Pyrazolate-Bridged NHC Cyclometalated [Pt2] Complexes and [Pt2Ag(PPh3)]+ Clusters in Electroluminescent Devices

The ionic transition metal complexes (iTMCs) [{Pt(C∧C*)(μ-Rpz)}2Ag(PPh3)]X (HC∧C* = 1-(4-(ethoxycarbonyl)phenyl)-3-methyl-1H-imidazole-2-ylidene, X = ClO4/PF6; Rpz = pz 1a/2a, 4-Mepz 1b/2b, and 3,5-dppz 1c/2c) were prepared from the neutral [{Pt(C∧C*)(μ-Rpz)}2] (Rpz = pz A, 4-Mepz B, and 3,5-dppz C) and fully characterized. The “Ag(PPh3)” fragment is in between the two square-planar platinum units in an “open book” disposition and bonded through two Pt–Ag donor–acceptor bonds, as shown by X-ray diffraction (dPt–Ag ∼ 2.78 Å, 1a–1c). 195Pt{1H} and 31P{1H} NMR confirmed that these solid-state structures remain in solution. Photoluminescence studies and theoretical calculations on 1a, were performed. The diphenylpyrazolate derivatives show the highest photoluminescence quantum yield (PLQY) in the solid state. Therefore, 2c and its neutral precursor C were selected as active materials on light-emitting devices. OLEDs fabricated with C showed a turn-on voltage of 3.2 V, a luminance peak of 21,357 cd m–2 at 13 V, and a peak current efficiency of 28.8 cd A–1 (9.5% EQE). They showed a lifetime t50 of 15.7 h. OLEDs using 2c showed a maximum luminance of 114 cd m–2, while LECs exhibited a maximum luminance of 20 cd m–2 and a current efficiency of around 0.2 cd A–1, with a t50 value of 50 min.


1.1
General procedures, materials and instrumentation.
IR spectra were recorded on a Perkin-Elmer Spectrum 100 FT-IR Spectrometer (ATR in the range 250-4000 cm-1).Mass spectral analyses were performed with a Microflex MALDI-TOF Bruker or an Autoflex III MALDI-TOF Bruker instruments.C, H, and N analyses were carried out in a Perkin-Elmer 2400 CHNS analyser or Thermo Flash 1112. 1 H, 13 C{ 1 H}, 31 P{ 1 H}, 195 Pt{ 1 H} NMR spectra were recorded on a Bruker NEO 400 and 500 MHz instruments using the standard references: SiMe4 for 1 H and 13 C, Na2PtCl6 in D2O for 195 Pt, 85 % H3PO4 for 31 P and CFCl3 for 19 F. J are given in Hz and δ are given in ppm; assignments are based on 1 H-1 H COSY experiments and 1 H- 13 C HSQC and HMBC experiments.Unless otherwise indicated, all measurement were performed at r.t.
UV-visible spectra were recorded on a Unicam UV4 spectrophotometer.Steady-state photoluminescence spectra were recorded on a Jobin-Yvon Horiba Fluorolog FL-3-11 Tau 3 spectrofluorimeter.Phosphorescence lifetimes were recorded with a Fluoromax phosphorimeter accessory containing a UV xenon flash tube.Nanosecond lifetimes were recorded with a Datastation HUB-B with a nanoLED controller and software DAS6.
NanoLEDs of 340 nm and 370 nm were employed for lifetimes measurements.The lifetime data were fitted using the Jobin-Yvon software package and the Origin Pro 8 program.Solid state Quantum Yields (QY) were measured using the Hamamatsu Absolute PL Quantum Yield Measurement System C11347-11.The absorbance and photoluminescence spectra of thin-films were measured with an Avantes AvaSpec-2048L spectrometer equipped with a Avantes AvaLight-DS-S-BAL deuterium halogen light source and optic fibres.For photoluminescence measurements, films were illuminated with a diode laser of integrated optics, with an emission wavelength of 365 nm.

