Publication date: 15th December 2025
High-throughput optical characterization and analysis are crucial for building digital twins of organic semiconductors, enabling accurate modeling of energetic landscapes and carrier dynamics to accelerate material discovery and device optimization. In this work, we fit time-resolved photoluminescence (TRPL) data of bulk heterojunction organic solar cells (BHJ-OSCs) using a rate-equation framework with seven free parameters, optimized through trust-region Bayesian optimization (TuRBO) combined with gradient-based methods. This approach provides detailed insights into the energies of charge-transfer (CT) and triplet local excited (T) states, as well as their associated rate constants and interfacial decay dynamics. By varying the molecular weight and halogenation of the materials, we find that although reducing the HOMO energy offset between the donor and acceptor increases the triplet formation rate to a level comparable to the charge generation rate, our analysis reveals that the dominant nonradiative losses in BHJ-OSCs originate from CT states, with only minor contributions from T states. Furthermore, reducing the energy offset between local excited (LE) and CT states enhances both PLQY and ELQY by shifting nonradiative decay from CT to LE states, but at the cost of charge generation efficiency. Therefore, comprehensive optimization of energetic and kinetic parameters is essential to balance performance—suppressing nonradiative losses from CT states while leveraging the long lifetime of T states as an energy reservoir.
