Abstract
This work presents a systematic topological analysis of the electronic wave functions for a series of multi-resonance thermally activated delayed fluorescence (MR-TADF) materials based on B,N-heteroarenes. Using Multiwfn and DFT/TD-DFT calculations combined with a graph convolutional network (GCN), a set of hole-electron descriptors was calculated and correlated with experimental photophysical parameters Lem, FWHM, EST, and kRISC. Theoretical kRISC values were estimated using Fermi’s golden rule and spin-orbit coupling. A newly developed regression model, integrating theoretical kRISC and the nitrogen Mayer bond order (MBO(N)), significantly enhances the prediction accuracy of RISC efficiency. The results identify nitrogen framework rigidity as a quantitative predictor for the data-driven design of high-performance MR-TADF emitters.