Probing Electron Excitation Characters of Carboline-Based Bis-Tridentate Ir(III) Complexes

In this work, we report a series of bis-tridentate Ir(III) metal complexes, comprising a dianionic pyrazole-pyridine-phenyl tridentate chelate and a monoanionic chelate bearing a peripheral carbene and carboline coordination fragment that is linked to the central phenyl group. All these Ir(III) complexes were synthesized with an efficient one-pot and two-step method, and their emission hue was fine-tuned by variation of the substituent at the central coordination entity (i.e., pyridinyl and phenyl group) of each of the tridentate chelates. Their photophysical and electrochemical properties, thermal stabilities and electroluminescence performances are examined and discussed comprehensively. The doped devices based on [Ir(cbF)(phyz1)] (Cb1) and [Ir(cbB)(phyz1)] (Cb4) give a maximum external quantum efficiency (current efficiency) of 16.6% (55.2 cd/A) and 13.9% (43.8 cd/A), respectively. The relatively high electroluminescence efficiencies indicate that bis-tridentate Ir(III) complexes are promising candidates for OLED applications.


Introduction
Organic light-emitting diodes (OLEDs) have been widely employed in the fabrication of flat panel displays and solid-state lighting luminaries. In this regard, Ir(III) phosphors have received special attention for their capability in harvesting both the singlet and triplet excited states formed in the devices [1]. The triplet states account for 75% of the total excited states generated; hence, the strong spin-orbit coupling exerted by the Ir(III) metal atom can reduce the radiative lifetime of triplet excited states, resulting in a significant improvement of the overall efficiency of OLEDs. This has triggered numerous studies on the quest of chemically and photochemically stable Ir(III) metal complexes, to which the efficient phosphorescence from the coupled ligand-centered (LC) ππ* and metal-to-ligand charge transfer (MLCT) excited states tend to fulfil the criteria for higher OLED efficiency [2][3][4][5][6][7].
Traditionally, these Ir(III) emitters were constructed using bidentate cyclometalates such as 2-phenylpyridine or functional analogues (CˆN) and/or monoanionic ancillary chelate, denoted as (LˆX). The tris-homoleptic and heteroleptic Ir(III) complexes [Ir(CˆN) 3 ] and [Ir(CˆN) 2 (LˆX)] have been extensively designed and studied [8]. In theory, both of them are capable of affording at least two stereoisomers, which are controlled by their intrinsic kinetic and thermodynamic factors. They are named as fac-(facial) and mer-(meridional) isomers in the case of homoleptic complexes [Ir(CˆN) 3 ]. Generally, these stereoisomers possess distinctive chemical and physical properties and, hence, their interconversion should be limited during preparation. One possible method in preventing the formation of multiple stereoisomers is to employ the bis-tridentate architectures, to which the planar

General Information
All solvents were dried and degassed before used, and commercially available reagents were used without further purification. 2,6-Dibromo-4-methoxypyridine [36,37], 2,6dibromo-N,N-dimethylpyridin-4-amine [38,39] and 6-(tert-butyl)-9H-pyrido [2,3-b]indole [40] were prepared using methods reported in literature. All reactions were conducted under N 2 atmosphere and monitored by precoated TLC plates (0.20 nm with fluorescent indicator F254). 1 H and 19 F spectra were recorded with Bruker 400 MHz AVANCE III Nuclear Magnetic Resonance System. Elemental analysis was performed by an elemental carbon-hydrogen-nitrogen analyzer (Elementar). Mass spectra were obtained on 4800 Plus MALDI TOF/TOF Analyzer (ABI), where 2,5-dihydroxybenzoic acid was applied as the matrix. TGA measurements were performed on a TA Instrument TGAQ50, at a heating rate of 10 • C min −1 under N 2 atmosphere. The X-ray intensity data were measured using phi and omega scan modes (APEX3) at 233 K on a Bruker D8 Venture Photon II diffractometer with microfocus X-ray sources.
