Investigating the effect of steric crowding in phosphorescent dendrimers

Article abstract of DOI:10.1021/ma050609q

A simple convergent procedure has been developed for the formation of sterically encumbered phosphorescent dendrimers. The procedure is demonstrated with the preparation of a first-generation dendrimer composed of a fac-tris(2-phenylpyridyl)iridium(III) c

Full text of DOI:10.1021/ma050609q

9564
Macromolecules 2005, 38, 9564-9570
Investigating the Effect of Steric Crowding in Phosphorescent
Dendrimers
Neil Cumpstey,† Raghu N. Bera,‡ Paul L. Burn,*,† and Ifor D. W. Samuel*,‡
Department of Chemistry, University of Oxford, Chemistry Research Laboratory, Mansfield Rd,
Oxford, OX1 3TA, UK, and Organic Semiconductor Centre, SUPA, School of Physics and Astronomy,
University of St. Andrews, North Haugh, St. Andrews, Fife, KY16 9SS, UK
Received March 23, 2005; Revised Manuscript Received May 5, 2005
ABSTRACT: A simple convergent procedure has been developed for the formation of sterically encumbered
phosphorescent dendrimers. The procedure is demonstrated with the preparation of a first-generation
dendrimer composed of a fac-tris(2-phenylpyridyl)iridium(III) core and three dendrons. Each dendron is
comprised of a branching phenyl unit with a further four phenyl groups attached. The lack of surface
groups on the dendrons was found to reduce solubility and also reduced the level of control over the
intermolecular interactions of the emissive and electroactive core in films. The 6-fold decrease in
photoluminescence quantum yield in going from solution (69%) to the solid state (11%) showed that there
were strong intermolecular interactions of the emissive cores in the solid state. Single-layer devices with
the dendrimer blended with 4,4′-bis(N-carbazolyl)biphenyl showed an external quantum efficiency of 1.7%
(5.4 cd/A) at 100 cd/m2 and 11.4 V, giving a power efficiency of 1.5 lm/W.
Introduction
dendrons attached to the core or the dendrimer genera-
tion. By judicious positioning of the dendrons this can
be done without changing the emissive properties of the
core chromophore.9,10 Recent reports have shown that
the dendritic approach can be used to develop solution
processable phosphorescent materials that give rise to
highly efficient green and red DLEDs.10,11 The green
emissive dendrimers developed had 2-ethylhexyl surface
groups, phenyl-based dendrons, and fac-tris(2-phenyl-
pyridyl)iridium(III) cores.10,12 It was found that with
these dendrimers increasing the dendrimer generation
decreased the charge mobility13 through the film but
improved the solid-state photoluminescence quantum
yield with a net improvement in DLED performance.14
However, the degree of branching in these dendrimers
close to the emissive chromophore is relatively low, and
it was therefore of interest to see whether the same level
of control over intermolecular interactions can be gained
by increasing the degree of branching but at lower
generation. In our work we have found that the rigidity
imparted by conjugated dendrons gives better control
over the intermolecular interactions and makes iden-
tifying the structure-property relationships more
straightforward. The most highly branched conjugated
dendrons that are easily accessed are those based on
biphenyl units, which have been developed by Mu¨llen
and co-workers.15 Given that the first branching point
in a “Mu¨llen” dendron is a phenyl unit as in the earlier
phosphorescent dendrimers,10,11 they are the natural
choice of dendron for comparing the effect of the level
of branching on controlling the intermolecular inter-
actions of fac-tris(2-phenylpyridyl)iridum(III) cored den-
drimers.
Intermolecular interactions play a key role in organic
light-emitting diodes (OLEDs). While strong inter-
molecular interactions between the electroactive chro-
mophores are good for charge transport, they can be
detrimental to light emission. Close interactions of
emissive chromophores can lead to emission from exci-
mers or aggregates leading to a change in emission color
and generally lower photoluminescence quantum yields
(PLQYs). The tradeoff between these two properties is
clearly seen with small phosphorescent molecules when
used in OLEDs. Reports on OLEDs based on phospho-
rescent small molecules have generally shown that
devices are more efficient when the phosphorescent
molecule is a guest in a host matrix.1,2 At low concen-
tration of phosphorescent molecules, there are few
intermolecular interactions of the emissive species and
the device efficiency is high. However, in blended
systems, whether evaporated or solution processed,
there is always the issue of how evenly distributed the
guest is in the host as localized high concentrations can
lead to poor device performance. The most efficient
method of controlling molecular interactions is by build-
ing in the control at the molecular level. This is difficult
to do with small molecules, and with polymeric materi-
als understanding the intermolecular interactions is
complicated by the often complex polymer morphol-
ogy.3,4 Dendrimer light-emitting diodes (DLEDs) are
developing into an exciting class of OLEDs.5-10 Light-
emitting dendrimers have mainly been designed with
a light-emitting chromophore at the core to which one
or more dendrons are attached. Surface groups are
generally placed at the distal ends of the dendrons to
provide the required processing properties. One of the
key features of light-emitting dendrimers is that by
having the emissive and electroactive chromophore at
the core the intermolecular interactions can be con-
trolled at the molecular level by either the number of
In this paper we describe the synthesis of a first-
generation dendrimer containing a Mu¨llen dendron
attached to each ligand of the fac-tris(2-phenylpyridyl)-
iridium(III) core (6, Scheme 1). In 6 the branching
phenyl has four phenyl groups attached, and there are
no surface groups at the distal ends of the dendrons. In
addition, we compare the physical, optoelectronic, and
device properties of 6 with first-generation dendrimers
† University of Oxford.
