Attractive versus truncated repulsive supercooled liquids: The dynamics is encoded in the pair correlation function.

Area of Science:

• Condensed matter physics
• Computational chemistry
• Statistical mechanics

Background:

• Glassy dynamics in liquids are complex and influenced by interparticle interactions.
• Understanding how potential energy landscapes affect liquid dynamics is crucial.

Purpose of the Study:

• To investigate the relationship between interaction potentials and glassy dynamics.
• To identify a method for predicting dynamical differences in liquids with similar structures.

Main Methods:

• Comparing two liquid systems with identical repulsive but differing attractive interactions.
• Analyzing pair correlation functions and their integrals.
• Employing a machine learning algorithm to determine weights for the integral.

Main Results:

• Despite nearly identical pair correlation functions, the two liquid systems exhibited significantly different dynamics.
• A weighted integral of the pair correlation function successfully captured these dynamical differences.
• Machine learning provided effective weights for the integral.

Conclusions:

• Subtle differences in attractive interactions significantly impact glassy dynamics.
• A weighted integral, informed by machine learning, serves as a powerful tool for predicting liquid dynamics.
• This approach offers a novel way to analyze complex liquid behavior.