Iridium(III) Sensitisers and Energy Upconversion: The Influence of Ligand Structure upon TTA-UC Performance

Article abstract of DOI:10.1002/chem.202004146

Six substituted ligands based upon 2-(naphthalen-1-yl)quinoline-4-carboxylate and 2-(naphthalen-2-yl)quinoline-4-carboxylate have been synthesised in two steps from a range of commercially available isatin derivatives. These species are shown to be effective cyclometallating ligands for Ir^III, yielding complexes of the form [Ir(C^N)₂(bipy)]PF₆ (where C^N=cyclometallating ligand; bipy=2,2′-bipyridine). X-ray crystallographic studies on three examples demonstrate that the complexes adopt a distorted octahedral geometry wherein a cis-C,C and trans-N,N coordination mode is observed. Intraligand torsional distortions are evident in all cases. The Ir^III complexes display photoluminescence in the red part of the visible region (668–693 nm), which is modestly tuneable through the ligand structure. The triplet lifetimes of the complexes are clearly influenced by the precise structure of the ligand in each case. Supporting computational (DFT) studies suggest that the differences in observed triplet lifetime are likely due to differing admixtures of ligand-centred versus MLCT character instilled by the facets of the ligand structure. Triplet–triplet annihilation upconversion (TTA-UC) measurements demonstrate that the complexes based upon the 1-naphthyl derived ligands are viable photosensitisers with upconversion quantum efficiencies of 1.6–6.7 %.

Full text of DOI:10.1002/chem.202004146

Accepted Article
Accepted Article
Title: Iridium(III) Sensitizers and Energy Upconversion: The Influence
of Ligand Structure Upon TTA-UC Performance
Authors: Simon J. Pope, Christopher E. Elgar, Haleema Y. Otaif, Xue
Zhang, Jianzhang Zhao, Joseph M. Beames, Peter N. Horton,
and Simon J. Coles
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202004146
Link to VoR: https://doi.org/10.1002/chem.202004146
01/2020

10.1002/chem.202004146
Chemistry - A European Journal
Iridium(III) Sensitizers and Energy Upconversion: The Influence of
Ligand Structure Upon TTA-UC Performance
Christopher E. Elgar,Ɨa Haleema Y. Otaif,Ɨa Xue Zhang,Ɨ b Jianzhang Zhao,b Peter N.
Horton,c Simon J. Coles,c Joseph M. Beames,*a and Simon J.A. Pope*a
aSchool of Chemistry, Main Building, Cardiff University, Cardiff CF10 3AT,
Cymru/Wales, UK; b State Key Laboratory of Fine Chemicals
Dalian University of Technology, Dalian 116024 (P.R. China);
cUK National Crystallographic Service, Chemistry, Faculty of Natural and
Environmental Sciences, University of Southampton, Highfield, Southampton, SO17
1BJ, England, UK.
Email: popesj@cardiff.ac.uk; beamesj@cardiff.ac.uk
Ɨ equal contributions
Abstract
Six substituted ligands based upon 2-(naphthalen-1-yl)quinoline-4-carboxylate and
2-(naphthalen-2-yl)quinoline-4-carboxylate have been synthesized in two steps from
a range of commercially available isatin derivatives. These species are
demonstrated to be effective cyclometalating ligands for Ir(III), yielding complexes of
the form [Ir(C^N)2(bipy)]PF6 (where C^N = cyclometalating ligand; bipy = 2,2’-
bipyridine). X-ray crystallographic studies on three examples demonstrate that the
complexes adopt a distorted octahedral geometry wherein a cis-C,C and trans-N,N
coordination mode is observed. Intra-ligand torsional distortions are evident in all
cases. The Ir(III) complexes display photoluminescence in the red part of the visible
region (668-693 nm) which is modestly tuneable via the ligand structure. The triplet
lifetimes of the complexes are clearly influenced by the precise structure of the
ligand in each case. Supporting computational (DFT) studies suggest that the
differences in observed triplet lifetime are likely due to differing admixtures of ligand-
centred versus MLCT character instilled by the facets of the ligand structure. Triplet-
triplet annihilation upconversion (TTA-UC) measurements demonstrate that the
complexes based upon the 1-naphthyl derived ligands are viable photosensitizers
with upconversion quantum efficiencies of 1.6-6.7 %.
1
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Introduction
Luminescent cyclometalated iridium(III) complexes are of major interest in
optoelectronic and photonic applications.1 Such complexes have demonstrated
applications in research fields including bioimaging,2 electroluminescence,3
photoredox catalysts,4 non-linear optics,5 and chemosensors.6 An intimate
understanding of the electronic characteristics of the complex7 can provide control
over the luminescence properties of these species and, thus, careful consideration
of the ligand design is highly desirable to maximise their applicability.8
The use of transition metal complexes as photosensitizers for energy
upconversion processes has also gathered increasing attention over the last
decade.9 In particular, the use of triplet sensitizers via a triplet-triplet annihilation
(TTA) mechanism has led to several studies on TTA upconversion (TTA-UC). This is
a bimolecular process that requires energy transfer between a sensitizer and an
annihilator species. As the Dexter energy transfer mechanism requires collisional
interactions, a long-lived excited state triplet lifetime is clearly advantageous.
Therefore, a variety of metal-based photosensitizers have been investigated in this
regard.
Castellano and co-workers have reported several important studies that
demonstrate the utility of phosphorescent metal complexes in TTA-UC. Early work
described the use of the classical metal-to-ligand charge transfer (MLCT) species
[Ru(dmb)3]2+ (dmb = 4,4,dimethyl-2,2’-bipyridine) as a photosensitizer for solution-
based TTA-UC with 9,10-diphenylanthracene (DPA) as the annihilator species.10 The
ease with which luminescent complexes can be functionalised, whilst retaining their
inherent photophysical attributes, has been demonstrated in the development of a
water soluble Ru(II) analogue, a bathophenanthroline disulfonate variant of
[Ru(N^N)3]2+, which was successfully utilised as a triplet photosensitiser for TTA-UC
under aqueous conditions.11 To optimise the photosensitizer absorption
characteristics, a multi-component system comprising a NIR absorbing Ru(II)
complex unit conjugated with a dimetallic Zn(II) di-porphyrin chromophore was
developed, which displayed TTA-UC when used in conjunction with a N,N-
bis(ethylpropyl)perylene-3,4,9,10-tetracarboxylicdiimide or tetracene annihilator
species.12 Importantly, earth abundant Cu(I) MLCT phosphorescent species based
2
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
on substituted 1,10-phenanthroline ligands have also been studied as viable
photosensitizers for TTA-UC.13
Organometallic iridium(III) complexes have also shown excellent application in
TTA-UC, benefiting from attributes such as good photostability, efficient intersystem
crossing (ISC) and long triplet lifetimes.14 Ma and co-workers have described a
number of energy upconverting systems, including tethered systems, based on tris-
cyclometalated Ir(III) photosensitizers with 2-phenylpyridine (ppy) type ligands.15,16
Some of the iridium-pyrene bichromophoric systems benefit from remarkable
millisecond triplet lifetimes, which enhances TTA-UC.16a Zhao and co-workers have
also reported a significant number of TTA-UC studies that exploit and optimise the
properties of Ir(III) photosensitizers.17 These include strategies to address the visible
light absorption characteristics, and elongate the excited state triplet lifetimes of the
photosensitizer species.