After that, the preparation of the bis-tridentate Ir(III) complexes Cb1-5 was conducted using a one-pot and two-step method. As a generalized protocol, the carboline chelate (cbF)H·HF 6 (or (cbB)H·HF 6 ) was first heated with [Ir(COD)Cl] 2 and sodium acetate in degassed acetonitrile. The intermediate was next reacted with a series of second chelate (phyzn)H 2 (n = 1, 2 and 3) in decalin to afford the desired Ir(III) complexes in moderate yields. The mass spectrometry and 1 H and 19 F NMR spectroscopies, together with a single crystal X-ray diffraction study on Cb1, were examined to offer the needed characterizations. Their structural drawings are depicted in Scheme 2 for scrutiny. Scheme 2. Structural drawings of the bis-tridentate Ir(III) complexes Cb1-5. Figure 1 depicts the molecular drawing of Cb1, with thermal ellipsoids drawn at a level of 30% probability. The crystal of Cb1 for X-ray diffraction was obtained via the slow diffusion of hexane into a saturated CH 2 Cl 2 solution of Cb1 at RT. The Ir(III) metal atom constituted a slightly distorted octahedral coordination arrangement with two mutually orthogonal tridentate chelates. The phyz1 chelate is essentially planar, while that of the tridentate chelate cbF underwent a slight distortion at the outer hexagonal ring of the carboline unit, which can be attributed to the unfavourable steric interaction between carboline and central benzene fragments. In agreement with the prediction of trans-influence [42], the carbene Ir-C distance (Ir-C(39) = 2.004(3) Å) is relatively shorter than the typical Ir-C distances observed in other bis-tridentate Ir(III) complexes bearing symmetrically arranged carbene pincer chelates (2.043 − 2.062 Å) [43,44]. Concomitantly, the Ir-C distance of central benzene group (Ir-C(31) = 2.011(3) Å) elongated slightly in comparison to that of the corresponding carbene pincer chelates (1.950-1.960 Å).   Table 1. All Ir(III) complexes give similar absorption patterns, and the higher energy bands above 380 nm are attributed to the spin-allowed ππ* transition, while those occurring at the longer wavelength regions of 380-450 nm are assigned to the singlet metal-to-ligand charge transfer ( 1 MLCT). The next lower absorption bands spanning the region from 450 nm up to the onset are ascribed to the mixed spin-forbidden ligand-centered ππ* transition and MLCT transition processes.

Photophysical and Electrochemical Properties
Molecules 2021, 26, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/molecules fostered stronger spin-orbital coupling and faster phosphorescence. This tendency was also observed by comparing the second set of the Ir(III) complexes Cb4 and Cb5, with the radiative rate constant being 2.9 × 10 5 s −1 and 3.3 × 10 5 s −1 , respectively. Furthermore, for Cb3 and Cb5, the bathochromic shift can also be rationalized with the electron-donating effect of NMe2 substituent at the 4-position of pyridinyl group, giving a higher-lying HOMO level and hence a narrower energy gap.  Figure 3 shows the electrochemical properties of bis-tridentate Ir(III) complexes Cb1−5, with numerical data listed in Table 2. All complexes present reversible oxidation and irreversible reduction waves. Replacing CF3 with the tert-butyl substituent in the monoanionic carbene pincer chelate induces a cathodic shift on the oxidation potential,  Upon photoexcitation, an intense green emission was observed among Cb1, Cb2 and Cb3 in the degassed CH 2 Cl 2 solution with the peak wavelength at 525, 521 and 529 nm, respectively. The slight shifting of peak indicates the substituent effects of the pyridinyl coordination unit. It is worth noting that the shoulder at the right of emission profile gradually vanished in accordance with the sequence of hydrogen, methoxy, dimethylamino presented, manifesting an increased MLCT contribution for a structureless profile. In addition, the radiative rate constant (k r ) for Cb1 to Cb3 (2.0, 3.2 and 3.4 × 10 5 s −1 ), which was calculated from quantum yield (Φ) divided by the observed lifetime (τ obs ), revealed an acending trend to the increased MLCT