a2c_ase

An ASE-friendly implementation of the amorphous-to-crystalline (a2c) workflow.

Installation

From PyPI

With uv (recommended):

uv pip install a2c-ase

pip install a2c-ase

From Source

With uv:

git clone https://github.com/abhijeetgangan/a2c_ase.git
cd a2c_ase
uv pip install .

git clone https://github.com/abhijeetgangan/a2c_ase.git
cd a2c_ase
pip install .

Usage

See example/Si64.py for basic usage.

To use a specific calculator you need to install the corresponding package.

In the example above, MACE is used as the calculator, so you need to install the corresponding package.

pip install mace-torch

Workflow Overview

1. Initial Structure: Generate a random atomic configuration with specified composition and volume.
2. Melt-Quench: Run MD simulation to create an amorphous structure.
3. Subcell Extraction: Identify potential crystalline motifs within the amorphous structure.
4. Structure Optimization: Relax subcells to find stable crystalline phases.
5. Analysis: Characterize discovered structures using symmetry analysis.

Development

Install dev dependencies:

# with pip
pip install -e ".[dev,test]"

Set up pre-commit hooks:

pre-commit install

Run checks:

ruff check         # lint
ruff format        # format
ty check           # type check
pytest             # test

Citation

If you use this software in your research, please cite it: DOI:https://doi.org/10.5281/zenodo.17355689

References

1. Aykol, M., Merchant, A., Batzner, S. et al. Predicting emergence of crystals from amorphous precursors with deep learning potentials. Nat Comput Sci 5, 105–111 (2025). DOI: 10.1038/s43588-024-00752-y
2. Reference implementation: a2c-workflow