Solubilised bright blue-emitting iridium complexes for solution processed OLEDs

Article abstract of DOI:10.1039/c6tc00151c

Combining a sterically bulky, electron-deficient 2-(2,4-difluorophenyl)-4-(2,4,6-trimethylphenyl)pyridine (dFMesppy) cyclometalating C^∧N ligand with an electron rich, highly rigidified 1,1′-(α,α′-o-xylylene)-2,2′-biimidazole (o-xylbiim) N^∧N ligand gives an iridium complex, [Ir(dFMesppy)₂(o-xylbiim)](PF₆), that achieves extraordinarily bright blue emission (Φ_PL = 90%; λ_max = 459 nm in MeCN) for a cationic iridium complex. This complex is compared with two reference complexes bearing 4,4′-di-tert-butyl-2,2′-bipyridine, and solution-processed organic light emitting diodes (OLEDs) have been fabricated from these materials.

Full text of DOI:10.1039/c6tc00151c

View Article Online
View Journal
Journal of
Materials Chemistry C
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: A. F. Henwood, A.
Bansal, D. B. Cordes, A. Slawin, I. Samuel and E. Zysman-Colman, J. Mater. Chem. C, 2016, DOI:
10.1039/C6TC00151C.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/materialsC

Page 1 of 27
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00151C
Solubilised Bright Blue-Emitting Iridium
Complexes for Solution Processed OLEDs
Adam F. Henwooda, Ashu K. Bansalb, David B. Cordesa, Alexandra M. Z. Slawina, Ifor D.W.
Samuelb, Eli Zysman-Colmana*
a Organic Semiconductor Centre, EaStCHEM School of Chemistry, University of St
Andrews, St Andrews, Fife, KY16 9ST, UK, Fax: +44-1334 463808; Tel: +44-1334 463826;
b Organic Semiconductor Centre, SUPA, School of Physics and Astronomy, University of St
Andrews, Fife, KY16 9SS, UK
E-mail: eli.zysman-colman@st-andrews.ac.uk;
URL: http://www.zysman-colman.com
Abstract. Combining a sterically bulky, electron-deficient 2-(2,4-difluorophenyl)-4-(2,4,6-
trimethylphenyl)pyridine (dFMesppy) cyclometalating C^N ligand with an electron rich,
highly rigidified 1,1’-(α,α’-o-xylylene)-2,2’-biimidazole (o-xylbiim) N^N ligand gives an
iridium complex, [Ir(dFMesppy)2(o-Xylbiim)](PF6), that achieves extraordinarily bright blue
emission (ΦPL = 90%; λmax = 459 nm in MeCN) for a cationic iridium complex. This
complex is compared with two reference complexes bearing 4,4’-di-tert-butyl-2,2’-
bipyridine, and solution-processed organic light emitting diodes (OLEDs) have been
fabricated from these materials.
Introduction. The world’s ever increasing demand for energy necessitates an overhaul in
virtually all areas of technology. One of the most pressing areas of concern in this problem is
that of artificial lighting, which constitutes approximately 19% of total global electricity
1

Journal of Materials Chemistry C
Page 2 of 27
consumption.1 This tremendous electricity consumption and its problematic societalVieawnAdrticle Online
DOI: 10.1039/C6TC00151C
environmental consequences stem from the poor efficiencies of commonly used lighting
technologies: incandescent light bulbs are typically only 5% efficient, and even compact
fluorescent lighting enforced by the EU is only about 20% efficient.1 There is thus an acute
need for new lighting technologies that can couple high lighting efficiencies with low
fabrication cost.
Organic light emitting devices (OLEDs) have been studied for some time as one of the
leading technologies for meeting these challenges. When iridium-based complexes are used
as the emissive materials, very high device efficiencies are possible, since these complexes
can harvest both singlet and triplet excitons generated from electron-hole recombination by
their strong spin-orbit coupling effects.2 Normally these devices are comprised of a number
of individual layers, which serve to carry out a single function, such as charge transport or
emission. They are frequently fabricated by vacuum sublimation of sequentially deposited
individual layers at sufficiently high temperatures and low pressures.3 Although processing in
this fashion has been shown to give devices with impressive performance metrics, drawbacks
such as the considerable production costs and poor scalability of this fabrication method has
impeded it from being adopted on an industrial scale for lighting.4 Furthermore, the high
temperatures associated with this technique mean only some materials are suitable, and there
are reports of thermal isomerisation of these iridium complexes during device fabrication that
can impact the overall device performance.5 Solution processing addresses these issues;
plastic electronics can be printed onto a flexible substrate in a roll-to-roll fashion.6 However,
typically the efficiencies of devices processed from solution are lower than their counterparts
fabricated by vacuum methods.4a, 7
2

Page 3 of 27
Journal of Materials Chemistry C
View Article Online
With respect to the performance of iridium-based emitters in OLEDs, a simiDlaOrI: 1p0.r1o03b9l/eCm6TC00151C
exists. Although exceptional performance red2b and green8 OLEDs based on iridium have
been reported, a corresponding deep blue emitter remains elusive.2a This is due in no small
part to the host-guest configuration of the emissive layers in OLEDs – the requirement thus
being that the high triplet energies required for blue emitters (>2.8 eV) necessitate even
higher triplet energies for the host materials. At these energy regimes, realising triplet host
energies that are compatible with the emitter and suitable functional device materials (charge
transport layers, electrodes) has become increasingly difficult to achieve. Thus, even devices
that show deep blue emission using iridium complexes show poorer efficiencies compared to
their red- and green-emitting counterparts (~ 30% EQE); a recent review on blue emitters in
OLEDs identified a champion true blue device based on iridium as having CIE coordinates of
(0.14, 0.10), with an EQE of 7.6%,2a while a recent report9 outlined the use of a tris-
cyclometalated NHC iridium complex that achieved CIE coordinates of (0.16, 0.09) at an
EQE of 10.1% – a different story from the higher efficiencies (>20%) reported for sky-blue
emitters.2a, 10
The problem of attaining bright blue emission for any iridium complex stems from the
increasing likelihood of thermal population of non-emissive metal-centred (MC) states as the
optical gap increases. This problem is particularly pertinent for cationic emitters, since their
rational design prohibits employing a third anionic strong field cyclometalating ligand to
efficiently destabilise the MC states. Instead, a combination of highly electron deficient
anionic C^N ligands, and electron rich neutral N^N ligands are required.
In our own research efforts, we have identified, among other motifs, nitrogen-rich
heterocyclic ancillary ligands as promising candidates in this endeavour.11 In particular, we
3