‡ University of St. Andrews.
10.1021/ma050609q CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/19/2005

Macromolecules, Vol. 38, No. 23, 2005
Steric Crowding in Phosphorescent Dendrimers 9565
Scheme 1a
a
Key: (A) CuI, Et3N, TMS-acetylene, tetrakis(triphenylphosphine)palladium(0), rt, 40 h, Ar; (B) TBAF, tetrahydrofuran, rt,
3 h, Ar; (C) diphenyl ether, 210 °C, 3 h, Ar; (D) IrCl3‚3H2O, water, 2-ethoxyethanol, ∆, 16 h, Ar then 5, silver trifluoromethyl-
sulfonate, diglyme, 130 °C, 2 d, Ar.
7 and 8 (Figure 1). The latter two dendrimers have the
same core and branching phenyl unit as in 6, but there
are only two phenyl groups attached to the branching
phenyl ring. 7 and 8 also have 2-ethylhexyloxy surface
groups attached to the distal ends of the dendrons.14
drimer 6 in a 63% yield for the two steps. Gel perme-
ation chromatography (GPC) showed that the dendrimer
was monodisperse, and the facial nature of the den-
1
drimer was confirmed by H NMR by comparison with
spectra reported for other dendrimers with fac-tris(2-
phenylpyridyl)iridium(III) cores.14
Results and Discussion
Physical Properties. The main comparison of the
physical, optoelectronic, and device properties is be-
tween dendrimers 6 and 7 as they have the dendrons
attached at the same point of the ligand and hence are
structurally the most similar. The first difference in the
physical properties between 6 and 7 was that 7 had
much greater solubility in polar aprotic solvents. While
7 could be spin-coated to form good quality thin films
either neat or as a blend with a host such as 4,4′-bis-
(N-carbazolyl)biphenyl (CBP), neat films of 6 tended to
be much thinner and poorer quality due to the low
solution concentrations associated with its much poorer
solubility. This clearly demonstrates the importance of
surface groups for solution processing dendrimers. The
hydrodynamic radius of 6 was determined by GPC from
the Mh v in combination with Hester-Mitchell equation
and the Mark-Houwink relationship.18 The Mh v of 6 was
954, and this corresponds to a hydrodynamic radius of
6.2 Å. The hydrodynamic radius of 6 is ≈25% smaller
than the value of 8.6 Å for 7 with the difference being
mainly due to the large bulky 2-ethylhexyloxy groups
on 7.14 Nevertheless, the degree of branching in 6 is
higher closer to the core, and it might be considered that
the extra steric encumbrance near the core could still
provide good control over the intermolecular interac-
tions. Thermal gravimetric analysis of 6 was carried out
at a scan rate of 10 °C/min under nitrogen, and the
Synthesis. The strategy for the convergent synthesis
of the highly branched first-generation dendrimer is
shown in Scheme 1. 2-(4-Iodophenyl)pyridine (1) was
prepared in a 40% yield by reaction of the monolithiated
1,4-diiodobenzene with 2-fluoropyridine. Trimethylsilyl-
acetylene was then coupled with 1 under Sonogashira
conditions to afford 2-(4-trimethylsilylethynylphenyl)-
pyridine (2) in an 89% yield. The acetylene was easily
deprotected with either tetra-n-butylammonium fluoride
in tetrahydrofuran or by treatment with aqueous potas-
sium hydroxide in a methanol/dichloromethane mixture.
Both methods gave 2-(4-acetylenylphenyl)pyridine (3)
in an isolated yield of 80%. The dendronized ligand 5
was then formed by reaction of 3 with 2,3,4,5-tetraphen-
ylcyclopentadienone (4).16 The reaction was heated at
210 °C for 2-3 h using diphenyl ether as the solvent,
after which 5 was isolated in a 93% yield. The complex-
ation of 5 to form the fac-tris(2-phenylpyridyl)iridium-
(III)-cored dendrimer utilized the standard two-step
procedure.17 In the first step 2.5 equiv of 5 was reacted
with iridium trichloride trihydrate in aqueous 2-ethoxy-
ethanol heated at reflux to give a mixture of the bis-
(iridium), bis(chloro) dimer and unreacted ligand. This
mixture was then reacted with an excess of 5 in the
presence of silver trifluoromethylsulfonate in diglyme
at 130 °C to give, after purification, the desired den-

9566 Cumpstey et al.
Macromolecules, Vol. 38, No. 23, 2005
Figure 1. Structures of 7 and 8.
