Abstract
Thermally activated delayed fluorescence (TADF) has emerged as a key mechanism for achieving nearly 100% internal quantum efficiency in organic light-emitting diodes (OLEDs). Conventional donor–acceptor (D–A) molecular architectures reduce the energy gap between the lowest excited singlet and triplet states, thereby promoting efficient reverse intersystem crossing and enhancing electroluminescence. However, these systems typically rely on relatively weak C–N donor–acceptor linkages, which may compromise their chemical and operational stability. In contrast, the multi-resonance TADF (MR-TADF) approach has proven highly effective in simultaneously achieving narrow emission spectra and high photoluminescence quantum yields. Nevertheless, most reported MR-TADF emitters incorporate boron atoms, whose high chemical reactivity presents challenges for material stability and practical device applications. Predictive theoretical modeling offers an efficient route for accelerating the discovery and rational design of high-performance OLED emitters. In this work, we present a multiscale computational framework for predicting the structural, energetic, and optical properties of both established and prospective TADF materials. The proposed methodology demonstrates high predictive accuracy and broad applicability to OLED material design. Furthermore, we introduce computational approaches for evaluating the chemical and structural stability of these compounds, providing additional guidance for the development of robust next-generation TADF emitters.