Faceting transition in aluminum as a grain boundary phase transition

Grain boundaries facet because of anisotropic grain boundary energies: While the faceted boundary has a larger area than the corresponding straight boundary, a significantly lower energy of the facets compared to a straight segment can drive the faceting. This picture is complicated by faceting/defaceting transitions where the free energy difference between the two states depends on the temperature. Such transitions have been observed and modeled before, but their exact nature was not fully understood. Here, we use atomistic computer simulations to show that a well-known faceting transition in <a:math xmlns:a="http://www.w3.org/1998/Math/MathML"><a:mrow><a:mi mathvariant="normal">Σ</a:mi><a:mn>3</a:mn><a:mspace width="4pt"/><a:mo>[</a:mo><a:mn>11</a:mn><a:mover><a:mn>1</a:mn><a:mo>¯</a:mo></a:mover><a:mo>]</a:mo></a:mrow></a:math> tilt grain boundaries in Al is in fact a grain boundary phase transition (also called complexion transition). This means that the faceted and defaceted boundaries are associated with different atomic structures, which have different thermodynamic stability ranges. At low temperatures, the grain boundary phase associated with faceting is stable, while at high temperatures the flat grain boundary phase is stable. We also report on our thorough tests of Al interatomic potentials for this purpose, which include comparisons to density-functional theory calculations. Our chosen potential performs well for our grain boundaries. As a consequence, we were able to obtain results that align with previous experimental results.