A rare case of dual emission in a neutral heteroleptic iridium(iii) complex

Article abstract of DOI:10.1039/c3dt33115f

Herein we describe the synthesis and characterization of a neutral heteroleptic iridium complex bearing an unusual 2-pyridyl-6-methylthiazine dioxide ligand (pythdo). X-ray crystallographic analysis reveals that in the complex, the pythdo ligand is twisted and puckered, resulting in very low photoluminescent quantum efficiency. The emission profile is structured. Excited state lifetime measurements along with oxygen quenching studies point to a rare case of dual emission from different excited states whereby the high energy bands possess significant ligand-centered (³LC) character while the lower energy bands are predominantly characterized as a mixture of charge transfer (³CT) states. A detailed computational analysis corroborates the unusual photophysical behavior.

Full text of DOI:10.1039/c3dt33115f

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
A rare case of dual emission in a neutral heteroleptic
iridium(III) complex†
Cite this: Dalton Trans., 2013, 42, 8838
Sébastien Ladouceur,a Loïc Donato,a Maxime Romain,a Bhanu P. Mudraboyina,b
Michael B. Johansen,b James A. Wisner*b and Eli Zysman-Colman*a
Herein we describe the synthesis and characterization of a neutral heteroleptic iridium complex bearing
an unusual 2-pyridyl-6-methylthiazine dioxide ligand (pythdo). X-ray crystallographic analysis reveals that
in the complex, the pythdo ligand is twisted and puckered, resulting in very low photoluminescent
quantum efficiency. The emission profile is structured. Excited state lifetime measurements along with
oxygen quenching studies point to a rare case of dual emission from different excited states whereby the
high energy bands possess significant ligand-centered (3LC) character while the lower energy bands are
predominantly characterized as a mixture of charge transfer (3CT) states. A detailed computational ana-
lysis corroborates the unusual photophysical behavior.
Received 31st December 2012,
Accepted 14th April 2013
DOI: 10.1039/c3dt33115f
www.rsc.org/dalton
observed in 77 K 2-MeTHF glasses results instead from
different vibronic transitions.
Introduction
The study of the spectroscopic and photophysical behavior of
To date only a limited set of reports assert dual phosphor-
iridium(III) complexes was pioneered by Watts and co-workers escence from distinct excited states of the same iridium(III)
in the late 1980s. In their study of [(ppy)2Ir(bpy)]+ and fac- luminophore at ambient temperature in solution.5 In the case
Ir(ppy)3 and related complexes, they originally attributed the of cationic complexes reported by Lo and co-workers,5b,c dual
structured emission observed at 77 K to dual emission result- emission was inferred from: (i) solvent-induced modulation of
ing from thermally non-equilibrated 3MLCT excited states the relative intensities of the low- and high-energy emission
(ppyH = 2-phenylpyridine and bpy = 2,2′-bipyridine).1 Güdel bands and (ii) the sensitivity of the emission intensity of only
and co-workers redefined the nature of the dual emission in the high-energy band in the presence of guest molecules such
[(ppy)2Ir(bpy)]+ as occurring only at low temperature in a as avidin. Chi and co-workers5a convincingly demonstrated
glassy matrix as proceeding from both 3MLCT and 3LCppy dual emission in a heteroleptic neutral iridium complex
states, each preferentially accessible in different sites within through a temperature dependence on the relative intensities
the inhomogenous environment of the glass;2 emission from of the structured emission observed and related relaxation
crystalline samples or in solution resulted from a single dynamics. Additionally, there are also some reports of dual
excited state. In detailed spectroscopic studies of fac-Ir(ppy)3 emission resulting from different chromophoric ligand units
and related neutral homo- and heteroleptic iridium(III) com- in neutral heteroleptic iridium(III) complexes.6 As well, in one
plexes, Yersin and co-workers3 along with Thompson and co- instance two distinct polymorphs were isolated with different
workers4 have contended that the structured emission spectroscopic profiles resulting from different sets of π-stack-
ing interactions in the solid state.7
Herein, we report the optoelectronic properties of a novel
aDépartement de Chimie, Université de Sherbrooke, 2500 Boul. de l’Université,
neutral heteroleptic iridium complex [(ppy)2Ir(pythdo)], 2,
containing an unusual pyridyl thiazine dioxide ligand
(pythdo), 1. We contrast the properties of 2 with those of the
benchmark complex [(ppy2)Ir(bpy)](PF6), 3. The asymmetric
complex is highly strained and thus is poorly luminescent in
Sherbrooke, QC, Canada, J1K 2R1. E-mail: eli.zysman-colman@usherbrooke.ca
bDepartment of Chemistry, The University of Western Ontario, 1151 Richmond Street,
London, ON, Canada, N6A 5B7
†Electronic supplementary information (ESI) available: Detailed synthesis pro-
cedures. Characterization of all compounds and complexes. 1H NMR and 13C
NMR for 1 and 2. UPLC-MS trace for 2. CIF Files for 1 and 2. Detailed photo- solution. The emission profile is complex and we have ascer-
physical protocol and supplementary absorption, emission and excitation
tained that 2 exhibits dual emission. The nature of this dual
emission is difficult to characterize but with the aid of a com-
spectra for 1–3. Detailed cyclic voltammetry protocol. Detailed computational
protocol, XYZ coordinates for S0 and T1 states of 2 and supplementary calcu-
bined DFT/TDDFT study, we have assigned the higher energy
lations including full TDDFT output. CCDC 917254 and 917255. For ESI and
emission to one dominated by ligand-centered (3LC) emission
crystallographic data in CIF or other electronic format see DOI: 10.1039/
c3dt33115f
on the pyridyl thiazine dioxide while lower energy emission
8838 | Dalton Trans., 2013, 42, 8838–8847
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
bands are mainly charge transfer (3CT) in nature. Excitation
spectra, excited state lifetime behavior and sensitivity to mole-
cular oxygen corroborate this assignment. Solid state, powder
and thin film emissive properties will also be discussed.
Ligand and complex synthesis
The ligand pythdo (1) is easily synthesized in three steps begin-
ning from α-mercaptoacetone8 and α-bromo-2-acylpyridine.9
The thioether (4), formed via substitution of the bromide by
the mercaptan in acetonitrile (ACN), was oxidized to the
sulfone (5) using meta-chloroperoxybenzoic acid (mCPBA). The
thiazine dioxide ring was formed by cyclization using NH4OAc
in refluxing dimethylformamide (DMF) rather than AcOH as
the latter solvent produces solely the undesired intramolecular
aldol condensation product. The iridium(III) dimer, [(ppy)2Ir-
(μ-Cl)]2, was cleaved in the presence of 1 and K2CO3 to afford
in good yield the heteroleptic neutral complex 2 (Scheme 1).
The purity and structure of 2 were established unequivocally
by 1D and 2D NMR experiments, UPLC-HRMS, elemental and
crystallographic analyses.
