University of Huddersfield Repository Luminescent biscyclometalated arylpyridine iridium(III) complexes with 4,4’-bi-1,2,3-triazolyl ancillary ligands

The synthesis, characterization and photophysical investigation of complexes of the form [Ir(R-ppy) 2 (btz)]PF 6 ( 1 to 3 ) are reported (btz = 1,1’-dibenzyl-4,4’-1,2,3-triazolyl, R-ppy = 4-(pyrid-2-yl)benzaldehyde ( 1 ), 2-phenylpyridine ( 2 ) and 2-(2,4-difluorophenyl)pyridine ( 3 )). Complexes 1 , 2 and 3 are luminescent and exhibit structured emission bands with vibronic progressions at 532 & 568 nm ( φ 10 0.28 %), 476 & 508 nm ( φ 0.82 %) and 454 & 483 nm ( φ 4.3 %) respectively. The structuring of these emission bands is indicative of cyclometalated ligand centred emissive states and is further corroborated by the nearly identical emission spectra for 2 and 3 to previously reported analogous complexes with 4-(pyrid-2-yl)-1,2,3-triazole based ancillary ligands. Computational density functional theory calculations on these complexes show that the LUMOs of 2 and 3 are largely btz-centred but with some 15 cyclometalated pyridine π * character. The LUMO of 1 on the other hand is localized primarily on the cyclometalated ligands. Spin population analysis of the lowest lying triplet excited states for these complexes indicate significant spin population over the iridium centres and the aryl and pyridyl moieties in these complexes with virtually no localization of unpaired electrons over the btz ancillary ligands. This is therefore in agreement with the assignment of the emissive state having largely cyclometalated 3 LC 20 character and being independent of the ancillary ligand. New biscyclometalated iridium(III) complexes with bitriazolyl (btz) ancillary ligands are reported. In contrast to previously reported d 6 metal btz complexes which show no emission, those described here are luminescent with quantum yields up to 4.3 %.


Introduction
Cyclometalated iridium(III) 1-3 and platinum(II) 4, 5 complexes have been the subject of a large amount of interest in the literature due to their attractive photophysical properties that 25 results in their potential application in biological imaging, as the basis of luminescent molecular sensors 6 and as the phosphors in organic light emitting diode (OLED) and light emitting electrochemical cell (LEEC) devices. 7-10 11 Efforts to tune the wavelengths of emission in these complexes have focused on the 30 modification of the cyclometalated and ancillary ligands in order to modulate the energies of the frontier orbitals. 12, 13 14, 15 Synthetic routes that provide access to a wide range of ligands and hence allow facile tuning of electronic properties are therefore of great interest. 35 The Huisgen-Sharpless copper catalysed alkyne/azide cycloaddition (CuAAC) to form 1,2,3-traizoles (commonly referred to as 'click' chemistry) 16,17 has attracted enormous interest over the past decade in organic synthesis, as a linking moiety in novel polymer and dendrimer systems [18][19][20][21][22] and in the 40 modification of biological macromolecules. [23][24][25] The past four years or so have since seen an explosion in the use of this versatile reaction in ligand design for metal complexes. Examples have appeared of monodentate N-donor triazole ligands, N-heterocyclic 'click' carbene complexes and triazole- 45 containing chelate systems. This area has recently been the subject of a comprehensive review. 26 Ligand architectures that have become ubiquitous in transition metal coordination chemistry due to the photophysical properties of their complexes include 2,2'bipyridyl (bpy) and 2,2';6',2"-50 terpyridyl (tpy). Several groups have reported analogous ligand systems constructed through CuAAC reactions where pyridyl moieties are replaced with N-donor 1,2,3-triazole rings. [27][28][29] Several examples of ruthenium(II), rhenium(I) and iridium(III) complexes bearing 4-(pyrid-2-yl)-1,2,3-traizole (pytz) based 55 ligands as bpy analogues have appeared and the resultant photophysical properties investigated. Replacement of bpy by pytz in hetero-and homoleptic complexes of the form [Ru(bpy) n (pytz) 3-n ] 2+ was shown to lead to a blue-shifting in absorption and emission maxima but with significant quenching 60 of emission intensity.
