A powerful computational crystallography method to study ice polymorphism

Area of Science:

• Computational physics
• Materials science
• Crystallography

Background:

• Investigating the structural properties of ice crystals is crucial for understanding water's complex behavior under varying conditions.
• Classical molecular dynamics (MD) simulations offer a powerful computational approach to study condensed matter systems.

Purpose of the Study:

• To investigate the structural properties of various ice crystal phases using classical molecular dynamics (MD) simulations.
• To develop and validate a computational protocol for identifying and analyzing ice polymorphs.
• To assess the relative stability of different ice phases.

Main Methods:

• Standard MD simulations were performed using a recently developed classical interaction potential for water.
• A computational protocol involving a clustering approach was applied to analyze simulated powder diagrams of ice polymorphs.
• The interaction potential was derived from neutron scattering data and accounts for phenomena like proton hopping.

Main Results:

• The study successfully described most known ice structures within the water phase diagram.
• The interaction potential accurately predicted the relative stability of 16 different ice phases.
• The proposed computational protocol demonstrated suitability for automated crystal structure identification.

Conclusions:

• The employed classical interaction potential effectively models various ice structures and their relative stabilities.
• The developed computational protocol offers an efficient method for automated identification of ice crystal structures.
• This work advances the computational study of water ice polymorphs.