Performing highly efficient Minima Hopping structure predictions using the Atomic Simulation Environment (ASE)

Read full text Summarise with VIRAT
Authors: Marco Krummenacher, Moritz Gubler, Jonas A. Finkler, Hannes Huber, Martin Sommer-Jörgensen, Stefan Goedecker
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The ase_mh package provides a robust, user-friendly, and efficient Python implementation of the Minima Hopping (MH) algorithm for the exploration of potential energy surfaces and the identification of stable low-energy structures in materials science. Over the past two decades, the Minima Hopping algorithm has proven to be a powerful method for global structure optimization. This implementation significantly enhances its performance by incorporating an MPI-based parallelization scheme that enables multi-level parallel execution. Multiple MH processes can run concurrently while communicating through a shared database, allowing them to exchange information about previously explored regions of the potential energy surface and thereby avoid redundant searches. The code integrates several advanced features inspired by existing MH implementations, including: Variable cell shape molecular dynamics Combined optimization of atomic positions and cell geometry for bulk systems Enhanced temperature feedback mechanisms Fragmentation detection and fixing for cluster simulations The package is built on top of the Atomic Simulation Environment (ASE), providing high flexibility and compatibility with a wide range of calculators for energy and force evaluations. Overall, ase_mh offers a scalable and versatile framework for efficient global optimization in atomistic simulations, suitable for both cluster and bulk systems.

Fields of Study & Subjects

Machine Learning in Materials Science Advanced Chemical Physics Studies Advanced Physical and Chemical Molecular Interactions

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
27 March 2026
Language
EN