Design strategies for alkaline HER catalysts with machine learning as an accelerator

Hydrogen from water electrolysis is central to low-carbon energy systems. Alkaline electrolysers offer flexible operation and cheaper component choices than acidic systems, yet the hydrogen evolution reaction (HER) remains slower because water dissociation is difficult and surface OH* can block active sites. This review maps recent progress and design rules for alkaline HER catalysts around three coupled levers - electronic structure, interfaces, and nanostructure - and link them to the rate-governing descriptors: near-thermoneutral ΔG_H*, a low Volmer barrier for water splitting, and moderated OH* adsorption. We assess electronic tuning (alloying, strain, and defects), interface engineering, and nanostructure designs (morphology, high-index-facet, and single/dual-atom modulation), highlighting the transport aware benchmarking that makes results transferable. Finally, we show how machine learning (ML) accelerates discovery by coupling theory-derived descriptors with standardized, condition-aware data, and the need for labels splits. Looking ahead, durable architectures that accelerate the water dissociation step with optimal hydrogen-binding sites, maintain that balance under bias, and operate at device-relevant current densities in strong alkali are prioritized.