Relationship between structure and dynamics of an icosahedral quasicrystal using unsupervised machine learning

Area of Science:

• Condensed Matter Physics
• Materials Science
• Computational Chemistry

Background:

• Quasicrystals exhibit unique long-range order without translational periodicity.
• Understanding the self-assembly mechanisms and dynamic behavior of quasicrystals is crucial for their application.
• Icosahedral quasicrystals (IQCs) represent a significant class with complex structural characteristics.

Purpose of the Study:

• To comprehensively investigate the structure, formation, and dynamics of a one-component system self-assembling into an IQC.
• To identify and characterize unique structural motifs and their evolution during IQC formation.
• To elucidate the relationship between structural order and dynamical processes in IQCs.

Main Methods:

• Molecular dynamics simulations to model system behavior.
• Unsupervised machine learning techniques to identify structural motifs and local environments.
• Development of a machine-learning-based order parameter to quantify structural order and its correlation with dynamics.

Main Results:

• Identified icosahedral and dodecahedral arrangements as key structural motifs.
• Revealed that IQC formation is driven by the emergence of distinct precursor clusters.
• Observed transitions from restricted motion to activated diffusion with increasing temperature.
• Correlated high structural order with suppressed self-diffusion and minimal dynamical heterogeneity (phason-like motion).
• Linked lower structural order to enhanced collective motion and increased dynamical heterogeneity.

Conclusions:

• Established a quantitative framework connecting structural organization and dynamical processes in IQCs.
• Provided new insights into the mechanisms governing IQC stability and dynamics.
• Demonstrated the utility of machine learning in analyzing complex self-assembling systems.