Cross-correlation for detecting and understanding patterns inab initiomolecular dynamics simulations of liquid metals

Area of Science:

• Computational materials science
• Physical chemistry
• Surface science

Background:

• Analyzing atomic motion in molecular dynamics simulations is challenging.
• Traditional methods for liquid metal systems rely on visual inspection, missing complex time-lagged correlations.
• Developing robust methods for pattern detection in atomic trajectories is crucial.

Purpose of the Study:

• To apply statistical cross-correlation techniques to molecular dynamics data.
• To identify and characterize time-lagged relationships between atomic properties (position, velocity, energy).
• To link observed atomic motion patterns to adsorbate interaction energies.

Main Methods:

• Utilized statistical cross-correlation analysis on time-series data.
• Analyzed trajectories of liquid Gallium-Indium (Ga-In) and Gallium-Tin (Ga-Sn) systems.
• Correlated atomic positions, velocities, and energies over time.

Main Results:

• Identified previously unrecognized temporal relationships in atomic motion for Ga-In and Ga-Sn.
• Successfully linked these atomic motion patterns to interaction energy trends with adsorbates.
• Demonstrated the capability of the method to reveal complex dynamics.

Conclusions:

• Statistical cross-correlation is a powerful framework for analyzing coupled atomic dynamics.
• The methodology provides new insights into liquid metal behavior in catalytic and surface science applications.
• This approach overcomes limitations of traditional 'by-eye' analysis for time-lagged motions.