High-throughput calculations combining machine learning to investigate the corrosion properties of binary Mg alloys
Magnesium (Mg) alloys have shown great prospects as both structural and biomedical materials, while poor corrosion resistance limits their further application. In this work, to avoid the time-consuming and laborious experiment trial, a high-throughput computational strategy based on first-principles calculations is designed for screening corrosion-resistant binary Mg alloy with intermetallics, from both the thermodynamic and kinetic perspectives. The stable binary Mg intermetallics with low equilibrium potential difference with respect to the Mg matrix are firstly identified. Then, the hydrogen adsorption energies on the surfaces of these Mg intermetallics are calculated, and the corrosion exchange current density is further calculated by a hydrogen evolution reaction (HER) kinetic model. Several intermetallics, e.g. Y3Mg, Y2Mg and La5Mg, are identified to be promising intermetallics which might effectively hinder the cathodic HER. Furthermore, machine learning (ML) models are developed to predict Mg intermetallics with proper hydrogen adsorption energy employing work function (W_f) and weighted first ionization energy (WFIE). The generalization of the ML models is tested on five new binary Mg intermetallics with the average root mean square error (RMSE) of 0.11 eV. This study not only predicts some promising binary Mg intermetallics which may suppress the galvanic corrosion, but also provides a high-throughput screening strategy and ML models for the design of corrosion-resistant alloy, which can be extended to ternary Mg alloys or other alloy systems.