Physics-Based Modeling and Multi-Objective Optimization of Fluorescent OLEDs Accounting for Dopant-Singlet Exciton Losses

This article presents a multi-objective optimization framework for fluorescent organic light-emitting diodes (OLEDs), combining numerical simulation with an analysis of exciton populations and the associated loss mechanisms. An ITO/NPB/Alq3:C545T/Alq3/LiF–Al structure was modeled, calibrated, and validated against experimental electro-optical characteristics. The influence of the emissive layer thickness (tEML) and the C545T doping concentration (Dp) was systematically studied across 25 configurations, with tEML= 20–40 nm and Dp=1–9%, under two operating conditions: J=0.15 A/cm2 and L=5000 cd/m2. Increasing the dopant concentration led to a marked deterioration in current efficiency (ηc) and power efficiency (ηp), together with a progressive localization of singlet excitons near the interface between the emissive layer (EML) and the hole transport layer (HTL) in structures with thinner EMLs. This exciton localization was accompanied by an increased contribution from non-radiative deactivation pathways, which were incorporated into a dopant singlet-exciton loss fraction, Floss,d. The systematic increase in this loss fraction with increasing dopant concentration and its overall inverse relationship with ηc and ηp motivated the development of a three-objective formulation that maximizes ηc and ηp while minimizing Floss,d. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) identified compromise solutions at the lowest dopant concentration, Dp=1%, with tEML=24–25 nm under both operating conditions. Multi-objective Particle Swarm Optimization (MOPSO) identified a trade-off region comparable to that obtained by NSGA-II, providing a cross-algorithm consistency check of the reported numerical results. The proposed methodology thus establishes a physically grounded link between device design, the spatial redistribution of excitons, the modeled loss pathways, and the macroscopic performance of OLEDs, thereby providing an interpretable framework for the multi-objective optimization of fluorescent OLEDs prior to fabrication.