Our own contribution to the area reported a series of cationic Ir(III) complexes
as triplet sensitizers which have shown the highest yet reported TTA-UC efficiency of
39% using DPA as the annihilator.18 These Ir(III) sensitizer complexes were based
19,20
upon a 2-phenylquinoxaline ligand framework,
which impart good absorption
characteristics in the visible region. Because of the increase in conjugation of the
chelating ligands, these characteristics also provide bathochromically shifted
emission compared to the ubiquitous Ir(III) complexes based on ppy.21,22
Indeed, red-emitting Ir(III) species are attracting significant attention, being
applicable to a number of other optoelectronic applications. A common strategy for
red-shifting the emission wavelengths of Ir(III) species is to increase the conjugation
of the chelating ligands. As an example, comparing [Ir(ppy)3] versus [Ir(piq)3]) (ppy =
2-phenylpyridine; piq = 1-phenylisoquinoline) shows a bathochromic shift in the
emission for the latter to around 620 nm.23 Relatively simple evolutions of this design
strategy can provide further extensions of em into the near-IR (>750 nm) region.24 Of
course, previous studies have investigated a variety of different aryl groups as
components of cyclometalating C^N ligands, including early, seminal25 studies on
luminescent Ir(III) complexes.26 Other recent examples of red phosphorescence from
Ir(III) species have been achieved through the use of conjugated triazole,27
conjugated phenazine,28 and cyclometalating phenylquinazoline29 ligands. Teets and
co-workers have also described a systematic study of a series of anionic ancillary
3
3
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
ligands which facilitate excellent control over the emitting state energies of their Ir(III)
complexes.30 The addition of bulky groups to the ancillary ligand was shown as a
useful strategy for improving the quantum yields of these deep red phosphorescing
species. We have also reported the use of substituted 2-phenylquinoline31
cyclometalating (C^N) ligands for red emitting Ir(III) complexes. It is noteworthy that
red-emitting luminophores are important in bioimaging disciplines which use confocal
fluorescence microscopy techniques.32
Herein we report the photophysical properties of a series of new
[Ir(C^N)2(bipy)]PF6 complexes formed from substituted 2-(naphthalen-1-yl)quinoline-
4-carboxylate and 2-(naphthalen-2-yl)quinoline-4-carboxylate as cyclometalating
ligands. Of direct relevance to the current study, Li et al have previously investigated
the influence of -conjugation in unsubstituted naphthyl-pyridine type ligands upon
the non-linear absorbing properties of a series of Ir(III) complexes;33 the complexes
luminesce in the yellow-orange region with some noted photophysical differences
that result from the isomeric form of the ligand. In our current work, we show that a
new naphthylquinoline ligand framework can further bathochromically shift the
absorption and emission properties of the Ir(III) complexes, and several species
have potential application as photosensitizers in TTA-UC.
4
4
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Results and Discussion
Synthesis and characterization of the ligands
The synthesis of the ligands was achieved using the Pfitzinger34 reaction, which
dates back to the 1880s.35,36 This reaction involves aryl acetyl and isatin reagents
and provides a convenient and efficient synthetic route towards substituted 2-aryl-
quinoline-4-carboxylic
acids.
Presently,
1-acetylnaphthalene
and
2-
acetylnaphthalene were employed as the aryl acetyl reagent; each was reacted with
three different, commercially available isatins (isatin, 5-methoxyisatin and 5-
fluoroisatin) to yield the subsequent 2-(naphthyl)quinoline-4-carboxylic acids. These
carboxylic acid derivatives were then esterified to improve solubility in common
organic solvents (Scheme 2).
1
For LH1 and LH4 and their corresponding carboxylic acids H NMR data was
consistent with the formulation of these previously reported compounds.37 Finally, to
optimize solubility for the subsequent Ir(III) coordination chemistry, the carboxylic
acids were esterified to give 2-(naphthalen-2-yl)quinoline-4-carboxylic esters, LH1,
LH2 and LH3 (where R = H, OMe and F, respectively), and 2-(naphthalen-1-
yl)quinoline-4-carboxylic esters, LH4, LH5 and LH6 (where R = H, OMe and F,
respectively). Thus, a series of structurally related, disubstituted ligands were
conveniently isolated in two synthetic steps (Scheme 1).
<Scheme 1>
1
All new ligands were fully characterised using a range of standard methods. H and
13C{1H} NMR spectroscopies were consistent with the proposed formulations. For
LH3 and LH6, 19F{1H} NMR was also used showing resonances at -110.4 and -109.9
ppm, respectively. Satisfactory HRMS data was obtained for each new esterified
ligand.
Coordination chemistry
The six ligands (LH1-6) were then explored as cyclometalating (C^N) agents for Ir(III).
The adopted approach for the Ir(III) precursor utilized standard reaction conditions
based upon those first reported by Nonoyama38 wherein reaction with hydrated IrCl3
proceeds to the dimeric species, [{Ir(C^N)2(-Cl)}2]. The target monometallic Ir(III)
5
5
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
complexes [Ir(C^N)2(bipy)]PF6 were then synthesised via the intermediate species
cis-[Ir(MeCN)2(C^N)2]BF4. All complexes (Scheme 2) were obtained as red coloured,
air-stable powders following chromatographic purification on silica gel (see
Experimental Section, SI).
<Scheme 2>
Characterization of the Complexes
1H NMR spectroscopic studies on the complexes showed shifts in the aromatic
protons of the ligands that were indicative of chelation. Firstly, the 1H NMR spectrum
of the reaction product with LH2 indicated quantitative conversion to the
ethoxyethanol ester analogue (supported by four resonances in the 13C NMR
spectrum at 70-55 ppm). The indicative transesterification in 2-ethoxyethanol solvent
at elevated temperatures has been observed previously, and the resultant complex
product is thus labelled [Ir(L2b)2(bipy)]PF6 (Scheme 2).39
In each complex, a
relatively upfield resonance between ca. 6.9 and 7.0 ppm was noted for the proton
adjacent to the cyclometalated carbon atom. Both [Ir(L2b)2(bipy)]PF6 and
[Ir(L5)2(bipy)]PF6 have an additional upfield signal at 6.58 and 6.60 ppm respectively.