Fig. 1 (a) Ball and stick representation of two molecules of ligand 1 forming
part of a hydrogen bonded chain in the solid state. (b) ORTEP representation of
one of two molecules of 2 in the asymmetric unit. All hydrogen atoms and the
solvent molecule have been removed for clarity.
Light-yellow needle-like crystals of 2 were grown by vapor
diffusion of Et2O into a MeOH–ACN solution (Fig. 1b).
ˉ
Complex 2 is triclinic and crystallizes in space group P1 with
two complexes and a MeOH molecule in the unit cell; the
solvent molecule does not form any relevant intermolecular
contacts with the complex. The complexes pack in the lattice
such that there are numerous weak face-to-face and edge-to-
face π–π interactions between phenyl and pyridyl rings of ppy
ligands on adjacent complexes. The iridium atoms in both
complexes possess a distorted octahedral geometry, with
X-ray and solution state structural elucidation
X-ray quality crystals of ligand 1 were grown by vapor diffusion
of diisopropyl ether into a concentrated chloroform solution
(Fig. 1a). The ligand crystallizes in space group C2/c as sheets
(oriented along the crystallographic a direction) of antiparallel
chains that are connected by a weak intermolecular hydrogen
N
pythdo–Ir–Npythdo and average Cppy–Ir–Nppy bond angles of 76°
and 85°, respectively. The nitrogen atoms on the two ppy
ligands are in a trans arrangement. Bond lengths to each of the
iridium metal centers are unremarkable compared to their ana-
logues found in [(ppy)2Ir(bpy)]PF6 (3)10 or analogous neutral Ir
complexes employing 2-(2′-pyridyl)imidazolate ligands.11 Com-
pared to the structure of the ligand, the thiazine dioxide rings
in 2 are both slightly puckered by displacement of the sulfur
atoms out of the heterocyclic planes (0.3976(23) and 0.4685(46)
Å from the least squares planes described by the remaining
C/N atoms of the thiazine rings). The pythdo ligand is now
further twisted in 2, with 13 and 18° N–C–C–N torsion angles,
likely as a consequence of sterics between the pythdo methyl
group and the neighboring ppy ligand. However, there remains
a CH–π interaction between the methyl hydrogen atoms and
the adjacent phenyl ring of the ppy ligand with distances of
3.0751(86) and 3.1014(125) Å between the methyl carbon atoms
and the least squares planes described by the rings. This inter-
action explains the large upfield shift of the methyl resonance
bond (N
The ligand conformation is the expected syn-arrangement that
provides favorable intramolecular electrostatic contact
between the thiazine NH proton and the adjacent N hydrogen
bond acceptor of the pyridine ring (N2–H2A⋯N1 = 2.1994(12)
̲2–H2A⋯O̲2 = 3.4128(15), N2–H2A⋯O2 = 154.413(74)°).
a
̲
̲
Å). The two heterocyclic rings of the ligand are planar (dis-
tances from least squares planes for C, N, S < 0.065 Å) and
nearly coplanar (angle between least squares planes
=
8.506(44)°). Molecules in adjacent antiparallel chains are
related by an inversion center and interact in a π–π slipstacked
manner (distance between least squares planes = 3.50 Å).
1
(Δδ = −0.97 ppm; see Fig. 2) in the H NMR spectra of 1 upon
ligation to the iridium in 2.
In a similar manner, the proton at the 6-position of the
pyridyl ring of the ligand undergoes an upfield shift (Δδ =
1
−0.33 ppm) in the H NMR spectrum of the complex due to its
positioning above the coordinated carbon atom (C44) of one of
the two ppy co-ligands. Disappearance of the ligand N–H reson-
ance in the 1H NMR of 2 upon complexation provides further evi-
dence that the solution and solid state structures are the same.
Computations
Given the unusual structure of 1 and its incorporation as a
Scheme 1 Synthesis of pyridylthiazinedioxide (pythdo) ligand 1 and its neutral
[(ppy)2Ir(pythdo)] complex 2.
ligand in 2,
a
combined DFT and TDDFT study was
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8838–8847 | 8839

View Article Online
Paper
Dalton Transactions
Fig. 3 Schematic representation of 2 in the (a) crystal structure; (b) computed
ground state structure; (c) computed triplet state structure.
Fig. 2 Stacked plots of the 400 MHz 1H NMR spectra of pythdo ligand 1,
complex 2 and iridium dimer [(ppy)2Ir(μ-Cl)]2 in CDCl3 at 298 K. Dashed lines
indicate shifts in labeled ligand proton resonances upon ligation to Ir in
complex 2.
thiazine dioxide (thdo) fragment, as evidenced by the two dis-
tinct Nthdo–C–C–Sthdo torsional angles. In the triplet state, the
pythdo ligand flattens to a nearly planar conformation as
denoted by the Npy–C–C–Nthdo dihedral angle of 4.6°. This flat-
tening is coupled to a conformational change in the thdo ring.
The computations overestimate the Ir–N bond lengths with
the pythdo ligand by around 0.05 Å compared to the crystal
structure. The Ir–Nppy and Ir–Cppy distances are overestimated
undertaken in order to gain greater insight into the photo-
physical behavior of 2. Table 1 lists selected bond lengths,
bond angles and dihedral angles for 2 from the X-ray data and
calculated ground (S0) and the triplet (T1) states. The complex
possesses Cs symmetry with a pseudo-octahedral geometry. slightly by 0.032 and 0.019 Å, respectively. Similarly, the C–C
The input geometry for the S0 geometric optimization was
taken from the X-ray structure (Fig. 3). Starting from a geo-
bond length between the pyridine (py) and thdo fragments is
also overestimated by 0.023 Å. Computed bond angles about
metry in which the pythdo ligand is completely planar results in the Ir metal reproduce those found in the crystal. Overall, the
a S0 geometry that is a configurational diastereomer analogous
to that resolved in the crystal structure. It should be noted that
there is an absence of stereocontrol during the cleavage of the
[(ppy)2Ir-(μ-Cl)]2 dimer with the pythdo and thus it is expected
that a stereoisomeric mixture will be present. However, the
computations accurately predict the important structural fea-
tures for 2.
Apart from the flattening of the pythdo ligand, important
deformations in the T1 state include a contraction of both the
Ir–Npythdo bonds as well as the C–C bond linking the thdo and
reported solution state spectroscopic and electrochemical data py rings. This strengthening of the interactions between the
are stereo-invariant. In the computed ground state, the two
rings of the pythdo ligand are twisted by approximately 22°,
iridium and the pythdo ligand is accompanied by a slight
lengthening of the Ir–Cppy bond, in part to compensate for the
slightly greater than the 13 and 17° torsions of the two com- enhanced steric interactions between the ancillary ligand and
plexes observed in the asymmetric unit of the crystal structure.