Table 1. Summary of the Thermal, Photophysical, and Electrochemical Properties of 6-8
dendrimer
T5% (°C)a
solutionb PLQY (%)
film PLQY (%)
E1/2 (V) (oxidation)
E1/2 (V) (reductions)
-2.83, -3.01, -3.18
6
7
8
450
69
11
0.21
40014
40014
7414
7014
2414
229
0.2714
0.2414
-
c,14
-2.91, -3.10, -3.2814
a The temperature at which 5% weight loss was observed. b Toluene. c Chemically reversible reductions were not observed.
thermal properties of the dendrimers are summarized
in Table 1. For 6 there was no appreciable decomposi-
tion below 400 °C. At 450 °C there was only a 5% weight
loss observed with significant decomposition occurring
above this temperature. This compares favorably with
the thermal properties of 7 and 8, which both had no
significant weight loss below 300 °C and only a 5% loss
by 400 °C.14 While 7 and 8 had glass transition tem-
peratures of 138 and 132 °C, respectively,14 6 did not
have an observable glass transition temperature.
Photophysical Properties. The addition of substit-
uents to the ligand of fac-tris(2-phenylpyridyl)iridium-

Macromolecules, Vol. 38, No. 23, 2005
Steric Crowding in Phosphorescent Dendrimers 9567
Figure 3. Solution and film photoluminescence spectra of 6
and 7. The solutions (in tetrahydrofuran) were excited at 360
nm and the films at 325 nm. The spectra have been normalized
for clarity.
Figure 2. Solution UV-vis absorption spectra of 6 and 7 in
tetrahydrofuran.
(III) complexes can change the optical properties of the
complex.19,20 For example, if the dendrons on 7 are
moved from being para to the pyridyl ring to meta as
in 8 (Figure 1), then the peaks and shoulders in the
UV-vis absorption spectrum are blue-shifted by about
20 nm.14 The 20 nm blue shift in the absorption
spectrum of 8 is also reflected in the photoluminescence
spectra. The reason for the difference in the absorption
and PL spectra between 8 and 7 is that in 7 the first
phenyl ring of the dendron is in conjugation with the
pyridyl ring, and hence the ligand, which is responsible
for the metal-to-ligand charge-transfer (MLCT) transi-
tions with the iridium(III), has a smaller HOMO-
LUMO energy gap. The solution UV-vis spectra of 6
and 7 are shown in Figure 2. The weaker long wave-
length absorptions of both 6 and 7 are due to the MLCT
states, whereas the stronger shorter wavelength ab-
sorptions between 260 and 350 nm are due to the ligand
π-π* transitions and the dendron absorptions.14 How-
ever, the key point from the absorption spectra is that
the wavelengths of the absorptions due to the MLCT
transitions of 6 and 7 are the same. While the extinction
coefficients of the absorptions due to the MLCT transi-
tions might be expected to be similar, dendrimer 6 has
more than two dendron phenyl rings that can be in
conjugation with the 2-phenylpyridyl ligands whereas
7 only has one. This should mean that the ligand
chromophore has a longer conjugation length and hence
lower HOMO-LUMO energy gap, which in turn would
lead to red-shifted MLCT absorptions. The reason why
the absorptions of the MLCT states of 6 are not shifted
to longer wavelengths is due to the ortho arrangement
of the “biphenyl” bonds, which causes the phenyl rings
to be twisted out of plane, reducing the π-orbital overlap
and delocalization pathway.
the solid state while for 7 the red shift was only around
4 nm, indicating that the attachment point of the
dendrons, whether meta or para to the pyridyl ring, was
important in controlling intermolecular interactions of
the cores.14 The fact that there is no change in the PL
spectrum for 6 in going from solution to the solid state
suggests that intermolecular interactions that lead to
quenching of the luminescence might be controlled. We
probed the intermolecular interactions of the emissive
cores by measuring the PL quantum yields (PLQYs)
with the results summarized in Table 1. The PLQY of
a material is particularly sensitive to interactions of the
emissive species, and if there are strong interactions of
the emissive chromophores in the solid state, then the
PLQY is generally observed to drop. The PLQY of 6, 7,
and 8 in solution were 69%, 70%,14 and 69%,14 respec-
tively, with the similarity showing that the attachment
of the Mu¨llen-type dendrons to the fac-tris(phenylpy-
ridyl)iridium(III) core does not have a detrimental effect
on the efficiency of emission. However, there is a
dramatic difference in the PLQYs when moving to the
solid state. For 7 and 8 the PLQY in the solid state
dropped to around 22-24%,14 while for 6 with the
Mu¨llen dendrons the PLQY was half this at 11%. This
indicates that the surface groups in 7 and 8 play a very
important role in controlling the intermolecular interac-
tions in addition to the branching in the dendrons.
