Luminescent iridium(III) complexes with N^C^N- coordinated terdentate ligands: Dual tuning of the emission energy and application to organic light-emitting devices

Article abstract of DOI:10.1021/ic202756w

A family of complexes (1a-3a and 1b-3b) was prepared, having the structure Ir(N^{^}C^{^}N)(N^{^}C)Cl. Here, N^{^}C ^{^}N represents a terdentate, cyclometallating ligand derived from 1,3-di(2-pyridyl)benzene incorporating CH₃ (1a,b), F (2a,b), or CF₃ (3a,b) substituents at the 4 and 6 positions of the benzene ring, and N^{^}C is 2-phenylpyridine (series a) or 2-(2,4-difluorophenyl) pyridine (series b). The complexes are formed using a stepwise procedure that relies on the initial introduction of the terdentate ligand to form a dichloro-bridged iridium dimer, followed by cleavage with the N^{^}C ligand. ¹H NMR spectroscopy reveals that the isomer that is exclusively formed in each case is that in which the pyridyl ring of the N ^{^}C ligand is trans to the cyclometallating aryl ring of the N ^{^}C^{^}N ligand. This conclusion is unequivocally confirmed by X-ray diffraction analysis for two of the complexes (1b and 3a). All of the complexes are highly luminescent in degassed solution at room temperature, emitting in the green (1a,b), blue-green (2a,b), and orange-red (3a,b) regions. The bidentate ligand offers independent fine-tuning of the emission energy: for each pair, the "b" complex is blue-shifted relative to the analogous "a" complex. These trends in the excited-state energies are rationalized in terms of the relative magnitudes of the effects of the substituents on the highest occupied and lowest unoccupied orbitals, convincingly supported by time-dependent density functional theory (TD-DFT) calculations. Luminescence quantum yields are high, up to 0.7 in solution and close to unity in a PMMA matrix for the green-emitting complexes. Organic light emitting devices (OLEDs) employing this family of complexes as phosphorescent emitters have been prepared. They display high efficiencies, at least comparable, and in some cases superior, to similar devices using the well-known tris-bidentate complexes such as fac-Ir(ppy)₃. The combination of terdentate and bidentate ligands is seen to offer a versatile approach to tuning of the photophysical properties of iridium-based emitters for such applications.

Full text of DOI:10.1021/ic202756w

Article
pubs.acs.org/IC
Luminescent Iridium(III) Complexes with N∧C∧N-Coordinated
Terdentate Ligands: Dual Tuning of the Emission Energy and
Application to Organic Light-Emitting Devices
Pierpaolo Brulatti,† Richard J. Gildea,† Judith A. K. Howard,† Valeria Fattori,‡ Massimo Cocchi,*,‡
and J. A. Gareth Williams*,†
†Department of Chemistry, University of Durham, South Road, Durham DH1 3LE, U.K.
‡Istituto per la Sintesi Organica e la Fotoreattivita,
̀
Consiglio Nazionale delle Ricerche, via P. Gobetti 101, 40129 Bologna, Italy
S
* Supporting Information
ABSTRACT: A family of complexes (1a-3a and 1b-3b) was
prepared, having the structure Ir(N∧C∧N)(N∧C)Cl. Here,
N∧C∧N represents a terdentate, cyclometallating ligand
derived from 1,3-di(2-pyridyl)benzene incorporating CH3
(1a,b), F (2a,b), or CF3 (3a,b) substituents at the 4 and 6
positions of the benzene ring, and N∧C is 2-phenylpyridine
(series a) or 2-(2,4-difluorophenyl)pyridine (series b). The
complexes are formed using a stepwise procedure that relies on
the initial introduction of the terdentate ligand to form a
dichloro-bridged iridium dimer, followed by cleavage with the
1
N∧C ligand. H NMR spectroscopy reveals that the isomer
that is exclusively formed in each case is that in which the
pyridyl ring of the N∧C ligand is trans to the cyclometallating
aryl ring of the N∧C∧N ligand. This conclusion is unequivocally confirmed by X-ray diffraction analysis for two of the complexes
(1b and 3a). All of the complexes are highly luminescent in degassed solution at room temperature, emitting in the green (1a,b),
blue-green (2a,b), and orange-red (3a,b) regions. The bidentate ligand offers independent fine-tuning of the emission energy: for
each pair, the “b” complex is blue-shifted relative to the analogous “a” complex. These trends in the excited-state energies are
rationalized in terms of the relative magnitudes of the effects of the substituents on the highest occupied and lowest unoccupied
orbitals, convincingly supported by time-dependent density functional theory (TD-DFT) calculations. Luminescence quantum
yields are high, up to 0.7 in solution and close to unity in a PMMA matrix for the green-emitting complexes. Organic light
emitting devices (OLEDs) employing this family of complexes as phosphorescent emitters have been prepared. They display
high efficiencies, at least comparable, and in some cases superior, to similar devices using the well-known tris-bidentate complexes
such as fac-Ir(ppy)3. The combination of terdentate and bidentate ligands is seen to offer a versatile approach to tuning of the
photophysical properties of iridium-based emitters for such applications.
cationic complexes containing two bidentate cyclometallating
ligands {e.g. [Ir(N∧C)2(L∧L)]+, where L∧L is a diimine ligand,
for example} are also proving to be of interest in other areas,
such as photocatalysts for water-splitting,5 as components of
assemblies for photoinduced energy-transfer processes,6 dyes
for solar cells,7 triplet-state generators for upconversion,8 and as
biocompatible luminescent probes, sensors, and imaging
agents.9
To date, almost all such work has focused on systems
containing at least two bidentate N∧C ligands, primarily of the
generic types indicated above. The usual synthetic strategy to
such compounds, established by Nonoyama, relies upon the
initial formation of chloro-bridged dimers of the type
[Ir(N∧C)2μ-Cl]2 upon reaction of the cyclometallating
INTRODUCTION
■
Interest in the photophysical properties of iridium(III)
complexes with aromatic heterocyclic ligands has soared over
the past decade, following the demonstration of their potential
as phosphors in organic light-emitting diodes (OLEDs).1 The
high efficiency of emission of complexes such as fac-[Ir(ppy)3],
together with their good chemical and photochemical stability
and charge neutrality, makes them suitable for harvesting the
otherwise nonemissive triplet states formed upon charge
recombination in such devices, in ratios as high as 3:1 over
the singlet states.1−3 The large increases in device efficiency
shown to be possible have proved to be a major impetus for the
synthesis and investigation of a vast range of derivatives of the
Ir(N∧C)3, Ir(N∧C)2(L∧X), and Ir(N∧N−)3 structural classes
{N∧C = bidentate cyclometallating ligands such as ppy; L∧X =
anionic ligands such as acac and 2-picolinate; N∧N− = 5-(2-
pyridyl)pyrazoles and 5-(2-pyridyl)triazoles}.4 Closely related
Received: December 29, 2011
Published: March 8, 2012
© 2012 American Chemical Society
3813
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
Scheme 1. (a) Synthesis of the 1,3-Dibrominated Aromatic Precursors to the Terdentate Ligands; (b) Synthesis of the
Terdentate Ligands by Palladium(II)-Catalyzed Cross-Coupling of the Products from (a) with 2-(Tri-n-butylstannyl)pyridine;
(c) Structural Formulae of the Mononuclear Iridium(III) Complexes Described in This Work, and Their Preparation via the
Chloro-Bridged Dimers
proligand with IrCl3·xH2O.10 Since this reaction introduces two
N∧C ligands simultaneously into the coordination sphere of
each metal ion, the products invariably contain two identical
cyclometallating ligands. As a result, despite the large synthetic
effort expended in producing a vast range of Ir(N∧C)2(L∧X)
and [Ir(N∧C)2(L∧L)]+ complexes, these studies have generally
not allowed complexes with two differently substituted
cyclometallated ligands to be investigated.11,12
di(2-pyridyl)benzene.13,14 Derivatives of this compound can
bind to iridium(III) as an N∧C∧N-coordinating ligand. We
previously demonstrated that compounds of the type [Ir-
(N∧C∧N)Cl(μ-Cl)]2 are accessible, from which a variety of
different classes of mononuclear iridium(III) complexes can be
obtained upon treatment with other terdentate ligands (e.g.,
N∧N∧N, N∧N∧C, N∧N∧O, C∧N∧C-coordinating ligands).15−17
The photophysical properties of these complexes, including the
very wide variation in luminescence quantum yields among
them, could be rationalized in terms of the nature of the
We have been investigating the complexation chemistry of
iridium with terdentate cyclometallating ligands based on 1,3-
3814
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
frontier orbitals.13 Meanwhile, Haga and co-workers have been
exploring structurally related complexes based on a different
type of N∧C∧N-coordinating ligand, namely, 1,3-bis-(1-methyl-
benzimidazol-2-yl)benzene (Mebib).18,19
conditions gave a much more soluble product, the H NMR
spectrum of which was consistent with metalation at C2 to give
[Ir(tfdpyb)Cl(μ-Cl)]2. Clearly, the presence of the CF3 groups
greatly improves the solubility of the compound.
1
Here, we report on the development of a class of complexes
of the type [Ir(N∧C∧N)(N∧C)X], that comprise one terdentate
cyclometallating ligand, one bidentate cyclometallating ligand,
and one monodentate ligand. We described the first example of
such a complex in our earlier study,16 while Haga and
colleagues also prepared such a complex with an asymmetric
pyridyl/imidazole N∧C∧N-binding ligand, which could be
resolved into its two constituent isomers.20 The current family
of new complexeswhich are readily accessible from the
[Ir(N∧C∧N)Cl(μ-Cl)]2 dimershave allowed a detailed
assessment of the effects of substitution in the two different
cyclometallating ligands to be made. We show that the emission
energy can be tuned over a wide range across much of the
visible spectrum, according to the substitution of the central
aryl ring of the terdentate ligand, while the bidentate ligand
independently offers further fine-tuning of the emission energy.
Selected complexes have been incorporated into OLEDs by
vacuum sublimation, leading to high-efficiency devices. Despite
the intense international activity in testing iridium complexes
with bidentate ligands as OLED phosphors, those with
terdentate ligands have been almost completely overlooked,21,19
even though they might be expected to offer advantages such as
improved chemical stability, for example. Here, we demonstrate
that good OLED performance is possible using such complexes,
successfully spanning a wide range of the visible spectrum.
2. Synthesis of Mononuclear Complexes by Cleavage
of Chloro-Bridged Dimers with N∧C Ligands. Subsequent
introduction of the bidentate, N∧C-coordinating cyclometallat-
ing ligand into the coordination sphere of the metal ion was
accomplished by treatment of the dichloro-bridged dimers with
phenylpyridine (ppyH) or difluorophenylpyridine (dfppyH) in
the presence of silver triflate as a chloride scavenger. The
reaction could be accomplished in toluene at 115 °C, under
which mild conditions, no scrambling of ligands occurs that
might otherwise lead to homoleptic products. This contrasts
with the scrambling that is typically encountered in the
preparation of heteroleptic Ir(N∧C)3 complexes.26 Using this
method, the family of six complexes shown in Scheme 1c were
prepared, the ground- and excited-state properties of which are
the subject of the sections that follow.
