Low-cost Vibrational Free Energies in Solid Solutions with Machine Learning Force Fields

Kasper Tolborg,Aron Walsh

DOI: https://doi.org/10.26434/chemrxiv-2023-wqcsq

2023-11-10

Abstract:The rational design of alloys and solid solutions relies on accurate computational predictions of phase diagrams. The cluster expansion method with effective cluster interactions fitted to energies from first principles calculations has proven to be a valuable tool for studying disordered crystals. However, the effects of vibrational entropy are commonly neglected due to their additional computational cost. Here, we devise a method for including vibrational free energy in cluster expansions at a very low computational cost by fitting a machine learning force field (MLFF) to the relaxation trajectories already available from the cluster expansion construction. We demonstrate our method for two (pseudo)binary systems, Na1-xKxCl and Ag1-xPd_x, for which accurate phonon dispersions and vibrational free energies are derived from the MLFF. For both systems, inclusion of vibrational effects results in significantly better agreement with miscibility gaps in experimental phase diagrams. This methodology can allow routine inclusion of vibrational effects in calculated compositional phase diagrams, and thus more accurate predictions of properties and stability for materials mixtures.

Chemistry

What problem does this paper attempt to address?

This paper focuses on a low-cost method for calculating vibrational free energy in solid solutions. Traditional cluster expansion methods neglect the influence of vibrational entropy when predicting the phase diagram of disordered crystals, as it increases computational cost. The researchers propose a new approach by incorporating vibrational free energy with minimal additional cost using a Machine Learning Force Field (MLFF) adapted from the relaxation trajectories obtained during the cluster expansion construction process. This method is applied to two exemplary systems (Na1-xKxCl and Ag1-xPdx), and the results demonstrate that considering the vibrational effects can significantly improve the consistency with experimental phase diagrams, particularly for the Na1-xKxCl system, where it better reproduces the observed immiscible gap in experiments. Furthermore, they discuss optimizing training strategies to reduce the number of required configurations and the potential applications of this machine learning method in solid solutions. This method has the potential to make vibrational effects a routine consideration in calculating phase diagrams, thereby improving the prediction accuracy of material mixture properties and stability.

Low-cost Vibrational Free Energies in Solid Solutions with Machine Learning Force Fields

Vibrational properties of metastable polymorph structures by machine learning

Vibrational Entropy and Free Energy of Solid Lithium using Covariance of Atomic Displacements Enabled by Machine Learning

Predicting Elastic Properties of Refractory High-Entropy Alloys Via Machine-Learning Approach

Exploring Phononic Properties of Two-Dimensional Materials using Machine Learning Interatomic Potentials

Accelerating ab initio melting property calculations with machine learning: Application to the high entropy alloy TaVCrW

Accelerating the discovery of low-energy structure configurations: a computational approach that integrates first-principles calculations, Monte Carlo sampling, and Machine Learning

The search for high entropy alloys: a high-throughput $\textit{ab-initio}$ approach

Machine learning approach for vibronically renormalized electronic band structures

Machine-Learned Fragment-Based Energies for Crystal Structure Prediction

The Effect of Lattice Vibrations on Substitutional Alloy Thermodynamics

Deciphering Chemical Ordering in High Entropy Materials: A Machine Learning-Accelerated High-throughput Cluster Expansion Approach

Alloy Informatics through Ab Initio Charge Density Profiles: Case Study of Hydrogen Effects in Face-Centered Cubic Crystals

Nudged-Elastic Band Calculations of Polymorph Transitions and Solid-State Reactions in Molecular Crystals

Machine learning, phase stability, and disorder with the Automatic Flow Framework for Materials Discovery

Towards machine learning for microscopic mechanisms: a formula search for crystal structure stability based on atomic properties

The effect of machine learning predicted anharmonic frequencies on thermodynamic properties of fluid hydrogen fluoride

Machine-Learning Surrogate Model for Accelerating the Search of Stable Ternary Alloys

Alkali-metal vibrations in bcc, fcc, hcp, and 9R structures: Implications for the energetics of Li and Na martensitic phases

Liquid-Vapor Phase Equilibrium in Molten Aluminum Chloride (AlCl3) Enabled by Machine Learning Interatomic Potentials

Phase Behavior and Atomic Dynamics in Rb x Na 1- x : Insights from Machine Learning Interatomic Potentials based on Ab Initio Molecular Dynamics