The Pressure-Composition- Temperature (PCT) analysis is an experimental procedure employed to assess the suitability of an alloy for hydrogen storage. PCT isotherms provide crucial information, like reversible storage capacity, hydrogen pressure for phase transition, and the plateau pressure. Unfortunately, PCT analysis of the materials is resource-intensive and time-consuming as it involves series of measurements to represents the relationship between hydrogen pressure, concentration and temperature of a sample at equilibrium. Hence, limiting the number of compositions that can be investigated. In light of this, we have built a machine learning model, Metal Hydride’s PCT isotherm predictior (MH-PCTpro) for multicomponent metal alloys. The model is trained on diverse family of metal alloy’s PCTs data and the feature set includes easily calculable elements’ periodic table properties, hydriding properties, and experimental parameters. The comprehensive feature set equips PCTpro to predict the PCT isotherms for any metal alloy based on its composition, hydrogen pressure, and temperature. The model is validated across diverse alloy families, agreeing with experimental results. The model also predicts temperature-dependent variations in plateau pressure, enabling the mapping of predicted PCTs onto Van’t Hoff plots to determine enthalpy and entropy of hydride formation.
.png)