Atomic positional disorder in fcc metal nanocrystalline grain boundaries

Abstract
We investigate the degree of grain boundary positional disorder in computer generated fcc nc-Ni samples, as a function of grain size. We find that these grain boundary atoms display a lack of a second peak in the pair distribution function by virtue of their classification, and find further that such a class of atoms cannot be easily selected via an energy criterion. Despite being positionally disordered, atomic visualization demonstrates that even these atoms display can a certain degree of regularity by belonging to, for example, localized regions of miss fit. This confirms once again, that in terms of the degree of disorder, the nature of the nanocrystalline grain boundary network is not fundamentally different from that of the polycrystalline regime.