Two worlds collide: image analysis methods for quantifying structural variation in cluster molecular dynamics.

Area of Science:

• Computational Chemistry
• Materials Science
• Statistical Physics

Background:

• Molecular dynamics (MD) simulations are crucial for understanding atomic and molecular behavior.
• Analyzing structural dynamics and variations in clusters remains a challenge.
• Existing methods may require a priori structural information, limiting their applicability.

Purpose of the Study:

• To introduce novel statistical techniques for analyzing structural variations in cluster MD simulations.
• To provide quantitative measures of structural stability and variation.
• To enable isomer identification and comparison across different cluster sizes.

Main Methods:

• Application of principal component analysis (PCA) to characterize cluster geometry over time.
• Utilizing Pearson Correlation Coefficient (PCC) to analyze bond structure variations.
• Employing atomic position data without requiring prior structural knowledge.

Main Results:

• PCA provides a quantitative assessment of structural stability and geometric shape changes.
• PCC effectively captures and compares bond structure variations within and between clusters.
• The methods are applicable to both classical and ab initio MD simulations.

Conclusions:

• PCA and PCC offer powerful, data-driven tools for cluster MD analysis.
• These techniques facilitate quantitative structural characterization and isomer identification.
• The methods are versatile, applicable to diverse cluster compositions and electronic configurations.