Computational methods.
Density functional calculations were carried out on the ground (S0) state with the Gaussian S3 kcal/mol 6 ) together with Grimme's D3 dispersion correction. 7The ECP-60-mwb for platinum and ECP-28-mwb, for silver, pseudopotential 8 was used, and the 6-31G(d) 9,10 basis sets were used for all other atoms.In order to facilitate the theoretical study, we have done a simplification on the real system, we have modelled the ethanoate substituent on the cyclometalated ligand as an acetate.General geometry optimizations were performed without any symmetry restriction and in THF using the polarizable continuum model (PCM). 11equency calculations were performed in order to determine the nature of the stationary points found in So no imaginary frequencies for minima.Mulliken population analysis was carried out as implemented in Gaussian 16 package. 4ChemissianLab program package was used for analysis and graphic representation of molecular structures and orbitals and for Mayer Bond Order analysis.Atomic charges were calculated by using the NBO analysis option as incorporated in Gaussian 16. 4 1.3 X-ray Structure determinations.Experimental procedures and refinement.
Crystal data and other details of the structure analyses are presented in Table S1.Suitable crystals of 1a 2Me2CO,1b 0.25Me2CO, 1c 0.35C7H8 for X-ray diffraction studies were obtained by slow diffusion of n-hexane into saturated solutions of 1a, 1b in acetone or toluene into saturated solution of 1c in THF.Crystals were mounted at the end of quartz fibers and the data collection was performed at 100 K temperature.The radiation used in all cases was graphite monochromated Mo Kα (λ = 0.71073 Å).X-ray intensity data were collected on an Oxford Diffraction Xcalibur diffractometer.The diffraction frames were integrated and corrected from absorption by using the CrysAlis RED program. 12The structures were solved by Patterson and Fourier methods and refined by full-matrix least squares on F 2 with SHELXL. 13All non-hydrogen atoms were assigned anisotropic displacement parameters and refined without positional constraints, except as noted below.All hydrogen atoms were constrained to idealized geometries and assigned isotropic displacement parameters equal to 1.2 times the Uiso values of their attached parent atoms.For 1a 2Me2CO, the CH3 fragment of one of the ethyl residues is disordered over two positions which were refined with 0.5 partial occupancy each.One of the perchlorate anions has three of hits oxygen atoms disordered over two positions which were refined with 0.6/0.4partial occupancy.One of the crystallization acetone molecules is disordered over two positions with 0.5/0.5 partial occupancy.Some soft restrains were used in the interatomic distances for some of the disordered atoms.For 1b 0.25Me2CO, the crystal quality is not very good and it did not diffract intensely at medium and high angles.This was probably due to poor packing caused by the shape of the molecules.Thus, voids appear in the structure, which are treated using the SQUEEZE procedure implemented in PLATON. 14One of the perchlorates has three of its oxygen atoms disordered in two positions which are refined with 0.6/0.4occupancies.The other perchlorate is disordered in two different, but close, positions that refine to 0.68/0.32 occupancies.An acetone molecule, although diffuse, can be modelled with occupancy 0.5.
Restrains were used in some of the geometrical parameters of the disordered moieties.For 1c 0.35C7H8, the solvent toluene molecule lies diffusely around an inversion center and the position of its methyl group is disordered.It was not possible to anisotropize the thermal parameters for the C atoms of the toluene moiety.Full-matrix least-squares refinement of these models against F2 converged to final residual indices given in Table S1.CCDC Nos.
2309669-2309671 contain the supplementary crystallographic data for 1a, 1b and 1c.(10 mg mL -1 ) to a thickness of 30 nm.These substrates were transferred to the vacuum chamber, where BmPyPhB, Ba and Ag were thermally deposited.The vacuum evaporation rates were 0.1 nm s -1 and 0.05 nm s -1 for the ETL and cathode, respectively, with the background pressure being around 3 × 10 −6 bar.Shadow masks were used during the metal evaporation to obtain a final active area of 6 mm 2 .As for OLEDs with C as emitting layer, the vacuum evaporation rates of TAPC, mCP, emitting layer and PO-T2T were 0.6 nm s -1 , 0.6 nm s -1 , 0.01 nm s -1 and 0.1 nm s -1 , respectively, with the background pressure being around 3 × 10 −6 bar.
After full device fabrication, the samples were introduced into a setup for a current density and luminance versus voltage (JVL) scan.For this we employed a Keithley 2400 Source-Meter and a photodiode coupled to a Keithley 6485 picoammeter.A LabVIEW program was used to control the Keithleys and to obtain the data.The photodiode was calibrated using a Konica Minolta LS-150 equipped with a 110 close-up lens for the measurement of small areas and controlled through the CS-S20 Data Management Software.Electroluminescence (EL) spectra were recorded by driving the cells with the Keithley 2400 Source-Meter and an optical fibre connected to the Avant spectrometer AvaSpec-2048L.The external quantum efficiency (EQE) of the devices was extracted from the current efficiency and the EL spectra.

LECs fabrication and characterization.
A solution of the emitter was mixed in a molar ratio of 3.14:1 with the ionic liquid (IL) 1-butyl3-methyl-imidazolium-hexafluorophosphate (BMIM + PF6 -).The solvent used was dichloromethane (DCM).The final concentration of the emitter in the solution was 20 mg/mL.The same substrates as for the fabrication of OLEDs were used and the same cleaning procedure was followed, but after the ozone treatment, a suspension of PEDOT:PSS CH8000 was used instead.An 80-nm thick film was obtained at 4000 rpm for 60 seconds, which was then annealed on a hotplate at 150 °C for 10 minutes.The active layer solution .

(
emitter:IL) was then spin-coated at 2000 rpm for 60 seconds, resulting in a thickness from 90 nm to 120 nm.Films were covered with a beaker for the first 10-15 seconds of spinning to reduce the fast solvent release and improve the morphology of the film.The films were annealed at 90°C for 45 minutes.Finally, an Al electrode (100 nm) was thermally evaporated on top of the active layer using a shadow mask under inert atmosphere.The final active area of the cells was 6 mm 2 .The thickness of the PEDOT:PSS and active layer was determined with an Ambios XP-1 profilometer.The devices were measured by applying a pulsed current density (50 A m -2 ) while monitoring the voltage and luminance versus time by using a True Color Sensor MAZeT (MTCSiCT sensor) with a Botest OLT OLED Lifetime-Test system.The applied pulsed current consisted of block waves at a frequency of 1000 Hz with a duty cycle of 50%.Hence, the average current density and voltage were obtained by multiplying the values by the time-on (0.5 ms) and dividing by the total cycle time (1 ms).Electroluminescence (EL) spectra were recorded by driving the cells with the Botest OLT system and an optical fibre connected to the Avant spectrometer AvaSpec-2048L.The external quantum efficiency (EQE) of the devices was extracted from the current efficiency and the EL spectra.

5 .
Figure S10. 1 H (a) and 31 P{1H} (b) NMR spectra of 1a in THF-d 8 before and after irradiation with λ= 365 nm for 15 min.Spectra of the precursors have been included for comparison.

Figure S14 .
Figure S14.(a) Electroluminescence spectra of OLEDs employing 2c as the emitting material driven at different voltage values, from 20 V to 25 V.(b) CIE 1931 Chromaticity Diagram showing the corresponding device's color points at the same voltage values.

Table S4 .
Population Analysis (%) of frontier MOs in the S0 in THF for 1a.

Table S5 .
Selection of the most significant and lowest-energy vertical singlet and triplet excitations calculated by TD-DFT for 1a at the S0 in solution of THF.