Journal of Materials Chemistry C
Page 4 of 27
View Article Online
have shown biimidazole-type N^N ancillary ligands to be effective motifs for invoking
DOI: 10.1039/C6TC00151C
significant destabilisation of the LUMO energies of these complexes (Figure 1). Crucially,
using these scaffolds we demonstrated that very bright blue emission could be realised
through alkylation of the distal nitrogen atoms with a rigid o-xylylene linker.11b Efficient non-
radiative decay processes observed for the protonated ([Ir(dFppy)2(H2biim)](PF6), where
dFppy is 2,4-phenylpyridinato and H2biim is 1H,1’H-2,2’-biimidazole) and especially the
methylated ([Ir(dFppy)2(dMebiim)](PF6), where dMebiim is 1,1’-dimethyl-2,2’biimidazole)
analogues could be strongly suppressed in ([Ir(dFppy)2(o-xylbiim)](PF6), where o-xylbiim is
1,1’-(α,α’-o-xylylene)-2,2’-biimidazole), to give exceptionally bright emission in MeOH
solution (ΦPL = 68%, λem = 450 nm). However, poor solubility of these complexes prevented
them from being used in solution-processed devices.
Figure 1. Biimidazole complexes previously studied by our group.11b A green arrow connotes
free rotation about the C-C bond, while the red arrow connotes the opposite.
We had previously demonstrated that aryl-substitution of the diimine ligand could
dramatically improve solution-state photoluminescence quantum efficiency, leading to
improved stability in the electroluminescent device.12 In further surveying the literature for
routes towards soluble analogues, we build on the work of Bryce and coworkers, in which
they demonstrate vastly improved solution-processed OLED performances using a mesityl-
functionalised FIrpic-type emitter (Figure 2, FIrpic = Bis[2-(4,6-difluorophenyl)pyridinato-
4

Page 5 of 27
Journal of Materials Chemistry C
C2,N](picolinato)iridium(III)).6b Crucially, the mesityl group accomplishes a number of rVoielweAsr:ticle Online
DOI: 10.1039/C6TC00151C
1) improved solubility to allow formation of high quality spin-coated thin films from
solution; 2) increased steric bulk for inhibiting intramoleculer quenching processes, giving
much brighter emission both in solution and in the device; 3) mutual orthogonality of the
mesityl group with respect to the cyclometalating ligand, which truncates any extension of
the π-conjugation which would invoke an unwanted red-shift in emission.
Figure 2. FIrpic and its mesityl functionalised analogue. The mesityl functionality confers
improved properties in virtually all important metrics, including solution state ΦPL, as well as
ηext values and CIE coordinates for comparable solution processed device architectures. This
comes only at the expense of a small red-shift in solution state λmax.
Herein, we report a highly soluble, very bright blue biimidazole analogue,
[Ir(MesdFppy)2(o-xylbiim)](PF6) (3), which can be used in solution-processed OLEDs and
LEECs (Chart 1).
A
previously studied green-emitting iridium complex,
[Ir(dFppy)2(dtBubpy)](PF6) (1) serves as the reference13 while its mesityl analogue,
[Ir(MesdFppy)2(dtBubpy)](PF6) (2) is included to probe the effect of mesityl substitution on
5

Journal of Materials Chemistry C
Page 6 of 27
View Article Online
the optoelectronic properties of the complex. We study the photophysical properties of these
DOI: 10.1039/C6TC00151C
three complexes in acetonitrile solution, and also the photophysics of 3 in methanol to
compare to [Ir(dFppy)2(o-xylbiim)](PF6) (4). OLEDs have been fabricated from complexes
1-3. To date, very few examples of OLEDs have been reported employing charged iridium
complexes.14
Chart 1. Complexes synthesised and characterised in this study.
Results and Discussion:
Synthesis.
Synthesis of the cyclometalating ligand, dFMesppy, proceeded through a modified
procedure to that reported previously.6b The reported method employed sequential Suzuki-
Miyaura15 cross coupling reactions starting from 2-chloro-4-iodopyridine with first
mesitylboronic acid and 2,4-difluorophenylboronic acid as the corresponding coupling
partners. While the second step is straightforward, isolation of the 2-chloro-4-mesitylpyridine
intermediate after the first step is more problematic. The original procedure stipulated
carrying out the reaction in a 1:1 stoichiometric ratio of mesitylboronic acid and 2-chloro-4-
iodopyridine, presumably to avoid secondary cross-coupling at the 2-position of the pyridine.
However, we found that mesitylboronic acid is prone to deborylation as a competitive
6

Page 7 of 27
Journal of Materials Chemistry C
undesired side reaction under these conditions due to the bulk of the methyl groups in theVi2ew,6Ar-ticle Online
DOI: 10.1039/C6TC00151C
positions. The starting pyridine was thus always recovered from the reaction when carried out
in this manner. Since the starting material has virtually the same Rf as the product, isolating
the desired precursor is not possible by column chromatography alone, requiring an
additional Kugelrohr distillation step to obtain the product in good purity. To facilitate the
purification process, we found that adding a large excess of boronic acid (1.5 – 1.8
equivalents) leads to complete consumption of the starting material. Any side products
generated from over cross coupling under our conditions were separable by column
chromatography. Aside from dFMesppy, all other ligands were synthesised in a similar
manner to that reported previously;11b dFppy was obtained by conventional Suzuki-Miyaura
conditions, while the synthesis of o-xylbiim proceeded first by a condensation reaction with
glyoxal in the presence of ammonium acetate to give H2biim, followed by alkylation with
1,2-bis(bromomethyl)benzene (Scheme 1).
a
Scheme 1. Synthesis of C^N and N^N ligands. Reagents and conditions: K2CO3, 1,4-
dioxane/water (2:1 v/v), Pd(PPh3)4 (5 mol%), 100 °C, 72 h. b Na2CO3, 1,4-dioxane/water (2:1
c
v/v), Pd(PPh3)4 (5 mol%), 100 °C, 19 h. NH4OAc, H2O, 40 °C, 8 h. MeCN, NaOH (35%
d
w/v), 82 °C, 16 h.
7