3. Regiochemistry, NMR Spectroscopy, and Molecular
Structures. Two regioisomeric products are in principle
possible from the above reaction, namely, that in which the
pyridine nitrogen of the N∧C ligand occupies the position trans
to the cyclometallating carbon atom of the N∧C∧N ligand
(Figure 1, isomer A), or in which the two cyclometallating
RESULTS AND DISCUSSION
■
1. Synthesis of the Terdentate Ligands and Their
Reaction with Iridium(III) Chloride. Previously, we showed
that 1,3-di(2-pyridyl)benzene undergoes metalation with
iridium primarily at position C4, to give bidentate coordina-
tion,15,16 rather than at C2 to give the terdentate structure that
is typically encountered when it binds to Ru(II) or
Pt(II).14,22,23 By introducing methyl substituents into the C4
and C6 positions [i.e., 1,3-di(2-pyridyl)-4,6-dimethylbenzene or
dpyxH], we showed that it is possible to block this competitive
metalation, and hence direct reaction to position C2 to give the
terdentately bound product.15 In principle, it should be possible
to use any noncoordinating, sterically undemanding substituent
for this purpose. In the present study, we have prepared and
investigated related ligands incorporating fluorine or trifluor-
omethyl groups at these positions. 1,3-Di(2-pyridyl)-4,6-
difluorobenzene (F2dpybH) and 1,3-di(2-pyridyl)-4,6-bis-
(trifluoromethyl)benzene (tfdpybH) were prepared by palla-
dium-catalyzed Stille cross-coupling of the corresponding
dibrominated aromatics with 2-(tri-n-butylstannyl)pyridine, as
shown in Scheme 1b. 1,3-Dibromo-4,6-difluorobenzene was
obtained by iron-catalyzed elemental bromination of 2,4-
difluorobromobenzene,24 while 1,3-dibromo-4,6-bis-
(trifluoromethyl)benzene was synthesized by bromination of
3,5-bis(trifluoromethyl)aniline followed by diazotization of the
amine and its subsequent reduction under acidic conditions
(Scheme 1a).25
Figure 1. Two isomers that are in principle possible for Ir(N∧C∧N)-
(N∧C)Cl complexes. Isomer A, in which the pyridyl ring of the N∧C
ligand is trans to the metalated aryl ring of the N∧C∧N ligand, is
exclusively formed, 1H NMR spectroscopic evidence for which
includes the NOE coupling shown.
1
carbon atoms are trans to one another (isomer B). H NMR
spectroscopy indicates that one of these two products forms
1
specifically, rather than a mixture, and two pieces of H NMR
1
data suggest it to be the former. A representative H NMR
spectrum of one of the new complexes (3b) is shown in Figure
2. The spectra are characterized by a very large shift to low-
frequency of proton H6′ adjacent to the metalated carbon of the
bidentate ligand (e.g., for 3b, this proton resonates at 5.21
ppm). An upfield shift is observed for the corresponding proton
in tris-bidentate complexes such as Ir(dfppy)3,27 which can be
rationalized in terms of the increase in electron density in the
aryl ring that accompanies cyclometalation. However, the
change in chemical shift is larger in 3b than in Ir(dfppy)3: Δδ =
−1.8 and −1.0 ppm, respectively, relative to the proligand
dfppyH. This effect can be understood if isomer A is formed
rather than isomer B, since H6′ is then positioned below the
plane of the central aryl ring of the terdentate ligand, where it
will experience the associated upfield-shifting effect of the
diamagnetic ring current. The pair of H6 protons in the
terdentate ligand are similarly observed to experience a
substantial upfield shift (e.g., Δδ = −1.5 ppm relative to the
The reaction of F2dpybH with IrCl3·xH2O in a mixture of
ethoxyethanol and water led to a yellow precipitate of low
solubility in all common solvents. By analogy with the reactivity
of dpyxH,15,16 this compound was assumed to be the dichloro-
bridged dimer [Ir(F2dpyb)Cl(μ-Cl)]2 (Scheme 1c). The
reaction of tfdpybH with IrCl3·xH2O under the same
3815
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
1
Figure 2. H NMR spectrum of Ir(tfdypb)(dfppy)Cl (3b) at 700 MHz in CDCl3 at 298 K.
precursor [Ir(tfdpyb)2Cl(μ-Cl)]2), being positioned on either
side of the pyridyl ring of the bidentate ligand. A further
characteristic feature of the spectra is the shift to high-
frequency of H6, the proton adjacent to the N atom of the
bidentate ligand (e.g., for 3b, Δδ = +1.5 ppm), strongly
suggesting that this proton is adjacent to the Ir−Cl bond, and
does not experience an upfield-shifting ring current effect, as
would be anticipated in isomer B. Finally, we note that a cross-
peak is observed in the NOESY spectra between H6(N∧C∧N)
and H6(N∧C), but not H6′(N∧C), which also supports the
assignment of isomer A.
center exhibits a distorted octahedral geometry, with an NNCN
−
Ir−NNCN angle of 161.9(1)°; this value is similar to those found
in the other two crystallographically characterized examples of
pyridyl-based N∧C∧N ligands bound to iridium(III) [159.9(2)°
and 160.3(1)° in Ir(dpyx)(dmso)Cl2 and [Ir(dpyx)(ttpy)]2+
The preference for isomer A can be understood in terms of
the strong trans influence of cyclometallating carbon atoms,
which typically disfavors the formation of compounds in which
two such ligating atoms are positioned trans to one another,
when an alternative configuration is possible. For example, the
compounds of the form [Ir(N∧C)2(μ-Cl)]2 discussed in the
introduction almost invariably display a cis rather than trans
disposition of the carbon atoms relative to one another.28
Likewise, it accounts for the instability of [Ir(N∧C∧N)-
(C∧N∧C)] compounds with respect to cleavage of the mutually
trans Ir−C bonds.15,16 Such an N−trans−C arrangement was
also observed crystallographically by Haga and co-workers in
the case of [Ir(Mebib)(Meppy)Cl].18
In the present instance, confirmation of the identity of the
isomer formed was obtained in two cases by X-ray diffraction
analysis of single crystals, namely, for Ir(dpyx)(dfppy)Cl (1b)
and Ir(tfdpy)(ppy)Cl (3a), which crystallized in the presence
of dichloromethane and diethyl ether respectively. The
molecular structures of these complexes in the crystals are
shown in Figure 3, and selected bond lengths and angles are
collated in Table 1. Most of the bond lengths and angles
involving the metal are essentially identical in the two cases,
given the standard deviations; for simplicity in the ensuing
discussion, we therefore refer to the values for only one
complex, Ir(tfdpy)(ppy)Cl, while noting that exactly the same
points apply to Ir(dpyx)(dfppy)Cl. As expected, because of the
geometric constraints associated with binding of a terdentate
ligand that forms 5-membered chelating rings, the iridium
Figure 3. Molecular structures of Ir(dpyx)(dfppy)Cl (1b, top) and
Ir(tfdpyb)(ppy)Cl (3a, bottom); T = 120 K. Hydrogen atoms have
been omitted for clarity.
3816
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
frequency calculations were performed to confirm the attain-
ment of energy minima on the potential energy surface. For the
two complexes for which experimental X-ray data are available
(1b and 3a), there is excellent agreement between the bond
lengths and angles predicted by DFT and those determined
experimentally (values are listed in Table 1). The frontier
orbitals (HOMO−1 to LUMO+1) for each of the three
complexes containing ppy (series a) are represented in Table 2,
Table 1. Selected Bond Lengths (Å) and Angles (deg) for
[Ir(dpyx)(dfppy)Cl] 1b and [Ir(tfdpyb)(ppy)Cl] 3a
Determined by X-ray Crystallography at 120 K, and
Corresponding Values from Optimized Geometries
Obtained by DFT Calculations
[Ir(dpyx)(dfppy)Cl] 1b
[Ir(tfdpyb)(ppy)Cl] 3a
X-ray data DFT data
X-ray data
DFT data
Ir−N(1)
Ir−C(1)
Ir−N(2)
Ir−N(3)
Ir−C(19)
Ir−Cl
2.046(5)
1.936(6)
2.039(5)
2.159(5)
1.993(5)
2.462(2)
2.065
1.951
2.065
2.190
2.030
2.545
2.046(3)
1.934(4)
2.047(3)
2.161(3)
2.018(4)
2.469(1)
2.054
1.945
2.054
2.187
2.038
2.549
Table 2. Frontier Orbitals of the ppy Series of Complexes 1a,
2a, 3a Obtained from TD-DFT Calculations, Together with
the Orbital Energies and the Proportion of Electron Density
a
on the Metal and the Three Ligands
N(1)−Ir−N(2)
N(1)−Ir−C(1)
N(2)−Ir−C(1)
N(3)−Ir−C(1)
C(1)−Ir−Cl
N(3)−Ir−Cl
C(19)−Ir−Cl
C(1)−Ir−C(19)
161.2(2)
80.5(2)
80.8(2)
174.2(2)
92.3(2)
93.2(2)
171.9(2)
95.8(2)
160.7
80.4
161.9(1)
80.8(1)
81.1(1)
176.2(1)
90.2(1)
93.6(1)
172.2.2(1)
97.0(1)
161.2
80.7
80.4
80.7
176.4
91.6
176.8
91.3
92.1
91.8
170.8
97.6
170.6
98.1
respectively16,15], and also to the corresponding angle of 160.2°
in [Ir(tpy)2]3+.29 The M−L bond lengths to the terdentate and
bidentate ligands reveal significant differences. Thus, the Ir−
C
NCN bond is shorter than the Ir−CNC [1.934(4) and 2.018(4)
Å respectively], presumably reflecting the constraints imposed
on the former by binding of the two flanking pyridyl rings. For
comparison, the Ir−C bond in fac-Ir(Meppy)3 is 2.024(6) Å
{Meppy = 2-(4′-methylphenyl)pyridine}.28b The Ir−NNCN
bonds (average 2.047 Å) are likewise shorter than the Ir−
N
NC [2.161(3) Å] and shorter than the Ir−N in fac-Ir(Meppy)3
{2.132(5) Å}. The main factor at play here is almost certainly
the fact that in the latter two cases, the N atoms are trans to
cyclometallated carbon atomswhich have a very strong trans
influencewhereas the N atoms of the N∧C∧N ligand are
necessarily trans to one another.
In the solid state, the compounds are stable indefinitely:
samples stored for three years show no evidence of
decomposition. In solution in CH3CN in the presence of
a
An analogous diagram for complexes 1b, 2b, and 3b is provided in
1
light, H NMR spectroscopy revealed that the complexes are
the Supporting Information.
partially transformed into different products over a period of
several hours, probably to the acetonitrile adducts [Ir(N∧C∧N)-
(N∧C)(MeCN)]+Cl−. Labilization of the Ir−Cl bond can be
understood on the basis of the high trans influence of the
cyclometallating carbon atom of the N∧C ligand. The effect was
most rapid for the complexes of the F2dpy ligand and least
significant for those of tfdpyb. The stability is higher in CH2Cl2.
A similar photosolvation reaction was observed by Haga and
co-workers for [Ir(Mebib)(Meppy)Cl] in solvents such as
acetonitrile.18
4. Frontier Orbital Description Based on DFT
Calculations. Density functional theory (DFT) calculations
using B3LYP have been carried out to predict the energy-
minimized, gas-phase structures of the complexes and to
identify the frontier orbitals. The approach employed is one
that is now well-established for third-row transition metal
complexes, using the LANL2DZ basis set for Ir(III), with the
inner core electrons replaced by a relativistic core potential, and
6-31G for the ligands.30 The ground-state geometries were fully
optimized without symmetry constraints, and harmonic
together with their energies and the % of the electron density
on the three ligands and the metal. A corresponding table for
the analogous complexes containing dfppy (series b) is shown
in the Supporting Information.
