We introduce a class of interatomic potential models that can be automatically generated from data consisting of the energies and forces experienced by atoms, as derived from quantum mechanical calculations. The models do not have a fixed functional form and hence are capable of modeling complex potential energy landscapes. They are systematically improvable with more data. We apply the method to bulk crystals, and test it by calculating properties at high temperatures. Using the interatomic potential to generate the long molecular dynamics trajectories required for such calculations saves orders of magnitude in computational cost.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1103/PhysRevLett.104.136403 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Long Time Scale Molecular Dynamics Simulation of Magnesium Hydride Dehydrogenation Enabled by Machine Learning Interatomic Potentials.
ACS Appl Energy Mater
January 2025
Advanced Materials Research Group, Faculty of Engineering, University of Nottingham, Nottingham NG7 2RD, United Kingdom.
Magnesium hydride (MgH) is a promising material for solid-state hydrogen storage due to its high gravimetric hydrogen capacity as well as the abundance and low cost of magnesium. The material's limiting factor is the high dehydrogenation temperature (over 300 °C) and sluggish (de)hydrogenation kinetics when no catalyst is present, making it impractical for onboard applications. Catalysts and physical restructuring (e.
View Article and Find Full Text PDFMetal-organic frameworks such as MOF-303 and MOF-LA2-1 have demonstrated exceptional performance for water harvesting applications. To enable a reticular design of such materials, an accurate prediction of the adsorption properties with chemical accuracy and fully accounting for the flexibility is crucial. The computational prediction of water adsorption properties in MOFs has become standard practice, but current methods lack the predictive power needed to design new materials.
View Article and Find Full Text PDFBifunctional NiCo-CuO Nanostructures: A Promising Catalyst for Energy Conversion and Storage.
Small Methods
January 2025
Department of Physics, Tamkang University, Tamsui, 25137, Taiwan.
This investigation explores the potential of co-incorporating nickel (Ni) and cobalt (Co) into copper oxide (CuO) nanostructures for bifunctional electrochemical charge storage and oxygen evolution reactions (OER). A facile wet chemical synthesis method is employed to co-incorporate Ni and Co into CuO, yielding diverse nanostructured morphologies, including rods, spheres, and flake. The X-ray diffraction (XRD) and Raman analyses confirmed the formation of NiCo-CuO nanostructure, with minor phases of nickel oxide (NiO) and cobalt tetraoxide (CoO).
View Article and Find Full Text PDFEnhanced sampling of robust molecular datasets with uncertainty-based collective variables.
J Chem Phys
January 2025
Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA.
Generating a dataset that is representative of the accessible configuration space of a molecular system is crucial for the robustness of machine-learned interatomic potentials. However, the complexity of molecular systems, characterized by intricate potential energy surfaces, with numerous local minima and energy barriers, presents a significant challenge. Traditional methods of data generation, such as random sampling or exhaustive exploration, are either intractable or may not capture rare, but highly informative configurations.
View Article and Find Full Text PDFAutomated and Efficient Sampling of Chemical Reaction Space.
Adv Sci (Weinh)
January 2025
Department of Chemistry, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul, 03722, Republic of Korea.
Machine learning interatomic potentials (MLIPs) promise quantum-level accuracy at classical force field speeds, but their performance hinges on the quality and diversity of training data. An efficient and fully automated approach to sample chemical reaction space without relying on human intuition, addressing a critical gap in MLIP development is presented. The method combines the speed of tight-binding calculations with selective high-level refinement, generating diverse datasets that capture both equilibrium and reactive regions of potential energy surfaces.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!