Journal of Materials Chemistry C
Page 8 of 27
View Article Online
Initially, the complexes were synthesized through cleavage of the dichloro-bridged iridium
DOI: 10.1039/C6TC00151C
dimer [Ir(C^N)2(µ-Cl)]2 with a small excess of ancillary ligand in refluxing DCM/MeOH
solution. However, we found that following Nonoyama’s method,16 in which the precursor
dimer complex is obtained from IrCl3.3H2O and a small excess of cyclometalating ligand,
meant that we were never able to isolate our desired complexes in good purity. Instead we
found that using [Ir(COD)(µ-Cl)]2 (where COD is 1,5-cyclooctadiene) as the initial iridium
source gave vastly improved results.17 Complexation conditions are summarised in Scheme 2.
a
Scheme 2. Synthesis of 1-3. Reagents and conditions: 2-EtOC2H4OH, 110 °C, N2, 3 h. b i.
CH2Cl2/MeOH (1:1 v/v), 50 °C, 19 h, N2; ii. Excess NH4PF6 (aq).
8

Page 9 of 27
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00151C
Characterisation.
13
19
Complexes 1-3 were characterised by NMR spectroscopy (1H, C, F), high resolution
1
mass spectrometry and elemental analyses. As with [Ir(dFppy)2(o-xylbiim)](PF6),11b the H
NMR spectrum of 3 presents itself as a complex mixture of broad signals. The complexity of
this spectrum comes from the conformational rigidity of the o-xylbiim ligand, which means
the fluxional ring flipping processes are slow on the NMR time scale at room temperature.
Since the iridium centre is itself chiral, coordination to this ligand results in diastereomeric
atropisomers. Similarly, the loss of pseudo-C2 symmetry on the NMR timescale leads to a 19F
spectrum in which each fluorine atoms is in its own unique magnetic environment, with the
spectrum presenting three peaks, two of which corresponding to one fluorine atom and
another broad multiplet integrating to double the intensity of the other two peaks (see the
1
ESI†). Heating the solution to 80 °C (372 K) resolved the H spectrum to one exhibiting the
expected C2-symmetry (Figure 3). Eyring analysis on the coalescing doublet at 6.7 ppm at 49
°C (322 K) gave a Gibbs free energy barrier to inversion of +72.29 kJ mol-1 similar to the
previously reported value for [Ir(dFppy)2(o-xylbiim)](PF6) (+82.97 kJ mol-1).11b
9

Journal of Materials Chemistry C
Page 10 of 27
View Article Online
DOI: 10.1039/C6TC00151C
Figure 3. 1H NMR temperature study of 3 in DMSO-d6 from room 298 K to 372 K.
X-ray Structure Analysis:
Suitable single crystals for X-ray analysis were obtained for all three complexes by vapour
diffusion of diethyl ether (1 and 3) or diisopropyl ether (2) into concentrated acetonitrile
solutions (Figure 4). Surprisingly, no single crystal data for 1 has been reported previously.
Analysis of crystal packing can be useful for providing an insight into how the molecules
might preferentially be arranged within the film. Although the spin-coating process is
intended to deposit amorphous films, intermediate-range ordering within spin-coated films of
[Ru(bpy)3](PF6)2 has nonetheless been observed,18 with such crystallinity suggested to impact
the performance of the corresponding electroluminescent devices.19
All three complexes display the expected pseudo-octahedral geometry, with the nitrogen
atoms of the C^N ligands oriented in a trans disposition. Bond lengths and bond angles are
10

Page 11 of 27
Journal of Materials Chemistry C
View Article Online
also as expected for this class of iridium complex. The crucial design feature that is the large
DOI: 10.1039/C6TC00151C
torsion of the mesityl ring with respect to the phenylpyridine moiety is observed for
complexes 2 and 3, with dihedral angles ranging from 69.2(12) °–88.6(19) °.
Figure 4. X-ray crystal structures of complexes 1-3. Hydrogen atoms, solvent molecules and
counterions have been removed for clarity.
An analysis of the extended crystal packing reveals features that explain several
photophysical phenomena observed in the solid state (vide infra). All three complexes show a
similar shortest Ir···Ir internuclear distance [1: 8.5148(9) Å, 2: 9.000(3) Å, 3: 8.8803(7) Å],
despite the other differences observed in intermolecular interactions. Complex 1 shows no
strong intermolecular interactions in the solid state, with only weak C–H···π hydrogen bonds
observed. In contrast, despite the added bulk resulting from the mesityl groups present in 2
and 3, these complexes show additional intermolecular interactions, with complexes
positioned such that two difluorophenyl rings of adjacent complexes are correctly positioned
to make a π-stacking interaction, with centroid···centroid distances of 3.628(8) and 3.531(7)
Å, respectively, facilitated by the propensity of fluorinated phenyl rings to more readily form
11

Journal of Materials Chemistry C
Page 12 of 27
View Article Online
π-stacking interactions than fluorine-free rings.20 These interactions are strengthened by
DOI: 10.1039/C6TC00151C
further mutual C–H···π contacts between the same adjacent complexes. The packing in 3 is
further ordered through a secondary π-stacking interaction between the xylyl rings of the o-
xylbiim ligand of adjacent complexes. This combination of π-interactions results in the
formation of strongly interacting chains of molecules running along the crystallographic ab-
diagonal axis, with alternating long [15.0779(8) Å] (Figure 5, left) and short [8.8803(7) Å]
iridium-iridium distances (Figure 5, right).
Figure 5. X-ray packing of 3 crystal viewed along the plane of the N^N ligand (left) and from
above the cyclometalating ligand N-Ir-N axes (right). For clarity, the iridium atoms have
been coloured red, the cyclometalating C^N ligands blue, the N^N ligands green and the PF6-
counterions yellow (fluorine) and orange (phosphorous). Hydrogens and solvent molecules
are omitted. The distance on the left reflects the iridium-iridium distance across the N^N
ligands, while the distance on the right reflects the iridium-iridium distance across the C^N
ligands.
Electrochemical Properties:
Cyclic voltammetry was undertaken to discern the energy levels of the complexes, which is
crucial for device fabrication purposes. The CV traces of 1 – 3 in acetonitrile are given in
Figure 6, and the relevant electrochemical parameters obtained for these complexes are
summarised in Table 1. Electrochemical data in dichloromethane had previously been
12