Considering Ir(dpyx)(ppy)Cl (1a) in the first instance, it can
be seen that the highest occupied molecular orbital (HOMO) is
delocalized across the metal, the halide ligand, and the
cyclometallating ring of the N∧C∧N ligand. The cyclo-
metallating ring of the bidentate ligand, in contrast, makes
only a minor contribution. The metal and halide similarly
contribute to HOMO−1 but, in this case, it is the bidentate
rather than the terdentate ligand that contributes. It should,
however, be noted that the difference in energy between these
orbitals is very small (0.01 eV), and the order at room
temperature and in the presence of solvent could be reversed.
Moving to the complexes Ir(F2dpyb)(ppy)Cl 2a and Ir-
(tfdpyb)(ppy)Cl 3a, the situation is similar to 1a but the
order is inverted; that is, the HOMO involves metal, halide, and
3817
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
the aryl ring of N∧C, while HOMO−1 involves metal, halide,
and aryl ring of N∧C∧N. This lowering in energy of molecular
orbitals involving the N∧C∧N ligand in 2a and 3a can be
rationalized intuitively in terms of the inductively electron-
withdrawing nature of the fluorine and trifluoromethyl
substituents in the N∧C∧N ligand in these two complexes,
respectively. The lowest unoccupied molecular orbitals
(LUMOs) of all six complexes have a very similar electron
distribution to one another, being delocalized across the
N∧C∧N ligands, but not significantly involving the other ligands
or the metal ion.
analogue in each case. In contrast, the substituents in the
N∧C∧N ligand are seen to have a larger influence on the
energies of corresponding bands; the order of energies is
F2dpyb (2a, 2b) > dpyx (1a, 1b) > tfdpyb (3a, 3b).
To assist in the assignment of bands in this region, time-
dependent DFT (TD-DFT) calculations have been carried out,
the results of which are summarized in Table 4 for the two
lowest-energy triplet states and five lowest-energy singlet
excited states. A quantitative comparison of experimental and
theoretical data should be treated with caution, as the
calculations apply to the gas-phase at 0 K and do not take
into account the spin−orbit coupling (SOC), while bands
arising from transitions within the singlet and triplet manifolds
may overlap. Nevertheless, they do provide some insight into
the electronic origins of the bands and the trend in energies.
Considering Ir(dpyx)(ppy)Cl (1a) as an example, the TD-DFT
suggests that the lowest-energy band, which has a relatively low
intensity (494 nm, ε = 1300 M−1 cm−1), is probably due to the
formally forbidden S0→T1 and S0→T2 transitions (calculated
values 487 and 483 nm), the oscillator strengths of which are
augmented by the SOC associated with the iridium ion.
(Further evidence in support of this assignment comes from
emission spectroscopy, see section 6). These transitions are
seen to be composed mainly of HOMO−1→LUMO and
HOMO→LUMO character, respectively, and thus can be
regarded as dIr/πNC→π*NCN charge-transfer (mixed MLCT/
LL′CT) and dIr/πNCN→π*NCN (mixed MLCT/IL) transitions,
respectively. Such assignments are quite typical for iridium(III)
complexes with cyclometallating ligands, where the high degree
of covalency in the C−Ir bond leads to heavily mixed d/π
character in the highest-energy filled orbitals. The substantial
participation of the metal would be expected to lead to
significant SOC and hence to relaxation of the spin selection
rule, as observed. Meanwhile, the lowest-energy, spin-allowed
singlet bands are seen to involve mostly the two highest-energy
filled orbitals and two lowest-energy virtual orbitals, and thus
also have a high degree of M→LNCN charge transfer character.
Very similar conclusions can be drawn for all the other
complexes using the data of Table 4 and the frontier orbitals
illustrated in Table 2. The qualitative trend in absorption
energies for the lowest-energy intense (i.e., spin-allowed)
absorption band among the complexes (i.e., 2b > 2a > 1b > 1a
> 3b > 3a) is seen to be essentially reproduced by the
calculations (see energies/wavelengths of the S0→S1 transitions
listed in Table 4). The trend can be rationalized in terms of the
electronic effects of the substituents as follows. First, the
introduction of the uniquely electron-withdrawing trifluoro-
methyl groups into the N∧C∧N ligand decreases the energy of
all orbitals, but the effect is larger for the LUMO (which is
based on this ligand) than for the HOMO (energies of orbitals
are given in Table 2). Thus, the transitions shift to lower
energy, and the absorption bands are red-shifted. Fluorine
atoms might similarly be expected to lower the energy of
orbitals through their inductively withdrawing nature. However,
fluorine has a mesomeric electron-donating effect, which will
tend to counteract the inductively withdrawing effect at
positions ortho and para (but not meta) to the F atom. Since
the metal is disposed meta to the F atoms, whereas the pyridyl
rings of the N∧C∧N ligand are ortho to F, one may thus
anticipate that the LUMO will be lowered in energy less than
the HOMO, leading to a net blue-shift, as observed. These
effects are represented schematically in the simplified energy
level diagram for frontier orbtials shown in Figure 5. The same
5. Ground State UV−Visible Absorption Spectra and
Time-Dependent DFT (TD-DFT) Analysis. UV−visible
spectral data of the complexes in CH2Cl2 at 298 K are
summarized in Table 3. All the complexes display very intense
Table 3. Absorption Data for the Six Iridium Complexes in
Dichloromethane Solution at 298 3 K
complex
absorption λmax / nm (ε/mol−1 dm3 cm−1
)
Ir(dpyx)(ppy)
Cl 1a
264 (35900), 288 (34900), 369 (7360), 397 (10100),
418 (11300), 457 (3330), 494 (1300)
Ir(dpyx)
(dfppy)Cl 1b
286 (36300), 363 (8310), 390 (9650), 412 (10800),
451 (3260), 488 (1200)
Ir(F2dpyx)
(ppy)Cl 2a
260 (44800), 280 (40300), 358 (9100), 381 (11300),
402 (14000), 438 (3640), 473 (1540)
Ir(F2dpyx)
(dfppy)Cl 2b
278 (42200), 351 (9660), 375 (11700), 393 (13600),
433 (2860), 467 (1520)
Ir(tfdpyb)(ppy) 259 (45600), 272 (45300), 373 (8940), 429 (11100),
Cl 3a
461 (8020), 506 (3080), 541 (2470)
Ir(tfdpyb)
(dfppy)Cl 3b
254 (43800), 266 (42900), 364 (7990), 418 (9920),
451 (6880), 492 (3040), 523 (2200)
bands at λ < 300 nm, as anticipated for π−π* transitions within
the aromatic ligands. There are moderately intense bands
around 400 nm (ε = 5000−10000 M−1 cm−1) with no
counterpart in the free ligands, together with some weaker
bands (ε ∼ 1000 M−1 cm−1) at longer wavelength. The λ > 300
nm region for the six complexes is shown in Figure 4.
Inspection of the figure reveals that the six complexes
essentially fall into three pairs, 1a,b; 2a,b; 3a,b. For each
pair, the a and b (ppy and dfppy) derivatives are similar to one
another, but the bands for the complex of the fluorinated dfppy
ligand are blue-shifted by around 500 cm−1 relative to its ppy
Figure 4. UV−visible absorption spectra of the iridium(III) complexes
in solution in CH2Cl2 at 298 3 K. The following key is used here
and in Figures 6 and 7: green = 1a,b, blue = 2a,b, red = 3a,b; solid line
= complexes with ppy, dashed line = complexes with dfppy.
3818
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
a
Table 4. TD-DFT B3LYP Excitation Energies for the Two Lowest-Lying Triplet States and Five Lowest-Lying Singlet States, ,
b
and Their Oscillator Strengths, f
1a
1b
2a
2b
3a
3b
T1
T2
S1
2.541 (487)
H−1→L
2.566 (483)
2.598 (477)
H−1→L
2.627 (472)
2.604 (476)
H→L
2.679 (463)
H→L
2.109 (588)
H−1→L
2.145 (578)
2.172 (571)
H−1→L
2.261 (548)
2.693 (460)
H→L+1
2.766 (448)
H→L
H→L
H−1→L
H→L
H→L
2.805 (442)
f = 0.015
H→L
2.885 (430)
f = 0.027
H→L
2.862 (433)
f = 0.0003
H→L+1
2.957 (419)
f = 0.004
H→L+1
2.322 (534)
f = 0.018
H→L
2.438 (509)
f = 0.017
H→L
S2
S3
S4
S5
2.859 (433)
f = 0.026
H−1→L
2.935 (422)
f = 0.015
H→L+1
2.959 (419)
f = 0.020
H−1→L
2.979 (416)
f = 0.0002
H→L+1
2.873 (432)
f = 0.019
H→L
2.985 (415)
f = 0.017
H→L
2.416 (513)
f = 0.040
H−1→L
2.930 (423)
f = 0.028
H→L+1
2.479 (500)
f = 0.042
H−1→L
3.018 (411)
f = 0.006
H−2→L
3.026 (410)
f = 0.021
H→L+1
3.010 (412)
f = 0.100
H−1→L
3.034 (409)
f = 0.015
H→L+2
3.083 (402)
f = 0.087
H−1→L
3.112 (398)
f = 0.014
H→L+2
2.954 (420)
f = 0.011
H−1→L+1
2.996 (414)
f = 0.096
H−1→L+2
3.024 (410)
f = 0.012
H−1→L+1
3.069 (404)
f = 0.093
H−1→L+2
2.942 (421)
f = 0.005
H−2→L
3.075 (403)
f = 0.019
H→L+2
3.119 (397)
f = 0.005
3.165 (392)
f = 0.009
H−1→L+2
3.130 (396)
f = 0.012
H−1→L+1
H−1→L+1
b
a
In units of eV (with corresponding wavelength in parentheses). The predominant nature of the excitation is also indicated (H = HOMO; L =
LUMO).
studies.27,31,32 Similarly, the absorption and emission bands of
Pt(F2dpyb)Cl have been found to be blue-shifted relative to
those of Pt(dpyb)Cl.33
6. Luminescence in Solution at Room Temperature
and in Dilute Glass at 77 K. All six complexes are intensely
photoluminescent in degassed dichloromethane solution at
ambient temperature. Data are compiled in Table 5, and the
emission spectra at 298 K are shown in Figure 6. The emission
color varies over a broad range from blue to orange/red (λmax
from 476 to 593 nm). The order of emission energies is 2b >
2a > 1b > 1a > 3b > 3a, mirroring the trend in absorption
discussed above. In each case, the spectra of the dfppy (b)
complexes resemble those of their ppy (a) analogues, but are
consistently blue-shifted by around 500 cm−1. This corresponds
closely to the trend in the absorption spectra, specifically with
regard to the lowest-energy absorption band which we assigned
to the S0→T1 transition. In the case of 1a,b and 2a,b in
particular, the Stokes shift between this lowest-energy
absorption band and the 0−0 vibrational emission band, is
very small, of the order of 400 cm−1, suggesting that the lowest-
energy absorption band and the emission involve the same
excited triplet state.
The quantum yields of emission in degassed solution are
high, in the range 0.2−0.7. The highest values are observed for
the green-emitting complexes of the dpyx ligand (Φlum = 0.71
and 0.66 for 1a and 1b respectively), which display
luminescence lifetimes of around 1.5 μs. The quantum yields
are marginally smaller for the orange/red-emitting complexes of
tfdpyb (Φlum = 0.46 and 0.57 for 3a and 3b, respectively),
although their lifetimes are longer (3.3 and 4.0 μs, respectively).
On the other hand, the quantum yields and lifetimes are
significantly reduced for the blue-emitting complexes of
F2dpyb, 2a and 2b (Φlum = 0.20 in each case).