The computations predict accurately the puckering of the
the rest of the complex as well as the weaker bonding of the
ppy ligand in the excited state.
Table 1 Selected structural parameter comparison for 2 between X-ray data, and calculated S0 and T1 geometries
Crystal structure
Bond distance (Å)
S0 state
T1 state
Structural
parameters
Average (Å)
Bond distance (Å)
Average (Å)
Bond distance (Å)
Average (Å)
Ir–N (py)
—
—
2.040
2.007
1.468
—
2.137
2.137
2.194
2.046
2.008
1.463
1.470
—
—
2.085
2.028
1.472
—
2.190
2.190
2.246
2.078
2.027
1.470
1.493
—
—
2.089
2.039
1.472
—
2.164
2.164
2.198
2.077
2.035
1.471
1.422
Ir–N (thdo)
Ir–N (ppy)
Ir–C (ppy)
2.194
2.051
2.009
1.457
1.470
2.246
2.070
2.027
1.469
1.493
2.198
2.065
2.031
1.469
1.422
C
ph–Cpy (ppy)
Cpy–Cthdo (pythdo)
Bond angle (°)
Average (°)
Bond angle (°)
Average (°)
Bond angle (°)
Average (°)
N–Ir–N (pythdo)
C–Ir–N (ppy)
—
89.7
75.9
80.2
75.9
85.0
74.8
80.2
74.8
80.3
76.4
80.1
76.4
80.3
80.4
Dihedral angle (°)
80.4
Dihedral angle (°)
Dihedral angle (°)
Average (°)
Average (°)
Average (°)
N
N
thdo–C–C–Sthdo
py–C–C–Nthdo
10.1
—
−10.4
13.2
−0.1
13.2
13.6
—
−15.1
21.8
−0.7
21.8
0.0
—
−24.7
4.6
−12.3
4.6
8840 | Dalton Trans., 2013, 42, 8838–8847
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
The UV-Vis spectrum was then modeled using TDDFT com-
πpythdo→π*pythdo ligand-centered (LC) character. Fig. 5 shows
putations and, as shown in Fig. 4, the principle absorption relative orbital energies along with selected qualitative rep-
bands match well with the absorption spectrum obtained in resentations of the electronic distribution from HOMO − 4 to
ACN solution. Table 2 summarizes the first ten significant low- LUMO + 4. The quantification of the electron distribution for
energy singlet–singlet transitions (f > 0.01), while Table S6† HOMO − 4 to LUMO + 4 is shown in Table S5.† The electron
details the hundred lowest energy singlet–singlet vertical tran- contributions for the HOMO and HOMO − 1 are distributed
sitions. Unlike many other iridium complexes where the across the phenyl ring of the ppy ligand, the Ir center and the
lowest energy absorption bands can be described using a pythdo ligand. The HOMO − 2 and HOMO − 3 are composed
small set of singlet–singlet transitions,12 here TDDFT analysis predominantly of Ir and Phppy-centred contributions. The
predicts a series of pseudoequi-permissible transitions. The LUMO and LUMO + 3 are mostly pythdo-centered but the
first seven singlet states combine to model the major low LUMO + 1, LUMO + 2 and LUMO + 4 are localized on the ppy
energy absorption band and these states are mixed character ligand. The complex nature of the character of the singlet–
that are difficult to describe but are composed of some singlet transitions in 2 is not unexpected, given the low sym-
dπ→π*pythdo metal-to-ligand charge transfer (MLCT), πppy
→
metry of the complex. Indeed, De Cola and co-workers had
π*pythdo ligand-to-ligand charge transfer (LLCT), intra-ligand
charge transfer (ILCT) involving both ligands and
Fig. 4 Experimental (red) and TDDFT simulated (blue) absorption spectra
(fwhm 1000 cm−1) of 2. The oscillator strengths (green) are represented by
unbroadened vertical lines of the 100 lowest-energy calculated singlet–singlet
transitions.
Fig. 5 Energy level diagram for selected Kohn–Sham molecular orbitals
(HOMO − 4 to LUMO + 4) of 2, along with the calculated HOMO–LUMO gap by
DFT and electron distribution representations (0.03 isovalue). Inset: spin density
distribution for the T1 state (0.0004 isovalue).
Table 2 Principal low energy singlet–singlet and singlet–triplet vertical transitions, with their corresponding oscillator strengths (f) and qualitative assignments
Energy
Oscillator
State
Nature of transition (major contribution, %)a,b
Primary character
(eV)
(nm)
strength (f)
S1
S2
S3
S4
S5
S6
S7
S8
S9
S10
HOMO→LUMO (77%)
HOMO→LUMO (17%), HOMO→L + 1 (71%)
HOMO→L + 2 (84%)
LCpythdo/MLCT/ILCTppy
LCpythdo/MLCT/ILCTppy
LLCT/MLCT
ILCTppy/LCpythdo/MLCT
ILCTppy/LCpythdo/MLCT
ILCTppy/ILCTpythdo
ILCTppy/LCpythdo
ILCTppy/LCpythdo/MLCT
LLCT/MLCT
ILCTppy/MLCT/ILCTpythdo
3.01
3.07
3.21
3.22
3.32
3.43
3.54
3.72
3.73
3.77
412
404
387
386
374
362
350
333
332
329
0.0337
0.0338
0.0389
0.0385
0.0374
0.0292
0.0103
0.0311
0.0817
0.0353
H − 1→LUMO (86%)
H − 1→L + 1 (81%)
H − 1→L + 2 (84%)
HOMO→L + 3 (96%)
H − 1→L + 3 (56%), HOMO→L + 4 (38%)
H − 3→LUMO (36%); H − 2→LUMO (38%)
H − 1→L + 3 (35%); HOMO→L + 4 (47%)
T1
T2
T3
T4
T5
HOMO→LUMO (59%)
LCpythdo/MLCT/ILCTppy
LCpythdo/MLCT/ILCTppy
MLCT/LLCT/LCpythdo
ILCTppy/LCpythdo/MLCT
ILCTppy/LCpythdo/MLCT
2.53
2.77
2.86
3.11
3.15
491
448
433
398
393
0
0
0
0
0
H − 1→L + 1 (15%); HOMO→L + 1 (42%)
H − 1→L + 2 (18%); HOMO→L + 2 (26%)
H − 1→L + 1 (39%); HOMO→L + 1 (18%)
H − 1→LUMO (46%); H − 1→L + 1 (24%)
a H = HOMO; L = LUMO. b Major contributions defined as >15% contribution to the transition.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8838–8847 | 8841

View Article Online
Paper
Dalton Transactions
Table 3 Electrochemical data obtained in ACN solution at 298 K for 1–3a
previously observed complex photophysical behavior that was
corroborated by computations in the related neutral
[(dFppy)2Ir(pytl)] (where dFppy is 2,4-difluorophenyl-2′-pyri-
dine and pytl is 2-(1H-1,2,3-triazol-5-yl)pyridine).13
Ligands
3
2
1
Epa,ox
Epa,ox
Epa,ox
ΔE
Epc,red
C^N N^N
(V)
(V)
(V)
(V)
(V)
On the triplet manifold, the five lowest energy triplet states
were calculated at the optimized S0 geometry by TDDFT
(Table 2). The T1 state lies at 2.53 eV, 0.48 eV stabilized from
the S1 state, and originates mainly from a HOMO→LUMO
excitation, which can be characterized as an admixture of
LCpythdo/MLCT/LLCT/ILCTppy, with a dominant ligand-cen-
tered component. The T2 and T3 states are of similar energies
at 2.77 and 2.86 eV, respectively. Each also exhibits mixed CT
and LC character that is difficult to deconvolute. However, the
T3 state clearly possesses a less pronounced LCpythdo character
due to contributions from the HOMO − 1→LUMO + 2 and
HOMO→LUMO + 2. The T4 and T5 states are pseudo-degener-
ate and are again multiconfigurational, making their descrip-
tion difficult. For the T4 state, the largest contribution (39%)
corresponds to HOMO − 1→LUMO + 1, which is a mixture of
ILCTppy, LCpythdo and MLCT. For the T5 state, the largest con-
tribution (46%) corresponds to HOMO − 1→LUMO, which is
also a mixture of ILCTppy and LCpythdo and MLCT, but with a
more dominant LC character.
