Data-driven discovery of intrinsic direct-gap 2D photocatalysts for overall water splitting

Intrinsic direct-gap two-dimensional (2D) materials hold great promise as photocatalysts advancing the application of photocatalytic water splitting for hydrogen production. However, the time- and resource-efficient exploration and identification of such 2D materials from a vast compositional and structural chemical space present a significant challenge within the realm of materials science research. To this end, we perform a data-driven study to find new 2D materials with intrinsic direct-gap and desirable photocatalytic properties for overall water splitting. By implementing a three-staged large-scale screening, which incorporates machine learning, high-throughput density functional theory (DFT) and hybrid-DFT calculations, we identify 16 new direct-gap 2D materials as promising photocatalysts. Subsequently, we conduct a comprehensive assessment of material properties that are related with the solar water splitting performance, which includes electronic and optical properties, solar-to-hydrogen conversion efficiencies, and carrier mobilities. Therefore, this study not only presents 16 new 2D photocatalysts but also introduces a rigorous data-driven approach for the future discovery of functional 2D materials from currently unexplored chemical spaces.