Arrested phase separation in double-exchange models: machine-learning enabled large-scale simulation

TL;DR

This study uses deep learning-enhanced large-scale simulations to reveal how electronic phase separation in double-exchange models is arrested due to hole trapping and magnetic interactions, impacting colossal magnetoresistance materials.

Contribution

It introduces a machine-learning approach for large-scale simulation of phase separation, uncovering the mechanism of arrested phase separation in double-exchange models.

Findings

Holes segregate from the insulating background during equilibration.

Ferromagnetic clusters stabilize hole aggregation via Hund's coupling.

Hole mobility reduction leads to arrested phase separation.

Abstract

We present large-scale dynamical simulations of electronic phase separation in the single-band double-exchange model based on deep-learning neural-network potentials trained from small-size exact diagonalization solutions. We uncover an intriguing correlation-induced freezing behavior as doped holes are segregated from half-filled insulating background during equilibration. While the aggregation of holes is stabilized by the formation of ferromagnetic clusters through Hund's coupling between charge carriers and local magnetic moments, this stabilization also creates confining potentials for holes when antiferromagnetic spin-spin correlation is well developed in the background. The dramatically reduced mobility of the self-trapped holes prematurely disrupts further growth of the ferromagnetic clusters, leading to an arrested phase separation. Implications of our findings for phase separation dynamics in materials that exhibit colossal magnetoresistance effect are discussed.