Workflows for "From electrons to phase diagrams with classical and machine learning potentials: automated workflows for materials science with pyiron"

Sarath Menon, Yury Lysogorskiy, Alexander L. M. Knoll, Niklas Leimeroth, Marvin Poul, Minaam Qamar, Jan Janssen, Matous Mrovec, Jochen Rohrer, Karsten Albe, Jörg Behler, Ralf Drautz, Jörg Neugebauer

Preprint at http://arxiv.org/abs/2403.05724 (2024)
Dataset at https://doi.org/10.17617/3.VKQ3ZM (2024)

This repository contains the workflows for the above publication.

Setting up the environment

The computational environment with all the necessary software can be installed using conda. First step is to create an environment:

conda env create -f binder/environment.yml

After the environment is created, it can be activated by

conda activate potentials

Once the environment is activated, the Tensorpot can be installed:

git clone --depth 1 --single-branch https://github.com/ICAMS/TensorPotential
cd TensorPotential
python setup.py install
cd ..

after this, jupyter lab can be started with

jupyter lab

Setting up pyiron

Create a configuration file called .pyiron in your home directory and add the following contents

[DEFAULT]
FILE = ~/pyiron.db
PROJECT_PATHS = ~/pyiron/projects
RESOURCE_PATHS = ~/pyiron/resources:~/<path-to-repo>/potential_publication/resources

Note that <path-to-repo> needs to be replaced with the actual path.

Interatomic potentials used in this work

The files for the interatomic potentials used in this work is available in the resources/lammps/potentials folder.

Contents

• 01_data_generation: workflows for data generation for parametrising interatomic potentials, and reproducing figures from the publication.
• 02_fitting: Workflows for fitting classical and ML potentials.
• 03_validation: Workflows for validating potentials including EV curves, elastic constants, and phonon density of states.
• 04_phase_diagram: Workflows for calcuting phase diagrams for EAM, HDNNP, and ACE potentials.