An Intelligent, User‐Inclusive Pipeline for Organic Semiconductor Design

Abstract

AbstractData‐guided methodologies for designing new materials are developing apace, yet advances for organic crystals have been infrequent. For the case of organic crystals, data‐guided design platforms involve two problems: the need to construct regression models that can predict solid‐state properties from molecular structures alone and the vast number of molecular structures that satisfy a targeted solid‐state property. In this paper, a new pipeline is presented for designing molecules with targeted solid‐state electronic properties. Here, the first problem is overcome by means of a novel visualization method for band structure data, which allows the user to specify their design preferences and simplify the creation of the required regression models. The second problem is overcome by allowing for user intervention in the selection of the final molecule for subsequent synthesis. The effectiveness of this pipeline is demonstrated by using it to design a simple new molecule with a targeted bandgap and impressive band dispersion. This pipeline is the first data‐guided method that can design new molecules on the basis of genuine solid‐state electronic properties, and with its emphasis on user‐inclusivity, it can easily be applied in collaborative or industry settings.