Cost-Effective Approach for Modeling of Multiresonant Thermally Activated Delayed Fluorescence Emitters

Multiresonant thermally activated delayed fluorescence (MR-TADF) emitters have recently attracted great interest for application in organic light-emitting diodes due to their remarkable electroluminescent efficiency and narrow emission spectra. It is therefore essential to establish computational methodologies that can accurately model the excited states of these materials at manageable computational costs. With regard to MR-TADF design and their associated photophysics, previous works have highlighted the importance of wave function-based methods, at much higher computational costs, over the traditional time-dependent density functional theory approach. Herein, we employ two independent techniques built on different quantum mechanical frameworks, highly correlated wave function-based STEOM-DLPNO-CCSD and range-separated double hybrid density functional, TD-B2PLYP, to investigate their performance in predicting the excited state energies in MR-TADF emitters. We demonstrate a remarkable mean absolute deviation (MAD) of ∼0.06 eV in predicting ΔE<inf>ST</inf> compared to experimental measurements across a large pool of chemically diverse MR-TADF molecules. Furthermore, both methods yield superior MAD in estimating S<inf>1</inf> and T<inf>1</inf> energies over earlier reported SCS-CC2 computed values [J. Chem. Theory Comput. 2022, 18, 4903]. The short-range charge-transfer nature of low-lying excited states and narrow fwhm values, hallmarks of this class of emitters, are precisely captured by both approaches. Finally, we show the transferability and robustness of these methods in estimating rates of radiative and nonradiative events with adequate agreement against experimental measurements. Implementing these cost-effective computational approaches is poised to streamline the identification and evaluation of potential MR-TADF emitters, significantly reducing the reliance on costly laboratory synthesis and characterization processes.