Machine Learning Guided the Discovery of Superionic Delafossite AgFeO₂.

Area of Science:

• Materials Science
• Solid-State Chemistry
• Computational Materials Science

Background:

• Solid-state electrolytes (SSEs) are key for safe, high-energy-density batteries.
• Superionic conductors (SICs) enable ion diffusion within solid lattices.
• Delafossite AgFeO2 is a potential material for SSEs, but its ionic transport needs detailed understanding.

Purpose of the Study:

• To elucidate the diffusion mechanism of Ag+ ions in delafossite AgFeO2.
• To investigate the influence of structural defects, specifically Ag vacancies, on ionic conductivity.
• To provide insights for designing improved solid-state electrolytes.

Main Methods:

• Molecular dynamics (MD) simulations utilizing machine learning force fields (MLFF).
• Analysis of atomic trajectories, mean square displacements (MSD), and radial distribution functions (RDF).
• Madelung energy analysis to understand interatomic interactions.

Main Results:

• Ag+ ions predominantly migrate via a concerted mechanism within the AgFeO2 lattice above 800 K.
• Madelung energy analysis indicates stronger Fe-O interactions than Ag-O interactions.
• Introduction of Ag vacancies significantly reduces the superionic transition temperature to 600 K.

Conclusions:

• The study clarifies the ion diffusion pathway in delafossite AgFeO2.
• Structural defects, like Ag vacancies, critically impact ionic behavior and transition temperatures.
• Findings pave the way for targeted material design in solid-state electrolytes.