from the holes flowing into the
electron transporting layer would probably result in a low-
er QE. Above all, the results indicate that Zn(sada)2 might
be a good electron transporting material for OLEDs. Fur-
ther optimization of the device structure is underway.
5. Supplementary material
Crystallographic data for the structure of Zn(sada)2 re-
ported in this paper has been deposited with the Cambridge
Crystallographic Data Center as Supplementary Publica-
tion No. CCDC-252862. Copies of the data can be ob-
tained free of charge on application to CCDC, 12 Union
Road, Cambridge CB21EZ, UK (fax: +44 1223 336 033;
e-mail: deposit@ccdc.cam.ac.uk).
Acknowledgments
This work was financially supported by the National
Natural Science Foundation of China (Nos. 50403001
and 50325310). The authors thank associate Professor Ruji
Wang for the analysis of the crystal structure.
References
[1] C.W. Tang, S.A. VanSlyke, Appl. Phys. Lett. 51 (1987) 913.
[2] Y. Hamada, T. Sano, M. Fujita, T. Fujii, Y. Nishio, K. Shibata, Jpn.
J. Appl. Phys. 32 (1993) L514.
[3] Y. Hamada, T. Sano, M. Fujita, T. Fujii, Y. Nishio, K. Shibata, Jpn.
J. Appl. Phys. 32 (1993) L511.
[4] T.A. Hopkins, K. Meerholz, S. Shaheen, M.L. Anderson, A. Schmidt,
B. Kippelen, A.B. Padias, H.K. Hall Jr., N. Peyghambarian, N.R.
Armstrong, Chem. Mater. 8 (1996) 344.
[5] M. Ghedini, M. La Deda, I. Aiello, A. Grisolia, Inorg. Chim. Acta
357 (2004) 33.
[6] G. Yu, Y. Liu, Y. Song, X. Wu, D. Zhu, Synth. Met. 117 (2001) 211.
[7] P. Wang, Z. Hong, Z. Xie, S. Tong, O. Wong, C.S. Lee, N. Wong,
L. Hung, S. Lee, Chem. Commun. (2003) 1664.
[8] N. Nakamura, S. Wakabayashi, K. Miyairi, T. Fujii, Chem. Lett.
(1994) 1741.
[9] Y. Hamada, T. Sano, H. Fujii, Y. Nishio, H. Takahashi, K. Shibata,
Jpn. J. Appl. Phys. 35 (1996) L1339.
[10] H. Tanaka, S. Tokito, Y. Taga, A. Okada, J. Mater. Chem. 8 (1998)
1999.

4458
J. Xie et al. / Inorganica Chimica Acta 358 (2005) 4451–4458
[11] T.S. Kim, T. Okubo, T. Mitani, Chem. Mater. 15 (2003) 4949.
[12] K.Y. Ho, W.Y. Yu, K.K. Cheung, C.M. Che, J. Chem. Soc., Dalton
Trans. (1999) 1581.
[13] W. Yang, H. Schmider, Q. Wu, Y.S. Zhang, S. Wang, Inorg. Chem.
39 (2000) 2397.
[14] S. Wang, Coord. Chem. Rev. 215 (2001) 79.
[15] C.J. Walsh, T. Sooksimuang, B.K. Mandal, J. Porphyr. Phthalocyan.
5 (2001) 803.
[16] C.F. Lee, K.F. Chin, S.M. Peng, C.M. Che, J. Chem. Soc., Dalton
Trans. (1993) 467.
[21] Y. Shao, Y. Qiu, W.H. Hu, X.Y. Hong, Adv. Mater. Opt. Electron.
10 (2000) 285.
[22] J. Qiao, Y. Qiu, L.D. Wang, L. Duan, Y. Li, D.Q. Zhang, Appl.
Phys. Lett. 81 (2002) 4913.
[23] M. Dreher, H. Elias, Z. Naturforsch. 42b (1987) 707.
[24] T. Sogo, J. Romero, A. Sousa, A. de Blas, M.L. Duran, Z.
Naturforsch. 43b (1988) 611.
[25] F.A. Bottino, P. Finocchiaro, E. Libertini, G. Mattern, Zeitschrift
fur Kristallographie 187 (1989) 71.
[26] X.L. Liu, H.L. Chen, F.M. Miao, Youji Huaxue 12 (1992) 522.
[27] F.A. Bottino, P. Finocchiaro, Polyhedron 4 (1985) 1507.
[28] G.M. Sheldrick, SHELXTL, Version 5.1, Bruker Analytical X-ray
System, Inc., Madison, WI, 1997.
[17] Y. Ma, T. Lai, Y. Wu, Adv. Mater. 12 (2000) 433.
[18] Y. Kai, M. Morita, N. Yasuoka, N. Kasai, Bull. Chem. Soc. Jpn. 58
(1985) 1631.
[19] L.S. Sapochak, F.E. Benincasa, R.S. Schofield, J.L. Baker, K.K.C.
Riccio, D. Fogarty, H. Kohlmann, K.F. Ferris, P.E. Burrows, J.
Am. Chem. Soc. 124 (2002) 6119.
[20] G. Yu, S. Yin, Y. Liu, Z. Shuai, D. Zhu, J. Am. Chem. Soc. 125
(2003) 14816.
[29] F.E. Lytle, D.R. Storey, M.E. Juricich, Spectrochim. Acta, Part A 29
(1973) 1357.
[30] R.R. Gange, C.A. Koval, G.C. Lisensky, Inorg. Chem. 19 (1980)
2854.
[31] P. Strohriegl, J.V. Grazulevicius, Adv. Mater. 14 (2002) 1439.