1
Using H-1H Correlation Spectroscopy on [Ir(L2b)2(bipy)]PF6 (Fig. S1), this was
assigned to the proton on C7 of the quinoline ring, which is adjacent to the methoxy
group. In the cases of [Ir(L2b)2(bipy)]PF6 and [Ir(L5)2(bipy)]PF6 the methoxy
substituent appears ca. 4 ppm.
For [Ir(L3)2(bipy)]PF6 and [Ir(L6)2(bipy)]PF6, 19F{1H} NMR spectra were also
obtained (Table S1), firstly showing the doublet (1JFP) feature associated with the
hexafluorophosphate anion at -72.6 ppm, and secondly, a resonance around -107.5
ppm attributed to the fluorine substituent of the quinoline ring. The latter represents a
subtle downfield shift upon coordination of the ligand to the Ir(III) centre.
HRMS data were obtained for all six complexes showing the correct isotopic
distribution for the complex cation in each case. For [Ir(L2b)2(bipy)]PF6 the MS data
again supported the isolation of the ethoxyethanol ester as evidenced by the NMR
studies.
6
6
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
X-ray Crystallographic Studies
Single, red plate-shaped crystals from three of the complexes, [Ir(L3)2(bipy)]PF6,
[Ir(L4)2(bipy)]PF6, and [Ir(L6)2(bipy)]PF6, were successfully isolated from vapour
diffusion of diethyl ether into an acetonitrile solution of the complex, and investigated
using X-ray diffraction. Although of lower quality, the data for [Ir(L6)2(bipy)]PF6 was
still sufficient to allow determination of the key features of the structure and is also
included here. Data collection parameters are in Table S2 (SI); selected bond
lengths and bond angles associated with the coordination spheres are in Table S3
(SI). In each case the structures confirmed the anticipated coordination sphere at
iridium, with the 2-(naphthalen-1-yl)quinoline-4-carboxylate and 2-(naphthalen-2-
yl)quinoline-4-carboxylate derived ligands coordinating in the expected
cyclometalating fashion.
Hypothetically, there are two available sites of
cyclometalation for the 2-(naphthalen-2-yl)quinoline-4-carboxylate ligand. The
structure of [Ir(L3)2(bipy)]PF6 shows the ligand coordinates at the 3 position of the
naphthyl ring. The reaction conditions employed here are known to promote the cis-
C,C and trans-N,N coordination mode of cyclometalation at Ir(III), which was
confirmed in all three structures.40 [Ir(L3)2(bipy)]PF6 also shows disorder over one of
the ethoxy groups on the ligand backbone.
A comparison of the Ir-C bond lengths suggests that there is little influence
from the particular isomeric form of the naphthyl donor, and the Ir-C and Ir-N bond
lengths are closely comparable to other relevant examples.19 In all cases significant
distortions from ideal octahedral geometry are noted, with N−Ir−N bond angles, that
originate from the trans-N,N arrangement, of 172.09(10), 174.57(12) and 175.2(4)
for [Ir(L3)2(bipy)]PF6, [Ir(L4)2(bipy)]PF6, and [Ir(L6)2(bipy)]PF6, respectively. Intra-
ligand distortions are also noted, with appreciable torsion angles between the
naphthyl and quinoline rings of the cyclometalating ligands, particularly for
[Ir(L4)2(bipy)]PF6.
<Figure 1>
<Figure 2>
<Figure 3>
7
7
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
The Continuous Shape Measures (CShM) software41 was used to evaluate the
polyhedral environment around the iridium ion and its deviation from an ideal
octahedron. The lowest CShM values define the best ideal polyhedron; values > 0.1
are considered chemically significant distortions, while > 3 indicate important
distortions. For these three complexes, the values against an ideal octahedron are
as follows: 1.952, 1.915, 1.619 for [Ir(L3)2(bipy)]PF6, [Ir(L4)2(bipy)]PF6
[Ir(L6)2(bipy)]PF6, respectively. Thus, although [Ir(L3)2(bipy)]PF6 has a different
isomeric form of ligand, it does not lead to a major comparative distortion.
Redox properties of the Ir(III) complexes
The electrochemical characteristics of the complexes were studied in deoxygenated
dichloromethane. The cyclic voltammograms (Fig. S2) were measured using a
platinum disc electrode (scan rate υ = 200 mV s-1, 1 × 10-3 M solutions, 0.1 M
[NBu4][PF6] as a supporting electrolyte). Each complex showed one fully reversible
oxidation (Table 1) at +1.26 to +1.39 V which is ascribed to the Ir3+/4+ couple. The
small differences in oxidation potential are likely due to two factors: the isomeric form
of the cyclometalating unit (2-naphthyl vs. 1-naphthyl), and the substitutent (H vs.
OMe vs. F) on the quinoline ring; both factors may influence the electron density at
the iridium centre. For a given ligand substituent, the 1-naphthyl derived complexes
showed a slightly higher Ir3+/4+ potential. For a given naphthyl isomer, the methoxy-
substituted complexes possess the lowest Ir3+/4+ potentials suggesting that the Ir3+ is
more easily oxidisable, whereas fluorinated complexes with the highest Eox values,
consistent with a more electron deficient quinoline ligand. The 1-naphthyl derivatives
[Ir(L4-6)2(bipy)]PF6 also showed two fully reversible reductions which were attributed
to one-electron, ligand-centred processes. These reduction features also varied
according to the nature of the substituent implying that both processes may be
associated with the cyclometalated ligand. The 2-naphthyl variants
[Ir(L1,2b,3)2(bipy)]PF6 appeared less electrochemically stable with non-reversible
features within the reduction couples.
8
8
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Table 1. Electrochemical properties of the iridium(III) complexes obtained from cyclic
voltammetry.a
Oxidation
Reduction
E(red 2) / V
Complex
E1/2 / V
E(red 1) / V
[Ir(L1)2(bipy)]PF6
[Ir(L2b)2(bipy)]PF6
[Ir(L3)2(bipy)]PF6
[Ir(L4)2(bipy)]PF6
[Ir(L5)2(bipy)]PF6
[Ir(L6)2(bipy)]PF6
+1.28
+1.26
+1.32
+1.37
+1.29
+1.39
-1.10b
-1.14b
-0.93b
-1.05c
-1.11c
-0.97c
-1.27b
-1.31b
-1.21b
-1.29c
-1.33c
-1.22c
a
potentials measured in CH2Cl2 solutions at 200 mVs–1 with 0.1 M [NBu4][PF6] as supporting
b
c
electrolyte calibrated with Fc/Fc+; cathodic peak of irreversible process;
E1/2 values for fully
reversible process.
