Ice phases under ambient and high pressure: Insights from density functional theory

Abstract

Water is common and plays a crucial role in biological, chemical, and physical processes, but its crystalline or ice state has a complicated structure. In this work, we study the lattice mismatch challenge for ice nucleation on silver iodide, the sublimation energy for different ice phases, and the structural phase-transition pressures of ice, with various density functionals. Our calculations show that the recently developed meta-generalized gradient approximation made simple (MGGA_MS) yields a lattice mismatch (3$%$) of hexagonal ice (ice Ih) with \ensuremath{\beta}-AgI in good agreement with experiment (2$%$), significantly better than the Perdew-Burke-Ernzerhof (PBE) GGA mismatch (6$%$). MGGA_MS is a computationally efficient semilocal functional that incorporates intermediate-range van der Waals (vdW) interaction, which, overall, performs well for ice and may be expected to improve upon PBE for liquid water. While MGGA_MS predicts the most realistic volumes and volume changes in the phase transitions of ice Ih to trigonal ice (ice II) and tetragonal ice (ice VIII), a more accurate description of some other properties of the higher-pressure phases (ice II and ice VIII) is provided by some functionals that include long-range vdW corrections (e.g., revised Tao-Perdew-Staroverov-Scuseria+vdW for sublimation energy and optB88-vdW for transition pressure).