Electrocatalysts and photocatalysts are key to a sustainable future, generating clean fuels, reducing the impact of global warming, and providing solutions to environmental pollution. Improved processes for catalyst design and a better understanding of electro/photocatalytic processes are essential for improving catalyst effectiveness. Recent advances in data science and artificial intelligence have great potential to accelerate electrocatalysis and photocatalysis research, particularly the rapid exploration of large materials chemistry spaces through machine learning. Here a comprehensive introduction to, and critical review of, machine learning techniques used in electrocatalysis and photocatalysis research are provided. Sources of electro/photocatalyst data and current approaches to representing these materials by mathematical features are described, the most commonly used machine learning methods summarized, and the quality and utility of electro/photocatalyst models evaluated. Illustrations of how machine learning models are applied to novel electro/photocatalyst discovery and used to elucidate electrocatalytic or photocatalytic reaction mechanisms are provided. The review offers a guide for materials scientists on the selection of machine learning methods for electrocatalysis and photocatalysis research. The application of machine learning to catalysis science represents a paradigm shift in the way advanced, next-generation catalysts will be designed and synthesized.

Machine Learning for Electrocatalyst and Photocatalyst Design and Discovery.

As a fundamental step of water splitting and a stepping stone toward exploring other multielectron transfer processes, the electrocatalytic hydrogen evolution reaction (HER) is an ideal model for both fundamental understanding and electrocatalyst design. Here, we review the fundamentals and recent developments of theoretical insights into HER, covering the mechanistic aspects, key activity descriptors, local environment considerations, and advances beyond the computational hydrogen electrode. Although it is experimentally challenging to explore the active sites and mechanisms in the electrocatalytic process, computational and theoretical advances show great potential in identifying active sites and reaction mechanisms. In this Review, we especially focus in depth on theoretical insights in revealing and designing the active sites for HER. Major challenges ahead will also be discussed at the end of the Review. 

Theoretical Advances in Understanding and Designing the Active Sites for Hydrogen Evolution Reaction

Photocatalytic oxygen evolution reaction (OER) by 2D semiconductors is a promising strategy for efficient energy conversion. The newly discovered 2D semiconductors MA2Z4 (M = transition metal, e.g., Mo, W; A = C, Si, and Ge; Z = N, P, and As) have shown characteristics of promising photocatalytic OER catalysts. Herein, an automated high‐throughput workflow is proposed to efficiently screen MA2Z4 photocatalytic OER catalysts, and further reveal the theoretical mechanism. The four criteria of semi‐conductivity, bandgap width/band edge position, structural stability, and free energy change, are proposed to screen MA2Z4 catalysts thermodynamically. The exciton properties and carrier mobility of these MA2Z4 are further studied to explore the high‐efficiency photocatalytic OER MA2Z4 with low exciton binding energy and high effective mass ratio. It is found that the bandgap width of MA2Z4 mainly depends on the strength of MZ bonding and the redistribution of electrons, while the catalytic effect is closely related to the adsorption capacity of O atom. In particular, β‐ZrSi2N4 and β‐HfSi2N4 are screened as efficient photocatalytic OER catalysts. This study develops a fully automated method and contributes to a complete framework for screening high‐efficiency photocatalytic OER catalysts.

Discovery of Efficient Visible‐light Driven Oxygen Evolution Photocatalysts: Automated High‐Throughput Computational Screening of MA2Z4

Modeling, optimization and understanding of adsorption process for pollutant removal via machine learning: Recent progress and future perspectives.

The discovery of 2D layered MoSi2N4 and WSi2N4 without knowing their 3D parents by chemical vapor deposition in 2020 has stimulated extensive studies of 2D MA2Z4 system due to its structural complexity and diversity as well as versatile and intriguing properties. Here, a comprehensive overview on the state‐of‐the‐art progress of this 2D MA2Z4 family is presented. Starting by describing the unique sandwich structural characteristics of the emerging monolayer MA2Z4, their versatile properties including mechanics, piezoelectricity, thermal transport, electronics, optics/optoelectronics, and magnetism is summarized and anatomized. The property tunability via strain engineering, surface functionalization and layered strategy is also elaborated. Theoretical and experimental attempts or advances in applying 2D MA2Z4 to transistors, photocatalysts, batteries and gas sensors are then reviewed to show its prospective applications over a vast territory. New opportunities are further discussed and prospects are suggested for this emerging 2D family. The overview is anticipated to guide the further understanding and exploration on 2D MA2Z4.

Emerging Versatile Two‐Dimensional MoSi2N4 Family

Two-dimensional (2D) materials have been developed into various catalysts with high performance, but employing them for developing highly stable and active nonprecious hydrogen evolution reaction (HER) catalysts still encounters many challenges. To this end, the machine learning (ML) screening of HER catalysts is accelerated by using genetic programming (GP) of symbolic transformers for various typical 2D MA2Z4 materials. The values of the Gibbs free energy of hydrogen adsorption (ΔGH*) are accurately and rapidly predicted via extreme gradient boosting regression by using only simple GP-processed elemental features, with a low predictive root-mean-square error of 0.14 eV. With the analysis of ML and density functional theory (DFT) methods, it is found that various electronic structural properties of metal atoms and the p-band center of surface atoms play a crucial role in regulating the HER performance. Based on these findings, NbSi2N4 and VSi2N4 are discovered to be active catalysts with thermodynamical and dynamical stability as ΔGH* approaches to zero (-0.041 and 0.024 eV). In addition, DFT calculations reveal that these catalysts also exhibit good deuterium evolution reaction (DER) performance. Overall, a multistep workflow is developed through ML models combined with DFT calculations for efficiently screening the potential HER and DER catalysts from 2D materials with the same crystal prototype, which is believed to have significant contribution to catalyst design and fabrication.

Jiaxi Hu

Papers

Symbolic Transformer Accelerating Machine Learning Screening of Hydrogen and Deuterium Evolution Reaction Catalysts in MA2Z4 Materials.