Figure 5. Schematic representation of the variation in the energies of
the orbitals involved in the lowest-energy spin-allowed (S0 → S1)
transitions, based on the results of TD-DFT calculations. The thick
lines refer to the a series (1a, 2a, 3a), and the thin lines to the b series.
All orbitals are stabilized by electron-withdrawing F and CF3
substituents but, in the former case, the LUMO is stabilized to a
lesser extent than the HOMO. The change from ppy (a series) to
dfppy (b series) stabilizes the HOMO more than the LUMO in each
case.
reasoning accounts for the slightly higher energy of the bands
in the spectra of the dfppy (b) complexes compared to their
ppy (a) analogues (thin versus thick lines in Figure 5). Indeed,
the blue-shifting influence of dfppy in tris-bidentate complexes
of iridium is already well established by numerous other
Some insight into the origin of these trends may be obtained
by considering the rate constants for radiative (kr) and
nonradiative (∑knr) decay, which can be estimated from the
quantum yields and lifetimes: kr = Φlum/τ and ∑knr = (1/τ) −
3819
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Table 5. Luminescence Data in Solution in Dichloromethane at 298 3 K and in EPA at 77 K
Article
a
77 K
λmax/nm
c
d
d
e
τ degassed (aerated)
Φlum degassed
kr /
∑knr
/
kQ /
b
complex
λmax/nm
/ns
(aerated)
104 s−1
104 s−1
108 M−1 s−1
τ/μs
Ir(dpyx)(ppy)Cl 1a
Ir(dpyx)(dfppy)Cl 1b
Ir(F2dpyx)(ppy)Cl 2a
503, 537
495, 529
487, 516
1400 (110)
1500 (160)
390 (120)
0.71 (0.06)
0.66 (0.07)
0.20 (0.06)
51
44
51
20
38
25
26
489, 529, 564, 613
481, 519, 553, 604
3.7
2.6
2.5
23
210
471, 508, 537, 548,
593
Ir(F2dpyx)(dfppy)Cl 2b
476, 508
410 (120)
0.20 (0.06)
49
190
27
464, 499, 528, 538,
578
2.0
Ir(tfdpyb)(ppy)Cl 3a
593
572
3300 (1100)
0.46 (0.14)
0.57 (0.24)
14
14
16
2.8
1.5
576, 621
41
46
Ir(tfdpyb)(dfppy)Cl 3b
4000 (1700)
11
563, 607, 668(sh)
a
b
c
d
EPA = diethyl ether/isopentane/ethanol (2:2:1 v/v). Lifetime of emission; λex = 374 nm. Luminescence quantum yield. Radiative (kr) and
e
nonradiative (∑knr) rate constants calculated from τ and ϕ values; estimated uncertainty 20%. Bimolecular rate constant for quenching by O2,
estimated from τ values in degassed and aerated solutions.
Figure 6. Normalized photoluminescence spectra of the iridium(III)
complexes in degassed dichloromethane solution at 298 3 K.
Figure 7. Normalized photoluminescence spectra of the iridium(III)
complexes in a frozen glass of composition diethyl ether/isopentane/
ethanol (2:2:1 v/v) at 77 K.
kr. The values thus obtained are summarized in Table 5. They
indicate that the reduced quantum yields and shorter lifetimes
of the blue-emitting complexes 2a and 2b are due to much
greater nonradiative decay, ∑knr being increased by an order of
magnitude compared to 1a and 1b. The tendency for blue-
emitting transition metal complexes to be subject to enhanced
nonradiative decay pathways compared to those that emit in
the green region is one that has been encountered in numerous
other studies with various metal ions including iridium(III) and
platinum(II).34,35 It is typically attributed to a decreased gap
between the energy of the emissive excited state and higher-
lying, deactivating d-d states, such that it becomes comparable
to kT at ambient temperature,36 or to enhanced coupling with
vibrational modes of the complex.37 In the present instance, the
photoinduced cleavage of the Ir−Cl bond mentioned earlier is
likely to be the limiting factor. This process presumably
proceeds via population of the d-d state, accounting for the
observation that it is most significant for the pair of blue-
emitting complexes, where the energy gap between the emissive
state and the d-d state is at its smallest. Despite this effect,
however, the efficiencies remain quite competitive with some of
the best blue phosphors reported to date. Upon cooling to 77 K
(Figure 7), the lifetimes of 2a and 2b become comparable to
those of 1a and 1b; at this temperature, the above deactivating
pathways are eliminated.
The increase in the luminescence lifetimes of the red-
emitting complexes 3a and 3b is intriguing. The data reveal that
kr is significantly decreased in these complexes, while ∑knr is
also slightly reduced, contrary to what might be expected if the
energy gap law were to hold for a series of complexes with the
same type of excited state.38 These observations suggest that
the nature of the emissive state in 3a and 3b is rather different
from 1a,b and 2a,b. Moreover, at 77 K (Figure 7), the
luminescence lifetime of 3a and 3b increases by an order of
magnitude to >40 μs. Clearly, this increase cannot simply be
due to reduced nonradiative decay, as the room temperature
quantum yields were already around 0.5. The nature of the
emissive state must be different under these conditions; the
long lifetimes under these conditions point to an excited triplet
state with a relatively poor SOC and thus perhaps of largely
ligand π−π* character, with relatively little metal character. At
room temperature, this state probably becomes heavily mixed
with one having greater MLCT character, promoting the
radiative decay rate. Charge-transfer states are typically more
likely to be destabilized by decreasing temperature and
rigidochromic effects than ligand-localized states, as noted
previously in other classes of Ir(III) complexes.39 Related
temperature-dependent behavior of iridium(III) complexes has
been reported recently by others.40,41
Complexes 1a,b and 2a,b are moderately quenched by
dissolved molecular oxygen in solution (lifetimes and quantum
3820
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
yields in air-equilibrated solution are given in parentheses in
Table 5). The bimolecular rate constants of quenching, kQ, for
these four complexes are similar to one another, of the order of
3 × 109 M−1 s−1 (Table 5), although the longer excited-state
lifetimes of 1a,b compared to their analogues 2a,b lead to a
more pronounced quenching effect for the former pair. On the
other hand, the kQ values for 3a,b are an order of magnitude
lower: these two complexes retain quite high luminescence
quantum yields and lifetimes in excess of 1 μs even in air-
equilibrated solution. A correlation between reduced oxygen
sensitivity and the presence of CF3 substituents has similarly
been observed recently in related platinum(II) complexes of
dipyridylbenzene ligands, and may be associated with a less
strongly reducing excited state diminishing the rate of Dexter
42
3
energy transfer to O2.
7. Photoluminescence in PMMA Matrix. The high
solution luminescence quantum yields, relatively short triplet
lifetimes, and color tunability across a wide portion of the
visible spectrum suggest that this new family of iridium(III)
complexes should be of interest for testing as possible triplet-
harvesting phosphors in OLEDs. To draw useful inferences as
to how these complexes will behave in an OLED emitting layer,
the luminescence properties of the complexes doped into a
poly(methylmethacrylate) (PMMA) matrix have been studied.
Films of composition 0.05 to 5% of Ir complex to PMMA by
weight were investigated to probe concentration-dependent
self-quenching effects. The photoluminescence emission wave-
lengths, quantum yields, and lifetimes in the PMMA films at the
two limiting concentrations are reported in Table 6. All
Figure 8. Architecture of the OLEDs produced, including molecular
structures of the materials used and an energy level diagram for
understanding the exciton and electronic traffic within the devices. The
positions of the levels are indicated by the numbers in eV relative to
the vacuum level at energy zero. The Fermi level positions for ITO
and Al electrode contacts are added for completion.
details of the materials and equipment used are given in the
Experimental Section. All of the OLEDs produced gave high
efficiency electroluminescence (EL) with colors ranging from
blue to red. Their EL spectra are shown in Figure 9. Practically
no contribution to the EL emission from the TCTA binder, nor
the TPD hole-transport layer, nor the TAZ electron-transport
layers is observed. The EL spectra closely match the PL spectra
Table 6. Photoluminescence of the Six Iridium Complexes in
a
PMMA Film
% complex by
complex
weight
λem/nm
ΦPL
τ/μs
Ir(dpyx)(ppy)Cl 1a
0.05
5
500, 535
500, 535
488, 521
488, 521
481, 513
481, 513
469, 501
469, 501
586
0.95
0.95
0.65
0.35
0.80
0.80
0.65
0.65
0.80
0.80
0.80
0.80
1.1
1.0
Ir(dpyx)(dfppy)Cl 1b
Ir(F2dpyb)(ppy)Cl 2a
Ir(F2dpyb)(dfppy)Cl 2b
Ir(tfdpyb)(ppy)Cl 3a
Ir(tfdpyb)(dfppy)Cl 3b
0.05
5
1.40
0.90
0.80
0.80
0.85
0.85
2.4
0.05
5
0.05
5
0.05
5
586
2.4
0.05
5
566
2.5
566
2.5
a
Estimated uncertainty in ΦPL is 20%; uncertainty in τ is 0.2 μs.
complexes are seen to preserve or even increase their quantum
yields in PMMA with respect to the solution values, even in the
more highly concentrated (5% by weight) samples, which
correspond to the doping levels typically required in the
emitting layer of a phosphorescent OLED. The high quantum
yields in PMMA, reaching almost 100% for compound 1a,
together with the relatively short lifetimes under these
conditions, clearly make these complexes good candidates as
OLED phosphors.
8. Electroluminescence and OLED Fabrication. Each of
the six complexes, blended with TCTA [4,4′,4″-tris(N-
carbazolyl)triphenylamine] at 5% by weight, was investigated
as the emitting layer in a multilayer OLED structure. The
device architecture employed is shown in Figure 8, and further
Figure 9. EL spectra of the OLEDs with 5 wt % complex-doped
TCTA emitters: (a) 1a−3a with ppy as the bidentate ligand; (b) 1b−
3b with dfppy as the bidentate ligand. The insets show the CIE
chromaticity coordinates of the devices.
3821
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
a
Table 7. OLED Performance Data
b
c
d
e
f
Ir complex in TCTA
bias (V)
j (mA/cm2)
QE (%ph/e)
PE (lm/W)
LE (cd/A)
CIE (x,y)
5% Ir(dpyx)(ppy)Cl (1a)
100% Ir(dpyx)(ppy)Cl (1a)
5% Ir(dpyx)(dfppy)Cl (1b)
5% Ir(F2dpyb)(ppy)Cl (2a)
5% Ir(F2dpyb)(dfppy)Cl (2b)
5% Ir(tfdpyb)(ppy)Cl (3a)
10.0
16.8
13
1.4
16.4
5.9
2.5
4.8
2.7
1.7
10.9
1.3
3.0
8.0
5.4
8.7
9.8
11.3
0.6
2.0
5.4
2.3
4.2
8.4
36.0
3.2
0.26,0.64
0.45,0.52
0.21,0.53
0.20,0.50
0.17,0.34
0.58,0.41
8.2
11.8
14.2
13.8
11
20.2
10.5
18.6
29.3
5% Ir(tfdpyb)(dfppy)Cl (3b)
0.51,0.49
a
b
c
Devices are listed according to the Ir complex in the emitting layer (data have been taken at ∼500 cd/m2). Compound # from Figure 1. QE =
d
e
f
external EL quantum efficiency. PE = power efficiency. LE = luminous efficiency. Commission Internationale d’Eclairage (CIE) coordinates.
of the corresponding complexes in PMMA. The absence of the
TCTA, TPD, and TAZ emissions indicates that the excitons are
confined to the EML, where they are localized on the
phosphorescent dopant molecules prior to radiative emission.
