Shry: Application of Canonical Augmentation to the Atomic Substitution Problem.

Area of Science:

• Materials Science
• Computational Chemistry
• Crystallography

Background:

• Studying solid solutions and disordered systems often requires creating supercells with substituted elements within a periodic ab initio framework.
• Identifying unique atomic substitution patterns is crucial, as the total number of possible substitutions can be astronomically large, while the number of symmetry-inequivalent patterns is much smaller.

Purpose of the Study:

• To develop and present Shry, a Python software package designed to efficiently select only symmetry-inequivalent atomic substitution patterns from a vast number of candidates.
• To provide a tool that simplifies the generation of models for disordered materials and solid solutions.

Main Methods:

• The Shry package utilizes the canonical augmentation algorithm to filter out symmetry-equivalent structures.
• It is implemented in Python 3 and employs the Crystallographic Information File (CIF) format for reading and writing crystal structure data.
• The software can be used as a standalone program or integrated as a module into other Python applications.

Main Results:

• Shry successfully identifies and selects only symmetry-inequivalent atomic substitution patterns, significantly reducing the dataset size for computational studies.
• Verification through comparison with existing codes confirmed the accuracy of Shry in determining the number and equivalency of unique structures.
• The crystal structure data used for verification can serve as a benchmark for future software development and algorithm testing.

Conclusions:

• Shry offers an efficient and reliable method for generating models of substituted crystal structures, crucial for high-throughput computational materials science.
• The package streamlines the process of studying disordered systems and solid solutions by managing the complexity of atomic substitutions.
• The availability of Shry and its verification data will aid researchers in computational materials discovery and algorithm development.