Experimental and Machine Learning-Assisted Discovery of 2D Materials for Hydrogen Evolution: From Fundamentals to Industrial Applications

Abstract

Water splitting to produce hydrogen is recognized as a green technology with significant potential to replace traditional non-renewable energy sources. Substantial progress has been made in the hydrogen evolution reaction (HER), with two-dimensional (2D) materials for both photocatalytic and electrocatalytic HER due to their unique structural features and favourable properties. Along with experimental materials design, the properties of the 2D materials have been complemented by computational methods such as density functional theory (DFT) over the past decade. However, these computational approaches face limitations in terms of time and cost efficiency. Consequently, data-driven approaches, particularly machine learning (ML), are emerging as powerful tools in materials science for identifying structure-activity relationships by learning from existing experimental and DFT calculation data. This review discusses the progress of 2D materials for hydrogen evolution, encompassing experimental advances, theoretical insights, and ML-assisted discovery. First, the fundamental principles of HER are examined, combining insights from photocatalysis and electrocatalysis. Next, an overview of 2D materials for HER is presented, including key challenges related to kinetics, stability, and scalability. Subsequently, ML strategies for 2D material discovery and screening are explored. Case studies on ML applications for various 2D photocatalysts and electrocatalysts, including graphene, g-C3N4, transition metal chalcogenides, MXenes, etc., are discussed. Finally, factors influencing large-scale applications and challenges associated with integrating materials science and ML approaches for HER are addressed.