We recently reported the preparation and characterization of the heteroleptic complexes [Ru(bpy) 2 (btz)] 2+ and 10 [Ru(bpy)(btz) 2 ] 2+ . 46 Our results revealed that replacement of bpy by btz yields an expected blue-shift in absorption bands and quenched emission. The quenching of luminescent emission in these systems is arises through destabilization of the 3 MLCT states of these complexes due to the much higher energy LUMO 15 of the btz compared to that of bpy. The reduced separation between 3 MLCT and 3 MC then results in thermal population of the latter from the former and non-radiative deactivation to the ground state.  47 Despite several examples of the use of pytz ligands, the btz framework has not been investigated as the ancillary ligand in 25 cationic biscyclometalated iridium(III) complexes (we note, however, a report independently detailing iridium btz complexes by Zysman-Colman and co-workers 48 has appeared whilst this manuscript was in the peer-review process). This is perhaps due to the presumption that this would similarly lead to deleterious 30 effects on the photophysical properties of the resultant complexes. However, we were hopeful that complexes of the general structure [Ir(R-ppy) 2 (btz)] + would make attractive luminescent complexes for LEEC device applications. The presence of a 5d metal and strongly donating anionic 35 cyclometalate ligands will result in a larger ligand field splitting and higher energy 3 MC states than in ruthenium complexes with neutral bpy ligands. Hence these may well become thermally inaccessible from 3 MLCT/ 3 LLCT and 3 MLCT/ 3 ILCT based excited states in these complexes thereby rendering them 40 emissive. Indeed, we show here that in stark contrast to btz complexes of the ruthenium and rhenium complexes discussed previously, complexes of the form [Ir(R-ppy) 2 (btz)]PF 6 do in fact show appreciable luminescent emission even in aerated solution at room temperature. 45
After concentration of the reaction mixtures under reduced pressure and treatment with aqueous 55 NH 4 PF 6 , the target complexes 1 to 3 were isolated as their hexafluorophosphate salts (Scheme 1).     Crystals of X-ray diffraction quality were obtained for 10 complexes 2 and 3 from acetonitrile solution with slow diffusion vapour diffusion of diethyl ether. Complex 2 crystallises in the space group P-1 and exhibits two crystallographically unique cations in the unit cell. An ORTEP plot of the structure of one of the cations is depicted in Figure 1 and selected bond distances 15 and angles are provided in Table 1. The cations adopt distorted octahedral geometries with the pyridine rings of the ppy ligands occupying mutually trans coordination sites. Bond lengths and angles for the cyclometalated ligands are unremarkable. Ir-N(btz) bond lengths lie between 2.149 and 2.175 Å for the two cations 20 with btz ligand bite angles of 75.77(7) and 75.27(7) ° which are comparable to those of ruthenium complexes of the same ligand. 40 An ORTEP plot of the structure of the cation for 3 is shown in Figure 2. The cation sits on an axis of symmetry such that only   UV-visible absorption spectra were recorded for 40 dichloromethane solutions of complexes 1 to 3 and are shown in Figure 3. A summary of the photophysical properties can be found in Table 2. Bands appear between 420 and 367 nm which are assigned to 1 MLCT-based transitions with intense absorptions below 300 nm assigned to ligand centred π→π* transitions. The 45 MLCT band of 1 appears at 420 nm and is red-shifted relative to that of 2 bearing unsubstituted ppy ligands (385 nm). This may due to the extended π-system associated with the formylsubstituted ligand which would result in stabilisation of the cyclometalated ligand centred unoccupied orbitals (vide infra) 50 and a reduced HOMO -LUMO separation. The electron withdrawing fluorine substituents in 3 would be expected to lead to stabilization of the HOMO relative to that of 2 resulting in blue-shifted absorption and indeed the MLCT band appears at 367 nm. In contrast to known rhenium and ruthenium btz complexes, 30, 5 45-47 complexes 1 to 3 exhibit luminescent emission upon excitation at 400 to 425 nm in aerated dichloromethane solutions at room temperature. Normalized emission spectra are shown in  33,37 In comparison, the analogous 2,2'-bipyridyl complex [Ir(ppy) 2 (bpy)] + exhibits a broad unstructured band with an emission maximum at about 591 nm, 25 significantly red-shifted relative to those of the corresponding pytz and btz complexes. These data confirm that the 3 LC emissive excited states in heteroleptic biscyclometalated iridium complexes of the type described here have little or no contribution from, and are therefore largely independent of, the 30 pytz or btz ancillary ligands. The excited state for the bpy complex is, on the other hand, predominantly localized on the ancillary ligand. We have previously shown that the LUMO of the btz ligand is significantly destabilized relative to that of bpy. 41 The triazole moiety in the pytz ligand could therefore similarly 35 lead to a large destabilization in the ligand-based LUMO bringing it to a comparable energy to those of vacant orbitals centred on the cyclometalate ligands leading to the observed switching in localization of the emissive state in these complexes. As mentioned above, previously reported ruthenium and 40 rhenium btz complexes were shown to be weakly or non-emissive in solution at room temperature, however, reasonably intense emission is observed at 77 K for [Ru(bpy) 2 (btz)] 2+ and [Ru(bpy)(btz) 2 ] 2+ . 