Interestingly, when the dendrimers are blended with
CBP, there is also a distinct difference between 6 and
the other two materials. For 7 and 8 as 20 wt % blends
in CBP the PLQYs, at 68% and 78%, were as high as
the solution PLQYs. This indicates that the dendrimers
were evenly distributed throughout the blended film.
In contrast, the PLQY of the blended 20 wt % film of 6
in CBP had a PLQY of only 49%, which is significantly
less than the solution PLQY. The lower PLQY in the
blend compared to the solution measurement indicates
that dendrimer 6 is not evenly distributed within the
film. It should be noted that at the concentration of 6
in solution required to give neat films of thickness of
order 50-60 nm the solutions were not clear and the
films formed were hazy. In contrast, the blended films
The solution and film photoluminescence (PL) spectra
are shown in Figure 3. In solution the PL peak maxima
are the same for 6 and 7, consistent with the absorption
spectra. In moving from solution to the solid state there
was no change in the PL spectrum for 6. This is in
contrast to the first-generation dendrimers 7 and 8
where a difference in the PL spectrum is seen in moving
from solution to the solid state. For 8 there was a red
shift of the emission peak of about 15 nm in moving to

9568 Cumpstey et al.
Macromolecules, Vol. 38, No. 23, 2005
were mostly clear. The PLQY results further illustrate
the important role the surface groups of the dendrimers
play in controlling film morphology and quality. A final
aspect of the photophysical study was to determine
whether energy transfer from the dendrons to the core
of 6 was efficient. No emission was observed from the
dendron, and the photoluminescence excitation spec-
trum followed the absorption spectrum, showing that
in common with other first-generation dendrimers
energy transfer was efficient.14,21
Electrochemical and Device Properties. To un-
derstand where charge would be injected into dendrimer
6 in a device, we studied its electronic properties using
electrochemistry with the results summarized in Table
1. The oxidation cyclic voltammagrams of 6 and 7 and
reduction cyclic voltammagram of 6 are shown in Figure
4. Both 6 and 7 show a single chemically reversible
oxidation corresponding to the oxidation of the core. The
oxidations occur at 0.21 and 0.27 V for 6 and 7,
respectively. For 6 three chemically reversible reduc-
tions were seen at -2.83, -3.01, and -3.18 V. The
oxidation of 6 occurred at a less positive potential and
reductions at slightly less negative potentials than 8
(Table 1), which is consistent with 8 giving green
emission and 6 green-yellow emitted light.14
Finally, simple single-layer devices with the den-
drimer containing film directly deposited onto the
indium-tin oxide (ITO) anode were prepared [ITO/20 wt
% of 6:CBP/Ca/Al], and the device characteristics are
shown in Figure 5. Devices with neat films of 6 could
not be prepared as the films were too thin due to its
low solubility in the processing solvent. The single-layer
devices showed reasonable performance. The turn-on
voltage (1 cd/m2) was 7.6 V, and the maximum external
quantum was 2.9% (9.3 cd/A), with a power efficiency
of 1.9 lm/W, at a brightness of 1545 cd/m2 and 15.6 V.
The fact that the highest efficiency occurs at a high drive
voltage suggests that there is an imbalance of charge
in the device. Previous reports on devices utilizing 7 and
8 in CBP as the emissive layers have shown electron
injection and transport is one of the factors limiting
device performance. The single-layer device performance
observed from the blend of 6 with CBP was similar to
that of 8 in the same device configuration. For the
devices containing 6 100 cd/m2 was achieved at 11.4 V
with an external quantum efficiency of 1.7% (5.4 cd/A),
while for devices with 8 100 cd/m2 was observed at 8.8
V with an external quantum efficiency of 3.3% (11.3 cd/
A).10
Conclusion. We have developed a simple synthetic
pathway to phosphorescent dendrimers comprised of
fac-tris(2-phenylpyridy)iridium(III) cores and highly
branched “Mu¨llen” dendrons. Although the first-genera-
tion dendrimer has a high level of branching close to
the emissive core, it was not sufficient to control the
intermolecular interactions in the solid state. By com-
paring it with dendrimers (7 and 8) with the same core
and branching unit, fewer branches, and surface groups,
it was found that surface groups play an important role
in controlling both solubility and intermolecular inter-
actions in conjugated dendrimers.
Figure 4. (a) Oxidation cyclic voltammograms of 6 and 7.
Scan rate ) 30 and 40 mV/s, respectively, solvent ) tetrahy-
drofuran for 6 and dichloromethane for 7, concentration ) 0.05
mM for 6 and 1 mM for 7, glassy carbon and platinum working
electrodes for 6 and 7, respectively. (b) Reduction cyclic
voltammogram of 6. Scan rate ) 30 mV/s, solvent ) tetrahy-
drofuran, concentration ) 0.05 mM, glassy carbon working
electrode.
trometers and were recorded as solutions in spectroscopic
grade dichloromethane or dry tetrahydrofuran. Infrared spec-
tra were recorded on a Perkin-Elmer Spectrum 1000 FT-IR
spectrometer. Mass spectra were recorded on a Fisons Platform
for ESI, a Micromass TofSpec 2E for matrix-assisted laser
desorption/ionization-time-of-flight (MALDI-TOF) from dithra-
nol (DITH) in reflectron mode, or a VG Platform for APCI.