The CIE (Commission Internationale d’Eclairage) coordinates
fall between the blue and red television standards (Table 7 and
insets to Figure 9).
The performance data of the OLEDs are summarized in
Table 7 and in Figures 10 and 11. The luminance (L) and the
Figure 11. Luminance versus driving current (a) and external EL QE
versus luminance (b) plots of the OLEDs. Key is as indicated in
caption to Figure 10.
(≥10 mA/cm2). Meanwhile, the blue- and green-emitting
OLEDs prepared with complexes 1a, 1b, 2a, and 2b, exhibit
external EL QEs between 4 and 10% at j ∼ 0.01 A/cm2 (the
luminance at this current level ranges from 500 cd/m2 for the
2a device to 1000 cd/m2 for the 1a device). These values are
comparable to those reported for OLEDs based on the
archetypal green-emitting tris-bidentate Ir(ppy)3 complex44 and
tetrazolate analogues in devices of an almost identical
architecture to that reported in Figure 9.45 All of the OLEDs
prepared have turn-on voltages around 3 V corresponding to
Figure 10. Plots showing luminance versus applied voltage (a) and
external EL QE versus driving current (b) for the OLEDs based on the
EMLs made of TCTA doped with 5 wt % of the iridium complexes.
Data for the OLED incorporating a neat film of compound 1a (100%)
current densities of ∼10−4−10−3 mA/cm2. Devices incorporat-
ing a neat film of complex as the emitting layer displayed a
pronounced lowering in EL QE across the entire current
density range compared with doped emitter devices (see, for
example, the data for 1a at 100% shown in Figures 10 and 11).
The trend is likely to be due to the formation of nonemissive
aggregates in the solid state. This behavior contrasts with that
observed for devices incorporating neat films of related
platinum(II) complexes with similar N∧C∧N-coordinating
ligands, wherein electroluminescence arising from quite
intensely emissive aggregates/excimers is observed at low
energy, in the red or near-infrared region of the spectrum.46
as the emitting layer are shown for comparison. Green squares: 1a,b;
blue triangles: 2a,b; red diamonds: 3a,b; black triangles: 1a doped at
100%.
external EL quantum efficiency (EL QE) as functions of driving
voltage (V) and current density (j) are displayed in Figure 11.
The maximum luminance reaches 10,000 cd/m2 at j ≈ 100
mA/cm2 for the red emitting complex 3b. The maximum
external quantum efficiency for the red devices (3a, 3b) (ELQE
≥ 10%) significantly exceeds previously reported efficiencies of
up to 7% for devices using the red iridium phosphor
Ir(btp)2(acac) (btp = 2,2′-benzothienylpyridine);43 indeed,
the difference exceeds a factor of 4 at high current densities
CONCLUSIONS
■
Terdentate followed by bidentate cyclometallating ligands can
be readily introduced into the coordination sphere of
3822
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
NaHCO3 solution (10%, 50 mL), and the resulting mixture extracted
with dichloromethane (2 × 100 mL). The organic phase was dried
over MgSO4, and the solvent was removed under reduced pressure to
leave a pale brown residue. Purification by column chromatography
(silica, hexane/ether gradient elution from 100/0 to 25/75) led to the
desired product, 1,3-di(2-pyridyl)-4,6-difluorobenzene, as a pale yellow
solid (0.57 g, 53%). Rf = 0.55 (silica, hexane/ether 25/75). Mp = 156−
iridium(III) in a stepwise manner. This contrasts with the
classical procedures for preparing tris-bidentate Ir(III) com-
plexes, where two ligands are introduced simultaneously. The
results reveal that substituents in the N∧C∧N-terdentate ligands
can have a significant influence on the excited state energies,
while the bidentate ligands independently offer additional fine-
tuning. The trends can be rationalized with the aid of TD-DFT,
which shows that the terdentate ligand contributes to both the
highest occupied and lowest unoccupied orbitals, whereas the
bidentate ligand plays little if any role in the latter.
1
3
4
158 °C. H NMR (CDCl3, 500 MHz): δ 8.72 (2H, dd, J = 5.0, J =
1.0, H6), 8.62 (1H, t, 4JH−F = 9.0, H2′), 7.76 (4H, m, H3 and H4), 7.26
(2H, m, H5), 7.03 (1H, t, 3JH−F = 10.5, H5′). 13C NMR (CDCl3, 125.7
1
3
MHz): δ 160.7 (dd, JC−F = 255, JC−F = 12, C4′), 152.8 (C2), 150.1
The complexes of the form Ir(N∧C∧N)(N∧C)Cl are highly
luminescent in solution at room temperature, with quantum
yields comparable to many of the best iridium-based emitters
reported to date across the blue to orange/red region. The
complexes have been successfully employed as triplet-state
emitters in the fabrication of highly efficient OLEDs. The
performance of these devices is at least comparable, and in
some cases superior, to that obtained with tris-bidentate Ir(III)
complexes, while offering a wider diversity of substitution and
extended scope for manipulation of their emissive properties.
Further possibilities are anticipated through replacement of the
chloride coligand by other monodentate ligands such as
acetylides. Such strategies could also open up the utility of
this type of complex for facile incorporation into multimetallic
assemblies offering well-defined excited-state energy gradients,
relevant to energy-harvesting applications.47
(C6), 136.7 (C4), 133.9 (t, JC−F = 4.5, C2′), 124.7 (dd, JC−F = 10,
4JC−F = 5.5, C1′), 124.5 (dd, 4JC−F = 6JC−F = 9.0, C3), 122.8 (C5), 105.2
3
2
(t, 2JC−F = 27, C5′). 19F NMR (CDCl3, 188.2 MHz): δ −113.3 (t, 3JF−H
= 10). HRMS (ES+) m/z = 269.0886 [M+H]+; calcd for C16H11F2N2
= 269.0890. Anal. Calcd for C16H10F2N2 = C, 71.63; H, 3.76; N,
10.44%. Found C, 71.52; H, 3.81; N, 10.43%.
1,3-Di(2-pyridyl)-4,6-bis(trifluoromethyl)benzene
(tfdpybH). Step 1: 2,6-Dibromo-3,5-bis(trifluoromethyl)aniline.
Dichloromethane (100 mL) was added to a mixture of
bis(trifluoromethyl)aniline (1.0 g, 4.4 mmol), Na2CO3 (467
mg, 5.6 mmol) and iron filings (128 mg, 2.3 mmol) in a 250
mL round-bottom flask. A solution of bromine (1 mL) in
dichloromethane was added with stirring, and the mixture then
heated at reflux temperature for 3 d with stirring. Upon cooling,
saturated Na2CO3(aq) (150 mL) was added, the mixture was
extracted with diethyl ether (3 × 100 mL), the combined
organic phase dried over anhydrous MgSO4, and the solvent
removed under reduced pressure. The crude product was
purified by chromatography on silica (100% hexane to 80%
hexane/20% ether), to give a yellow solid (1.30 g, 77%). NMR
(CDCl3, 400 MHz): δ 7.37 (1H, s, H4), 5.24 (2H, s, NH2).
Step 2: 1,3-Dibromo-4,6-bis(trifluoromethyl)benzene. 2,6-Dibro-
mo-3,5-bis(trifluoromethyl)aniline (1.30 g, 3.4 mmol) from step (i)
was dissolved in hot ethanol (20 mL) and concentrated sulfuric acid
(2.55 mL, 95%) was added to the boiling solution. The mixture was
then cooled to 0 °C, and sodium nitrite (2.56 g, 0.037 mol) was added.
The mixture was allowed to warm to room temperature and stirring
continued for 15 min. It was then refluxed until gas evolution ceased.
The solvent was removed under reduced pressure, and the crude
product was purified by chromatography on silica using hexane as the
eluant to give the product as a colorless solid (460 mg, 37%). 1H NMR
(CDCl3, 400 MHz): δ 8.11 (1H, s, H5), 7.95 (1H, s, H2).
EXPERIMENTAL SECTION
■
1H and 13C NMR spectra, including NOESY and COSY, were
recorded on a Varian 500 MHz instrument. Chemical shifts (δ) are in
parts per million (ppm), referenced to residual protio-solvent
resonances, and coupling constants are in hertz (Hz). Electrospray
ionization mass spectra were acquired on a time-of-flight Micromass
LCT spectrometer. All solvents used in preparative work were at least
Analar grade, and water was purified using the Purite system. Solvents
used for optical spectroscopy were HPLC grade. 1,3-Di(2-pyridyl)-4,6-
dimethylbenzene (dpyxH) and Ir(dpyx)(ppy)Cl (1a) were prepared
as reported previously.15,16 2-(2,4-Difluorophenyl)pyridine (dfppyH)
was prepared from 2,4-difluorobenzene-boronic acid and 2-bromopyr-
idine by standard Suzuki cross-coupling methods.
Synthesis of Ligands. 1,3-Di(2-pyridyl)-4,6-difluoroben-
zene (F2dpybH). Step 1: 1,3-Dibromo-4,6-difluorobenzene. 2,4-
Difluorobromobenzene (7.53 g, 39 mmol) and iron filings (100
mg) were placed in a two-necked round-bottomed flask
equipped with an addition funnel and a condenser holding a
drying tube filled with KOH pellets. The flask was heated to 60
°C. Bromine (9.33 g, 59 mmol) was added dropwise via the
funnel over 1 h, and the mixture then allowed to stir for an
additional 1.5 h. The solution was poured into aqueous NaOH
(10%, 30 mL) with stirring. The organic layer was collected,
and the aqueous layer was extracted twice with toluene. The
combined organic layers were combined and dried with
anhydrous Na2SO4. The solvent was removed under reduced
pressure to give the product as a light cream solid (7.24 g,
68%). 1H NMR (CDCl3, 300 MHz): δ 7.75 (1H, td, 3J = 7.0, 5J
Step 3: 1,3-Di(2-pyridyl)-4,6-bis(trifluoromethyl)benzene
(tfdpybH). Toluene (25 mL) was added to a mixture of 2-tri-n-
butyl-stannyl-pyridine (2.2 g, 6.0 mmol), 2,6-dibromo-3,5-bis-
(trifluoromethyl)benzene (840 mg, 2.3 mmol), Pd(PPh3)2Cl2 (200
mg, 0.29 mmol), and lithium chloride (1.2 g, 28 mmol). The mixture
was degassed via five freeze−pump−thaw cycles and then heated at
reflux under an atmosphere of nitrogen for 24 h. After cooling to room
temperature, the mixture was added to a saturated aqueous solution of
KF (20 mL) and stirred for 30 min. The solution was filtered, and the
solid residue washed with water and toluene. The combined filtrate
and washings was added to a solution of NaHCO3 (100 mL, 10% w/
w) and extracted into dichloromethane (2 × 100 mL). The organic
phase was dried over MgSO4 and the solvent then removed to yield a
brown residue, which was purified by silica gel separation
chromatography (gradient elution from 100% hexane to 20%
hexane/80% diethyl ether), leading to the product as a pale yellow
solid (210 mg, 25%). 1H NMR (CDCl3, 700 MHz): δ 8.70 (2H, dd, 3J
= 5.0, 4J = 1.5, H6), 8.17 (1H, s, 8.17, H5′), 7.78 (2H, ddd, 3J = 7.0, 7.0,
1.5, H4), 7.71 (1H. s, H2′), 7.49 (2H, d, 3J = 7.5, H3), 7.35 (2H, ddd, 3J
= 7.5, 5.0, 4J = 1.5, H5). 13C NMR (CDCl3, 175.9 MHz): δ 156.0 (C2),
149.5 (C6), 143.2 (C1′), 136.2 (C4), 135.3 (C2′), 128.4 (q, 2JC−F = 31,
C4′), 125.0 (septet, 3JC−F = 5, C5′), 124.0 (s, C3), 123.3 (q, 1JC−F = 274,
CF3), 123.2 (s, C5). 19F NMR (CDCl3, 282.2 MHz): δ −57.5 (s). MS
(ES+): m/z 369 [M+H]+. HRMS (ES+): m/z 369.0818. Calcd for
C18H11N2F6: m/z 369.0826. Elemental analysis: C, 57.7; H, 3.9; N,
6.2%. Calcd for C18H10N2F6: C, 58.7; H, 2.7; N, 7.6%.