Density functional theory (DFT)
Discussion below focuses on two of the complexes that represent the key isomeric
variants within the series, [Ir(L1)2(bipy)]+ and [Ir(L4)2(bipy)]+. These were optimized in
their lowest electronic singlet and triplet states at the DFT//B3LYP/6-31G* level of
theory, with an implicit CH3CN solvent (IEFPCM). The optimized geometries of both
complexes obtained by DFT method were confirmed through harmonic vibrational
frequency calculations and are shown in Fig. S3. The modelled structures exhibit
near C2 symmetry, with a distorted octahedral geometry around the iridium metal
center. The DFT-derived optimized bond lengths and angles are in good agreement
with the X-ray experimental data, with typically less than 0.05 Å (bond distances) and
1° (bond angles) difference between optimized geometric parameters and those
derived from experiment (Table S4). It is worth noting that the DFT bond lengths are
typically longer than those reported in the X-ray crystal structure(s).
9
9
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Molecular orbital decomposition analysis (Tables 2, 3 and 4) indicates that the
HOMO for both Ir(III) complexes is made predominantly of contributions from the
naphthyl moiety of both C^N ligands, and a significant, but smaller contribution from
the iridium centre. For example, in the case of the 2-naphthyl derivative
[Ir(L1)2(bipy)]+, the HOMO is mainly localized over the π-orbitals of the naphthyl
moiety of both C^N ligands (36%, 34%) and the 5d-orbitals of Ir(III) (28%), with just a
2% contribution from the bipyridine. The prediction, that a metal d-orbital contribution
to the HOMO is less than ligand contributions is similar to other Ir(III) quinoxaline
systems reported previously. The 1-naphthyl derivative [Ir(L4)2(bipy)]+ also predicts
an Ir contribution to the HOMO of 21%, with equal contributions from naphthyl
moieties on both C^N ligands (39%). The HOMO-1 of the two complexes is divided
between the naphthyl moieties of both ligands (around 49%) and a negligible
contribution from the metal centre (≤ 3%).
In both cases, the LUMO is predicted to reside predominantly on the quinoline
part of the C^N ligands. It is noticeable that the two ligands form pairs of pseudo-
degenerate orbitals with alternating contributions from each of the C^N ligands: e.g.
the [Ir(L1)2(bipy)] LUMO is comprised of contributions of 26% and 71% from the
quinolines, Q1 and Q2 respectively, whilst the LUMO+1 shows the reverse (Q1 =
71%; Q2 = 26%). This effect is due to the near C2 symmetry of the system, with
subtle geometrical distortions leading to a lowering of the total symmetry and thus a
splitting in the orbitals. The metal center and bipyridine make little contribution to the
LUMO or LUMO+1; it is only in the higher energy LUMO+2 that a dominant
contribution (97%) of the bipyridine ligand is predicted. The full list of calculated MO
contributions and excited state descriptions and associated transitions for all of the
complexes are presented in the Supplementary Information (Tables S5-S15).
TD-DFT was used in conjunction with the long-range corrected CAM-B3LYP
functional to investigate the predicted UV-Vis electronic transitions of these
complexes. All calculations were performed from the optimized ground state singlet
(S0) and lowest triplet (T1) state geometries. The calculated transition energies are in
reasonable agreement with the experimental UV-Vis absorption spectra discussed
later (Fig. 4). The lowest spin-forbidden singlet-triplet transition is predicted to occur
3
at 570 nm for [Ir(L1)2(bipy)]+, attributed to a MLCT contribution, and correlates well
10
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
with low intensity absorption tail at 550 nm in the experimental absorption spectra.
The lowest singlet excited state is predicted to be an intense HOMO to LUMO
transition at 407 nm, followed at higher energy by an intense HOMO to LUMO+1
transition at 402 nm (f = 0.0375), both of which suggest a combination of 1MLCT and
C^N ligand-centered character. These transitions are in good agreement with the
experimental absorption bands at ca. 350 nm <  < 425 nm. The TD-DFT also
predicts intense transitions between 350 nm <  < 300 nm, including transitions with
1
much more mixed ILCT/1LLCT/1MLCT character. In general, as one moves to
shorter wavelength (higher energy), the TD-DFT calculations suggest an increase in
ligand-centered transitions, as might be expected for quinoline, which has
unperturbed UV-Vis band onsets at ca. 310 nm. The simulations for [Ir(L4)2(bipy)]+
predict (see also Table S12, SI) similar features albeit with a bathochromic shift
compared to [Ir(L1)2(bipy)]+; again this is supported by the experimental data
discussed later.
11
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Table 2. Calculated Kohn-Sham molecular orbitals for [Ir(L1)2(bipy)]+. Calculated spin
densities are reported in the SI (Fig. S4 and Table S5).
HOMO
HOMO-1
HOMO-2
HOMO-3
LUMO
LUMO+1
LUMO+2
LUMO+3
Table 3. A description of the calculated MO contributions, excited state descriptions
and their associated transitions for [Ir(L1)2(bipy)]X (Q1 and Q2 are the different
naphthylquinoline ligands; bipy = bipyridine).
Moiety Contribution to Orbital
Orbital Contribution to Excited State
(%)
Orbital
Ir (5d)
bipy
23
44
97
0
Q1 Q2 Excited State Contributing Transitions (> 10%)
LUMO +4
LUMO +3
1
1
17
26
1
59 1 (407 nm f=0.0039) HOMO → LUMO (83.84%)
29
LUMO +2
LUMO +1
LUMO
2 (403 nm f=0.0375) HOMO → LUMO +1 (82.53%)
2
1
3
71
26
36
48
45
45
35
26
71
34
49
45
45
2
1
HOMO
3 (356 nm f=0.4075) HOMO -2 → LUMO +1 (22.16%)
HOMO -1 → LUMO (31.29%)
28
3
2
HOMO -1
HOMO -2
HOMO -3
HOMO -4
1
4 (352 nm f=0.3415) HOMO -2 → LUMO (31.29%)
HOMO -1 → LUMO +1(40.03%)
10
9
41
0
1
5
19 5 (344 nm f=0.0015) HOMO → LUMO+2 (84.16%)
12
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Table 4. A description of the calculated MO contributions, excited state descriptions
and their associated transitions for [Ir(L4)2(bipy)]+ (Q1 and Q2 are the different
quinoline ligands; bipy = bipyridine).
