Kinetic pathways of coesite densification from metadynamics

Area of Science:

• Mineral Physics
• Computational Materials Science
• Geochemistry

Background:

• Coesite, a high-pressure polymorph of SiO2, undergoes significant structural changes under extreme pressures.
• Previous experimental studies identified metastable coesite-II and coesite-III, followed by reconstructive transformations into octahedral phases, coesite-IV, and coesite-V above 30 GPa.
• Simulating these complex reconstructive transformations computationally presents a significant challenge due to the large structural rearrangements and coordination changes.

Purpose of the Study:

• To computationally investigate the high-pressure compression of coesite and elucidate the atomistic mechanisms of its structural transformations.
• To validate a computational approach combining metadynamics, machine learning potentials, and collective variables for studying complex phase transitions.
• To identify and describe the distinct transformation pathways leading to observed experimental outcomes.

Main Methods:

• Utilized metadynamics simulations with Si-O coordination number and volume as collective variables.
• Employed an advanced machine-learning based atomistic potential (ACE potential) for accurate energy and force calculations.
• Analyzed the atomistic mechanisms of transformation pathways, including amorphization, formation of coesite-IV, and octahedral phases.

Main Results:

• The computational approach successfully reproduced all experimentally observed transformation pathways: amorphization, coesite-IV, and octahedral phases.
• Detailed atomistic mechanisms for the transformation to coesite-IV and the two-step process for octahedral phases were described.
• The pathway to coesite-IV was predicted to be favored at room temperature, while octahedral phase formation is more likely at 600 K.

Conclusions:

• Machine learning potentials combined with metadynamics provide a powerful tool for studying complex reconstructive phase transitions in minerals.
• The study clarifies the atomistic details of coesite compression, offering insights into Earth's deep interior and materials under extreme conditions.
• Temperature-dependent pathway selection was predicted, highlighting the influence of thermal energy on high-pressure mineral transformations.