EXPERIMENTAL METHODS

Theoretical calculation methods

All the DFT and TD‐DFT calculations were carried out using Gaussian 16 A.03 package. The optimized ground‐state geometries, energy levels, and FMO distributions were calculated on the basis of DFT by using the B3LYP/6‐31G (d, p) method. The geometry of the S 1 state was optimized by TD‐DFT at the B3LYP/6‐31G (d, p) level. Multifwn 3.8 was utilized to analyze the NTOs of excited states. SOC matrix elements were calculated by TD‐DFT and the ORCA 4.1.1 package at B3LYP/6‐31G (d, p).

Device fabrication and measurement

Patterned ITO glasses with a sheet resistance of 20 Ω per square were ultrasonically treated in detergents and deionized water and then dried for 30 mins at 120°C. After being treated by oxygen plasma for 7 mins, Clean ITO substrates were transferred into the vacuum deposition system. The target compounds used for the device have been sublimated. When the pressure was < 2 × 10−4 Pa, the devices were fabricated. The evaporation rates of organic materials LiF and Al were 1–1.5, 0.2 and 5–10 Å s−1 via a shadow mask, which were detected by a frequency counter and calibrated by a Dektak 6 M profiler (Veeco). The emitting area (3 × 3 mm2) was determined by the overlap between the ITO and Al electrodes. The current density–luminance–voltage characteristics were performed by a computer controlled Keithley 2450 series digital source‐meter and LS160 luminancemeter. EL spectra at different voltages were recorded by the optical analyzer FlAME‐S‐VIS‐NIR photometer. Supposing that the light emitted by the devices is accorded with the Lambertian distribution, the EQEs can be calculated from the EL spectra, luminance, and current density.