The geometry and energy of the T1 state were fully opti-
mized at the unrestricted UB3LYP level of theory. The spin
density for the T1 state, which mirrors the topology of the
HSOMO, is shown in Fig. 5 and is distributed on the pythdo
and the iridium center. As a result, emission from the T1 state
can be described as an admixture of LCpythdo/MLCT/LLCT that
is predominantly ligand-centered, which would explain the
structured bands observed in the emission spectrum
(Fig. S11†). The predicted emission energy as determined by
the ΔSCF approach is 545 nm, which compares favorably with
the λmax at 77 K in MeOH–EtOH glass at 520 nm (vide infra).
Both the ground and excited state dipoles point away from
the thdo ring, with little change in the vectoral direction
between S0 and T1 states. The magnitude of the excited state
dipole (9.30 D) is somewhat lower than that of the ground
state (14.01 D, see Fig. S17†).
1
2
3
pythdo 1.77
pythdo 1.39
1.33
1.16
—
—
0.95
3.62 −1.85
2.93 −1.98
ppy
ppy
bpy
—
1.27 [56]b 2.65 −1.38 [55]b
a Measured in ACN (ca. 1.5 mM) with NBu4PF6 (ca. 0.1 M) as the
supporting electrolyte. Potentials (V) are reported vs. SCE and were
calibrated using an internal standard Fc/Fc+ redox couple (0.40 V in
ACN).14 All peaks were either irreversible or quasi-irreversible except
where noted. b Reversible process. The value reported is E ,ox instead of
1
2
1
Epa,ox for oxidation and and E ,red instead of Epc,red for the reduction
process. The values reported in2brackets refer to ΔEp in mV.
analogous process in 3. A similarly large destabilization of
570 mV of the first oxidation wave was noted when comparing
[(dFppy)2Ir(bpy)]PF6, synthesized by us,12b to [(dFppy)2Ir(pytl)]
obtained by De Cola and co-workers,13 where pytl chelates to Ir
in an analogous fashion to the pythdo in that the ligand con-
tains a pyridine ring, is asymmetric and contains a similarly
hybridized anionic nitrogen atom.
A single irreversible
reduction wave occurs in 2 at −1.98 V and can be associated
with the reduction of the pythdo moiety, which is 130 mV
cathodically shifted compared to the irreversible reduction
wave found in 1. The resulting electrochemical gap in 2 is
2.93 V, 280 mV larger than that found in 3. This corroborates
well with the spectroscopic measurement as the emission of 2
in ACN solution at 298 K is 523 meV more energetic than 3
(vide infra).
Photophysical characterization
The absorption and emission spectra for 1, 2 and 3 are shown
in Fig. 6 and the photophysical data listed in Table 4. All
measurements at 298 K were recorded in ACN except where
specifically mentioned while 77 K measurements were con-
ducted in
a 1 : 1 MeOH–EtOH glass. The spectroscopic
Electrochemical characterization
The electrochemical potentials for 1–3 were measured by
means of cyclic voltammetry (CV) in ACN with NBu4PF6 as the
supporting electrolyte solution using an Fc/Fc+ internal stan-
dard (0.40 V in ACN) and are reported versus SCE (Table 3, cf.
Fig. S16† for CV traces). Whereas archetypal complex 3 exhibits
reversible one-electron oxidation and reduction waves at 1.27
and −1.38 V, respectively, 2 exhibits more complex electroche-
mical behavior. In this case, there are three irreversible oxi-
dation processes at 0.95, 1.16 and 1.39 V. The oxidation waves
at 1.16 and 1.39 V can be assigned to the oxidations of the
pythdo moiety, which are shifted cathodically by 170 and
380 mV, respectively, compared to the two oxidation waves
observed in the ligand (1) itself. The oxidation wave at 0.95 V
can then be assigned to a process implicating the IrIII → IrIV
Fig. 6 Absorption (solid lines) and emission (dashed lines) spectra at 298 K in
deaerated ACN and emission (dotted lines) at 77 K in MeOH–EtOH (1/1) of
couple, cathodically shifted by 320 mV compared to an ligand 1 (blue), complexes 2 (red) and 3 (orange).
8842 | Dalton Trans., 2013, 42, 8838–8847
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 4 Spectroscopic and photophysical data for 1–3 in ACN
Emission
τe
Absorbance (nm) [ε (×104 M−1 cm−1)]
77 Ka,b (nm)
298 Kb (nm)
77 Ka (μs)
298 K (ns)
ΦPL (%)
1
2
234 [0.23]; 273 [0.12]; 300 [0.03]
261 [4.79]; 292 [2.38]; 332 [0.97]; 365 [0.74]; 420 [0.30]
390
—
478 [0.88]
508 [1.00]
600 [0.47]
0.12c
430d
522 [0.95]
555 [1.00]
600 [0.53]
—
—
15.1
14.8
15.0
1.75 × 103
1.93 × 103
1.8e
568 [78%]
53.7 [22%]
275
3
254 [4.17]; 310 [1.29]; 332 [0.96]; 375 [0.60]; 420 [0.26]
542
602
4.77
9.3e
a Measured in MeOH–EtOH (1/1) glass state at 77 K. b Values in brackets refer to the relative intensity of the emission to λmax
using quinine bisulfate (ΦPL = 54.6% in 0.5 M H2SO4).15 d 5.1 × 10−3
.
c Measured at 298 K
.
e Measured at 298 K using [Ru(bpy)3](PF6)2 (ΦPL = 9.5% in ACN).16
behavior and photophysical properties measured for bench- and to solution concentration; 2 also exhibited good photo-
mark complex 3 matched those reported in the literature.12a stability in ACN solution over several days; however solutions
Individual absorption and emission spectra at 298 K for 2 and in CDCl3 turned to black precipitates within the same time
the pythdo ligand (1) in different solvents (ACN, DCM, frame.