Moiety Contribution to Orbital
(%)
Orbital Contribution to Excited State
Orbital
Ir (5d)
Bipy
Q1 Q2 Excited State
Contributing Transitions (> 10%)
LUMO +4
LUMO +3
2
2
6
23
46
34
46 1 (432 nm f=0.1466) HOMO-1 → LUMO (11.70%)
HOMO → LUMO (63.65%)
41
HOMO → LUMO +1 (11.52%)
LUMO +2
LUMO +1
LUMO
2
4
4
21
2
16
37
97
1
1
1
0
1
1
17
78
2 (430 nm f=0.1722)
HOMO -1 → LUMO+1 (11.03%)
HOMO → LUMO (12.50%)
HOMO → LUMO +1 (62.32%)
79
16
39
49
42
51
HOMO
39 3 (350 nm f=0.1746)
49
41
HOMO -2 → LUMO+1 (11.62%)
HOMO -1 → LUMO (50.34%)
HOMO → LUMO +2 (13.23%)
HOMO -1
HOMO -2
HOMO -3
1
4
8
4 (347 nm f=0.1456) HOMO -2 → LUMO (14.66%)
HOMO -1 → LUMO +1 (58.09%)
HOMO -4
36
2
10
52 5 (343 nm f=0.0480) HOMO → LUMO+2 (70.56%)
TD-DFT was also used to compute the vertical and adiabatic spin-forbidden T1 → So
emission energies from the triplet state minimum energy geometries (Table S16).
These calculations suggest that phosphorescence bands should exist at long
wavelengths ( = 649 nm and 705 nm for [Ir(L1)2(bipy)]+ and [Ir(L4)2(bipy)]+,
respectively) which is in good qualitative agreement with the experimental data; the
band positions are overestimated for the calculated adiabatic emission profiles. As
3
with the So → Tn absorption, this band is predicted as possessing some MLCT
character. The molecular orbital decomposition analyses shows that the contribution
of the iridium to the HOMO orbital may be subtly dependent upon the nature of the
C^N ligand.42
Absorption and Emission properties of the Ir(III) complexes
UV-vis. absorption spectra were obtained on MeCN solutions of the complexes. The
spectra (Fig. 4) reveal a composite of bands between 220-550 nm. Ligand centered
bands from the various aromatic components are anticipated to contribute to the
more intense absorptions between 220-400 nm. Previous studies on related Ir(III)
complexes suggest that spin allowed metal-to-ligand charge transfer (1MLCT)
3
absorptions contribute at 350-450 nm with spin forbidden absorptions to MLCT
13
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
excited states likely to result weaker features >450 nm.18 The complexes herein
demonstrate a similar pattern where the MLCT absorptions tail to around 600 nm,
which is attributed to the extended conjugation of the cyclometalating ligands. Within
the series it is clear that [Ir(L4-6)2(bipy)]PF6 possess relatively bathochromically
shifted MLCT features compared to [Ir(L1,2b,3)2(bipy)]PF6. This is consistent with the
predictions from the DFT calculations. Within these structural variants, the addition of
substituents (H vs. OMe vs. F) to the quinoline ring induces only minor changes in
the position of these absorption bands (Table 5).
Table 5. Absorbance and emission data for the complexes in aerated MeCN at room
temperature.a
λem / nmb
λabs ( 104 M-1 cm-1) / nma
 /nsc
  d
Complex
[Ir(L1)2(bipy)]PF6
464 (0.2), 365 (2.9), 315 (3.3), 283
(5.4), 265 (6.6), 218 (7.4)
480 (0.2), 401 (1.6), 367 (1.7), 321
(3.5), 280 (4.9), 221 (5.7)
483 (0.2), 399 (1.6), 365 (2.1), 268
(5.5)
494 (0.5), 405 (1.4), 366 (2.0), 271
(9.2)
495 (0.6), 417 (1.2), 367 (2.1), 267
(5.0), 223 (7.6)
673
36
0.3
[Ir(L2b)2(bipy)]PF6
[Ir(L3)2(bipy)]PF6
[Ir(L4)2(bipy)]PF6
[Ir(L5)2(bipy)]PF6
[Ir(L6)2(bipy)]PF6
668
675
679
693
687
89
17
0.4
0.3
0.6
0.2
1.0
240
223
217
492 (0.5), 408 (1.3), 367 (2.2), 264
(5.0), 216 (8.2)
a aerated acetonitrile, 10-5 M; b λex = 510 nm; c λex = 295 nm; d Quantum yield with [Ru(bipy)3][PF6]2 as
standard (ΦΔ = 0.018 in acetonitrile).43
<Figure 4>
Aerated MeCN solutions (10-5 M) of the complexes were used to obtain
photophysical data (Table 5). Following irradiation at 510 nm (i.e. in the region of the
MLCT absorption bands) each of the complexes demonstrated emission in the deep
red region of the spectrum, peaking at 668-693 nm and tailing to 800 nm (Fig. 5).
The combination of the naphthyl and quinoline units within the C^N ligand framework
leads to a significant bathochromic shift in emission wavelength when compared to
either phenyl-quinoline or naphthyl-pyridine type systems reported by Sun and co-
workers.38 The peak appearances of [Ir(L4-6)2(bipy)]PF6 were broad and featureless,
but for [Ir(L1,2b,3)2(bipy)]PF6 the higher energy portion of the peak possessed some
14
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
weaker vibronic features. Corresponding excitation spectra were also recorded and
showed a close matching with the absorption spectral profiles (Fig. S5). The data
suggests that the emitting state of the complexes can be populated via both MLCT
and LC excited states.
As predicted by the DFT calculations, the 1-naphthyl species, [Ir(L4-
6)2(bipy)]PF6, were the most bathochromically shifted with subtle variations in em
according to the presence of the different quinoline substituents (H vs. OMe vs. F).
Under aerated conditions the observed lifetimes were significantly longer for the
[Ir(L4-6)2(bipy)]PF6 species (217-240 ns) compared to [Ir(L1,2b,3)2(bipy)]PF6 (17-89 ns).
The emission spectra were also collected in a range of solvents of varying polarity
(Fig. S6). The results suggest that the excited states of the complexes are relatively
insensitive to the polarity of the medium, unlike a classical MLCT species such as
[Ru(bipy)3]2+.43 Therefore in an admixture of LC and MLCT excited states the latter is
unlikely to be the dominant contribution in these complexes.
