Automated first-principles mapping for phase-change materials.

Area of Science:

• Computational materials science
• Solid-state chemistry

Background:

• Materials plotting on bi-coordinate maps is a long-standing technique in computational solid-state science.
• Previous maps for phase-change materials (PCMs) had limitations in accurately charting all industrially relevant materials.

Purpose of the Study:

• To generate an improved 'treasure map' for phase-change materials (PCMs).
• To identify new potential PCM candidates using novel computational methods.
• To enable automated generation of materials maps based on first-principles calculations.

Main Methods:

• Development of new ab initio techniques for materials mapping.
• Inclusion of structural information into ab initio descriptors for sp³ orbital mixing.
• Introduction of a quantum-mechanical ionization measure incorporating structural data.

Main Results:

• An improved PCM map accurately charts all industrially used PCMs.
• Seven novel PCM candidates (SiSb₄Te₇, Si₂Sb₂Te₅, SiAs₂Te₄, PbAs₂Te₄, SiSb₂Te₄, Sn₂As₂Te₅, PbAs₄Te₇) are identified.
• The developed tools allow for automated materials map generation from first-principles calculations.

Conclusions:

• The new computational tools provide a more accurate and comprehensive approach to PCM discovery.
• This work facilitates the identification of new materials with potential applications.
• The methodology relies solely on first-principles calculations, enhancing predictive power.