4
5
= 2.0, H2), 6.97 (1H, td, J = 8.0, J = 2.0, H5). 19F NMR
3
4
(CDCl3, 400 MHz): δ −103.8 (dd, JF−H = 7.5, JF−H = 7.5).
Step 2: 1,3-Di(2-pyridyl)-4,6-difluorobenzene (F2dpybH). 2-
(Tributyltin)pyridine (4.17 g, 85% purity by H NMR, 9.62 mmol),
1
1,3-dibromo-4,6-difluorobenzene (1.09 g, 4.01 mmol), Pd(PPh3)2Cl2
(0.104 g, 0.15 mmol), LiCl (1.24 g, 29.3 mmol), and toluene (6 mL)
were placed in a Schlenk tube. The mixture was degassed by four
freeze−pump−thaw cycles, then heated under a nitrogen atmosphere
to 120 °C for 48 h. After cooling to room temperature, saturated
aqueous KF (10 mL) was added, and the mixture stirred for 30 min.
The precipitated solid was removed by filtration and washed with
water (25 mL) and toluene (30 mL). The filtrate was added to
3823
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
3
4
3
4
Synthesis of Chloro-Bridged Dimers. [Ir(F2dpyb)Cl(μ-Cl)]2.
A suspension of F2dpybH (250 mg, 0.93 mmol) and
IrCl3·3H2O (327 mg, 0.93 mmol) in a mixture of 2-
ethoxyethanol (14 mL) and water (6 mL) was refluxed for
24 h under a nitrogen atmosphere. After cooling to room
temperature, the yellow solid that formed was separated by
centrifugation, washed successively with water, ethanol and
diethyl ether (3 × 5 mL of each), and finally dried under
vacuum. This solid (383 mg, 78%) has low solubility in all
common solvents, and was used in subsequent reactions
without further purification.
13, JH−F = 9.0, J = 2.5, H4′-NC), 5.44 (1H, dd, JH−F = 9.0, J = 2.5,
H6′-NC). 19F NMR (CDCl3, 376.3 MHz): δ −107.7 (2F, d, 3JF−H = 12,
NCN), −107.9 (1F, ddd appears as q, JF−H = JF−F = 9, F5′-NC),
−109.4 (1F, dd, appears as t, net J = 11, F3′-NC). MS (ES+): m/z 683
[M]+, 648 [M−Cl]+. HRMS (ES+): m/z 689.1073 [M−Cl
+CH3CN]+. Calcd for C29H18N4F4191Ir: m/z 689.1077.
3
4
[Ir(tfdpyb)(ppy)Cl] 3a. A mixture of [Ir(tfdpyb)Cl(μ-Cl)]2 (30
mg, 0.028 mmol), ppyH (0.25 mL, 270 mg, 1.74 mmol) and AgOTf
(30 mg, 0.12 mmol) in toluene (4 mL) was heated at reflux under an
atmosphere of nitrogen for 24 h. After cooling to room temperature,
the precipitated AgCl was removed by centrifuge and washed with
acetonitrile. The filtrate and washings were combined, and the solvent
removed under reduced pressure. The residue was taken up into
CH2Cl2 (15 mL), and the solution washed with HCl(aq) (1M, 3 × 10
mL), followed by water (10 mL). After drying over MgSO4, the
solvent was removed under reduced pressure, and the residue purified
by chromatography (silica, CH2Cl2/MeOH, gradient elution from
100/0 to 95/5), to a give the product as an orange-red solid (19 mg,
[Ir(tfdpyb)Cl(μ-Cl)]2. A suspension of tfdpybH (155 mg, 0.42
mmol) and IrCl3·3H2O (145 mg, 0.42 mmol) in a mixture of 2-
ethoxyethanol (14 mL) and water (6 mL) was refluxed for 24 h under
a nitrogen atmosphere. The solvent was removed under reduced
pressure, to yield the product as an orange solid which was used in
1
subsequent reactions without further purification (173 mg, 77%). H
NMR (CD3CN, 200 MHz): δ 9.25 (2H, d, 3J = 5.0, H6), 8.48 (2H, dd,
J = 8.0, H3), 8.08 (2H, m, H4), 7.97 (1H, s, H4′), 7.64 (1H, m, H5). 19
NMR (CD3CN, 188 MHz): δ −60.1 (s).
F
1
3
4
91%). H NMR (CDCl3, 500 MHz): δ 10.12 (1H, dd, J = 5.5, J =
1.0, H6−NC), 8.39 (2H, d, J = 8.0, H3−NCN), 8.11 (1H, H3−NC),
3
8.06 (1H, s, H4′-NCN), 8.04 (1H, td, 3J = 8.0, 4J = 1.0, H4−NC), 7.83
(2H, dd, 3J = 6.0, 4J = 1.0, H6−NCN), 7.71 (2H, ddd, 3J = 8.0, 8.0, 4J =
1.0, H4−NCN), 7.61 (2H, m, H3′-NC and H5−NC), 6.99 (2H, ddd, 3J
Synthesis of Mononuclear Complexes. [Ir(dpyx)(dfppy)Cl]
1b. A mixture of [Ir(dpyx)Cl(μ-Cl)2] (50 mg, 0.048 mmol),
dfppyH (19 mg, 0.096 mmol), and silver triflate (57 mg, 0.22
mmol) in toluene (4 mL) was refluxed for 24 h under a
nitrogen atmosphere. The precipitated AgCl was removed by
centrifuge and washed with acetonitrile. The filtrate and
washings were combined, the solvent removed under reduced
pressure, and the residue purified by chromatography (silica gel,
CH2Cl2/MeOH, gradient elution from 100/0 to 98/2), leading
= 8.0, 5.5, J = 1.0, H5−NCN), 6.76 (1H, ddd, J = 7.5, 7.5, J = 1.0,
4
3
4
H4′-NC), (1H, ddd, J = 7.5, 7.5, J = 1.0, H5′-NC), (1H, dd, 3J = 5.5,
4J = 1.0, H6′-NC). 19F NMR (CDCl3, 376.4 MHz): δ −59.1. HRMS
(ES+): m/z 753.1198 [M−Cl+CH3 CN]+ . Calcd for
C31H20N4ClF6191Ir: m/z 753.1201.
3
4
[Ir(tfdpyb)(dfppy)Cl] 3b. This complex was prepared from
[Ir(tfdpyb)Cl(μ-Cl)]2 (42 mg, 0.033 mmol) and dfppyH (13 mg,
0.066 mmol) in the presence of silver triflate (35 mg, 0.14 mmol) in
toluene (4 mL), as described above for [Ir(dpyx)(dfppy)Cl]. The
crude product was purified by chromatography (silica gel, CH2Cl2/
MeOH, gradient elution from 100/0 to 99/1), giving an orange solid
(18 mg, 35%). 1H NMR (CDCl3, 700 MHz): δ 10.17 (1H, d, 3J = 5.5,
1
to the product as a yellow solid (61 mg, 94%). H NMR (700
3
4
MHz, CDCl3): δ 10.16 (1H, dd, J = 5.0, J = 1.0, H6−NC),
8.42 (1H, d, J = 8.5, H3−NC), 8.01 (2H, d, J = 8.5, H3−
3
3
NCN), 7.97 (1H, ddd, J = J ∼ 8.0, J = 1.0, H4−NC), 7.58−
7.52 (5H, m, H6−NCN, H4−NCN, H5−NC), 6.88 (1H, s, H4
),
6.78 (2H, ddd, 3J = 7.0, 3J = 6.0, 4J = 1.5, H5−NCN), 6.18 (1H,
3
3
4
′
ddd, JH−F = 13, JH−F = 9.0, J = 2.5, H4
′-NC), 5.50 (1H, dd,
3
3
4
3
3
4
H6−NC), 8.49 (1H, d, J = 8.5, H3−NC), 8.42 (2H, d, J = 8.0, H3−
3JH−F = 9.0, J = 2.5, H6
′
-NC), 2.82 (6H, s, CH3). 19F NMR
NC), 8.09 (1H, s, H4′-NCN), 8.07 (1H, dd, J = 8.0, J = 7.0, H4−
3 3
3
(CDCl3, 376.3 MHz): δ −109.7 (1F, ddd appears as q, JF−H
=
NC), 7.76 (4H, m, H4−NCN and H6−NCN), 7.62 (1H, dd, J = 7.0,
3
4JF−F = 10, F5 -NC), −110.5 (1F, dd, appears as t, net J = 12, F3
′ ′-
3J = 5.5, H5−NC), 7.04 (2H, dd, J = 7.0, J = 5.5, H5−NCN), 6.25
3
3
NC). MS (ES+): m/z 642, [M−Cl]+, 682 [M−Cl+CH3CN]+.
HRMS (ES+): m/z 642.1314. Calcd for C29H21N3F2191Ir: m/z
642.1337.
(1H, ddd, 3JH−F = 12, 3JH−F = 9.0, 4JH−H = 2.5, H4′-NC), 5.21 (1H, dd,
3JH−F = 9.0, JH−H = 2.5, H6′-NC). 19F NMR (CDCl3, 376.4 MHz):
4
3
4
−59.2 (6F, s, CF3), −108.0 (1F, ddd appears as q, JF−H = JF−F = 9,
[Ir(F2dpyb)(ppy)Cl] 2a. This complex was prepared from
[Ir(F2dpyb)Cl(μ-Cl)]2 (50 mg, 0.047 mmol) and ppyH (15 mg,
0.095 mmol) in the presence of AgOTf (50 mg, 0.18 mmol) in toluene
(4 mL), using the procedure described above for [Ir(dpyx)(dfppy)Cl].
The product was purified by chromatography (silica, CH2Cl2/MeOH,
gradient elution from 100/0 to 96/4), leading to the product as a
F5′-NC), −109.4 (1F, dd, appears as t, net J = 11, F3′-NC). H NMR
1
(CD3CN, 500 MHz): δ 10.02 (1H, d, 3J = 5.5, H6−NC), 8.49 (1H, d,
3J = 8.5, H3−NC), 8.43 (2H, d, 3J = 9.0, H3−NCN), 8.18 (1H, dd, 3J =
7.5, 3J = 7.5, H4−NC), 8.11 (1H, s, H4′-NCN), 7.91 (2H, ddd, 3J = 9.0,
3J = 9.0, J = 1.5, H4−NCN), 7.87 (2H, d, J = 5.5, H6−NCN), 7.75
4
3
1
yellow solid (54 mg, 88%). H NMR (CDCl33, 500 MHz): δ 10.12
(1H, dd appears as t, 3J = 6.5, H5−NC), 7.16 (2H, dd appears as t, 3J =
3
4
(1H, dd, J = 5.5, J = 1.0, H6), 8.10 (3H, m, J = 5.0, H3−NC and
H3−NCN), 8.01 (1H, ddd, 3J = 8.0, 3J = 8.0, 4J = 1.5, H4−NC), 7.64−
7.56 (6H, m, H4−NCN, H6−NCN, H3′-NC, H5−NC), 6.84 (2H, ddd,
6.5, H5−NCN), 6.35 (1H, ddd, JH−F = 13, JH−F = 9.0, JH−H = 2.5,
H4′-NC), 5.20 (1H, dd, 3JH−F = 9.0, 4JH−H = 2.5, H6′-NC). MS (ES+):
m/z 791 [M−Cl+CH3CN]+. HRMS (ES+): m/z 789.1042 [M−Cl
+CH3CN]+. Calcd for C31H18N4F8Ir: m/z 789.1013. Elemental
analysis: Found C, 43.3; H, 2.5; N, 4.61%. Calcd for
C29H15N3F8IrCl·H2O: C, 43.4; H, 2.1; N, 5.2%.
