Answers are grounded in this paper and cite their sources.
Metal hydrides (MHs) couple reversible hydrogen absorption and desorption with substantial reaction heat and are therefore candidates for thermochemical thermal energy storage (TES) and metal hydride heat pump (MHHP) systems. This review synthesizes publications identified through topic-oriented searches of multiple academic databases and examines the relationships among material thermodynamics and kinetics, reactor behavior, system configuration, and reported performance. The comparative analysis indicates that no hydride family is universally optimal. AB5 and AB2-type alloys are generally best suited to low- and medium-temperature, rapid-cycling MHHP duties because of their tunable equilibrium pressures and comparatively favorable kinetics, whereas Mg-based and complex hydrides offer greater reaction enthalpy for high-temperature TES but require higher regeneration temperatures and stronger heat transfer enhancement. Open-loop TES offers simpler active bed architecture when external hydrogen infrastructure is available; closed-loop TES enables internal hydrogen recovery but requires compatible paired beds and additional hardware. Thermally driven MHHP systems can exploit low-grade or waste heat but are constrained by hydride pair matching, hysteresis, sensible heat, and incomplete hydrogen transfer. Compressor-driven MHHP systems provide direct pressure control and may use the same hydride in both reactors, although compressor work, overheating, and parasitic losses reduce practical performance. Reported coefficient of performance (COP) values are not directly comparable across the two driving modes because their energy inputs differ and experiments remain below the predictions. Persistent gaps include low effective bed conductivity, limited long-term cycling and scale-up data, inconsistent performance boundaries, insufficient auxiliary energy accounting, and scarce techno-economic validation. The review provides an integrated material to the system framework for selecting configurations and interpreting evidence.
No takes yet. Share an insight, caveat, or question.
Li et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: