Computational Tools

Phonon Thermal and Mechanical Properties via MLPs

Package Overview

This comprehensive toolkit seamlessly integrates advanced Machine Learning Potentials (MLPs) with the Quasi-Harmonic Approximation (QHA). It provides an automated workflow for calculating phonon spectra, elastic constants, and thermodynamic properties, enabling efficient exploration of finite-temperature effects and negative thermal expansion behaviors.

GitHub Repository

Key Features

Structure Relaxation: Efficient crystal structure optimization.
Mechanical Properties: Elastic constants (K, G, E, Poisson's ratio).
Lattice Dynamics: Phonon dispersion calculation and stability analysis.
Thermodynamics: Full thermodynamic properties evaluation.
Thermal Expansion: Accurate NTE prediction based on QHA theory.

Supported ML Potentials

CHGNet DPA3 GPTFF GRACE M3GNet MACE MatterSim NEP NequIP ORB SevenNet UPET

Please cite our work if you use this package in your research.