46 This is explained by the presence of the btz ligand resulting in a higher energy LUMO and hence elevated 45 3 MLCT state from which non-radiative 3 MC states can be thermally populated. The intense emission observed for the complexes described in the present study is a likely consequence of the presence of the strongly donating anionic cyclometalated ligands that result in a larger ligand field splitting. The 3 MC 50 states in complexes 1 to 3 are therefore elevated such that they become thermally inaccessible for population from the emissive excited states of these complexes. Luminescent lifetimes were determined for each complex in aerated dichloromethane solutions at room temperature. 55 Complex 2 exhibits a lifetime of 18 ns whereas that of 3 is slightly elongated at 24 ns. In contrast, the lifetime of 1 was measured as 792 ns. Quantum yields were determined from their integrated emission intensities and referenced to [Ru(bpy) 3 ][PF 6 ] 2 in aerated acetonitrile (φ = 1.8 %). 49 Quantum yields of 2 and 3 60 are 0.82 and 0.28 % respectively. Consistent with the much longer lifetime the quantum yield of 1 is much larger than those of the other btz complexes at 4.3 %.

[Ir(ppy)2(btz)]PF6
Cyclic and square-wave voltammetry were performed on solutions of complexes 1 to 3. All three complexes show 65 reversible oxidations at 1.05, 0.85 and 1.19 V respectively (referenced against ferrocene/ferrocenium, E = 0 V). The data for 3 is indicative of the commonly observed stabilization of the HOMO on inclusion of electron withdrawing substituents on the aryl rings. The observed stabilization in 1 compared to 2 is 70 similarly likely due to the presence of the -M formyl groups. Complexes 2 and 3 exhibit irreversible reductions at -1.96 and -1.85 V, however, 1 displays a reversible reduction at much more positive potential centered at -1.38 V. We tentatively assign the reductions for 2 and 3 as arising from btz centred LUMOs and the 75 reduction for 1 as arising from a cyclometalated ligand centered LUMO which is stabilized through the larger π-system provided by the formyl groups. This then accounts for the red shifting of absorption and emission in 1 relative to 2.
DFT studies were undertaken in order to further understand the 80 photophysical and electronic properties of the ground and excited states of complexes 1 to 3. The ground state geometry of each complex was optimized without symmetry constraints in the gas phase at the B3LYP level of theory using the Stuttgart-Dresden relativistic small core potential for iridium and 6-311G* basis 85 sets for all other atoms (optimized xyz coordinates for the complexes may be found in the Supporting Information). In order to minimize the computational expense required in these calculations the benzyl substituents of the btz ligands were simplified to methyl. In addition, the ground state geometries of 90 the pytz and bpy complexes [Ir(ppy) 2 (pytz)] + and [Ir(ppy) 2 (bpy)] + were also calculated for comparison. Energies of the frontier molecular orbitals were determined and plots of the HOMO and LUMO orbitals for each complex are depicted in Figure 5. Associated energies are provided in Table 2   95 with a comparative energy level diagram shown in Figure 6. As expected for complexes of this type, the HOMO of the parent complex 2 (which appears at -7.74 eV) has primarily phenyl πcharacter in an anti-bonding combination with an iridium dorbital and is common for those of both 1 and 3 (-8.17 and -8.22 100 eV respectively). The stabilization of the HOMOs in 1 and 3 relative to that of 2 therefore mirrors the measured oxidation potentials for these complexes. The closely-spaced LUMO (-4.01 eV) and LUMO + 1 (-3.97 eV) orbitals of 2 are primarily centered on the ancillary btz ligand but have some additional 105 pyridyl π* contribution. A similar involvement of the pyridyl moiety in the LUMO (-4.24 eV) and LUMO + 1 (-4.19 eV) of 3 is also observed. For both of these complexes, LUMO + 2 and LUMO + 3 appear solely btz-centered. The resultant HOMO-LUMO gap for 3 of 3.98 eV is larger than that for 2 (3.74 eV) in 110 agreement with expectations and the observed blue-shift in experimental absorption and emission spectra.
