A Workflow for Identifying Viable Crystal Structures with Partially Occupied Sites Applied to the Solid Electrolyte...

Area of Science:

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

Background:

• The ground-state configuration of cubic lithium lanthanum zirconium oxide (c-LLZO), a promising solid electrolyte, remains elusive.
• Accurate computational studies of c-LLZO require reliable low-energy reference structures.
• Previous theoretical and experimental efforts have not definitively determined the stable structure of c-LLZO.

Purpose of the Study:

• To develop a robust methodology for identifying energetically favorable configurations of c-LLZO.
• To establish a reliable set of feasible and stable structures for c-LLZO research.
• To provide a database of generated structures for improved accuracy and reproducibility in future studies.

Main Methods:

• A novel computational approach combining geometric filtering and machine learning was employed.
• Structures were filtered by eliminating those with overlapping Li atoms based on nearest neighbor analysis.
• Symmetry images were removed, and machine learning was used for energetic ordering of remaining configurations.

Main Results:

• The methodology successfully identified energetically favorable configurations of c-LLZO.
• A significant portion of previously reported structures were found to be geometrically infeasible or unstable.
• A comprehensive database of generated c-LLZO structures was created.

Conclusions:

• The developed methodology provides a systematic way to determine stable configurations for c-LLZO.
• This approach enhances the accuracy and reproducibility of computational research on solid electrolytes.
• The methodology can be extended to investigate other ion-conducting materials.