3
3
4
3
4
3
3J = 7.0, J = 6.0, J = 1.5, H5−NCN), 6.79 (1H, t, JH−F = 11, H4′-
NCN), 6.78 (1H, m, H4′-NC), 6.63 (1H, ddd, J = 7.5, J = 7.5, J =
3
3
4
1.0, H5′-NC), 6.07 (1H, dd, J = 7.5, J = 1.0, H6′-NC). 19F NMR
3
4
3
(CDCl3, 376 MHz): δ −109.0. (d, JF−H = 12). MS (ES+): m/z 655
Photophysical Measurements in Solution and in Low-
Temperature Glass. Absorption spectra were measured on a Biotek
Instruments XS spectrometer, using quartz cuvettes of 1 cm path
length. Steady-state luminescence spectra were measured using a Jobin
Yvon FluoroMax-2 spectrofluorimeter, fitted with a red-sensitive
Hamamatsu R928 photomultiplier tube; the spectra shown are
corrected for the wavelength dependence of the detector, and the
quoted emission maxima refer to the values after correction. Samples
for emission measurements were contained within quartz cuvettes of 1
cm path length modified with appropriate glassware to allow
connection to a high-vacuum line. Degassing was achieved via a
minimum of three freeze−pump−thaw cycles while connected to the
vacuum manifold; final vapor pressure at 77 K was <10−2 mbar, as
monitored using a Pirani gauge. Luminescence quantum yields were
determined by the method of continuous dilution, using [Ru(bpy)3]-
[M−Cl+CH3CN]+. HRMS (ES+): m/z 653.1267 [M−Cl+CH3CN]+.
Calcd for C29H20N4F2191Ir: m/z 653.1265.
[Ir(F2dpyb)(dfppy)Cl] 2b. Prepared from [Ir(F2dpyb)Cl(μ-Cl)]2
(50 mg, 0.047 mmol) and dfppyH (19 mg, 0.095 mmol) in the
presence of AgOTf (50 mg, 0.18 mmol) in toluene (4 mL), using the
procedure described above for [Ir(dpyx)(dfppy)Cl]. The product was
purified by chromatography (silica, CH2Cl2/MeOH, gradient elution
from 100/0 to 96/4), leading to the product as a yellow solid (25 mg,
1
3
39%). H NMR (CDCl3): δ (CDCl3, 700 MHz): δ 9.52 (1H, d, J =
5.0, H6−NC), 8.49 (1H, d, J = 7.5, H3−NC), 8.12 (2H, d, J = 7.5,
3
3
H3−NCN), 8.08 (1H, dd, J = 7.5, J = 7.0, H4−NC), 7.72 (2H, ddd,
3
3
3J = 7.5, J = 7.5, J = 1.5, H4−NCN), 7.68 (1H, m, H5−NC), 7.50
3
4
(2H, d, 3J = 5.0, H6−NCN), 6.94 (2H, ddd, 3J = 7.5, 3J = 5.0, 4J = 1.5,
H5−NCN), 6.84 (1H, t, 3JH−F = 12, H4′-NCN), 6.27 (1H, ddd, 3JH−F
=
3824
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
Cl2 in aqueous solution (Φ = 0.02848) as the standard, and cross-
checked against Pt(dpyb)Cl in degassed CH2Cl2 (Φ = 0.6023b);
estimated uncertainty in Φ is 20% or better.
and refined using a riding model. All geometric parameters were
calculated using the Olex2 software. Crystal data and structure
refinement parameters: Ir(dpyx)(dfppy)Cl (1b): C31H25Cl5F2IrN3; M
= 847.04; triclinic, a = 9.0990(7) Å, b = 11.7466(9) Å, c =
15.0771(12) Å, α = 102.711(1)°, β = 93.462(1)°, γ = 102.191(1)°; U
The luminescence lifetimes of the complexes were measured by
time-correlated single-photon counting (TCSPC), following excitation
at 374.0 nm with an EPL-375 pulsed diode laser. The emitted light was
detected at 90° using a Peltier-cooled R928 PMT after passage
through a monochromator. The estimated uncertainty in the quoted
lifetimes is 10% or better. The lifetimes of 3a and 3b at 77 K were
determined by multichannel scaling using the same detector following
excitation with a pulsed xenon lamp. Bimolecular rate constants for
quenching by molecular oxygen, kQ, were determined from the
lifetimes in degassed and air-equilibrated solution, taking the
concentration of oxygen in CH2Cl2 at 0.21 atm O2 to be 2.2 mmol
dm−3.49
= 1526.9(2) Å3; T = 120 K; space group P1 (no. 2); Z = 2; μ(Mo Kα)
̅
= 4.850 mm−1; 21361 reflections measured, 8050 unique (Rint
=
0.0527) which were used in all calculations. The final wR(F2) was
0.1113 (all data). Ir(tfdpyb)(ppy)Cl (3a): C33H27ClF6IrN3O; M =
823.26; triclinic, a = 8.860(2) Å, b = 12.017(3) Å, c = 14.947(4) Å, α =
85.021(4)°, β = 89.893(4)°, γ = 75.995(4)°; U = 1536.4(4) Å3; T =
120 K; space group P1 (no. 2); Z = 2; μ(Mo Kα) = 4.496 mm−1;
̅
20540 reflections measured, 7386 unique (Rint = 0.0373) which were
used in all calculations. The final wR(F2) was 0.0638 (all data).
Photoluminescence in PMMA Films. The samples in PMMA
matrix were prepared by drop casting dichloromethane solutions of Ir
complex and PMMA. Two films were prepared for each complex with
0.05% and 5% Ir complex-to-PMMA weight ratios. The films
thicknesses were about 500 μm for the 0.05% samples and about 5
μm for the 5% samples, their maximum absorption in the 350−450 nm
region being 0.2−0.3. Absorption and emission spectra of the films
were recorded using a Perkin-Elmer Lambda 950 UV/vis spectropho-
tometer and an Edinburgh FLS920 spectrofluorimeter, respectively.
The photoluminescence quantum yields were measured by a custom
integrating sphere system applied to the same spectrofluorimeter,
using the method of DeMello et al.50 Lifetimes were obtained with an
IBH 5000F time-correlated single-photon counting device.
Electroluminescence. OLEDs were prepared by growing a
sequence of thin layers on clean glass substrates precoated with a
150 nm-thick layer of indium tin oxide (ITO) with a sheet resistance
of 20 Ω per square. A 70 nm-thick hole-transporting layer consisting of
a TPD:PC blend was deposited on top of the ITO by spin-coating
from a 10 mg/mL dichloromethane solution at room temperature. All
remaining organic layers were deposited in succession by thermal
evaporation under vacuum of ∼10−6 hPa, followed by the cathode
layer consisting of 0.5 nm thick LiF and by a 100 nm thick Al cap (see
Figure 8). The current−voltage characteristics were measured with a
Keithley Source-Measure unit, model 236, under continuous operation
mode, while the light output power was measured with an EG&G
power meter and the EL spectra were acquired with a Stellarnet
spectroradiometer. All measurements were carried out at room
temperature under argon atmosphere and were reproduced for many
runs, excluding any irreversible chemical and morphological changes in
the devices. The performance of the emissive layer has been optimized
by locating the EML between exciton blocking layers of TCTA and
TAZ, the latter acting also as an electron-transporting and hole-
blocking layer.
ASSOCIATED CONTENT
■
S
* Supporting Information
Crystallographic data in CIF format. Table S1 with frontier
orbitals of the dfppy series of complexes 1b, 2b, 3b together
with the orbital energies and the proportion of electron density
on the metal and the three ligands. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: j.a.g.williams@durham.ac.uk (J.A.G.W.), cocchi@isof.
cnr.it (M.C.).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Durham University and EPSRC for Ph.D. student-
ships (to P.B. and R.J.G.) and E.U. COST D35 for facilitating
international co-operation between our laboratories.
REFERENCES
■
(1) (a) Baldo, M. A.; Thompson, M. E.; Forrest, S. R. Nature 2000,
403, 750. (b) Baldo, M. A.; Lamansky, S.; Burrows, P. E.; Thompson,
M. E.; Forrest, S. R. Appl. Phys. Lett. 1999, 75, 4. (c) Adachi, C.; Baldo,
M. A.; Forrest, S. R.; Thompson, M. E. Appl. Phys. Lett. 2000, 77, 904.
(d) Adachi, C.; Baldo, M. A.; Forrest, S. R.; Thompson, M. E. J. Appl.
Phys. 2001, 90, 5048.
(2) Highly Efficient OLEDs with Phosphorescent Materials; Yersin, H.,
Ed.; Wiley-VCH: Berlin, Germany, 2007.
DFT and TD-DFT Calculations. B3LYP density functional
calculations were performed using the Gaussian03 software package.
Double-ζ quality basis sets were employed for the ligands (6-31G) and
the Ir ion (LANL2DZ). The inner core electrons of Ir were replaced
with a relativistic effective core potential, leaving the outer core
[(5s)2(5p)6] electrons and the (5d)6 valence electrons of Ir(III). The
geometries were fully optimized without symmetry constraints.
Crystallography. The single crystal X-ray diffraction experiments
were carried out at 120 K, using graphite monochromated Mo Kα X-
radiation (λ = 0.71073 Å) on a Bruker SMART 1K CDD area detector
diffractometer, equipped with an Oxford Cryosystems N2 open-flow
cooling device. Series of narrow ω-scans (0.3°) were performed at
several φ-settings in such a way as to cover a sphere of data to a
maximum resolution between 0.70 and 0.77 Å. Cell parameters were
determined and refined using the Bruker SMART software,51 and raw
frame data were integrated using the Bruker SAINT program.52 The
structures were solved with olex2.solve53 using the charge flipping
method, and refined by full-matrix least-squares on F2 using
olex2.refine. Reflection intensities were corrected for absorption
using SADABS.54 All non-hydrogen atoms (with the exception of
disordered solvent atoms) were refined with anisotropic displacement
parameters, and the hydrogen atoms were positioned geometrically
(3) Selected examples: (a) Grushin, V. V.; Herron, N.; LeCloux, D.
D.; Marshall, W. J.; Petrov, V. A.; Wang, Y. Chem. Commun. 2001,
1494. (b) Tsuboyama, A.; Iwawaki, H.; Furugori, M.; Mukaide, T.;
Kamatani, J.; Igawa, S.; Moriyama, T.; Miura, S.; Takiguchi, T.; Okada,
S.; Hoshino, M.; Ueno, K. J. Am. Chem. Soc. 2003, 125, 12971.
(c) Ragni, R.; Plummer, E. A.; Brunner, K.; Hofstraat, J. W.; Babudri,
F.; Farinola, G. M.; Naso, F.; De Cola, L. J. Mater. Chem. 2006, 16,
1161. (d) Chou, P.-T.; Chi, Y. Chem.Eur. J. 2007, 13, 380.