Melting points were recorded on a Gallenkamp melting point
apparatus and are uncorrected. Microanalyses were carried
out in the Inorganic Chemistry Laboratory, Oxford, UK. Gel
permeation chromatography was carried out using PLgel
Mixed-A columns (600 mm + 300 mm lengths, 7.5 mm
diameter) from Polymer Laboratories calibrated with polysty-
rene narrow standards (Mh p ) 580-3.2 × 106) in tetrahydro-
furan with toluene as flow marker. The tetrahydrofuran was
degassed with helium and pumped with a rate of 1 mL/min at
30.0 °C. Thermal gravimetric analysis was performed on a
Experimental Section
Measurements. NMR spectra were recorded on Bruker
DPX 400 and AV 400 MHz spectrometers: sp ) surface
phenyl; bp ) branch phenyl; L PhH ) ligand phenyl H; PyH
) pyridyl H. All J values are in hertz. UV-vis spectra were
recorded on a Perkin-Elmer UV-vis Lambda 14 or 25 spec-

Macromolecules, Vol. 38, No. 23, 2005
Steric Crowding in Phosphorescent Dendrimers 9569
Solution photoluminescence quantum yields (PLQYs) were
measured by a relative method using quinine sulfate in 0.5 M
sulfuric acid which has a photoluminescence quantum yield
of 0.546 as the standard.23 The dendrimers were dissolved in
toluene and freeze-thaw degassed. Photoluminescence spectra
were recorded in a JY Horiba Fluoromax 2 fluorimeter, with
the dendrimer solutions excited at 360 nm. The optical
densities of the standard and sample were similar and small
(less than/equal to 0.1). The accuracy of these measurements
is estimated to be (10%.
Neat films of 6 were spin-coated from chloroform solutions
with a dendrimer concentration of 10 mg/mL at 800-1200 rpm
for 1 min to give a thickness of about 50-60 nm. Blend films
were spin-coated from solutions of total solute concentration
of 20 mg/mL at 800 rpm for 1 min to give films of thickness
≈150 nm. The PLQY of the films was measured using an
integrating sphere in accordance with Greenham et al.24 using
a helium cadmium laser (Kimmon) as the excitation source.
The excitation power was 0.2 mW at 325 nm, and the sphere
was purged with nitrogen.
DLEDs were fabricated by spin-coating a chloroform solu-
tion of dendrimer and CBP (20:80 weight ratio) onto cleaned
ITO-coated glass substrates. The ITO coated glass substrates
were cleaned by ultrasound in acetone and propan-2-ol,
followed by treatment with an oxygen plasma. The concentra-
tion of the dendrimer/CBP solution was around 20 mg/mL, and
films were spin-coated for 1 min at a speed of 1500 rpm. The
resulting thickness of the organic layer was ≈180 nm. The
devices were then completed by deposition at a pressure of
10-6 mbar of a 30 nm layer of calcium followed by a 100 nm
layer of aluminum. Current-voltage-brightness measure-
ments were performed using a Keithley 2400 source meter and
Keithley 2000 multimeter with a calibrated Si photodiode
(303-674 of RS) to collect the forward emission, and efficien-
cies were deduced assuming Lambertian emission from the
device.25 EL spectra were recorded with a charge-coupled
device (CCD) spectrograph.
2-(4-Iodophenyl)pyridine (1). A solution of 1,4-diiodo-
benzene (46.8 g, 142 mmol) in dry ether (450 mL) under argon
was cooled in a salt-ice bath. n-Butyllithium (2.0 M in
pentane, 72 mL, 142 mmol) was added, and the solution was
stirred for 10 min. 2-Fluoropyridine (13 mL, 150 mmol) was
added, and the reaction was allowed to warm to room tem-
perature over 2 h. Water (500 mL) was added, and the organic
layer was separated. The aqueous layer was extracted with
ether (3 × 200 mL). The combined organic layers were washed
with brine (1000 mL), dried over anhydrous magnesium
sulfate, and filtered, and then the solvent was completely
removed. The residue was purified in two steps: first by
column chromatography over silica using a dichloromethane/
light petroleum mixture (1:1) followed by recrystallization of
the main fraction from a dichloromethane/light petroleum
mixture to give 1 (17.0 g, 40%); mp 86-87 °C. Anal. Calcd for
C11H8NI: C, 47.0; H, 2.9; N, 5.0. Found: C, 47.15; H, 2.8; N,
5.0. λmax(CH2Cl2)/nm 262 (log ꢀ/dm3 mol-1cm-1 4.24), 284 (4.31).
δH (400.2 MHz; CDCl3): 7.27 (1 H, m, PyH), 7.70-7.85 (6 H,
m, PyH and PhH), 8.70 (1 H, d, J 5 Hz, PyH). δC (100.6 MHz;
CDCl3): 95.3, 120.3, 122.5, 128.6, 136.9, 137.8, 138.8, 149.7,
156.3. m/z (APCI) 281.7 (MH+, 100%).
