Artificial kagome spin ice magnetic phase recognition from the initial magnetization curve

TL;DR

This paper presents a new method to identify magnetic phases in artificial kagome spin ice using initial magnetization curves and pattern recognition, offering a simpler alternative to traditional imaging techniques.

Contribution

The study introduces a novel approach to determine spin ice phases from magnetization curves using micromagnetic simulations and pattern recognition, enhancing phase identification efficiency.

Findings

Phase-specific features in magnetization curves enable phase prediction.

Pattern recognition accurately classifies different spin ice phases.

Method offers a more accessible alternative to magnetic imaging.

Abstract

Artificial spin ices (ASIs) are designable arrays of interacting nanomagnets that span a wide range of magnetic phases associated with a number of spin lattice models. Here, we demonstrate that the phase of an artificial kagome spin ice can be determined from its initial magnetization curve. As a proof of concept, micromagnetic simulations of these curves were performed starting from representative microstates of different phases of the system. We show that the curves are characterized by phase-specific features in such a way that a pattern recognition algorithm predicts the phase of the initial microstate with good reliability. This achievement represents a new strategy to identify phases in ASIs, easier and more accessible than magnetic imaging techniques normally used for this task.