Page 13 of 27
Journal of Materials Chemistry C
reported for 1. The previously reported values (Eo1x/2: 1.17 V; E : -1.80 V vs SCE;DΔOIE: 10.10:3V9i2e/wC.96A7TrtiCcl0e0O1n5l1inCe
red
redox
1/2
red
V)13b differ marginally from our own (Eo1x/2: 1.22 V; E : -1.74 V vs SCE; ΔEredox: 2.96 V); the
1/2
modest shifting to more positive potential in MeCN is due to its increased polarity, which confers a
greater stabilization of both the HOMO and LUMO. DFT calculations for 1 have shown the
electron density of the HOMO to be localised on the metal centre and on the C^N ligands
while the LUMO is largely localised on the N^N ancillary, distributions characteristic of
cationic iridium complexes.21 Thus, we assign the HOMO for all three complexes to the
IrIII/IrIV redox couple, with contributions from the phenyl components of the C^N ligands,
while the reduction process occurs on the dtBubpy in the case of 1 and 2. Indeed, the
introduction of the mesityl groups in 2 have a negligible effect on its electrochemistry
compared to 1.
Figure 6. CV traces of complexes 1-3 in MeCN solution, reported versus SCE (Fc/Fc+ = 0.38
V in MeCN)22 redox couple. Scan rates were at 100 mV s-1, and are in the positive scan
direction.
13

Journal of Materials Chemistry C
Page 14 of 27
Table 1. Relevant electrochemical data for complexes 1-3.a
View Article Online
DOI: 10.1039/C6TC00151C
EHOMO (eV) b ELUMO (eV) b
E!!/"! (V)
!"#
Complex
ΔE (V)
E
(V)
!/!
1
2
1.60
1.59
1.37
-1.36
-1.36
-1.99
2.96
2.95
3.36
-6.03
-6.01
-5.79
-3.06
-3.06
-2.43
3
Measurements were carried out in MeCN at a scan rate of 100 mV s-1 with
Fc/Fc+ employed as an internal standard, and reported vs SCE (Fc/Fc+ = 0.38
V in MeCN).22 Reported potentials are referenced to the Fc/Fc+ redox couple.
b EHOMO/LUMO = -[Eox/red vs Fc/Fc+ + 4.8] eV.23
a
The electrochemical properties of 3 differ markedly from 1 and 2. Both the oxidation and
reduction potentials of 3 are shifted cathodically compared to 1 and 2. The reduction wave in
3 shows much poorer reversibility compared to the corresponding reduction waves of 1 and
2. The destabilisation of the HOMO of 3 (5.79 eV) compared with 1 and 2 (ca. -6.0 eV) is
offset by the significant destabilisation of the LUMO (-2.37 eV vs -1.74 eV for 1 and 2)
resulting in a much larger electrochemical gap of 3.36 V. Although the oxidation process in 3
is not likely to involve the o-xylbiim ligand directly, its strong electron-releasing potential
nevertheless strongly influences the largely metal-localised HOMO. The electrochemical data
for [Ir(dFppy)2(o-xylbiim)](PF6), 4, in MeCN demonstrated an oxidation at 1.06 V, although
a reduction process was not observed within the solvent window for that complex.11b
UV-Vis Absorption:
UV-vis absorption spectra for 1-3 are shown in Figure 7 with the data summarised in Table
2. The profile observed for 1 generally reproduces that reported in the literature, with the
principal band at 249 nm and a shoulder at 298 nm assigned as typical π-π* transitions
associated with this family of complexes. The bands at energies lower than 360 nm are
assigned as a combination of singlet and triplet metal-to-ligand and ligand-to-ligand charge
3
transfer (1MLCT/1LLCT and MLCT/3LLCT) transitions.24 The mesityl groups in 2 confer
14

Page 15 of 27
Journal of Materials Chemistry C
View Article Online
only a small change in the profile of the spectrum with increased molar absorptivities across
DOI: 10.1039/C6TC00151C
the spectrum; the shoulder at 261 nm present in 1 is not observed in 2.
Figure 7. UV-vis absorption spectra of complexes 1-3 in MeCN solution.
Similar to that observed in 2, the presence of the mesityl substituents in 3 have minimal
impact on the absorption profile with only modest increases in the molar absorptivities
compared to [Ir(dFppy)2(o-xylbiim)](PF6), 4. The shoulder at 281 nm for 3 is considerably
more pronounced than the shoulder at 279 nm for 4, which is likely the result of π-π*
transitions localised on the mesityl ring. Although the measurements for 4 were carried out in
MeOH while those for 3 are reported in MeCN, the highly ligand-centred (1LC) nature of the
excited state of these complexes results in negligible solvatochromic effects.
15

Journal of Materials Chemistry C
Page 16 of 27
a
View Article Online
Table 2. Absorption maxima and their corresponding molar absorptivities for comDpOleI: x10e.1s0319-/C36.TC00151C
λabs (nm) [ε (×104 M-1 cm-1)]
249 [6.00], 261(sh) [5.46], 298 [3.02], 307(sh) [2.74], 365 [0.48], 420 [0.08], 450
[0.04]
257 [6.25], 297(sh) [3.60], 309(sh) [3.21], 330(sh) [1.66], 368 [0.65], 420 [0.11],
449 [0.04]
258 [4.84], 281(sh) [4.33], 306(sh) [2.73], 321(sh) [2.13], 371 [0.77], 422 [0.12],
450 [0.04]
Complex
1
2
3
a Measurements were carried out in MeCN.
Emission Spectroscopy:
The photoluminescence properties of these complexes were studied in MeCN solution and
are collected in Table 3 and Figure 8. On the basis of the relatively long lifetimes observed
for all three complexes (τe > 1.3 µs), the emission is ascribed as phosphorescence, as is
typical of many cyclometalated iridium(III) complexes. In order to render a more accurate
comparison with 4, the photophysics of 3 was also studied in MeOH. The photophysical
behaviour measured for 1 generally reproduced that described previously (avg. ΦPL = 70-
71%; τe = 1.25-1.40 µs), although two different λmax values have been reported previously
(512 and 524 nm):13 we observe bright green emission (λem = 515 nm; ΦPL = 72%; τe = 1.36
3
µs) from a mixed MLCT/3LLCT state from our own sample, as evidenced by the broad
unstructured emission profile.
16