2-(4-Trimethylsilylethynylphenyl)pyridine (2). Tri-
methylsilylacetylene (4.1 mL, 28.8 mmol) and tetrakis(tri-
phenylphosphine)palladium(0) (832 mg, 0.72 mmol) were
added to a suspension of 1 (4.06 g, 14.4 mmol) and copper(I)
iodide (274 mg, 1.44 mmol) in triethylamine (120 mL) that had
been deoxygenated with argon. The mixture was deoxygenated
with argon and then stirred under argon at room temperature
for 40 h. Aqueous hydrochloric acid (3 M, 320 mL) and dichloro-
methane (250 mL) were added. The organic layer was sepa-
rated, and the aqueous layer was extracted with dichloro-
methane (3 × 50 mL). The combined organic layers were
washed with aqueous hydrochloric acid (3 M, 30 mL), water
(250 mL), and brine (250 mL), dried over anhydrous magne-
sium sulfate, and filtered, and then the solvent was completely
removed. The residue was purified by column chromatography
over silica using a dichloromethane/light petroleum mixture
Figure 5. Device characteristics of a single-layer device (ITO/
6:CBP 20 wt %/Ca/Al): (a) current-voltage and luminance-
voltage; (b) external quantum efficiency-voltage and power
efficiency-voltage.
Perkin-Elmer thermogravimetric analyzer TGA 7. Glass tran-
sition temperatures were measured on a Perkin-Elmer dif-
ferential scanning calorimeter Pyris 1. Light petroleum refers
to the fraction of boiling point 40-60 °C, unless otherwise
stated. When solvent mixtures are used for chromatography
over silica, the proportions are given by volume.
Electrochemistry was performed using an EG&G Princeton
Applied Research potentiostat/galvanostat model 263A. All
measurements were made at room temperature on samples
dissolved in dichloromethane or tetrahydrofuran, with 0.1 M
tetra-n-butylammonium tetrafluoroborate as the electrolyte.
The tetrahydrofuran was passed through a column of alumina
and then distilled from lithium aluminum hydride under
argon. HPLC grade dichloromethane or tetrahydrofuran were
used for the oxidation studies. The electrolyte was purified by
recrystallization from ethyl acetate. The solutions were deoxy-
genated with argon. A glassy carbon working electrode,
platinum wire counter electrode, and Ag/0.1 M AgNO3 in
acetonitrile reference electrode was used. The ferricenium/
ferrocene couple was used as standard,22 and the ferrocene was
purified by sublimation. All potentials are quoted relative to
the ferricenium/ferrocene couple. In all cases several scans
were carried out to confirm the chemical reversibility of the
redox processes.

9570 Cumpstey et al.
Macromolecules, Vol. 38, No. 23, 2005
(1:4) as the eluent. The main band was isolated and the solvent
completely removed to give 2 (2.59 g, 72%); mp 56-58 °C. Anal.
Calcd for C16H17NSi: C, 76.4; H, 6.8; N, 5.6. Found: C, 76.5;
H, 6.8; N, 5.6. λmax(CH2Cl2)/nm 272 sh, (log ꢀ/dm3 mol-1 cm-1
4.67) and 294 (4.90). δH (400.2 MHz; CDCl3): 0.27 (9 H, s,
SiMe3), 7.25 (1 H, m, PyH), 7.58 and 7.96 (4 H, AA′BB′, PhH),
7.73 (2 H, m, PyH), and 8.70 (1 H, m, PyH). δC (100.6 MHz;
CDCl3): -0.1, 95.5, 104.9, 120.6, 122.4, 123.7, 126.6, 132.3,
136.8, 139.2, 149.7, and 156.5. m/z (ESI+) 252 (MH+, 100%).
2-(4-Ethynylphenyl)pyridine (3). Tetrabutylammonium
fluoride (1 M in tetrahydrofuran, 36 mL, 36 mmol) was added
to a solution of 2 (2.26 g, 8.97 mmol) in tetrahydrofuran (50
mL), and the solution was stirred at room temperature under
argon for 3 h. Dichloromethane (100 mL) and water (150 mL)
were added. The organic layer was separated, and the aqueous
layer was extracted with dichloromethane (3 × 15 mL). The
combined organic layers were washed with brine (200 mL),
dried over anhydrous magnesium sulfate, and filtered, and
then the solvent was completely removed. The residue was
purified by passing through a silica plug using dichlo-
romethane as eluent to give 3 (1.44 g, 90%); mp 53 °C. Anal.
