Organic Photodetectors Operating under Strong Sunlight: Combining Machine-Learning and Time-Dependent Density Functional Theory for Molecular Design of Diarylethene-Type Photochromic n-Type Dopants Mixed with p-Type Organic Semiconductors

Abstract

Linear dynamic ranges (LDR) of organic photodetectors, which are typically narrow due to the low charge mobilities of organic semiconductors, have been extended by doping them with diarylethene (DAE) photochromic switches. The speculated mechanism is that DAE acts as n-type trap only in its aromatic closed form which is predominant under strong sunlight. This mechanism involves photo-switchable change transfer between p-type donor polymer and DAE. We thus herein design optimal DAE dopants according to a set of design rules embracing the roles of their HOMO/LUMO energies. We first identified two optimal dopants out of 133 candidates, using time-dependent density functional theory (TDDFT) to calculate the HOMO/LUMO energies of their open/closed isomers (532 data). We then predicted those of 40,000 candidates through machine learning, identified additional optimal dopants, and confirmed them with TDDFT. One of the designed dopants indeed succeeded in LDR extension in real experiments.