Machine learning-based closed-loop for optimizing HOMO-LUMO gap in diarylethene.

Area of Science:

• Materials Science
• Computational Chemistry
• Machine Learning Applications

Background:

• Diarylethenes are photochromic molecular switches crucial for photoresponsive applications.
• Optimizing the Highest Occupied Molecular Orbital-Lowest Unoccupied Molecular Orbital (HOMO-LUMO) gap is essential for diarylethene performance.
• Rational structural modification is key to tuning the HOMO-LUMO gap and expanding diarylethene functionality.

Purpose of the Study:

• To introduce an automated, closed-loop optimization framework for designing diarylethene derivatives.
• To efficiently identify molecules with a target HOMO-LUMO gap using machine learning.
• To accelerate the discovery of photoresponsive materials with tailored electronic properties.

Main Methods:

• Development of a machine learning model trained on existing data to act as a surrogate predictor.
• Implementation of a closed-loop optimization framework for automated exploration of chemical space.
• Validation of top-predicted candidate molecules using density functional theory (DFT) calculations.

Main Results:

• The automated framework efficiently identified diarylethene candidates optimized for a target HOMO-LUMO gap.
• The machine learning approach significantly reduced the need for costly experimental measurements.
• The proposed strategy demonstrated superior performance compared to existing methods in identifying optimal molecules.

Conclusions:

• The study presents a general methodology for integrating molecular structure data with machine learning.
• The developed framework and tools accelerate the design and discovery of advanced photoresponsive materials.
• This approach enables the creation of diarylethene-based systems with precisely tailored electronic properties.