2-MeTHF and MeOH–EtOH (1/1) can be found in the ESI
(Fig. S4–S11†).
The emission λmax in ACN is distinctly blue-shifted com-
pared to that in both DCM (Fig. S6†) and 1 : 1 MeOH–EtOH
The absorption spectra for 1 in ACN solution show moder- (Fig. S5†) but similar to 2-MeTHF (Fig. S7†), whose emission
ately intense π → π* absorption bands at 234, 273 and 300 nm. profiles remain complex. At first glance this “inverted solvato-
When excited at 310 nm, the organic ligand is poorly fluo- chromism” could in part be rationalized through a modest
rescent (ΦPL = 0.12%) in ACN solution with an emission decrease in the magnitude of the computed excited state
maximum at 390 nm along with a small shoulder at 510 nm. dipole compared to that in the ground state in ACN solution
Unfortunately, due to instrument limitations, we could not from 14.01 and 9.30 Debye (Fig. S17†). Upon closer inspection
measure the excited state lifetime but taking into account the however, the choice of the solvent influences the dominance
nature of the molecule, the luminescence can be ascribed to of one emission band over the other (emission intensity)
fluorescence.
rather than influencing the emission energy. Similar “solvato-
The absorption spectrum for 2 shows two important bands chromic” behavior was reported by Lo and co-workers for their
centered at 261 and 365 nm with molar absorptivity coeffi- dual emissive cationic iridium complex in ref. 5c. From the
cients (ε) on the order of 4.79 and 0.74 × 104 M−1 cm−1
,
computations, we assign the high-energy structured bands to
respectively. Similar to the profile for 3, the high-energy band an emission with rich LC character while the low-energy less
can be assigned to a spin-allowed ligand centered (1LC) π → π* structured emission centered around 600 nm contains a large
transition while the latter is predominantly charge-transfer in CT component. The precise nature of the CT is difficult to
origin. Weaker singlet–singlet and singlet–triplet CT bands are characterize based on the electronic distribution of the frontier
found between 400 and 500 nm. From DFT calculations (vide MOs.
supra), the nature of this transition is a mixture of metal-to-
Though we cannot rule out solvent interactions in the
ligand (1MLCT), ligand-to-ligand (1LLCT) and intra-ligandpythdo excited state, the UV-Vis spectrum is invariant to changes in
and intra-ligandppy (1ILCT) charge transfers. A small red-shift the solvent. Additionally no concentration effects were noted
1
1
of ca. 7 nm (1050 cm−1) for the LC bands in 3 compared to 2 on the emission shape and maxima. H NMR titration experi-
is noted while a much larger red-shift of 24 nm (4240 cm−1) is ments were undertaken in order to rule out direct solvent
observed compared to the principle absorption band for the interactions with 2 in the ground state as a contributing factor
pythdo ligand. The absorption profile of 2 is insensitive to to the emission energy. Titration of up to 50 equivalents of
changes in solvent polarity (Fig. S9†) and to concentration.
CD3CN into a solution of 2 in CDCl3 or titration of CD3OH
The 298 K emission spectra for 2 and 3 were measured in into a solution of 2 in CD3CN did not produce any notable
deaerated ACN. Complex 2 shows structured emission with changes in the 1H NMR profile for 2, indicating that neither
strong bands at 480 and 510 nm and shoulders at 540 and ACN nor MeOH is directly interacting with 2.
600 nm. This emission is blue-shifted by 3000 cm−1 compared
The relative intensities of the two emission bands at 478
to benchmark 3, according to their respective λmax. The emis- and 508 nm are different in ACN and 2-MeTHF. In 2-MeTHF
sion for 2 is also significantly red-shifted by 6030 cm−1 com- the intensity of the high energy band decreased measurably
pared to 1 and the absence of higher energy peaks in the compared to that in ACN solution while the lower energy band
emission spectrum of 2 further points to an analytically pure was much less affected (Fig. S9†). By contrast, the intensities
sample not contaminated by 1. The structured emission of both these high energy emission bands decreased consider-
profile for 2 is insensitive to changes in excitation wavelength ably compared to the emission centered at 600 nm and are
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8838–8847 | 8843

View Article Online
Paper
Dalton Transactions
better characterized as high-energy shoulders. There are two the following manner. Under deaerated conditions, emission
principal emission bands at 592 and 600 nm in 1 : 1 MeOH– from the 478 and 508 bands is much more intense than that
EtOH while a single emission λmax of 599 nm exists in DCM. In from the 600 nm band. Thus, there is an overlap of the emis-
comparison with 2, 3 shows unstructured emission with λmax sion spectra from these two states and independent probing of
of 602 nm and only a minor solvatochromic effect. Excitation the lifetime of the CT state at 600 nm is impossible. The 568
spectra obtained for 2 (Fig. S12–S13†) at emission wavelengths ns component is due to the residual emission from the high
ranging from 445 to 525 nm were superimposable to one energy LC state while the lifetime at 600 nm is 53.7 ns. In
another and matched well the absorption spectrum, signifying point of comparison, the excited state lifetime for 3 is 275 ns.
that emission is occurring from a unique excited species and Together with the shape of the different emission bands, these
confirming the purity of the sample.
results point to a dual emission process wherein the structured
The emission spectrum for 2 recorded at 77 K in a rigid high-energy longer-lived emission is significantly 3LC in
matrix of 1 : 1 MeOH–EtOH was also found to exhibit fine nature while the broad low-energy shorter-lived emission is
structure. Usually, as is the case observed with 3, the emission principally 3CT in nature.
at 77 K is significantly hypsochromically shifted (λmax
=
Inspection of the τe’s measured in MeOH–EtOH at 77 K
542 nm) compared to that obtained in 298 K ACN solution, reveals similar values of 15.1, 14.8 and 15.0 μs measured at
typical of characteristic rigidochromic blue-shifts of CT 522, 555 and 600 nm, respectively. These lifetimes are about
states.17 The gross features in the 77 K emission profile for 2 three times longer than that measured for 3 (4.77 μs). Similar
are similar to those observed at 298 K in ACN but with emis- but shorter excited state lifetimes were measured in 2-MeTHF
sion maxima at 522 and 555 nm; there is an additional very at 77 K and showed a bi-exponential character (Table S2†).
weak emission band at 430 nm (Fig. S11†). Measurement at
low temperature in a 2-MeTHF glass state gave similar results
with structured emission (λmax at 540 nm) (Fig. S11†). Under
Cyclometallated iridium(III) complexes are known to be highly
both sets of conditions, the 77 K emission spectra are batho-
chromically shifted to those obtained in ACN solution at 298 K
(Fig. S5 and S7†).