The LUMO of the analogous bpy complex [Ir(ppy) 2 (bpy)] + , calculated using the same exchange correlation functional and basis sets, is primarily bpy-based and resides at -5.24 eV, some respectively. This results in a much smaller HOMO -LUMO gap of 2.78 eV consistent with the largely red-shifted emission 5 band for this complex relative to that of 2. Here, LUMO + 1 (-4.41 eV) is also localized on the ancillary bpy ligand without involvement of the cyclometalate ligands with the first unoccupied orbital with ppy pyridine π* character being LUMO + 2 (-4.22 eV). 10 In the case of the pytz complex [Ir(ppy) 2 (pytz)] + the HOMO is similarly localized on the metal and the cyclometalated phenyl rings with a greater contribution from the ppy ligand trans to the pyridyl donor of the ancillary ligand. The LUMO and LUMO + 1 orbitals are both pytz π* in character and are localized to a 15 greater extent on the pyridyl ring. Here, the LUMO lies between those of 2 and [Ir(ppy) 2 (bpy)] + , some 0.47 eV above that of the bpy analogue. LUMO + 2 (-4.11 eV) and LUMO + 3 (-4.03 eV) are the lowest ppy-based pyridyl π* orbitals. This change in the localization of the unoccupied frontier 20 orbitals in complexes 2 and 3, in which there is some localization on the cyclometalated ligands, when compared to those of [Ir(ppy) 2 (bpy)] + is due to the much higher energy of the LUMO of the btz ligand compared to that of bpy. Indeed, in separate calculations on the free btz and bpy ligands, the LUMO of btz is 25 some 1.02 eV higher in energy relative to that of bpy. 41 Hence, the btz LUMO lies close in energy to orbitals of the cyclometalated ligands allowing mixed R-ppy / btz character in the unoccupied frontier orbitals of these complexes. In contrast to 2 and 3, the LUMO and LUMO + 1 orbitals of 1 30 are primarily centered on the pyridyl rings and formyl groups of the cyclometalated ligand and have no significant contribution from the btz ligand. Consistent with experimentally observed red-shifted spectra, the HOMO-LUMO gap for 1 is smaller than that of 2 at 3.57 eV. Here, LUMO + 2, some 0.32 eV above the 35 LUMO, is the first btz-centered unoccupied orbital.  Time-dependent DFT (TDDFT) calculations were performed 45 at the ground state geometries each complex to derive vertical excitation energies and hence simulated optical absorption spectra. TDDFT derived spectra for 1 to 3 (with experimental spectra overlaid) are presented in Figure 7. From a simple visual analysis of the positions of the major transitions depicted in 50 Figure 3 it can be seen that the energies of the calculated transitions are in good agreement with the experimentally recorded spectra. Consistent with experimental data these major transitions are also observed to blue-shift from complex 1 to 3. Indeed, the S 1 states appear at 426 nm (f = 0.054), 409 nm (f = 55 0.030) and 381 nm (f = 0.014) for 1, 2 and 3 respectively, are relatively intense and are primarily HOMO → LUMO in character. Hence, the nature of these transitions is R-ppy-based MLCT/LC for 1 and MLCT/LLCT to btz for 2 and 3.
