Crystalline 'Genes' in Metallic Liquids

TL;DR

This paper introduces a genetic algorithm-based method to identify and compare structural motifs in metallic liquids and crystalline phases, providing new insights into order-disorder transitions in condensed matter.

Contribution

It presents a novel approach combining genetic algorithms and cluster-alignment to analyze structural motifs in metallic liquids and crystalline states, surpassing traditional methods.

Findings

Identified common energetic packing motifs near Al-10 at.% Sm composition.

Compared motifs in liquids and crystals revealing average topology.

Provided new insights into order-disorder transitions in metals.

Abstract

The underlying structural order that transcends the liquid, glass and crystalline states is identified using an efficient genetic algorithm (GA). GA identifies the most common energetically favorable packing motif in crystalline structures close to the alloy's Al-10 at.% Sm composition. These motifs are in turn compared to the observed packing motifs in the actual liquid structures using a cluster-alignment method which reveals the average topology. Conventional descriptions of the short-range order, such as Voronoi tessellation, are too rigid in their analysis of the configurational poly-types when describing the chemical and topological ordering during transition from undercooled metallic liquids to crystalline phases or glass. Our approach here brings new insight into describing mesoscopic order-disorder transitions in condensed matter physics.