The deaerated ACN photoluminescence quantum yield
(ΦPL) of 2 is weak at 1.8%, much lower than that measured for
3 at 9.3%. Quantum yield measurements in other solvents
were not measured but the intensity of the emission spectra
obtained from isoabsorptive solutions of 2 (λexc = 390 nm) in
both alcoholic and chlorinated solvents are markedly
decreased compared to that in ACN (Table S3 and Fig. S14†).
solution. Qualitatively, the emission maxima at higher energy
Indeed, the integrated emission areas in degassed solutions of
1 : 1 MeOH–EtOH and DCM are 21% and 36%, respectively, of
Influence of 3O2 on the photophysical behaviour of 2
sensitive to the presence of dioxygen (3O2), which acts as a
triplet excited state quencher.21 As such, many examples of
iridium complexes have been reported as ratiometric oxygen
sensors both in solution and solid state,22 particularly for bio-
logical applications.23 Notably, a color change in the emission
of 2 was observed upon measurement in aerated ACN solution.
This led us to investigate more thoroughly the influence of
dioxygen concentration on the emission properties of 2. Fig. 7
shows the impact of increasing concentrations of O2 in ACN
are quenched much more quickly than the emission centred at
600 nm, resulting in the perceived red-shift. Excited state life-
that determined in ACN. The inefficient phosphorescence in 2
times in 20% O2 measured at each of 480, 510 and 600 nm
can be rationalized due to: (i) the puckering of the thiazine
dioxide ring and increased torsional strain present in 2 requir-
ing a twisted conformation between the pyridine and thiazine
were found to be similar at 55.8, 55.5 and 57.9 ns, respectively.
The τe measured at 600 nm is almost unchanged compared to
that measured in N2-saturated solution (53.7 ns) while
dioxide rings both favoring non-radiative pathways; (ii) the geo-
metric deformation that occurs between the S0 and T1 states
(vide supra). Both [(ppy)2Ir(Hpbpy)](PF6) and [(ppy)2Ir(dpbpy)]-
(PF6), obtained by Bolink and co-workers, were found to be
less luminescent than 3 in solution (Hpbpy = 6-phenyl-2,2′-
bipyridine; dpbpy = 6,6′-diphenyl-2,2′-bipyridine).18 Qiu and
co-workers later rationalized the lower ΦPL for these latter two
complexes in terms of configurational distortions.19 Bolink
and co-workers also noted that conformational changes
between the T1 and S0 states promote temperature indepen-
dent nonradiative decay processes associated with vibrational
deactivation.20
The excited state lifetimes (τe) for 2 measured at 478 and
508 nm are 1.75 and 1.93 μs, respectively, indicating that the
nature of the emission is similar for these two emission
bands. By contrast, τe measured at 600 nm was found to be bi-
exponential and significantly shorter at 568 (78%) and 53.7
(22%) ns. This bi-exponential behavior can be rationalized in Fig. 7 Emission spectra of complex 2 at 298 K in ACN at various pO2.
8844 | Dalton Trans., 2013, 42, 8838–8847
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
emission at 480 and 510 nm each decays monoexponentially
with τe’s that are significantly shorter than under degassed
conditions. The Stern–Volmer (KSV) and bimolecular quench-
ing constants (kq) at 478 and 508 nm are 61.6 and 59.4 atm−1
,
and 3.52 × 107 and 3.08 × 107 atm−1 s−1, respectively. Based on
a 53.7 ns lifetime measured under deaerated conditions (vide
supra), those obtained for emission at 600 nm are much lower
at 10.3 atm−1 and 1.91 × 108 atm−1 s−1 (cf. Fig. S15†). These
results suggest that the longer-lived and more pronounced
3LC-based emission is much more sensitive to O2 than the
3CT-dominated emission. The decreased sensitivity to decay in
the low energy emission to dioxygen can be rationalized due to
(i) shorter excited state lifetimes of the 3CT emission; (ii) more
significant metal contribution to the CT, which is more geo-
metrically shielded to interactions with dioxygen than an LC
emission (Table 5).
Fig. 8 Emission spectra of 2 in thin films and as a powder at 298 K under air.
Table 6 Photophysical characterization of 2 in the solid state
Though 2 is poorly luminescent with relatively short life-
times, to the best of our knowledge wavelength-dependent
oxygen sensitivity for an iridium complex has not yet been
reported.
Sample
λmax (nm)
τ (ns)
Φ (%)
Neat film
610
579
575
588
12.0 (25%); 61.2 (75%)
621 (43%); 2675 (57%)
654 (42%); 2790 (58%)
493 (41%); 3279 (59%)
166 (63%); 378 (37%)
—
—
—
—
Thin film and solid state measurements
33% in PMMA film
50% in PMMA film
90% in PMMA film
Powder
Motivated by the interesting solution state photophysical be-
havior of 2, its characterization in the solid state as a powder
and in thin films, both neat and 33, 50 and 90% w/w doped in
PMMA, was investigated. The thin films were spin-coated from
an ACN solution and the corresponding emission curves and
data can be found in Fig. 8 and Table 6. The emissions
measured in the thin films are broad and structureless. The
emission maximum for 2 doped to different levels in PMMA
did not appreciably change, ranging from 575 to 588 nm. A
larger red-shift to 610 nm was observed for the neat film, prob-
ably due to increased intermolecular interactions. The emis-
sion profiles for the thin films, which were measured in air,
504, 536
0.13
emission spectrum is considerably blue-shifted compared to
those in the thin films and modestly red-shifted compared to
the degassed emission spectrum in ACN solution. Bi-exponen-
tial decay of the emission was observed with lifetimes of inter-
mediate value compared to those of the neat and doped thin
films. The quantum yield for the powder was determined to be
0.13%.
are similar to those of oxygenated samples in solution (Fig. 7). Dual emission in 2
The thin films were poorly luminescent and their corres-
Synthesis of the results obtained from the different spectro-
ponding quantum yield could not be measured with the inte-
grating sphere available. In all cases, bi-exponential decay
kinetics was observed, with significantly longer lifetimes
measured in the doped films compared to the neat film.
