Prediction of 2D Materials for Energy Applications Using Computational Methods

The recent application of data-driven methods for molecule and material discovery has shown significant promise. In this context, we have developed an artificial intelligence (AI)-aided virtual screening recipe for two-dimensional (2D) materials discovery, enhancing the search for new materials with specific physical and chemical properties. As part of this effort, we have established the Virtual 2D Materials Database (V2DB), a publicly available resource that includes potentially stable 2D materials along with their AI-predicted key physicochemical properties [1].

Our subsequent focus is on pinpointing the functional 2D materials constituted by abundant chemical elements for energy applications, particularly those suitable for solar-driven photocatalytic water splitting to produce hydrogen (H2). The challenge lies in efficiently navigating the vast chemical space to identify promising 2D materials. To tackle this, we utilize a data-centric approach, screening the V2DB to find stable candidates with appropriate band gaps and optimal photocatalytic properties. This robust virtual screening process incorporates machine learning (ML), high-throughput density functional theory (HT-DFT), hybrid-DFT, and GW calculations.

Through this approach, we have identified 27 new 2D materials that show good potential for photocatalytic water splitting [2,3]. We then performed a detailed analysis of their solar water splitting properties, including electronic and optical features, solar-to-hydrogen conversion efficiency, and carrier mobility. These studies not only introduce new 2D photocatalysts but also highlight the efficiency of a data-driven strategy in systematically exploring materials in an extensive chemical space.

Our approach is versatile in identifying materials with specific properties for various renewable energy applications, including photovoltaic systems and electro/photo-catalytic conversion of feedstock molecules like H₂O, CO₂, and N₂ into valuable fuels and products, thereby exploring previously uncharted chemical spaces of 2D materials.

References:
[1] M.C. Sorkun, S. Astruc, J.M.V.A. Koelman, S. Er, npj Computational Materials 6, 106 (2020).
[2] Y. Wang, M.C. Sorkun, G. Brocks, S. Er, The Journal of Physical Chemistry Letters 15, 4983 (2024).
[3] Y. Wang, G. Brocks, S. Er, ACS Catalysis 14, 1336 (2024).