Page 17 of 27
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00151C
Figure 8. Normalised emission spectra for complexes 1-3 in deaerated MeCN solution. λexc:
360 nm.
The electronics of the system remain unchanged between 1 and 2, analogous to that
observed between FIrpic and its mesityl-functionalised analogue.6b This observation is
consistent with the absorption spectroscopy and the electrochemistry. However, the excited
state kinetics are noticeably different. The ΦPL observed for 2 is higher than for 1, despite
exhibiting virtually the same τe; the result of a combination of slightly increased kr and
decreased knr. By comparison, the study of 3 and 4 in MeOH show this increased brightness
(ΦPL = 82% for 3 vs 68% for 4) is almost entirely due to an enhancement of kr resulting from
the reduced triplet-triplet quenching promoted by the bulky mesityl groups. Complex 3 is
extremely bright in MeCN solution with a ΦPL of 90%. The improved brightness of 3
compared with 1 and 2 is likely to be due to a combination of energy gap law effects arising
from the strongly blue-shifted emission of 3 as well as the strongly rigidifying effects of the
o-xylbiim ligand. Ultimately, we have demonstrated that a blue emitter with near unitary ΦPL
17

Journal of Materials Chemistry C
Page 18 of 27
View Article Online
is achievable through the combination of mesityl substituents on the C^N ligands (which
DOI: 10.1039/C6TC00151C
inhibit intermolecular quenching) and the use of the o-xylbiim ancillary ligand (which
strongly inhibits intramolecular non-radiative processes).
Table 3. Relevant solution state photophysical data for complexes 1-3.a
Complex
1
2
λem (nm) b
515
kr × 105 s-1 knr × 105 s-1
ΦPL (%) c
τe (µs) d
1.36
1.37
2.19
2.26
72
80
90
82
68
5.29
5.84
4.11
3.63
1.77
2.06
1.46
0.46
0.80
0.83
515
3
459, 487
458, 489
457, 487
3 e
4 e,f
3.84
a
Measurements at 298 K in deaerated MeCN unless otherwise stated
otherwise. b λexc: 360 nm. c Quinine sulfate used as the reference (ΦPL = 54.6%
e
in 0.5 M H2SO4 at 298 K).25 d λexc: 375 nm. Measured at 298 K in deaerated
MeOH. f data taken from Ref 11b
.
To the best of our knowledge this is the highest photoluminescence quantum yield for a
blue cationic complex reported to date; all cationic iridium emitters that have a bluer λmax are
significantly less efficient. While this is no doubt in part due to more thermally accessible
MC states in these emitters as a function of a larger HOMO-LUMO gap, this result
nevertheless demonstrates the potency of our ligand design. For example, Baranoff et al
report a deep blue-emitting iridium complex (440 nm in MeCN) based on a bis-NHC
ancillary ligand, but this complex has a ΦPL of only 13%.26 Similarly, we recently reported a
series of complexes bearing multiple 1,2,3-triazole heterocycles, and these too demonstrated
moderately bluer emission (452 nm for the best example) but a much lower ΦPL of 5%.27 On
the other hand, the most efficient example among these 1,2,3-triazole complexes was also the
least blue (487 nm, ΦPL = 30%) and still lower in ΦPL than 3. The group of Qiu et al have
focussed on incorporating pyrazole and imidazole heterocycles into their ligand scaffolds
with a view to blue-shifting the emission of their complexes. Although they have achieved
significant results, none of their complexes have been reported to be more efficient than 3
18

Page 19 of 27
Journal of Materials Chemistry C
(highest ΦPL of 54%), while their bluest is only moderately bluer than our own D(OλIm: 1a0x.10=3V9i4e/wC56A6rTtCicl0e0O1n5l1inCe
nm).7, 27-28
Table 4. Relevant solid-state photophysical data for complexes 1-3.
Complexa
λem (nm)b
520
518
508
474, 502
465, 492
462, 492
ΦPL (%)c
62
90
54
97
τe (µs)d
0.41 (41%), 0.85 (59%)
1.75
0.39 (68%), 1.23 (34%)
1.63
0.19 (29%), 1.23 (71%)
1.92
1 (neat film)
1 (doped film)
2 (neat film)
2 (doped film)
3 (neat film)
43
89
3 (doped film)
a
Neat films were dip coated from MeCN solution while doped films were dip coated
from a DCM solution of 5 wt% of the complex in PMMA. b λexc at 360 nm. c Measured
using an integrating sphere, with λexc for the neat film performed at 360 nm and for
the doped films at 300 nm. c λexc at 378 nm. d λexc at 378 nm.
Photophysical measurements in the solid state were also performed with a view to
correlating these measurements to the crystal packing observations made above.
Measurements were carried out both in neat and doped (5 wt% in PMMA) films, and the
results collated in Table 4 and the spectra shown in Figure 9. We find that the ΦPL values
decrease across the series from 1 (62%) to 2 (54%) to 3 (43%). The decrease in ΦPL may be
linked to the propensity of 2 and particularly 3 to form more ordered crystal packing arrays
than 1, which may make these complexes more susceptible to concentration quenching
processes by means of aggregate formation.29 By contrast, the photoluminescence quantum
yields of the doped films, as expected, are independent of crystal packing effects, with
exceptionally high values for all three complexes. The ΦPL for 1 was previously reported to
be 96%,13a while by comparison we measured 90% following excitation at 300 nm. Under
identical conditions, the ΦPL of 2 is essentially unitary at 97%! Finally, complex 3 is virtually
as bright in the solid state (ΦPL: 89%) as in MeCN solution (ΦPL: 90%), indicating that the
rigidifying effects in the solid state are limited for this complex since this has already largely
19

Journal of Materials Chemistry C
Page 20 of 27
View Article Online
been achieved by the o-xylbiim ligand. These are remarkably high ΦPL values for a charged
DOI: 10.1039/C6TC00151C
iridium complex in the solid state.
Figure 9. Normalised emission spectra for neat (a) and doped (5 wt% in PMMA, b) films of
complexes 1-3. λexc: 360 nm.
We have extensively analysed reported ΦPL values for ionic iridium(III) emitters in the
solid state, and found that very few exceed the values we report here. We note that although
complexes 2 and 3 show significantly diminished ΦPL in neat film compared to solution or
20