By contrast, the emission spectrum for the powder obtained
in air was more complex, with two intense bands found at 504
and 536 nm and a very large tail extending until 800 nm. The
scopic and photophysical measurements, notably disparate
solvatochromic behavior in 2 compared to 3, distinct wave-
length-dependent excited state lifetimes and oxygen quench-
ing sensitivity, leads us to propose the following energy
diagram (Fig. 9). Upon excitation into the singlet manifold and
rapid internal conversion back to the S1 state, intersystem
crossing (ISC) to the triplet surface occurs readily promoted by
the large spin–orbit coupling on iridium. The CV measure-
ments and computations indicate that the triplet state first
accessed is one that is predominantly ligand-centered (3LC) on
the pythdo. Emission from this state is observed (τe = 1.75 ×
103 ns; 2.58 eV). As well, a second lower energy charge-transfer
(3CT) state exists and internal conversion between these two
triplet states is possible though difficult with emission from
the 3CT state also observed (τe = 53.7 ns; 2.07 eV). At low temp-
Table 5 O2 quenching dynamic parametersa
Lifetime (τ) 298 K
Emission
KSV
kq
bandsb (nm) N2 (ns)
Airb (ns) τ0/τ
(atm−1
)
(atm−1 s−1
)
480
510
600
1.75 × 103
55.8
55.5
57.9
31.4
34.8
61.6
59.4
3.52 × 107
3.13 × 107
1.93 × 103
568 [78%]
53.7 [22%]
0.93c 10.3
1.91 × 108 c
3
eratures, the CT state is not readily thermally accessible and
‡
the energy barrier, ΔEa , cannot be overcome. At 298 K, the
a For a detailed experimental protocol and calculation of KSV and kq,
see ESI. b Measured under 20% O2 (medical grade air) at 298 K in ACN.
c Calculated using the 53.7 ns lifetime.
3CT state is now thermally accessible and the relative intensity
of the emission from this state increases. Generally, the
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8838–8847 | 8845

View Article Online
Paper
Dalton Transactions
(réseau québécois de calculs de haute performance). EZ-C
acknowledges CFI (Canadian Foundation for Innovation),
NSERC (The Natural Sciences and Engineering Research
Council of Canada), FQRNT (Le Fonds québécois de la
recherche sur la nature et les technologies) and the Université
de Sherbrooke for financial support. SL acknowledges FQRNT
for a PhD scholarship. BPM, MBJ and JAW would also acknowl-
edge NSERC for its financial support of this research. We
thank Professor Christian Reber for helpful discussions.
References
1 (a) F. O. Garces, K. A. King and R. J. Watts, Inorg. Chem.,
1988, 27, 3464–3471; (b) Y. Ohsawa, S. Sprouse, K. A. King,
M. K. DeArmond, K. W. Hanck and R. J. Watts, J. Phys.
Chem., 1987, 91, 1047–1054; (c) K. A. King and R. J. Watts,
J. Am. Chem. Soc., 1987, 109, 1589–1590; (d) A. P. Wilde,
K. A. King and R. J. Watts, J. Phys. Chem., 1991, 95, 629–
634; (e) P. Didier, I. Ortmans, A. Kirsch-De Mesmaeker and
R. J. Watts, Inorg. Chem., 1993, 32, 5239–5245.
Fig. 9 Schematic energy diagram of the different states playing a role in the
complex luminescence of 2. 3LC refers to the high-energy, structured emission
with high ligand-centered character while 3CT refers to the low-energy, feature-
less emission with high charge-transfer character. The photophysical parameters
noted in this figure are taken from Table 4 in ACN at 298 K ACN under deaer-
ated conditions. See the text.
magnitude of the stabilization of the CT state is solvent depen-
dent. The degree of stabilization could explain the different
observed intensity ratios between the different emission bands
when changing from ACN, to MeOH–EtOH (1/1) and to DCM.
Computations, specifically TDDFT calculations of the five
lowest-energy triplet states, confirmed that the T1–T3 states are
close in energy to one another, supporting this explanation.
Importantly, the rigorous micro-analysis confirms the iden-
tity and purity of 2 during our studies. Additionally, excitation
spectra were invariant to emission wavelength and had fea-
tures that superimposed with the absorption spectrum of 2,
indicating that the luminescence originates from a single
excited species.
2 M. G. Colombo, A. Hauser and H. U. Güdel, Top. Curr.
Chem., 1994, 171, 143–171.
3 (a) W. J. Finkenzeller and H. Yersin, Chem. Phys. Lett., 2003,
377, 299–305; (b) T. Hofbeck and H. Yersin, Inorg. Chem.,
2010, 49, 9290–9299; (c) W. J. Finkenzeller, T. Hofbeck,
M. E. Thompson and H. Yersin, Inorg. Chem., 2007, 46,
5076–5083.
4 J. Li, P. I. Djurovich, B. D. Alleyne, M. Yousufuddin,
N. N. Ho, J. C. Thomas, J. C. Peters, R. Bau and
M. E. Thompson, Inorg. Chem., 2005, 44, 1713–1727.
5 (a) Y.-S. Yeh, Y.-M. Cheng, P.-T. Chou, G.-H. Lee,
C.-H. Yang, Y. Chi, C.-F. Shu and C.-H. Wang, Chem-
PhysChem, 2006, 7, 2294–2297; (b) K. K.-W. Lo, C.-K. Chung
and N. Zhu, Chem.–Eur. J., 2006, 12, 1500–1512; (c) K. K.-
W. Lo, K. Y. Zhang, S.-K. Leung and M.-C. Tang, Angew.
Chem., Int. Ed., 2008, 47, 2213–2216.
Conclusions
6 (a) D. N. Kozhevnikov, V. N. Kozhevnikov, M. Z. Shafikov,
A. M. Prokhorov, D. W. Bruce and J. A. Gareth Williams,
Inorg. Chem., 2011, 50, 3804–3815; (b) Y. You, Y. Han,
Y.-M. Lee, S. Y. Park, W. Nam and S. J. Lippard, J. Am.
Chem. Soc., 2011, 133, 11488–11491; (c) Y. You, S. Lee,
T. Kim, K. Ohkubo, W.-S. Chae, S. Fukuzumi, G.-J. Jhon,
W. Nam and S. J. Lippard, J. Am. Chem. Soc., 2011, 133,
18328–18342.
7 C. H. Shin, J. O. Huh, M. H. Lee and Y. Do, Dalton Trans.,
2009, 6476–6479.
8 A. P. Gaywood and H. McNab, J. Org. Chem., 2009, 74,
4278–4282.
A
new neutral heteroleptic iridium complex, [(ppy)2Ir-
(pythdo)], has been synthesized and characterized spectro-
scopically and electrochemically and the structure of which has
been ascertained unequivocally by 2D NMR and X-ray crystallo-
graphic analyses. The incorporation of the pythdo ligand sig-
nificantly blue-shifts the emission maximum for this
compound compared to [(ppy)2Ir(bpy)]PF6, presenting a much
more complex spectroscopic profile. The nature of the struc-
tured luminescence is the result of a dual emission process
resulting from a higher energy ligand-centered pythdo-domi-
nated state that was very sensitized to oxygen quenching and a
lower energy charge-transfer state. This complex represents a
rare example of dual emission from an iridium(III) species.
9 Y. Katoh, Y. Tsujimoto, C. Yamamoto, T. Ikai, M. Kamigaito
and Y. Okamoto, Polym. J., 2011, 43, 84–90.