contribution, as it fostered stronger spin-orbital coupling and faster phosphorescence. This tendency was also observed by comparing the second set of the Ir(III) complexes Cb4 and Cb5, with the radiative rate constant being 2.9 × 10 5 s −1 and 3.3 × 10 5 s −1 , respectively. Furthermore, for Cb3 and Cb5, the bathochromic shift can also be rationalized with the electron-donating effect of NMe 2 substituent at the 4-position of pyridinyl group, giving a higher-lying HOMO level and hence a narrower energy gap. Figure 3 shows the electrochemical properties of bis-tridentate Ir(III) complexes Cb1−5, with numerical data listed in Table 2. All complexes present reversible oxidation and irreversible reduction waves. Replacing CF 3 with the tert-butyl substituent in the monoanionic carbene pincer chelate induces a cathodic shift on the oxidation potential, e.g., Cb1 (0.56 V) to Cb4 (0.35 V). For Cb1, Cb2 and Cb3, the oxidation potentials experience a decrease from 0.56 V and 0.53 V to 0.45 V, with changing 4-hydrogen atom on the pyridinyl fragment to methoxy and dimethylamino substituents. A similar trend is also observed between Cb4 and Cb5, which varied from 0.35 V to 0.25 V, after the introduction of the dimethylamino group. Meanwhile, the reduction potentials are also influenced by the substituent effect as mentioned earlier. Among Ir(III) complexes Cb1-3, Cb3 exhibits the most destabilized LUMO by giving the most negative potential at −2.48 V, which can be explained by the strongest electron-donating ability of the dimethylamino group. Moreover, both the Ir(III) complexes Cb4 and Cb5 (−2.50 V and −2.56 V, respectively) with the tert-butyl substituent on the monoanionic tridentate chelate display more negative reduction potentials than that of the CF 3 substituted counterparts Cb1, Cb2 and Cb3 (−2.42 V, −2.45 V and −2.48 V, respectively), showing that the LUMO is not associated with this pyridinyl coordination unit.   Table 2. Electrochemical data of the Ir(III) metal complexes Cb1−5 in acetonitrile at RT.

Theoretical Calculation
We then conducted the density functional theory (DFT) calculations at PBE0/LANL2DZ (Ir) and PBE0/6-31g(d,p) (H, C, N, F, O) levels using CH2Cl2 as the solvent to optimize the ground-state (S0) geometries of all molecules. In addition, timedependent (TD) DFT calcualtions at the same levels were performed to optimize the geometries of the excited states and to probe the transition characteristics of the studied Ir(III) complexes. The calculated transition energies and major assignments of Ir(III)

Theoretical Calculation
We then conducted the density functional theory (DFT) calculations at PBE0/LANL2DZ (Ir) and PBE0/6-31g(d,p) (H, C, N, F, O) levels using CH 2 Cl 2 as the solvent to optimize the ground-state (S 0 ) geometries of all molecules. In addition, time-dependent (TD) DFT calcualtions at the same levels were performed to optimize the geometries of the excited states and to probe the transition characteristics of the studied Ir(III) complexes. The calculated transition energies and major assignments of Ir(III) complexes Cb1-5 in CH 2 Cl 2 solution are summarized in Tables 3 and S1-S5, respectively. The frontier molecular orbitals involved in the major transitions were also depicted in Figures 4 and S1 and 587.7 nm and Cb5: 520.2 and 598.9 nm, respectively. For Cb1-5, the calculated S 1 → S 0 wavelengths were all close to the onset of the emission spectra while the T 1 → S 0 wavelengths were akin to the experimental emissive peaks as recorded in Figure 2. The trends of S 0 → S 1 absorption and T 1 → S 0 emission were in good agreement with their corresponding absorption and phosphorescence spectra, respectively.  