Molecular design for high-performance organic photosensitizers

Abstract

Organic photosensitizers (PSs) are central to light-driven photochemical and photobiological processes across biomedicine, energy, environment and synthetic photochemistry, owing to their tunable molecular structures and versatile excited-state dynamics. The photosensitization mechanisms of organic PSs can be broadly classified into oxygen-dependent type II, less oxygen-dependent type I and emerging oxygen-independent pathways. Building on these mechanistic insights, diverse molecular design strategies have been developed to fine-tune excited-state processes and to improve photosensitization efficiency. However, the incomplete understanding of structure–property relationships has motivated the integration of machine learning as a data-driven and predictive framework to streamline candidate discovery and optimization. In this Review, we first summarize the molecular design guidelines for type II, type I and oxygen-independent PSs, followed by a discussion of strategies that modulate the photosensitization pathways. We then examine the machine-learning approaches for predictive PS discovery. Finally, we outline the challenges in designing high-performance PSs and future directions for machine learning towards more reliable and intelligent PS discovery. Organic photosensitizers enable diverse light-driven processes in biomedicine, energy, environment and synthetic photochemistry. This Review outlines molecular design strategies and photosensitization pathway modulation across type II, type I and oxygen-independent mechanisms and discusses machine-learning approaches for predictive and data-driven photosensitizer discovery.