a2c_ase

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

Installation

From Source

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.

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