Moreover, the S 0 → S 1 absorption was derived mainly from HOMO → LUMO+1 for Cb1 and Cb4 and HOMO → LUMO for Cb2, Cb3 and Cb5, respectively ( Table 3). The S 1 → S 0 and T 1 → S 0 emission were all assigned to LUMO → HOMO for Cb1-5. For the ground state S 0 of Cb1-5, the electron density distribution of the HOMO was mainly localized at the central Ir(III) metal atom (31-34%) and delocalized over the chromophoric chelate 2-phenyl-6-(3-(trifluoromethyl)-1H-pyrazol-5-yl)pyridine (phyz) and carbene-benzene-carboline (cb), while the electron density distribution of the LUMO and LUMO+1 was mainly localized at the cb or phyz chelate, respectively, accompanying a little contribution at the Ir(III) atom (1-3%) (Figures 4 and S1-S5). For the excited states S 1 and T 1 of Cb1-5, the electron density distribution of the HOMO was mainly localized at the central Ir(III) metal atom (29-36%) and delocalized over the phyz and cb fragment, while the electron density distribution of the LUMO was mainly localized at the cb or phyz chelate, together with a few contribution at the Ir(III) atom (2-4%). Moreover, it is notable that LUMO is partially shifted to carboline moiety in Cb3, while completely moved to carboline moiety as observed in Cb5. We attributed this to the introducing of the dimethylamino substituent at the pyridinyl unit of the dianionic chelate that greatly increased the associated π* orbital energy, such that the LUMO is now dominated by the relatively unaffected carboline π* orbital. Overall, the S 0 → S 1 , S 1 → S 0 and T 1 → S 0 transitions were all mainly ascribed to the metal-to-ligand charge transfer (MLCT) process (19-31%), accompanied by minor ligand-to-ligand charge transfer (LLCT) or intraligand charge transfer (ILCT). These high MLCT characters were in nice relevance to the moderate emission quantum yield (41-69%) of the emissive complexes Cb1-5 in Tables 1 and 3. Furthermore, with regard to the calculated HOMO energy levels of S 0 , S 1 and T 1 , Cb3 was higher than Cb1 and Cb2 due to the electron-donating effect of NMe 2 substituent at the 4-position of pyridinyl group in Cb3. Additionally, Cb5 is higher than that of Cb4 (Table 2 and Figures S1-S5). The trend of calculated HOMO energy levels is in good agreement with the experimental results (vide supra).

Fabrication of OLED Devices
All these new Ir(III) complexes showed a high decomposition temperature (>283 • C, Figure S6), which is suitable for conducting device fabrication via thermal deposition. In view of their better photophysical properties, Cb1 and Cb4 were selected as the dopant emitter in fabrication of OLED devices with architecture: ITO/TAPC (40 nm)/TCTA (10 nm)/mCP (10 nm)/8 wt.% dopant in mCP (20 nm)/TmPyPB (45 nm)/LiF (1 nm)/Al. Figure 5 presents the chemical structures of the employed materials and device configuration. The obtained device characteristics and key parameters are summarized in Figure 6 and Table 4 for scrutiny. Here, 1,1-bis((di-4-tolylamino)phenyl)cyclohexane (TAPC) and tris(4-carbazoyl-9-ylphenyl)amine (TCTA) are taken as the hole-transporting and electronblocking layer. 1,3-Bis(N-carbazolyl)benzene (mCP) serves as both the hole-blocking layer and host in the emissive layer. 1,3,5-Tri(3-pyridyl-3-phenyl)benzene (TmPyPB), LiF and Al are acting as the electron-transporting layer, electron injection layer and cathode, respectively.   As showed in Figure 6, their normalized EL spectra resemble the PL spectra recorded in the degassed CH 2 Cl 2 solution, confirming that the emission is solely generated from the emitters, from which EL of Cb4 is also red-shifted compared to that of Cb1. Moreover, the Cb4-based device shows a relatively lower current density at the same voltage compared to that of the Cb1-based device, which can be ascribed to the carrier trapping effect of Cb4 with a narrower energy gap than that of Cb1 [45,46]. In contrast, the Cb1-based device exhibited a bright green emission with EL peak at 530 nm and a maximum luminance of 12,420 cd/m 2 at 11.5 V, while the Cb4-based device delivered a yellow EL peak centered at 559 nm with a maximum luminance of 21,480 cd/m 2 at 13.0 V. A maximum external quantum efficiency (current efficiency) of 16.6% (55.2 cd/A) and 13.9% (43.8 cd/A) was also observed for Cb1-and Cb4-based devices, respectively. More importantly, both OLED devices present a small efficiency roll-off at 1000 cd/m 2 (15.4% and 12.1% for Cb1 and Cb4-based devices, respectively), evidencing good carrier balance during device operation.