10 R. D. Costa, E. Ortì, H. J. Bolink, S. Graber, S. Schaffner,
M. Neuburger, C. E. Housecroft and E. C. Constable, Adv.
Funct. Mater., 2009, 19, 3456–3463.
Acknowledgements
Computational time on the Mammouth supercomputer at the 11 (a) C.-J. Chang, C.-H. Yang, K. Chen, Y. Chi, C.-F. Shu,
Université de Sherbrooke was underwritten by the RQCHP
M.-L. Ho, Y.-S. Yeh and P.-T. Chou, Dalton Trans., 2007,
8846 | Dalton Trans., 2013, 42, 8838–8847
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
1881–1890; (b) A. J. Hallett, B. D. Ward, B. M. Kariuki and
S. J. A. Pope, J. Organomet. Chem., 2010, 695, 2401–2409;
(c) M. Xu, R. Zhou, G. Che, G. Wang, Z. Wang and Q. Xiao,
J. Lumin., 2011, 131, 909–914.
D. Klink, O. J. Schmitz, L. González, M. Schäferling and
E. Holder, Eur. J. Inorg. Chem., 2010, 2010, 4875–4885;
(f) A. Ruggi, M. Mauro, F. Polo, D. N. Reinhoudt, L. De
Cola and A. H. Velders, Eur. J. Inorg. Chem., 2012, 2012,
1025–1037; (g) P. Brulatti, R. J. Gildea, J. A. K. Howard,
V. Fattori, M. Cocchi and J. A. G. Williams, Inorg. Chem.,
2012, 51, 3813–3826; (h) M. Marín-Suárez, B. F. E. Curchod,
I. Tavernelli, U. Rothlisberger, R. Scopelliti, I. Jung, D. Di
Censo, M. Grätzel, J. F. Fernández-Sánchez, A. Fernández-
Gutiérrez, M. K. Nazeeruddin and E. Baranoff, Chem.
Mater., 2012, 24, 2330–2338.
12 (a) S. Ladouceur, D. Fortin and E. Zysman-Colman, Inorg.
Chem., 2010, 49, 5625–5641; (b) S. Ladouceur, D. Fortin
and E. Zysman-Colman, Inorg. Chem., 2011, 50, 11514–
11526.
13 E. Orselli, R. Q. Albuquerque, P. M. Fransen, R. Frohlich,
H. M. Janssen and L. De Cola, J. Mater. Chem., 2008, 18,
4579–4590.
14 N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96, 22 (a) L. Huynh, Z. Wang, J. Yang, V. Stoeva, A. Lough,
877–910.
I. Manners and M. A. Winnik, Chem. Mater., 2005, 17,
4765–4773; (b) Y. Li, Y. Liu and M. Zhou, Dalton Trans.,
2012, 41, 2582–2591; (c) M.-L. Ho, Y.-A. Chen, T.-C. Chen,
P.-J. Chang, Y.-P. Yu, K.-Y. Cheng, C.-H. Shih, G.-H. Lee and
H.-S. Sheu, Dalton Trans., 2012, 41, 2592–2600; (d) M. C.
DeRosa, P. J. Mosher, G. P. A. Yap, K. S. Focsaneanu,
R. J. Crutchley and C. E. B. Evans, Inorg. Chem., 2003, 42,
4864–4872; (e) Z. Xie, L. Ma, K. E. deKrafft, A. Jin and
W. Lin, J. Am. Chem. Soc., 2009, 132, 922–923; (f) M. Marin-
Suarezdel Toro, J. F. Fernandez-Sanchez, E. Baranoff,
M. K. Nazeeruddin, M. Graetzel and A. Fernandez-Gutier-
rez, Talanta, 2010, 82, 620–626; (g) G. Di Marco, M. Lanza,
M. Pieruccini and S. Campagna, Adv. Mater., 1996, 8, 576–
580; (h) Y. Amao, Y. Ishikawa and I. Okura, Anal. Chim.
Acta, 2001, 445, 177–182; (i) M. E. Köse, R. J. Crutchley,
M. C. DeRosa, N. Ananthakrishnan, J. R. Reynolds and
K. S. Schanze, Langmuir, 2005, 21, 8255–8262.
15 S. R. Meech and D. Phillips, J. Photochem., 1983, 23, 193–217.
16 H. Ishida, S. Tobita, Y. Hasegawa, R. Katoh and K. Nozaki,
Coord. Chem. Rev., 2010, 254, 2449–2458.
17 F. Neve, M. LaDeda, A. Crispini, A. Bellusci, F. Puntoriero
and S. Campagna, Organometallics, 2004, 23, 5856–5863.
18 (a) R. D. Costa, E. Ortí, H. J. Bolink, S. Graber, C. E.
Housecroft, M. Neuburger, S. Schaffner and E. C.
Constable, Chem. Commun., 2009, 2029–2031; (b) H. J.
Bolink, E. Coronado, R. D. Costa, E. Ortì, M. Sessolo,
S. Graber, K. Doyle, M. Neuburger, C. E. Housecroft and
E. C. Constable, Adv. Mater., 2008, 20, 3910–3913.
19 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.
20 R. D. Costa, F. Monti, G. Accorsi, A. Barbieri, H. J. Bolink,
E. Ortí and N. Armaroli, Inorg. Chem., 2011, 50, 7229–7238.
21 (a) J. N. Demas, E. W. Harris and R. P. McBride, J. Am.
Chem. Soc., 1977, 99, 3547–3551; (b) A. Ruggi, M. B. Alonso, 23 (a) T. Yoshihara, Y. Yamaguchi, M. Hosaka, T. Takeuchi
D. N. Reinhoudt and A. H. Velders, Chem. Commun., 2010,
46, 6726–6728; (c) C. S. K. Mak, D. Pentlehner, M. Stich,
O. S. Wolfbeis, W. K. Chan and H. Yersin, Chem. Mater.,
2009, 21, 2173–2175; (d) P. I. Djurovich, D. Murphy,
M. E. Thompson, B. Hernandez, R. Gao, P. L. Hunt and
M. Selke, Dalton Trans., 2007, 3763–3770; (e) N. Tian,
D. Lenkeit, S. Pelz, L. H. Fischer, D. Escudero, R. Schiewek,
and S. Tobita, Angew. Chem., Int. Ed., 2012, 51, 4148–4151;
(b) A. L. Medina-Castillo, J. F. Fernandez-Sanchez, C. Klein,
M. K. Nazeeruddin, A. Segura-Carretero, A. Fernandez-
Gutierrez, M. Graetzel and U. E. Spichiger-Keller, Analyst,
2007, 132, 929–936; (c) S. Zhang, M. Hosaka, T. Yoshihara,
K. Negishi, Y. Iida, S. Tobita and T. Takeuchi, Cancer Res.,
2010, 70, 4490–4498.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8838–8847 | 8847