Page 21 of 27
Journal of Materials Chemistry C
View Article Online
doped films, these values still compare well with both ‘classic’ iridium complexes which do
DOI: 10.1039/C6TC00151C
not possess large, bulky substituents, such as fac-Ir(ppy)3 (ΦPL = 97% in a 1.5 mol% CBP
film compared to ΦPL = 3% in neat film)30 and FIrpic (ΦPL = 99% in 1.4 mol% mCP film
compared to ΦPL = 15% in neat film),30 as well as with other iridium complexes bearing bulky
ligand functionalities such as Wong’s 4,5-diaza-9,9’-spirobifluorene (ΦPL = 67% in a 1.5
mol% mCP film compared to ΦPL = 32% in neat film).14d
In doped film, aside from the ΦPL of 96% reported previously for 1¸13a only nine cationic
complexes have been reported previously with ΦPL of similar magnitude. For instance, a
series of emitters based on dFppz and varying N^N ligands were reported by Baranoff and
co-workers with three within the study showing ΦPL ranging from 89% to 100%, but with
greener emission than 3 (λmax = 500 – 510 nm).31 Tordera and co-workers in the same year
reported fluorinated green emitters with similar photophysical behaviour (ΦPL ranging from
82% to 93%) at similar wavelengths (503 – 519 nm).32 Housecroft et al reported a number of
complexes characterised by cyclometalating ppz ligands decorated with electron-withdrawing
sulfone groups on the phenyl moiety and found that they were all near-quantitative emitters in
the solid state, with ΦPL ranging from 86 – 94% and λmax 487 – 505 nm.33 Qiu’s complex
bearing a tritylphenyl functionalised phenylimidazole ancillary ligand is the bluest (λmax
=
474 nm) complex amongst those we could find with reported solid state emission data, with a
high ΦPL of 79%;34 it is nevertheless less bright and more red-shifted in emission than 3. It
should be noted that all the values reported for these complexes were from samples in doped
PMMA films (5 wt%) similar to our own. Of particular note, however, is the blue emitting
(λmax = 474 nm) complex [Ir(dFppz)2(dtbubpy)](PF6), which demonstrates a brighter ΦPL
(75%) in neat film than any of 1-3.35
21

Journal of Materials Chemistry C
Page 22 of 27
View Article Online
DOI: 10.1039/C6TC00151C
Device Characterization
Electroluminescent devices: Organic Light-Emitting Diodes (OLEDs).
In view of their promising thin film photophysical properties 1, 2 and 3 were investigated
as emitters in OLEDs (Figure 10). The emitter was embedded in an OXD7/mCP host, and
sandwiched between organic layers of PVK and TPBI. PVK facilitates the injection of holes,
while the electron transport layer TPBI blocks the holes from penetration into the cathode due
to a deep lying HOMO, reducing current leakage. Such a multi-layer structure helps to
confine the excitons within the emitting layer as is needed for good OLED performance.
Except for TPBI and the contacts, all the layers were deposited by solution-processing
methods. Device performance is summarized in Table 5.
Figure 10. Schematic for OLED device. The architecture consisted of the following structure:
ITO/PEDOT:PSS (40 nm)/PVK (35 nm)/(mCP + Active layer + OXD7) (30 nm)]/TPBI (60
nm)/Ca (20nm)/Al (200nm) structure, where 1, 2 or 3 act as the active layer.
22

Page 23 of 27
Journal of Materials Chemistry C
View Article Online
Figure 11a shows the electroluminescence spectra of the three devices. TheDsOpI:e10c.1t0ra39/aCr6eTC00151C
broad for 1, modestly structured for 2, and structured for 3. The EL spectra exhibit similar
profiles to the film photoluminescence spectra, except for the change in the relative intensity
of the vibronic peaks for 3 and a small blue shift for 1 and 2. The photographs for actual
working OLEDs for 1, 2 and 3 are shown in Figures 11b, 11c and 11d, respectively. The
Commission International de L’Éclairage (CIE) colour coordinates of the OLED using
complex 1 are (0.25, 0.48). As observed with the neat PL films, incorporation of the mesityl
groups in 2 results in a blue-shift of the CIE coordinates to (0.21, 0.40). The replacement of
the dtBubpy ligand in 2 with the o-xylbiim ligand in 3 unfortunately does not lead to
substantially bluer CIE coordinates, which were found to be (0.21, 0.37). In the EL spectra,
the emission from mCP:OXD-7 (expected around 410 nm) disappears completely indicating
that complete energy transfer occurs from mCP: OXD-7 mixed host system to all the
emissive complexes.
Figure 11. (a) Normalized electroluminescent emission spectra of 1 (black), 2 (red) and 3
(blue) in thin films. Photograph of the working OLED with (b) 1 (c) 2 and (d) 3. (e) Current
density and luminance versus applied voltage for OLEDs. (f) External quantum efficiency
(solid circles) and Power efficiency (open circles) versus applied voltage for OLEDs made
with 1 (black) and 2 (red) and 3 (blue).
Figure 11e shows the current-voltage-luminance characteristics of the three devices. The
turn-on voltage is lowest for the device made with 3 as the active layer, which is 4.8 V. The
23