Conclusions
In summary, by introducing varied substituents at the 4-position of central pyridinyl fragment of dianionic chelate or on the central phenyl coordination unit of carboline-based monoanionic pincer chelate, a series of five bis-tridentate Ir(III) complexes were successfully designed and synthesized, with an isolation yield higher than 50% and absence of any isomeric product. This result is consistent with those documented in literature [37,43]. The addition of methoxy and dimethylamino substituents at the 4-position of central pyridinyl fragment of dianionic chelate effectively increased the electron density at the Ir(III) metal center, which increased the MLCT contribution at the excited states, and gave a structureless emission profile. As for Ir(III) complexes Cb4 and Cb5, the tert-butyl substituent on the 4-position of the phenyl ring also red-shifted the emission and exhibited slightly reduced emission quantum yields. Next, Cb1 and Cb4 were doped into the emission layer for fabrication of OLEDs, achieving a maximum external quantum efficiency (current efficiency) of 16.6% (55.2 cd/A) and 13.9% (43.8 cd/A), respectively. The wellperformed electroluminescence efficiencies indicate that the studied bis-tridentate Ir(III) complexes and their future derivations are promising candidates for OLED applications.
Supplementary Materials: The following are available online. General experimental procedures of all measurements and calculations, synthetic protocol of chelates, original electrochemical data and detailed TD-DFT results of studied Ir(III) metal complexes. Scheme S1. Synthetic protocol given the employed dianionic chelates (phyz)H 2 ; Scheme S2. Synthetic protocol given the employed carboline chelates (cbF)H·HF 6 and (cbB)H·HF 6 ; Figure S1. Frontier molecular orbitals pertinent to the optical transitions for the ground state S 0 , excited state T 1 and S 1 of Ir(III) complex Cb1. The electron density distributions of Ir atoms in each molecular orbital are shown; Figure S2. Frontier molecular orbitals pertinent to the optical transitions for the ground state S 0 , excited state T 1 and S 1 of Ir(III) complex Cb2. The electron density distributions of Ir atoms in each molecular orbital are shown; Figure S3. Frontier molecular orbitals pertinent to the optical transitions for the ground state S 0 , excited state T 1 and S 1 of Ir(III) complex Cb3. The electron density distributions of Ir atoms in each molecular orbital are shown; Figure S4. Frontier molecular orbitals pertinent to the optical transitions for the ground state S 0 , excited state T 1 and S 1 of Ir(III) complex Cb4. The electron density distributions of Ir atoms in each molecular orbital are shown; Figure S5. Frontier molecular orbitals pertinent to the optical transitions for the ground state S 0 , excited state T 1 and S 1 of Ir(III) complex Cb5. The electron density distributions of Ir atoms in each molecular orbital are shown; Figure S6. Thermal gravimetric analysis of studied Ir(III) complexes Cb 1-5 with a decomposition temperature (T d ) showing a loss of 5% in weight; Table S1. The calculated wavelengths, transition probabilities and charge transfer character of the optical transitions for Ir(III) complex Cb1 in CH 2 Cl 2 ; Table S2. The calculated wavelengths, transition probabilities and charge transfer character of the optical transitions for Ir(III) complex Cb2 in CH 2 Cl 2 ; Table S3. The calculated wavelengths, transition probabilities and charge transfer character of the optical transitions for Ir(III) complex Cb3 in CH 2 Cl 2 ; Table S4. The calculated wavelengths, transition probabilities and charge transfer character of the optical transitions for Ir(III) complex Cb4 in CH 2 Cl 2 ; Table S5. The calculated wavelengths, transition probabilities and charge transfer character of the optical transitions for Ir(III) complex Cb5 in CH 2 Cl 2 .