Discovery of Two-Dimensional Multinary Component Photocatalysts Accelerated by Machine Learning.

Area of Science:

• Materials Science
• Chemistry
• Nanotechnology

Background:

• Discovering novel two-dimensional (2D) materials is crucial for advancing photocatalysis.
• Multinary 2D compounds offer greater structural and property diversity than binary materials but are less explored.

Purpose of the Study:

• To develop an efficient machine learning (ML) approach for predicting 2D multicomponent photocatalysts.
• To identify promising multinary compounds for photocatalytic applications, particularly overall water splitting.

Main Methods:

• Utilized a machine learning (ML) technique combined with high-throughput screening.
• Examined over 4000 2D monolayers to predict photocatalytic properties.

Main Results:

• Identified 75 multinary compounds suitable for photocatalytic applications.
• Highlighted ternary and quaternary compounds (A2P2X6 and ABP2X6) with superior properties for water splitting.
• Specific promising compositions include A = Cu/Zn/Ge/Ag/Cd, B = Ga/In/Bi, and X = S/Se.

Conclusions:

• The developed ML approach provides an efficient pathway for exploring novel 2D photocatalysts.
• The identified multinary compounds show significant potential for overall water splitting.
• This work encourages further theoretical and experimental research into 2D multinary compounds for photocatalysis.