General-purpose neural network interatomic potential for the <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mi>?</mml:mi></mml:math> -iron and hydrogen binary system: Toward atomic-scale understanding of hydrogen embrittlement

To understand the physics of hydrogen embrittlement at atomic scale, a general-purpose neural network interatomic potential (NNIP) for $\ensuremath{\alpha}$-iron and binary system is presented. It trained using an extensive reference database produced by density functional theory (DFT) calculations. The NNIP can properly describe interactions with various defects in $\ensuremath{\alpha}$-iron, such as vacancies, surfaces, grain boundaries, dislocations; addition to basic properties itself, also handles defect behavior hydrogen-hydrogen vacuum, including molecule formation dissociation surface. These are superb challenges existing empirical potentials, like embedded-atom method based system. In this study, was applied several key phenomena necessary understanding embrittlement, charging discharging transportation defective trapping desorption defects, hydrogen-assisted cracking boundary. Unlike simulations quantitatively described atomistic details DFT accuracy.