Machine learning-assisted exploration of a versatile polymer platform with charge transfer-dependent full-color emission

The development of color-tunable fluorescent materials with sim-ple chemical compositions that are easy to synthesize is highly desirable but practically challenging. Here, we report a versatile yet simple platform based on through-space charge transfer (TSCT) polymers that has full-color-tunable emission and was developed with the aid of predictive machine learning models. Us-ing a single-acceptor fluorophore as the initiator for atom transfer radical polymerization, a series of electron donor groups contain-ing simple polycyclic aromatic moieties (e.g., pyrene) are intro-duced either by one-step copolymerization or by end-group func-tionalization of a pre-synthesized polymer. By manipulating donor-acceptor interactions via controlled polymer synthesis, continuous blue-to-red emission color tuning was easily achieved in solid polymers. Theoretical investigations confirm the structur-ally dependent TSCT-induced emission redshifts. We also exem-plify how these TSCT polymers can be used as a general design platform for solid-state stimuli-responsive materials with high -contrast photochromic emission by applying them to proof-of -concept information encryption.