Accelerating OLED development with machine learning: advances and prospects

Abstract

Organic light-emitting diodes (OLEDs) have garnered significant attention and demonstrated robust application potential across diverse fields. Traditional OLED material and device research and development—reliant on empirical experimentation and heuristic knowledge—faces inherent limitations in efficiency and scalability. In recent years, machine learning (ML) has emerged as a transformative data-driven paradigm in accelerating OLED innovation. This review comprehensively examines ML's role in advancing OLED technology, focusing on (i) OLED luminescent material property prediction, (ii) quantitative structure–property relationship (QSPR) construction and high-throughput virtual screening (HTVS) of OLED materials, and (iii) device structural optimization strategies. Through critical case studies, the efficacy and constraints of universal ML models and deep learning (DL) in OLED contexts are evaluated, offering prospective directions for future research. Collectively, this review provides an in-depth academic foundation to catalyze sustained innovation in ML-accelerated OLEDs and related optoelectronic technologies.