In order to probe the nature of the emissive states of these 60 complexes the lowest triplet excited states of 1 to 3 were optimized, along with those of [Ir(ppy) 2 (pytz)] + and [Ir(ppy) 2 (bpy)] + , starting from their S 0 geometries using the constraint of the spin multiplicity of 3. Table 4 collates the calculated energies of the T 1 states for complexes 1 to 3 and those 65 of [Ir(ppy) 2 (pytz)] + and [Ir(ppy) 2 (bpy)] + quote relative to their respective S 0 ground state energies. Consistent with the experimentally observed spectroscopic data which shows a blueshift in emission maxima, the T 1 state is destabilized on going from complex 1 to 3. The T 1 state of [Ir(ppy) 2 (pytz)] + (2.59 eV) 70 is noted to have almost the same energy as that of 2 (2.60 eV) consistent with emission occurring from the same ppy-based 3  the geometries of the T 1 states for the complexes, the energies of the singlet ground states at these T 1 geometries were calculated in 5 single point calculations. The difference in energy between the optimized T 1 states and these non-equilibrium S 0 * states at the same geometries are therefore the calculated emission energies which are then used to derive the calculated emission maximum. As can be seen from Table 4  Mulliken population analyses were carried on to determine the localization of spin density for these T 1 states and summed atomic spin populations for the iridium atom and the aryl, pyridyl and btz moieties of 1 to 3 are listed in Table 6. The data clearly 25 show significant spin populations on the iridium atom and the cyclometalated ligands but which do not extend to the btz ligands. Hence, the emissive triplet states in these complexes are predicted to be largely 3 LC in character with little or no involvement of the ancillary btz ligand. Hence, tuning of the 30 emissive properties of these complexes is entirely dependent on the nature of the substituents of the cyclometalated ligands. Similarly, and in agreement with the near identical emission characteristics to those of complex 2, analysis of the spin population in [Ir(ppy) 2 (pytz)] + shows that there is no significant 35 population of the unpaired electrons on the pytz ligand and that again the T 1 state of this complex is also 3 LC in character. On the other hand, the spin populations for [Ir(ppy) 2 (bpy)] + reveal that the excited electron is localized on the bpy ligand confirming the 3 MLCT/ 3 LLCT character of the emissive state for this complex. The difference in the frontier orbital localization for 1 50 compared to those of 2 and 3 along with the 3 LC character in the emissive states of these btz complexes may account for the greater quantum yield of emission observed for 1 and the greatly elongated luminescent lifetime. Since both the HOMO and LUMO in this complex are largely centered in the cyclometalate 55 ligands, redistribution of the unpaired electron density occurs in 1 to a much lesser extent during the process of the excitation of an electron to the LUMO, subsequent inter-system crossing and relaxation of the resultant T 1 state than would be the case for 2 and 3. This may therefore lead to the greater rate of non-radiative 60 decay rate through interactions with ligand vibrational oscillators for the latter complexes relative to their radiative decay, k r , when compared to 1.
The computational data presented here is therefore in agreement with the 3 LC centered assignment of the emissive 65 states for the pytz complexes [Ir(ppy) 2 (pytz)] + and [Ir(dfppy) 2 (pytz)] + 33, 37 and complexes 2 and 3 based on the near identical emission spectra and the observed vibronic structure therein. Hence, the tuning of the photophysical properties of these btz complexes and their pytz analogues are entirely 70 dependent on the nature of cyclometalated ligands. These complexes provide the basis of further development of LEEC phosphors.

Conclusions
We have reported the synthesis, characterization and the 75 photophysical and theoretical study of iridium(III) cyclometalated complexes with 4,4'-bi-1,2,3-triazolyl ancillary ligands. We have shown that these are luminescent with emission wavelengths are tunable via through variation of the aryl substituents of the cyclometalated ligands but are independent of the ancillary btz 80 ligand. Complexes of this type are promising candidates for further development as phosphors in light emitting devices and in other applications such as biological imaging. resources used in this work. We also acknowledge the University of Huddersfield Centre for High Performance Computing. We also thank Dr Marcus Chadha, Innovative Physical Organic Solutions, University of Huddersfield, for mass spectrometry assistance.

General methods
The btz ligand, 46 6 ] 53 were all prepared by previously reported procedures. NMR 15 spectra were recorded on Bruker 500 Avance and 400 AVIII spectrometers and mass spectrometry data were obtained on a Bruker Micro-Q-TOF instrument. UV-visible absorption data were recorded on a Varian Cary 4000 UV-visible spectrophotometer and emission spectra were recorded on a 20 Jobin-Yvon Flouromax instrument. Excited state lifetimes were obtained using an Edinburgh Instrument Mini-tau spectrometer.