Artificial Intelligence Designer for Highly-Efficient Organic Photovoltaic Materials.

Area of Science:

• Materials Science
• Organic Electronics
• Computational Chemistry

Background:

• Designing efficient organic photovoltaic (OPV) materials is complex and slow.
• Identifying fundamental structural units and their property relationships is crucial for material development.

Purpose of the Study:

• To develop an automated design framework for high-performance OPV donor molecules.
• To establish a substructure-property relationship using machine learning.
• To accelerate the discovery of novel OPV materials.

Main Methods:

• Utilized an in-house La FREMD Fingerprint and machine learning (ML) algorithms.
• Generated a library of 18,960 new potential OPV donor molecules.
• Employed Density Functional Theory (DFT) for material property analysis.

Main Results:

• Identified key molecular building blocks for OPV performance.
• Proposed design guidelines for efficient OPV materials.
• Predicted power conversion efficiencies (PCEs) exceeding 15% for promising candidates when paired with acceptor Y6.
• DFT confirmed excellent charge carrier transport potential in candidate materials.

Conclusions:

• The developed framework effectively designs new materials based on ML-derived substructure-property relationships.
• This methodology offers an alternative approach for applying ML in new material discovery.
• The study provides a pathway for accelerating the development of next-generation OPV materials.