<Figure 5>
Low temperature measurements (77 K) of each complex were also obtained (Fig. 6)
using frozen glasses (4:1, EtOH:MeOH). For [Ir(L1,2b,3)(bipy)]PF6 the dominant peak
maxima were hypsochromically shifted from the room temperature spectra and
characterized by stronger vibronic structure, suggesting the possibility of a ligand-
centered contribution to the emitting states. Interestingly, for the fluoro-substituted
species, [Ir(L3)(bipy)]PF6, the peak appearance was broadened. In comparison, the
spectra for [Ir(L4-6)(bipy)]PF6 presented a well-defined peak maximum at 650-675 nm
with a much weaker shoulder feature to longer wavelength. Evidently, the vibronic
features in [Ir(L1,2b,3)(bipy)]PF6 were stronger than for [Ir(L4-6)(bipy)]PF6. The onset of
the low temperature steady state emission spectra give an approximation of the
triplet energy levels of the Ir(III) complexes. Using this approach the triplet levels of
[Ir(L1,2b,3)(bipy)]PF6 are ca. 17900 cm-1 and those of [Ir(L4-6)(bipy)]PF6 are ca. 16400
cm-1 (see later discussion). The emission lifetimes obtained at 77 K for
[Ir(L1,2b,3)2(bipy)]PF6 were extended to ca. 12 s, but for [Ir(L4-6)2(bipy)]PF6 they were
ca. 3 s. These observations, again, may point to differing and subtle admixtures of
LC/MLCT character to the emitting states of these complexes.
15
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
<Figure 6>
Transient Absorption Spectroscopy
Time-resolved transient absorption (TA) spectra were obtained for MeCN solutions
of [Ir(L4-6)2(bipy)]PF6. Figures 7-9 show the obtained TA spectra, together with the T1
→ S0 phosphorescence decay kinetics of the complexes. The appearance of the
spectra of [Ir(L4)2(bipy)]PF6 and [Ir(L6)2(bipy)]PF6 are comparable, with strong
positive features at 400 nm <  < 500 nm and 550 nm <  < 700 nm regions, which
are the excited state absorption (ESA) bands of the T1 state (T1 → Tn transitions). A
bleach at 350 <  < 400 nm is also evident. This general appearance is comparable
to that observed previously for Ir(III) 2-phenylquinoxaline complexes.18 Each feature
is associated with similar lifetime characteristics which suggests that each peak
relates to the same excitation, ISC process and deactivation. For [Ir(L4)2(bipy)]PF6,
the triplet state lifetime is 687 ns under N2 atmosphere, the triplet state is quenched
by O2 in the air, where experiments show the triplet state lifetime was shortened to
240 ns. Similar results were observed for [Ir(L5)2(bipy)]PF6 and [Ir(L6)2(bipy)]PF6,
where for both complexes the triplet state lifetimes were shortened in aerated
solution compared to deaerated solution (Figs 9 and 10).
<Figure 7>
<Figure 8>
<Figure 9>
Interestingly, the TA spectrum of the methoxy-substituted variant [Ir(L5)2(bipy)]PF6 is
slightly different, with strong positive features (ESA bands) at 400 nm <  < 500 nm
and also across the 550 nm <  < 850 nm region. Again, the decay characteristics
suggest these features are likely to belong to the same excitation and deactivation
16
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
processes. To support these studies, the TA spectrum of [Ir(L4)2(bipy)]+ was
successfully simulated (Fig. S8, SI) by combining admixtures of positive going TD-
DFT ESA simulated spectrum for Tn – T1 (n = 1-30) transitions with negative going
contributions from a ground state bleach (the negative going UV-vis absorption
spectrum) and small contributions of luminescence emission. The simulated
spectrum is in excellent agreement with that obtained experimentally (Fig. 7).
Unfortunately, despite numerous attempts using varying concentrations of
sample and experimental setup, the very short triplet lifetime characteristics of [Ir(L1-
3)(bipy)]PF6 at room temperature precluded detailed TA measurements on these
species. In contrast to [Ir(L4-6)2(bipy)]PF6, for [Ir(L1,2b,3)2(bipy)]PF6 the absorbance at
the excitation wavelength was very weak and insufficient excited states were
generated at low concentration. However, monitoring the triplet signal at higher
concentrations (7  10−5 M) enabled their detection, albeit of weak intensity, and the
corresponding time-resolved TA spectra are included for qualitative comparison
(Figs. S9-11, SI).
Triplet-Triplet Annihilation Energy Upconversion (TTA-UC)
As highlighted earlier, a relatively small number of transition metal complexes have
been investigated as photosensitizers in TTA-UC. Well known cyclometalated Ir(III)
complexes typically possess quite weak molar absorption properties in the visible
region and thus rely upon UV or blue light excitation and therefore need adaption for
TTA-UC purposes.44 A long lived triplet excited state is also desirable feature of the
photosensitizer. Previous work by Zhao has shown that one strategy for addressing
these twin challenges is to conjugate organic chromophores to the ligand framework
of the Ir(III) complex moiety.45 The judicious choice of chromophore enhances
3
3
absorption in the visible region and can facilitate the MLCT  IL equilibrium
leading to a large extension in observed lifetime. In our previous work, red emitting
Ir(III) photosensitizers based on cyclometalating 2-phenylquinoxaline ligands also
achieved very good visible absorption characteristics. And despite rather modest
triplet lifetimes ca. 2
s, these Ir(III) complexes demonstrated exceptional
performance in TTA-UC (up to 39% efficiency). As discussed earlier, [Ir(L4-
6)2(bipy)]PF6 possess good absorption characteristics in the visible region, but rather
17
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
short triplet lifetimes. Therefore, we were interested to assess the performance of
[Ir(L4-6)2(bipy)]PF6 as sensitizers in TTA-UC experiments.
Firstly, the triplet levels of the three complexes were approximated from the
band onset of the low temperature emission spectra. Obtained values of 16530 (L4),
16260 (L5) and 16390 cm-1 (L6) compare to the reported triplet state energy of 9,10-
diphenylanthracene (DPA)46 of 14300 cm-1 suggesting that forward TTET is feasible,
but that back energy transfer is a possibility where E  2000cm-1. Secondly,
solution state (deaerated acetonitrile) TTA-UC measurements were conducted using
[Ir(L4-6)2(bipy)]PF6 as the photosensitizers and DPA as the annihilator (see
Experimental Section for details). In these measurements, 532 nm excitation was
utilized which is selective for the long wavelength absorption of the complexes, and
primarily correlates with the spin forbidden (S0 → T1) MLCT-based transition.
Critically, DPA does not absorb at this excitation wavelength.