Calcd for C13H9N: C, 87.1; H, 5.1; N, 7.8. Found: C, 86.7; H,
5.5; N, 7.7. νmax(Nujol)/cm-1 2098 (CtC). λmax(CH2Cl2)/nm 306
sh, (log ꢀ/dm3 mol-1 cm-1 4.05) and 289 (4.48), 257 (4.48); δH-
(400.2 MHz; CDCl3): 3.17 (1 H, s, CtC-H), 7.26 (1 H, m,
PyH), 7.61 and 7.98 (4 H, AA′BB′, PhH), 7.72-7.80 (2 H, m,
PyH), and 8.71 (1 H, m, PyH); δC(100.6 MHz; CDCl3): 78.3,
83.5, 120.6, 122.5, 122.6, 126.7, 132.5, 136.8, 139.5, 149.7, and
156.4. m/z (ESI+) 179.8 (MH+, 55%).
2-[4-(2,3,4,5-Tetraphenylphenyl)phenyl]pyridine (5). A
solution of 2,3,4,5-tetraphenylcyclopentadienone (4)16 (1.14 g,
2.96 mmol) and 3 (354 mg, 1.98 mmol) in diphenyl ether (3
mL) was deoxygenated and heated at 210 °C under argon for
3 h. The reaction mixture was purified by column chromatog-
raphy over silica using a dichloromethane/light petroleum
mixture (1:1) as eluent to give 5 (988 mg, 93%); mp 207-208
°C. Anal. Calcd for C41H29N: C, 91.9; H, 5.45; N, 2.6. Found:
C, 91.5; H, 5.4; N, 2.6. λmax(CH2Cl2)/nm 283 (log ꢀ/dm3 mol-1
cm-1 4.38) and 259 (4.44); δH(400.2 MHz; CDCl3): 6.80 (2 H,
m, sp H), 6.85-6.96 (13 H, m, sp H), 7.15-7.24 (6 H, sp H
and PyH), 7.29 and 7.83 (2 H, AA′BB′, L PhH), 7.63 (1 H, s,
G1-bp H), 7.68-7.77 (2 H, m, PyH), and 8.67 (1 H, d, J ) 4.5,
PyH). δC (100.6 MHz; CDCl3): 120.5, 122.0, 125.3, 125.6, 125.7,
126.2, 126.3, 126.4, 126.9, 127.1, 127.6, 129.9, 130.4, 131.35,
131.45, 131.5, 136.7, 137.1, 139.3, 139.5, 139.8, 139.9, 140.2,
140.25, 140.8, 141.6, 141.9, 142.5, 149.6, 157.1. m/z (ESI+):
536.1 (MH+, 100%).
fac-Tris[2-{4-[(2,3,4,5-tetraphenyl)phenyl]phenyl}-
pyridinato-N,C2′]iridium(III) (6). A mixture of 5 (1.00 g,
1.87 mmol), iridium trichloride trihydrate (263 mg, 0.747
mmol), water (8 mL), and 2-ethoxyethanol (25 mL) was
deoxygenated and then heated at reflux with stirring under
argon for 16 h. The reaction mixture was filtered, and the
residue was washed with water (3 × 50 mL). The residue was
collected and dissolved in dichloromethane (200 mL), washed
with water (3 × 100 mL) and brine (100 mL), dried over
anhydrous magnesium sulfate, and filtered, and then the
solvent was completely removed to leave a yellow solid (1.19
g) containing a mixture of bis(iridium), bis(chloro) dimer and
unreacted 5. The mixture, 5 (1.32 g, 2.46 mmol), and silver
trifluoromethylsulfonate (190 mg, 0.74 mmol) in diglyme (4
mL) was heated at 130 °C under argon for 2 days. The diglyme
was removed, and the crude product was purified by column
chromatography over silica using a dichloromethane/light
petroleum mixture (1:1) as eluent. The yellow band was
collected, and the solvent was completely removed to give a
bright yellow solid (1.05 g). The residue was dissolved in
dichloromethane and precipitated into methanol to give 6 (841
mg, 63%); mp 274 °C. Anal. Calcd for C123H84IrN3: C, 82.25;
H, 4.7; N, 2.3. Found: C, 82.05; H, 4.7; N, 2.3. λmax(CH2Cl2)/
nm 275 sh (log ꢀ/dm3 mol-1 cm-1 4.97), 308 (4.92), 388 (4.23)
and 467sh (3.58); δH (400.2 MHz; CDCl3): 6.32 (3 H, bt, J 7,
sp H), 6.43-6.62 (12 H, m, L PhH, and sp H), 6.65-6.77 (12
H, m, Py H, L PhH and/or sp H), 6.77-6.95 (27 H, m, L PhH
and/or sp H), 7.08 (15, bs, sp H), 7.20 (3 H, d, J 8, Py H), 7.28-
7.34 (6 H, m, G1-bp H, L PhH), 7.52 (3 H, t, J 8, Py H), 7.71
(3 H, d, J 8, Py H). m/z [MALDI: DITH] Anal. Calcd for
C123H84IrN3: 1793.6 (37%), 1794.6 (51%), 1795.6 (96%), 1796.6
(100%), 1797.6 (65%), 1798.6 (28%), 1799.6 (9%). Found:
1793.2 (38%), 1794.1 (48%), 1795.1 (92%), 1796.1 (100%),
1797.2 (62%), 1798.2 (40%), 1799.3 (13%); GPC: Mh w ) 970 and
Mh v ) 954.
