Skip to content
/ heaict Public

High entropy alloy (intermetallic) ctalytic tools (HEAICT). Design high-entropy alloy eNRR catalysts through multi-objective optimization.

License

Notifications You must be signed in to change notification settings

jchddd/heaict

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

64 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

HEAICT

This repository contains a collection of tools and scripts for catalyst design in high-entropy alloys (HEAs) and high-entropy intermetallics (HEIs).


Installation

Copy the Folder heaict into the Lib folder under your Python directory, or run the code at the same level as Folder heaict.

Requirment

Software Version Requirement
torch 2.3.0 required
torch_geometric 2.5.3 required
pymatgen 2024.5.1 required
ase 3.22.1 required
pymoo 0.6.1.3 required
scikit-learn 1.4.2 required
periodictable 2.0.2 required
numpy 1.26.4 required
scipy 1.13.0 required
matplotlob 3.9.0 required
pandas 2.2.2 optional
tqdm 4.66.4 optional
gpflow 2.10.0 optional
tensorflow 2.19.0 optional
tensorflow-probability 0.25.0 optional
pygco 0.0.16 optional
optuna 4.0.0 optional

Note:
pygco is related to the graph cut algorithm. It can be ignored by disabling the sparse_approx parameter in the ParetoDiscovery class.
gpflow and related TensorFlow libraries are used for Gaussian Process Regression (GPR) based on gpflow. If they are not installed, you may use GPR based on the scikit-learn by importing the relevant models from heaict.ml.GPR_scikit.
tqdm and pandas are primarily used for training data preprocessing (heaict.data related methods). If you perform this step using other methods or scripts, you may skip the installation of them.
optuna is mainly for hyperparameter tuning in ML training, but other methods can be used as well.

Overview

  • heaict/cats: An extended surface model that incorporates versatile sites to account for site-blocking effects. A microkinetic framework for evaluating NRR performance, which explicitly accounts for site coverage and the influence of applied potential U. And a problem class that integrates the above components, enabling the prediction of catalytic performance based on alloy composition.
  • heaict/data: Identifying adsorption sites, filtering anomalous structures, and building the ML training dataset.
  • heaict/hea: Determining the thermodynamic stability of high entropy alloys, and analyzing the site occupation propensity in high entropy intermetallics.
  • heaict/ml: Implementations of ASGCNN for adsorption energy prediction and GPR surrogate models (using either scikit-learn or gpflow) for simulating the composition–performance potential energy surface.
  • heaict/mobo: Scripts and functions related to multi-objective optimization.
  • Data: Data and results related to the paper.

Tutorials

Citation

If you are interested in our work, you can read our literature, and cite us using

Submission in progress

About

High entropy alloy (intermetallic) ctalytic tools (HEAICT). Design high-entropy alloy eNRR catalysts through multi-objective optimization.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published