Chalcogenide phase-change materials: unveiling new horizons with big data and machine learning

Abstract

Chalcogenide phase-change materials (PCMs) have been the subject of extensive research and development due to their unique electrical and optical properties. PCMs have been successfully applied in various optical discs and have made significant strides in data storage, such as in phase-change random access memory (PCRAM) devices. Moreover, PCMs have found promising applications in photonics for controlling light propagation and interaction, as well as in neuromorphic computing systems that mimic the functionality of the human brain. This review comprehensively summarizes the research on PCMs assisted by big data analytics and machine learning (ML) methods. Computational data exploration involves screening optimal dopants and predicting material properties through high-throughput calculations and ML models. Large-scale simulations enabled by machine learning potential (MLP) have deepened the understanding of phase transition dynamics and thermodynamic properties. In device-scale simulation and design, ML has been crucial in optimizing memory devices and exploring the potential of PCMs in neuromorphic computing. Finally, the future research directions and current challenges of PCMs are summarized.