A Cost-Effective Approach to Precisely Estimate Singlet−Triplet Energy Gaps in MR-TADF Molecules: Combining Delta Self-Consistent Field (ΔSCF) and Time-Dependent Density Functional Theory (TD-DFT) Methods

As a novel class of purely organic fluorescent materials, multiple resonance thermally activated delayed fluorescence (MR-TADF) compounds hold significant promise for next-generation display technologies. The efficiency of exciton utilization and the overall performance of organic light-emitting devices are closely linked to the singlet-triplet energy gap (∆EST) of MR-TADF emitters. Identifying an economic and accurate theoretical approach to predict ∆EST would be beneficial for high-throughput screening and facilitate the inverse design of MR-TADF molecules. In this study, we evaluated the S1 state energy (E(S1)), T1 state energy (E(T1)), and ∆EST using three different physical interpretations: adiabatic excitation energy, vertical absorption energy, and vertical emission energy. We employed the TDDFT and ΔSCF methods to calculate E(S1), E(T1) and ∆EST for 20 MR-TADF molecules reported in the literature. We compared these calculated values with experimental data obtained from fluorescence spectroscopy at room-temperature (or 77 K) and phosphorescence spectroscopy conducted at 77 K. Our findings indicate that the vertical absorption energy at the S0 state minimum, determined by the ΔSCF method, accurately predicts the S1 state energy. Similarly, the vertical absorption energy at the S0 state minimum, calculated using the TDDFT method, effectively predict the T1 state energy. The ∆EST derived from the difference between these two excitation energies exhibited the smallest mean absolute error of only 0.039 eV compared to the experimental values. This combination represents the most accurate and cost-effective method reported to date for predicting the ∆EST of MR-TADF molecules, and can be integrated into AI-driven inverse design workflows for new emitters.