Journal of Materials Chemistry C
Page 24 of 27
turn-on voltages for other devices are similar at around 6.3 V. The maximum luminescVeienwcAerticle Online
DOI: 10.1039/C6TC00151C
of 2900 cd/m2 is achieved for devices using 2 as the active layer at a driving voltage of 11.8
V. Figure 11f shows the external quantum efficiency (EQE) of the devices together with the
power efficiency. The maximum EQE is achieved for the reference complex 1. The
efficiencies of these devices are reduced at high brightness. This is likely due to the
deterioration of charge carrier balance in the device at high current density and the increase
of non-radiative quenching processes, including triplet–triplet annihilation (TTA).36 These
results suggest the performance of the devices can be optimized further by improving the
charge balance.
Table 5. OLED performance data.
Complex
Turn-on
Voltage (V)
Luminance
(cd/m2)
max
EQE (%)
@100
cd/m2
EQE (%)
max
Power
efficiency
(lm/W)
@100 cd/m2
Power
efficiency
(lm/W) max
CIE
Coordinates
(x, y)
1
2
3
6.4
6.2
4.8
1790
2935
1086
3.74
3.12
2.86
5.42
3.95
3.42
2.92
2.62
2.42
6.23
4.81
4.42
0.25, 0.48
0.21, 0.40
0.21, 0.37
Compared to our devices, the device reported by Bryce demonstrates superior overall
efficiency (EQE: 10.4% for the optimised device).6b However, we note that charged iridium
emitters are a rarity for OLEDs, and it is useful to compare our own devices to what has been
reported using emitters of this family. To the best of our knowledge, only a single report
exists employing charged iridium complexes as the emitters in a vacuum deposited OLED.
The emitters utilised a conjugated diphenylamine-fluorenylpyridine C^N ligand that allowed
the inherently poor volatility of cationic iridium complexes to be overcome. These devices
show reasonable external quantum efficiencies, peaking at an EQE of 6.5%.14d Our solution-
24

Page 25 of 27
Journal of Materials Chemistry C
View Article Online
processed emitters are not quite as efficient but offer the advantage of simpler processing and
DOI: 10.1039/C6TC00151C
the possibility of being patterned by printing processes.
Conclusions.
A series of cationic green- to blue-emitting, fluorinated iridium complexes have been
studied. We have shown that both highly efficient and deep blue emission can be achieved in
both MeCN solution and in doped PMMA thin films by rational design of the ligand
scaffolds that simultaneously suppress inter- and intramolecular quenching pathways.
Solution processed OLEDs were fabricated, giving sky blue devices with favorable
performances compared with other OLEDs employing charged iridium emitters to date.
Acknowledgements. EZ-C acknowledges the University of St Andrews for financial
support. We thank Johnson Matthey and Umicore AG for the gift of materials. IDWS and
AKB acknowledge support from EPSRC (EP/J01771X). We would like to thank the
Engineering and Physical Sciences Research Council for financial support for Adam
Henwood: EPSRC DTG Grants: EP/J500549/1; EP/K503162/1; EP/L505097/1 We thank the
EPSRC UK National Mass Spectrometry Facility at Swansea University for analytical
services.
Supporting information. Synthesis; CIF of the crystal structures, CCDC Nos. 1440897-
1440899.
References
(1). C. J. Humphreys, MRS Bull., 2008, 33, 459-470.
(2). (a) X. Yang, X. Xu and G. Zhou, J. Mater. Chem. C, 2014, 3, 913-944; (b) K.-H. Kim, S.
Lee, C.-K. Moon, S.-Y. Kim, Y.-S. Park, J.-H. Lee, J. Woo Lee, J. Huh, Y. You and J.-J.
Kim, Nat Commun, 2014, 5.
(3). (a) H. Sasabe and J. Kido, J. Mater. Chem. C, 2013, 1, 1699-1707; (b) G. M. Farinola
and R. Ragni, Chem. Soc. Rev., 2011, 40, 3467-3482.
25

Journal of Materials Chemistry C
Page 26 of 27
View Article Online
(4). (a) M. Cai, T. Xiao, E. Hellerich, Y. Chen, R. Shinar and J. Shinar, Adv Mater, 2011, 23,
DOI: 10.1039/C6TC00151C
3590-3596; (b) R. D. Costa, E. Ortí, H. J. Bolink, F. Monti, G. Accorsi and N. Armaroli,
Angew. Chem. Int. Ed., 2012, 51, 8178-8211.
(5). E. Baranoff, S. Suàrez, P. Bugnon, C. Barolo, R. Buscaino, R. Scopelliti, L. Zuppiroli,
M. Graetzel and M. K. Nazeeruddin, Inorg. Chem., 2008, 47, 6575-6577.
(6). (a) S. R. Forrest, Nature, 2004, 428, 911-918; (b) V. N. Kozhevnikov, Y. Zheng, M.
Clough, H. A. Al-Attar, G. C. Griffiths, K. Abdullah, S. Raisys, V. Jankus, M. R. Bryce and
A. P. Monkman, Chem. Mater., 2013, 25, 2352-2358; (c) P. Matyba, H. Yamaguchi, G. Eda,
M. Chhowalla, L. Edman and N. D. Robinson, ACS Nano, 2010, 4, 637-642.
(7). L. He, J. Qiao, L. Duan, G. Dong, D. Zhang, L. Wang and Y. Qiu, Adv. Funct. Mater.,
2009, 19, 2950-2960.
(8). Y.-S. Park, S. Lee, K.-H. Kim, S.-Y. Kim, J.-H. Lee and J.-J. Kim, Adv. Funct. Mater.,
2013, 23, 4914-4920.
(9). J. Lee, H.-F. Chen, T. Batagoda, C. Coburn, P. I. Djurovich, M. E. Thompson and S. R.
Forrest, Nat Mater, 2016, 15, 92-98.
(10). J.-H. Lee, G. Sarada, C.-K. Moon, W. Cho, K.-H. Kim, Y. G. Park, J. Y. Lee, S.-H. Jin
and J.-J. Kim, Advanced Optical Materials, 2015, 3, 211-220.
(11). (a) S. Evariste, M. Sandroni, T. W. Rees, C. Roldan-Carmona, L. Gil-Escrig, H. J.
Bolink, E. Baranoff and E. Zysman-Colman, J. Mater. Chem. C, 2014, 2, 5793-5804; (b) A.
F. Henwood, S. Evariste, A. M. Z. Slawin and E. Zysman-Colman, Faraday Discuss., 2014,
174, 165-182.
(12). (a) S. Ladouceur, D. Fortin and E. Zysman-Colman, Inorg. Chem., 2010, 49, 5625-
5641; (b) L. Sun, A. Galan, S. Ladouceur, J. D. Slinker and E. Zysman-Colman, J. Mater.
Chem., 2011, 21, 18083-18088.
(13). (a) H. J. Bolink, E. Coronado, R. n. D. Costa, N. Lardiés and E. Ortì, Inorg. Chem.,
2008, 47, 9149-9151; (b) D. Tordera, M. Delgado, E. Ortí, H. J. Bolink, J. Frey, M. K.
Nazeeruddin and E. Baranoff, Chem. Mater., 2012, 24, 1896-1903.
(14). (a) G. Nasr, A. Guerlin, F. Dumur, L. Beouch, E. Dumas, G. Clavier, F. Miomandre, F.
Goubard, D. Gigmes, D. Bertin, G. Wantz and C. R. Mayer, Chem. Commun., 2011, 47,
10698-10700; (b) R. Tao, J. Qiao, G. Zhang, L. Duan, L. Wang and Y. Qiu, J. Phys. Chem.
C, 2012, 116, 11658-11664; (c) J. O. Huh, M. H. Lee, H. Jang, K. Y. Hwang, J. S. Lee, S. H.
Kim and Y. Do, Inorg. Chem., 2008, 47, 6566-6568; (d) W. Y. Wong, G. J. Zhou, X. M. Yu,
H. S. Kwok and Z. Lin, Adv. Funct. Mater., 2007, 17, 315-323; (e) F. Zhang, L. Duan, J.
Qiao, G. Dong, L. Wang and Y. Qiu, Org. Electron., 2012, 13, 1277-1288; (f) N. M.
Shavaleev, G. Xie, S. Varghese, D. B. Cordes, A. M. Z. Slawin, C. Momblona, E. Ortí, H. J.
Bolink, I. D. W. Samuel and E. Zysman-Colman, Inorg. Chem., 2015, 54, 5907-5914.
(15). (a) N. Miyaura, T. Yanagi and A. Suzuki, Synth. Commun., 1981, 11, 513-519; (b) N.
Miyaura and A. Suzuki, Chem. Rev., 1995, 95, 2457-2483.
(16). M. Nonoyama, Bull. Chem. Soc. Jpn., 1974, 47, 767-768.
(17). E. Baranoff, B. F. E. Curchod, J. Frey, R. Scopelliti, F. Kessler, I. Tavernelli, U.
Rothlisberger, M. Grätzel and M. K. Nazeeruddin, Inorg. Chem., 2011, 51, 215-224.
(18). D. R. Blasini, J. Rivnay, D.-M. Smilgies, J. D. Slinker, S. Flores-Torres, H. D. Abruna
and G. G. Malliaras, J. Mater. Chem., 2007, 17, 1458-1461.
(19). C.-Y. Liu and A. J. Bard, J. Am. Chem. Soc., 2002, 124, 4190-4191.
(20). L. He, D. Ma, L. Duan, Y. Wei, J. Qiao, D. Zhang, G. Dong, L. Wang and Y. Qiu,
Inorg. Chem., 2012, 51, 4502-4510.
(21). S. Ladouceur and E. Zysman-Colman, Eur. J. Inorg. Chem., 2013, 2013, 2985-3007.
(22). V. V. Pavlishchuk and A. W. Addison, Inorg. Chim. Acta, 2000, 298, 97-102.
(23). C. M. Cardona, W. Li, A. E. Kaifer, D. Stockdale and G. C. Bazan, Adv. Mater., 2011,
23, 2367-2371.
26