<Figure 10>
The spectroscopic representation of the TTA-UC measurements for each
complex is shown in Figure 10. It is clear that in each case addition of DPA to the
solution of the Ir(III) sensitizer resulted in a quench of the sensitizer
phosphorescence at 600-800 nm and an evolution of DPA-based fluorescence at
400-500 nm. Additional data regarding the dependence of TTA-UC upon DPA
concentration and excitation power are shown in Figs S12-S17, SI. Upconversion
luminescence intensity was shown to reach a plateau at a DPA concentration of 5 ×
10−4 M. Thus, a photosensitiser/DPA ratio of 1:5 was utilized for the upconversion
quantum yield measurements. The upconversion also showed a linear relationship
with excitation power density. At a laser power intensity of 76.5 mW cm-2, the relative
efficiencies of this process were given by upconversion quantum yields (UC) of
6.7%, 2.2% and 1.6% for [Ir(L4)2(bipy)]PF6, [Ir(L5)2(bipy)]PF6 and [Ir(L6)2(bipy)]PF6,
respectively. The visual phenomenon of energy upconversion in these systems can
be seen in Figure 11 where photographs clearly illustrate the effect of mixing the
DPA annihilator into the different solutions of the Ir(III) sensitizers.
<Figure 11>
18
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Conclusions
A series of structurally related cyclometalated Ir(III) complexes that incorporate
substituted 2-(naphthalen-1-yl)quinoline-4-carboxylate and 2-(naphthalen-2-
yl)quinoline-4-carboxylate type ligands have been synthesised. The complexes are
all red emitters with subtle variance according to ligand structure. A range of
spectroscopic and analytical techniques suggest that the position of chelation at the
naphthyl unit strongly influences the emission properties, more so than the role of
the ligand substituent. Supporting TD-DFT calculations suggest the emission
properties are described by a mixture of ligand-centred and MLCT character
contributing to the emitting state. Within the series, [Ir(L4-6)2(bipy)]PF6 possess much
more favourable absorption characteristics in the visible region and longer triplet
lifetimes. These twin attributes appear to enable [Ir(L4-6)2(bipy)]PF6 to be operative
sensitizers for TTA-UC with efficiencies up to ca. 7%. Therefore while these
conjugated naphthylquinoline ligands represent a viable strategy to bathochromically
shift the absorption and emission properties of the resultant Ir(III) complexes, these
species are unexpectedly compromised with respect to their triplet lifetimes under
deaerated conditions.
19
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Experimental
All reactions were performed with the use of vacuum line and Schlenk techniques.
1
Reagents were commercial grade and were used without further purification. H, 13C
and 19F NMR spectra were recorded on a Bruker Avance dpx 400, or 500 MHz
1
spectrometer, and were recorded in CDCl3, CD3CN or d6-DMSO solutions. H and
13C{1H} NMR chemical shifts (δ) were determined relative to internal
tetramethylsilane, Si(CH3)4 and are given in ppm. Low-resolution mass spectra were
obtained by the staff at Cardiff University. High-resolution mass spectra were carried
out by the staff at Cardiff University and the EPSRC National Mass Spectrometry
Service at Swansea University, UK. All photophysical data was obtained on a
JobinYvon-Horiba Fluorolog-3 spectrometer fitted with a JY TBX picosecond
photodetection module in CHCl3 or MeCN solutions. Emission spectra were
uncorrected and excitation spectra were instrument corrected. The pulsed source
was a Nano-LED configured for 295, 372 or 459 nm output operating at 500 kHz or 1
MHz. Luminescence lifetime profiles were obtained using the JobinYvon–Horiba
FluoroHub single photon counting module and the data fits yielded the lifetime
values using the provided DAS6 deconvolution software. IR spectra were recorded
on an ATR equipped Shimadzu IRAffinity-1 spectrophotometer. UV-vis data were
recorded as solutions on a Perkin Elmer Lamda20 spectrophotometer. TTA Energy
Upconversion measurements utilized c(sensitiser) = 1 × 10–5 M, c(DPA) = 5 × 10–4
M, 20 C; deaerated MeCN; and the upconversion quantum efficiencies were
obtained using Bengal Rose as a standard (F = 0.08 in methanol).
Triplet−triplet annihilation upconversion
A 532 nm continuous wave (CW) diode-pumped solid state laser was used as the
excitation light source for the TTA upconversion experiments. The upconverted
fluorescence was recorded with a RF 5301PC spectrofluorometer (Shimadzu Ltd.,
Japan). The mixed solution of triplet energy donor and acceptor were purged with N2
for 15 min before measurement. The upconversion quantum yields were determined
with Rose Bengal (F = 0.08 in methanol) as the luminescence quantum yield
standard. The upconversion quantum yields were calculated with the following
equation (Eq. 1), the subscript with ‘std’ and ‘sam’ stand for the corresponding
20
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
parameter for the standard and sample (Eq. x), , A, I and  represents the quantum
yield, absorbance, integrated photoluminescence intensity and the refractive index,
respectively. Factor 2 is used in the equation, thus the maximal upconversion
quantum yield should be 100%.
2
1 − 10−퐴
1 − 10−퐴
퐼sam
std
sam
std
UC = 2std
sam
Eq. 1
퐼std
Spin statistics predicts that the upconversion quantum yield should be less than
11.1% (given the maximal upconversion quantum yield is 50%). However, the recent
studies indicated that the upconversion quantum yields of some systems are up to
30%, because higher spin states can also be involved in the TTA upconversion.47
Nanosecond transient absorption spectroscopy
The nanosecond transient absorption spectra were studied on LP980 Laser Flash
Photolysis Spectrometer (Edinburgh Instruments, U.K.). Before the measurements,
the sample solution was purged with N2 for 15 min. The spectra were recorded in
collinear mode and excited with a nanosecond pulsed laser (Opolette 355II+UV
nanosecond pulsed laser, OPOTEK). The typical laser power is ca. 5 mJ per pulse.
The data was analyzed with software of L900.
X-ray crystallography
Data collection and processing
Suitable crystals were selected and data collected following a standard method,48 on
either a a Rigaku FRE+ diffractometer equipped with VHF Varimax confocal mirrors
and an AFC12 goniometer and HyPix 6000 detector, equipped with an Oxford
Cryosystems low-temperature device operating at T = 100(2) K (for [Ir(L3)2(bipy)]PF6
and [Ir(L4)2(bipy)]PF6) or a Rigaku 007HF diffractometer equipped with Varimax
confocal mirrors and an AFC11 goniometer and HyPix 6000 detector equipped with
an Oxford Cryosystems low-temperature device operating at T = 100(2) K (for
[Ir(L6)2(bipy)]PF6).
21
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Cell determination, data collection, data reduction, cell refinement and absorption
correction were carried out using CrysAlisPro49 The structures were solved with the
ShelXT50 structure solution program using the Intrinsic Phasing solution method and
by using Olex251 as the graphical interface. The models were refined with version
2018/3 of ShelXL52 using Least Squares minimisation.
