High‐Efficiency Solution‐Processable OLEDs by Employing Thermally Activated Delayed Fluorescence Emitters with Multiple Conversion Channels of Triplet Excitons

Abstract The state‐of‐the‐art luminescent materials are gained widely by utilizing thermally activated delayed fluorescence (TADF) mechanism. However, the feasible molecular designing strategy of fully exploiting triplet excitons to enhance TADF properties is still in demand. Herein, TADF emitters with multiple conversion channels of triplet excitons are designed by concisely halogenating the electron acceptors containing carbonyl moiety. Compared with the chlorinated and brominated analogues, the fluorinated emitter exhibits distinguishing molecular stacking structures, participating in the formation of trimers through integrating C—H···F and C═O···H hydrogen bonds together. It is also demonstrated that the multiple channels can be involved synergistically to accelerate the spin‐flip of triplet excitons, and to take charge of the relatively superior reverse intersystem crossing constant rate of 6.20 × 105 s–1, and thus excellent photoluminescence quantum yields over 90% can easily be achieved. Then the solution‐processable organic light emitting diode based on fluorinated emitter can achieve a record‐high external quantum efficiency value of 27.13% and relatively low efficiency roll‐off with remaining 24.74% at 1000 cd m−2. This result manifests the significance of enhancing photophysical properties through constructing multiple conversion channels of triplets excitons for high‐efficiency TADF emitters and provides a guideline for the future study.

platinum wire auxiliary electrode, and an Ag/AgNO 3 pseudo-reference electrode were used in the conventional three-electrode system. Cyclic voltammograms were performed from 0 to 1.5 V at scan rate of 100 mV s -1 . Then HOMO and LUMO energy levels have been calculated according to the internal reference ferroceneredox couple in acetonitrile by using the following formulas: [1] ( ( ⁄ ) ) ( ( ⁄ ) ) UV/Vis absorption spectra were recorded on a Hitachi U-2910 spectrophotometer. The PL spectra, including fluorescence at 77 K, and phorescence at 77 K, were recorded on a Hitachi F-7000 fluorescence spectrophotometer, and the energy gap (ΔE ST ) between lowest singlet (S 1 ) and triplet excited states (T 1 ) was determined from the difference values of the onset positions of fluorescencent and phorescencent spectra. To gain the proportion of delayed fluorescence (DF), the steady-state PL spectra in vacuum and in air were performed using FLS-980 spectrometer from Edinburgh Instruments Limited with Xe lamp source. Transientstate PL spectra in vacuum and in air, including prompt fluorescence (PF) and delay fluorescence (DF) spectra, and the temperature dependence of transient PL decay curves, were determined using nanosecond gated luminescence and lifetime measurements with a high-energy pulsed Nd:YAG laser emitting at 320 nm to fit the transient decay curves using following equation: ⁄ ⁄ . PLQY was directly obtained from Hamamatsu Absolute PL quantum yield spectrometer C11347 series in air. The TA spectra were recorded by a commercial TA system equipped with a ns pulse laser as the pump source and a broadband laser-driven light source as the probe source. The detector output was fed to a digital oscilloscope, which acquired the waveform and stored it for eventual data processing. The pump wavelength was 330 nm, and detected range is 400-780 nm All the measurements were carried out at room temperature.

S 1.2 Molecular Simulation and Calculation
The following procedures were performed to optimize the molecular structures, calculate the energy properties and analyze the excited states of polymer representative fragments.
Firstly, the molecular structures were optimized in Gaussian 09 by performing a density functional theory (DFT) calculation in the CAM-B3LYP mode with a 6-31G (d, p) basis set in the ground state. Then according to the optimized results, highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) levels, and ground-state (S 0 ) dipole moments can be obtained. After that, the excited states energy levels and the energy properties of these polymers in the excited states were determined using Gaussian 09 with with a 6-31G (d, p) basis in a time dependent (TD) mode. To boost the calculation precision of the singlet and triplet energy levels, the number of calculated states was set to 10. Hence, the difference of dipole moments between S 0 and S 1 , and transition dipole moments from S 1 to S 0 can be obtained from electron excitation analysis of TD-DFT calculations.
In order to get better insight of the natures of excited states, natural transition orbital (NTO) analysis was performed using Multiwfn 3.6 based on Gaussian output results. [2] NTO analysis was selected to analyze the singlet and triplet excited states in which the sum of square of excitation coefficients was close to ideal value of 0.5, and the eigenvalues of occupied NTOs and maximal NTO pairs were also shown. The distributions of hole and electron can be found from the NTO results and the dominate nature, charge transfer (CT) or locally excited state (LE), can be found out. In order to quantitatively determine the natures of singlet and triplet excited states, the overlap integral, S r , was defined as the proportion of LE characters in the excited states by using the following equation: [3][4][5] √ Where the and are the distribution of hole and electron, respectively.
So with the intention of analyzing the impact of nature flip quantitatively, the SOC matrix elements for above three fragments were calculated using the ORCA 4.1 package with CAM-B3LYP/G-6-31G (d, p) method, and five states were considered in the analog computations. The compound of BD-F was obtained following the literature [6] . The compound of BD-Cl was obtained following the literature [6] .    was added to the reaction system in batches and stirred for 5 h at 0 ºC. After that, F-BP-Br (1395.6 mg, 5 mmol) was added to the reaction system slowly, and then the mixture was stirred for another 12 hours after the reaction system returned to room temperature naturally.
Then the system temperature was raised to 60 ºC and continued to be stirred for 12 hours.
When the reaction system was returned to room temperature naturally, the reactants were poured into a solution of dilute hydrochloric acid to quench the reaction. After washing for three times by water, the organic phase was extracted with dichloromethane. Then the organic phase was dried with anhydrous magnesium sulfate, filtered and concentrate to obtain the crude product. The crude product was further purified by column chromatography on silica gel using petroleum ether/dichloromethane (v/v = 5/1) as eluent to obtain the compound of BD-Br as a yellow-green powder (1.69 g, 72% yield).