Abstract
We recently leveraged the FORMED repository made of 116,687 synthesizeable molecules to deploy fragment-based high-throughput virtual screening (HTVS) and genetic algorithm (GA) searches of singlet fission (SF) molecular candidates. With these approaches, both prototypical (e.g., acenes, boron-dipyrromethane (BODIPY)) and unknown (e.g., heteroatom-rich mesoionic) classes of chromophore candidates fulfilling stringent SF energetic requirements were identified. Yet, the reliance on pre-defined fragments limits chemical space exploration and, thus, the discovery of truly unforeseen molecular cores. Here, we exploit a generative learning framework driven by reinforcement learning and property predictions. The generative model rediscovers a diverse range of previously reported SF chromophore classes, including polyenes, benzofurans, fulvenoids, and quinoidal systems, but also suggests a previously unreported SF scaffold not found in the training data, neocoumarin (2-benzopyran-3-one), characterized by two endocyclic double bonds in an ortho arrangement and capped by a lactone group. An in-depth investigation reveals a diradicaloid behavior over the conjugated core comparable to 2-benzofuran, a widely-known SF compound. This work highlights the potential of inverse design pipelines using both generative and property prediction models to discover candidates beyond derivatives of known chemistry for tailored material applications.
Supplementary materials
Title
Supporting Information: Generative Design of Singlet Fission Materials by Revisiting the Use of a Fragment-oriented Database
Description
Supporting Information.
Actions
Supplementary weblinks
Title
SF_generative
Description
Github repository containing scripts, configuration files, and trained models.
Actions
View