Page 27 of 27
Journal of Materials Chemistry C
View Article Online
(24). For recent reviews see: L. Flamigni, A. Barbieri, C. Sabatini, B. VenDtuOrI:a10.a10n3d9/CF6T.C00151C
Barigelletti, Top. Curr. Chem., 2007, 281, 143-203.
(25). W. H. Melhuish, J. Phys. Chem., 1961, 65, 229-235.
(26). S. B. Meier, W. Sarfert, J. M. Junquera-Hernandez, M. Delgado, D. Tordera, E. Orti, H.
J. Bolink, F. Kessler, R. Scopelliti, M. Gratzel, M. K. Nazeeruddin and E. Baranoff, J. Mater.
Chem. C, 2013, 1, 58-68.
(27). J. M. Fernandez-Hernandez, S. Ladouceur, Y. Shen, A. Iordache, X. Wang, L. Donato,
S. Gallagher-Duval, M. de Anda Villa, J. D. Slinker, L. De Cola and E. Zysman-Colman, J.
Mater. Chem. C, 2013, 1, 7440-7452.
(28). (a) C. D. Sunesh, K. Shanmugasundaram, M. S. Subeesh, R. K. Chitumalla, J. Jang and
Y. Choe, ACS Appl Mater Interfaces, 2015, 7, 7741-7751; (b) L. He, L. Duan, J. Qiao, R.
Wang, P. Wei, L. Wang and Y. Qiu, Adv. Funct. Mater., 2008, 18, 2123-2131.
(29). (a) J. Mei, Y. Hong, J. W. Y. Lam, A. Qin, Y. Tang and B. Z. Tang, Adv. Mater., 2014,
26, 5429-5479; (b) Y. Q. Zhang, G. Y. Zhong and X. A. Cao, J. Appl. Phys., 2010, 108,
083107.
(30). Y. Kawamura, K. Goushi, J. Brooks, J. J. Brown, H. Sasabe and C. Adachi, Appl. Phys.
Lett., 2005, 86, 071104-071103.
(31). E. Baranoff, H. J. Bolink, E. C. Constable, M. Delgado, D. Haussinger, C. E.
Housecroft, M. K. Nazeeruddin, M. Neuburger, E. Orti, G. E. Schneider, D. Tordera, R. M.
Walliser and J. A. Zampese, Dalton Trans., 2013, 42, 1073-1087.
(32). D. Tordera, J. J. Serrano-Pérez, A. Pertegás, E. Ortí, H. J. Bolink, E. Baranoff, M. K.
Nazeeruddin and J. Frey, Chem. Mater., 2013, 25, 3391-3397.
(33). D. Tordera, A. M. Bünzli, A. Pertegás, J. M. Junquera-Hernández, E. C. Constable, J.
A. Zampese, C. E. Housecroft, E. Ortí and H. J. Bolink, Chem. Eur. J., 2013, 19, 8597-8609.
(34). L. He, L. Duan, J. Qiao, G. Dong, L. Wang and Y. Qiu, Chem. Mater., 2010, 22, 3535-
3542.
(35). H.-C. Su, H.-F. Chen, Y.-C. Shen, C.-T. Liao and K.-T. Wong, J. Mater. Chem., 2011,
21, 9653-9660.
(36). M. A. Baldo, M. E. Thompson and S. R. Forrest, Nature, 2000, 403, 750.
TOC Graphic
27