(e) Baranoff, E.; Bolink, H. J.; De Angelis, F.; Fantacci, S.; Di Censo,
D.; Djellab, K.; Gratzel, M.; Nazeeruddin, M. K. Dalton Trans. 2010,
̈
39, 8914.
(4) For recent reviews of the development of luminescent
iridium(III) complexes, see, for example: (a) Flamigni, L.; Barbieri,
A.; Sabatini, C.; Ventura, B.; Barigelletti, F. Top. Curr. Chem. 2007,
281, 143. (b) Liu, Z.; Bian, Z.; Huang, C. Top. Organomet. Chem. 2010,
28, 113. (c) Chi, Y.; Chou, P.-T. Chem. Soc. Rev. 2010, 39, 638.
(5) Lowry, M. S.; Bernhard, S. Chem.Eur. J. 2006, 12, 7970.
(6) (a) Arm, K. J.; Williams, J. A. G. Chem. Commun. 2005, 230.
(b) Arm, K. J.; Williams, J. A. G. Dalton Trans. 2006, 2172.
(7) Baranoff, E.; Yum, J. H.; Gratzel, M.; Nazeeruddin, M. K. J.
̈
Organomet. Chem. 2009, 694, 2661.
3825
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826

Inorganic Chemistry
Article
(8) Sun, J. F.; Wu, W. H.; Guo, H. M.; Zhao, J. Z. Eur. J. Inorg. Chem.
2011, 3165.
Williams, J. A. G.; Murphy, L. Appl. Phys. Lett. 2009, 94, 073309.
́
(c) Mroz, W.; Botta, C.; Giovanella, U.; Rossi, E.; Colombo, A.;
Dragonetti, C.; Roberto, D.; Ugo, R.; Valore, A.; Williams, J. A. G. J.
Mater. Chem. 2011, 21, 8653.
(9) (a) Lo, K. K. W.; Tsang, K. H. K.; Sze, K. S.; Chung, C. K.; Lee.,
T. K. M.; Zhang, K. Y.; Hui, W. K.; Li, C. K.; Lau, J. S. Y.; Ng, D. C.
M.; Zhu, N. Coord. Chem. Rev. 2007, 251, 2292. (b) Fernandez-
Moreira, V.; Thorp-Greenwood, F. L.; Coogan, M. P. Chem. Commun.
2010, 46, 186. (c) Murphy, L.; Congreve, A.; Palsson, L. O.; Williams,
J. A. G. Chem. Commun. 2010, 46, 8743. (d) Zhao, Q.; Huang, C.; Li,
F. Chem. Soc. Rev. 2010, 39, 3007. (e) Zhao, Q.; Huang, C.; Li, F.
Chem. Soc. Rev. 2011, 40, 2508. (f) Lo, K. K. W.; Zhang, K. Y.; Li, S. P.
Y Pure Appl. Chem. 2011, 83, 823.
(10) Nonoyama, M. Bull. Chem. Soc. Jpn. 1974, 47, 767.
(11) The few instances reported to date rely on the initial formation
of mixtures, from which the pure products are separated chromato-
graphically in low yields; see, for example: Felici, M.; Contreras-
Carballada, P.; Smits, J. M. M.; Nolte, R. J. M.; Williams, R. M.; De
Cola, L.; Feiters, M. C. Molecules 2010, 15, 2039. Edkins, R. M.;
Wriglesworth, A.; Fucke, K.; Bettington., S. L.; Beeby, A. Dalton Trans.
2011, 40, 9672.
(34) See, for example: Lo, S.-C.; Shipley, C. P.; Bera, R. N.; Harding,
R. E.; Cowley, A. R.; Burn, P. L.; Samuel, I. D. W. Chem. Mater. 2006,
18, 5119.
(35) (a) Williams, J. A. G.; Develay, S.; Rochester, D. L.; Murphy, L.
Coord. Chem. Rev. 2008, 252, 23. (b) Kalinowski, J.; Fattori, V.;
Cocchi, M.; Williams, J. A. G. Coord. Chem. Rev. 2011, 255, 2401.
(36) Rausch, A. F.; Homeier, H. H.; Yersin, H. Top. Organomet.
Chem. 2010, 29, 193.
(37) Harding, R. E.; Lo, S.-C.; Shipley, C. P.; Burn, P. L.; Samuel, I.
D. W. Org. Electron. 2008, 9, 377.
(38) Caspar, J. V.; Meyer, T. J. J. Phys. Chem. 1983, 87, 952.
(39) See, for example: Colombo, M. G.; Hauser, A.; Gudel, H. U.
̈
Top. Curr. Chem. 1994, 171, 143.
(40) Sajoto, T.; Djurovich, P. I.; Tamayo, A. B.; Oxgaard, J.;
Goddard, W. A.; Thompson, M. E. J. Am. Chem. Soc. 2009, 131, 9813.
(41) Costa, R. D.; Monti, F.; Accorsi, G.; Barbieri, A.; Bolink, H. J.;
Orti, E.; Armaroli, N. Inorg. Chem. 2011, 50, 7229.
(42) Rossi, E.; Murphy, L.; Brothwood, P. L.; Colombo, A.;
Dragonetti, C.; Roberto, D.; Ugo, R.; Cocchi, M.; Williams, J. A. G. J.
Mater. Chem. 2011, 21, 15501.
(43) Adachi, C.; Baldo, M. A.; Forrest, S. R.; Lamansky, S.;
Thompson, M. E.; Kwong, R. C. Appl. Phys. Lett. 2001, 78, 1622.
(44) Baldo, M. A.; Lamansky, S.; Burrows, P. E.; Thompson, M. E.;
Forrest, S. R. Appl. Phys. Lett. 1999, 75, 4.
(12) An improved synthetic scheme for tris-heteroleptic complexes
has appeared a few days prior to submission of this article: Baranoff,
E.; Curchod, B. F. E.; Frey, J.; Scopelliti, R.; Kessler, F.; Tavernelli, I.;
Rothlisberger, U.; Gratzel, M.; Nazeeruddin, Md. K. Inorg. Chem.
̈
2012, 51, 215.
(13) Williams, J. A. G.; Wilkinson, A. J.; Whittle, V. L. Dalton Trans.
2008, 2081.
(14) Williams, J. A. G. Chem. Soc. Rev. 2009, 38, 1783.
(15) Wilkinson, A. J.; Goeta, A. E.; Foster, C. E.; Williams, J. A. G.
Inorg. Chem. 2004, 43, 6513.
(16) Wilkinson, A. J.; Puschmann, H.; Howard, J. A. K.; Foster, C. E.;
Williams, J. A. G. Inorg. Chem. 2006, 45, 8685.
(17) Whittle, V. L.; Williams, J. A. G. Inorg. Chem. 2008, 47, 6596.
(18) Obara, S.; Itabashi, M.; Okuda, F.; Tamaki, S.; Tanabe, Y.; Ishii,
Y.; Nozaki, K.; Haga, M.-A. Inorg. Chem. 2006, 45, 8907.
(19) Kuwabara, J.; Namekawa, T.; Haga, M.-A.; Kanbara, T. Dalton
Trans. 2012, 41−44.
(20) Ashikawa, M.; Yang, L.; Kobayashi, K.; Sato, H.; Yamagishi, A.;
Okuda, F.; Harada, T.; Kuroda, R.; Haga, M. Dalton Trans. 2009, 1700.
(21) Choi, D.; Kim, T.; Reddy, S. M.; Kang, J. Inorg. Chem. Commun.
2009, 12, 41.
(22) Beley, M.; Collin, J.-P.; Louis, R.; Metz, B.; Sauvage, J.-P. J. Am.
Chem. Soc. 1991, 113, 8521.
(45) Cocchi, M.; Kalinowski, J.; Stagni, S.; Muzzioli, S. Appl. Phys.
Lett. 2009, 94, 083306.
(46) (a) Cocchi, M.; Virgili, D.; Fattori, V.; Williams, J. A. G.;
Kalinowski, J. Appl. Phys. Lett. 2007, 90, 023506. (b) Cocchi, M.;
Kalinowski, J.; Fattori, V.; Williams, J. A. G.; Murphy, L. Appl. Phys.
Lett. 2009, 94, 073309.
(47) Whittle, V. L.; Williams, J. A. G. Dalton Trans. 2009, 3929.
(48) Nakamaru, K. Bull. Chem. Soc. Jpn. 1982, 55, 2697.
(49) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of
Photochemistry, 2nd ed.; Marcel Dekker: New York, 1993.
(50) DeMello, J. C.; Wittmann, H. F.; Friend, R. H. Adv. Mater.
1997, 9, 230.
(51) SMART, Data Collection Software, version 5.625; Bruker
Analytical X-ray Instruments Inc.: Madison, WI, 2001.
(52) SAINT, Data Reduction Software, version 6.45A; Bruker
Analytical X-ray Instruments Inc.: Madison, WI, 2003.
(53) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.;
Puschmann, H. J. Appl. Crystallogr. 2009, 42, 339.
́
(23) (a) Cardenas, D. J.; Echavarren, A. M.; Ramírez de Arellano, M.
C. Organometallics 1999, 18, 3337. (b) Farley, S. J.; Rochester, D. L.;
Thompson, A. L.; Howard, J. A. K.; Williams, J. A. G. Inorg. Chem.
2005, 44, 9690. (c) Develay, S.; Blackburn, O.; Thompson, A. L.;
Williams, J. A. G. Inorg. Chem. 2008, 47, 11129.
(54) Sheldrick, G. M. SADABS, Version 2006/1; University of
Gottingen: Gottingen, Germany, 2006.
̈
̈
(24) Wang, Z.-X.; Duan, W.; Wiebe, L. I.; Balzarini, J.; De Clercq, E.;
Knaus, E. E. Nucleosides, Nucleotides Nucleic Acids 2001, 20, 11.
(25) Rajca, A.; Rajca, S.; Wongsriratanakul, J.; Ross, C. R. Polyhedron
2001, 20, 1669.
(26) Dixon, I. M.; Collin, J.-P.; Sauvage, J.-P.; Flamigni, L.; Encinas,
S.; Barigelletti, F. Chem. Soc. Rev. 2000, 29, 385.
(27) 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.
(28) (a) Garces, F. O.; King, K. A.; Watts, R. J. Inorg. Chem. 1988, 27,
3464. (b) Garces, F. O.; Dedeian, K.; Keder, N. L.; Watts, R. J. Acta
Crystallogr., Sect. C: Cryst. Struct. Commun. 1993, 49, 1117.
(29) Collin, J.-P.; Dixon, I. M.; Sauvage, J.-P.; Williams, J. A. G.;
Barigelletti, F.; Flamigni, L. J. Am. Chem. Soc. 1999, 121, 1493.
(30) Hay, P. J. J. Phys. Chem. A 2002, 106, 1634.
(31) Coppo, P.; Plummer, E. A.; De Cola, L. Chem. Commun. 2004,
1774. Lafolet, F.; Welter, S.; Popovic, Z.; De Cola, L. J. Mater. Chem.
2005, 15, 2820.
(32) Sabatini, C.; Barbieri, A.; Barigelletti, F.; Arm, K. J.; Williams, J.
A. G. Photochem. Photobiol. Sci. 2007, 6, 397.
(33) (a) Yang, X.; Wang, Z.; Madakuni, S.; Li, J.; Jabbour, G. E. Adv.
Mater. 2008, 20, 2405. (b) Cocchi, M.; Kalinowski, J.; Fattori, V.;
3826
dx.doi.org/10.1021/ic202756w | Inorg. Chem. 2012, 51, 3813−3826