Acknowledgment. We thank CDT Oxford Ltd.,
EPSRC, and SHEFC for financial support.
References and Notes
(1) Baldo, M. A.; Lamansky, S.; Burrows, P. E.; Forrest, S. R.;
Thompson, M. E. Appl. Phys. Lett. 1999, 75, 4.
(2) Yang, X. H.; Neher, D. Appl. Phys. Lett. 2004, 84, 2476.
(3) Nguyen, T.-Q.; Martini, I. B.; Liu, J.; Schwartz, B. J. J. Phys.
Chem. B 2000, 104, 237.
(4) Mitchell, W. J.; Burn, P. L.; Thomas, R. K.; Fragneto, G.;
Markham, J. P. J.; Samuel, I. D. W. J. Appl. Phys. 2004, 95,
2391.
(5) Halim, M.; Pillow, J. N. G.; Samuel, I. D. W.; Burn, P. L.
Adv. Mater. 1999, 11, 371.
(6) Wang, P. W.; Liu, Y. J.; Devadoss, C.; Bharathi, P.; Moore,
J. S. Adv. Mater. 1996, 8, 237.
(7) Freeman, A. W.; Koene, S. C.; Malenfant, P. R. L.; Thompson,
M. E.; Fre´chet, J. M. J. J. Am. Chem. Soc. 2000, 122, 12385.
(8) Adronov, A.; Fre´chet, J. M. J. Chem. Commun. 2000, 1701.
(9) Lo, S.-C.; Male, N. A. H.; Markham, J. P. J.; Magennis, S.
W.; Burn, P. L.; Salata, O. V.; Samuel, I. D. W. Adv. Mater.
2002, 14, 975.
(10) Markham, J. P. J.; Lo, S.-C.; Magennis, S. W.; Burn, P. L.;
Samuel, I. D. W. Appl. Phys. Lett. 2002, 80, 2645.
(11) Anthopolous, T. D.; Frampton, M. J.; Namdas, E. B.; Burn,
P. L.; Samuel, I. D. W. Adv. Mater. 2004, 16, 557.
(12) Anthopolous, T. D.; Markham, J. P. J.; Namdas, E. B.;
Samuel, I. D. W.; Lo, S.-C.; Burn, P. L. Appl. Phys. Lett. 2003,
82, 4824.
(13) Markham, J. P. J.; Samuel, I. D. W.; Lo, S.-C.; Burn, P. L.;
Weiter, M.; Ba¨ssler, H. J. Appl. Phys. 2004, 95, 438.
(14) Lo, S.-C.; Namdas, E. B.; Burn, P. L.; Samuel, I. D. W.
Macromolecules 2003, 36, 9721.
(15) Berresheim, A. J.; Mu¨ller, M.; Mu¨llen, K. Chem. Rev. 1999,
99, 1747.
(16) Morgenroth, F.; Mu¨llen, K. Tetrahedron 1997, 53, 15349.
(17) (a) Sprouse, S.; King, K. A.; Spellane, P. J.; Watts, R. J. J.
Am. Chem. Soc. 1984, 106, 6647. (b) Colombo, H. G.; Brunold,
T. C.; Riedener, T.; Gu¨del, H. W. Inorg. Chem. 1994, 33, 545.
(c) Tamayo, A. B.; Alleyne, B. D.; Djurovich, P. I.; Lamansky,
S.; Tsyba, I.; Ho, N. N.; Bau, R.; Thompson, M. E. J. Am.
Chem. Soc. 2003, 125, 7377.
(18) Burn, P. L.; Beavington, R.; Frampton, M. J.; Pillow, J. N.
G.; Halim, M.; Lupton, J. M.; Samuel, I. D. W. Mater. Sci.
Eng., B 2001, 85, 190.
(19) Grushin, V. V.; Herron, N.; LeCloux, D. D.; Marshall, W. J.;
Petrov, V. A. Chem. Commun. 2001, 1494.
(20) Coppo, P.; Plummer, E. A.; De Cola, L. Chem. Commun. 2004,
1774.
(21) Frampton, M. J.; Namdas, E. B.; Lo, S.-C.; Burn, P. L.;
Samuel, I. D. W. J. Mater. Chem. 2004, 14, 2881.
(22) Gritzner, G.; Kuta, J. Electrochim. Acta 1984, 29, 869.
(23) Demas, J. N.; Crosby, G. A. J. J. Phys. Chem. 1971, 75, 991.
(24) Greenham, N. C.; Samuel, I. D. W.; Hayes, G. R.; Phillips,
R. T.; Kessener, Y. A. R. R.; Moratti, S. C.; Holmes, A. B.;
Friend, R. H. Chem. Phys. Lett. 1995, 241, 89.
(25) Greenham, N. C.; Friend, R. H.; Bradley, D. D. C. Adv. Mater.
1994, 6, 491.
MA050609Q