DFT studies
All calculations were performed within the Gaussian 09 suite of programmes.53
Geometry optimisations were carried out without constraints using the DFT//B3LYP
level of theory.54 The Stuttgart-Dresden basis set was used for the iridium atoms,55
and was invoked with pseudo-potentials for the core electrons, with a 6-31G* basis
set for all remaining atoms.56 All calculations included the use of a polarized
continuum model (IEFPCM) approach for the treatment of the MeCN solvent.57 All
stationary point geometries obtained by DFT method were confirmed through
harmonic vibrational frequency calculations
All absorption spectra predictions, orbitals and other ground state properties were
computed using the ground state minimum energy geometry. Excitation energies
(absorption spectra predictions) were computed in the same manner as ground state
properties, but using the long range corrected CAM-B3LYP functional for improved
accuracy. For the prediction of emission energies, the triplet state was allowed to
relax to its optimal geometry, prior to single point energy calculations of the ground
state at this set of geometric parameters. Molecular orbital decomposition was
performed using the GaussSum package. 58
Acknowledgements
Cardiff University (Knowledge Economy Skills Scholarship, via the Welsh
Government’s European Social Fund, to C.E.E.) and STG Aerospace (Dr Sean
O’Kell and Dr Andrew Hallett) are thanked for financial support. We thank the staff of
the Engineering and Physical Sciences Research Council (EPSRC) Mass
Spectrometry National Service (Swansea University) for providing MS data and the
Engineering and Physical Sciences Research Council (EPSRC) UK National
Crystallographic Service at the University of Southampton.
22
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Keywords: Iridium complex • Spectroscopy • Upconversion • Phosphors • DFT
Notes and references ‡ CCDC 2023982, 2023983, 2023984 contains supplementary
X-ray crystallographic data for [Ir(L3)2(bipy)]PF6, [Ir(L4)2(bipy)]PF6, and
[Ir(L6)2(bipy)]PF6 respectively. This data can be obtained free of charge via
http://www.ccdc.cam.ac.uk/conts/retrieving.html,
or
from
the
Cambridge
Crystallographic Data Centre, Union Road, Cambridge, CB2 1EZ; fax(+44) 1223-
336-033 or email: deposit@ccdc.cam.ac.uk.
Corresponding Authors: Prof. Simon J. A. Pope: E-mail: PopeSJ@cardiff.ac.uk Dr
Joseph M. Beames: E-mail: BeamesJ@cardiff.ac.uk
23
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Figures and Captions
OEt
O
O
O
R
R
LH1 R = H
i)
LH2 R = OMe
LH3 R = F
O
ii)
N
N
H
OEt
O
O
O
R
LH4 R = H
LH5 R = OMe
LH6 R = F
R
i)
ii)
O
N
N
H
Scheme 1. Synthesis of the ligands: (i) NaOH, EtOH; (ii) EtOH, HCl.
Scheme 2. Top: Structures of the synthesized cationic Ir(III) complexes. Bottom:
general synthetic route.
24
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Figure 1. X-ray structure of [Ir(L3)2(bipy)]PF6. Note the disorder on one of the ester
groups.
Figure 2. X-ray structure of [Ir(L4)2(bipy)]PF6. Solvent not shown. Note the
intraligand distortion.
25
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Figure 3. X-ray structure of [Ir(L6)2(bipy)]PF6. Solvent not shown.
26
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
Figure 4. UV-vis. absorption spectra of the complexes (10-5 M, MeCN).
Figure 5. Normalized room temperature luminescence spectra of the complexes
(aerated MeCN; ex = 510 nm).
Figure 6. Luminescence spectra of the complexes recorded as a frozen glass (1:4,
MeOH/EtOH) (ex = 510 nm).
27
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
0.020
0.015
0.010
0.005
0.000
Air
N2
0 s
a
b
0.02
0.01
0.00
-0.01
0.16 s
...

= 686.8 ns
N2
3.68 s
3.84 s

Air = 239.9 ns
400 500 600 700 800
Wavelength / nm
0
2000
Time / ns
4000
Figure 7. (a) Nanosecond time-resolved transient absorption spectra of compound
[Ir(L4)2(bipy)]PF6 upon pulsed laser excitation in deaerated acetonitrile (ex = 510 nm,
c = 5 × 10–5 M) and (b) decay trace of compound [Ir(L4)2(bipy)]PF6 at 625 nm in
acetonitrile (c = 5 × 10–5 M) under air and N2 atmosphere, respectively, 20 C.
0 s
a
0.16 s
0.02
...
3.68 s
3.84 s
0.01
0.00
-0.01
400 500 600 700 800
Wavelength / nm
Figure 8. (a) Nanosecond time-resolved transient absorption spectra of compound
[Ir(L5)2(bipy)]PF6 upon pulsed laser excitation in deaerated acetonitrile (ex = 510 nm,
c = 5 × 10–5 M) and (b) decay trace of compound [Ir(L5)2(bipy)]PF6 at 645 nm in
acetonitrile (c = 5 × 10–5 M) under air and N2 atmosphere, respectively, 20 C.
28
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202004146
Chemistry - A European Journal
0.020
0.015
0.010
0.005
0.000
0.02
0.01
0.00
-0.01
0 s
0.12 s
......
2.64 s
2.76 s
Air
N2
b
a

= 435.2 ns
N2

Air = 216.8 ns
0
1000 2000 3000
Time / ns
400 500 600 700 800
Wavelength / nm
Figure 9. (a) Nanosecond time-resolved transient absorption spectra of compound
[Ir(L6)2(bipy)]PF6 upon pulsed laser excitation in deaerated acetonitrile (ex = 510 nm,
c = 5 × 10–5 M) and (b) decay trace of compound [Ir(L6)2(bipy)]PF6 at 645 nm in
acetonitrile (c = 5 × 10–5 M) under air and N2 atmosphere, respectively, 20 C.
Figure 10. Triplet-triplet annihilation upconversion emission spectra with the
photosensitizers, (a) [Ir(L4)2(bipy)]PF6, (b) [Ir(L5)2(bipy)]PF6, and (c) [Ir(L6)2(bipy)]PF6,
in deaerated acetonitrile where DPA was used as the annihilator. Excitation was
achieved with a continuous laser at  = 532 nm (power density of 5.6 mW) under a
deaerated atmosphere. c(sensitizer) = 1 × 10–5 M, c(DPA) = 5 × 10–4 M, 20 C. The
asterisks indicate the scattered laser signal.
29
2
This article is protected by copyright. All rights reserved.