Decohesion Energy of \(\Sigma 5(012)\) Grain Boundaries in Ni as Function of Hydrogen Content

Degradation due to hydrogen embrittlement (HE) is initiated at the nanoscale, and the development of a predictive understanding of HE will have to consider the range all the way from atomic to macro scale. Density functional theory (DFT) can be used to calculate the effect of hydrogen in metals on the atomic scale. It is, however, limited to small models consisting of \(\sim 1000\) atoms, which puts the relevance for real-life scenarios into question. In the current study, the effect of model size is explored for two grain boundary (GB) systems, \(\Sigma 3[110](111)\) and \(\Sigma 5[100](012)\). For the latter, the preferred sites for H occupation and the sequence in which sites are filled upon increasing H contents are investigated, and decohesion energies are calculated as function of H content. Emphasis is laid on identifying the most physical model for these studies, by evaluating effects of model size and structural relaxation procedures. It is found that the decohesion energy does not decrease linearly with H content and has to be calculated explicitly for full H coverage to find out how much H absorption weakens the GB.