Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1103/physrevb.47.5571 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Automated 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 PDFSuitability of Available Interatomic Potentials for Sn to Model Its 2D Allotropes.
J Comput Chem
January 2025
Institute of Fundamental Technological Research Polish Academy of Sciences, Warsaw, Poland.
The suitability of a range of interatomic potentials for elemental tin was evaluated in order to identify an appropriate potential for modeling the stanene (2D tin) allotropes. Structural and mechanical features of the flat (F), low-buckled (LB), high-buckled (HB), full dumbbell (FD), trigonal dumbbell (TD), honeycomb dumbbell (HD), and large honeycomb dumbbell (LHD) monolayer tin (stanene) phases, were gained by means of the density functional theory (DFT) and molecular statics (MS) calculations with ten different Tersoff, modified embedded atom method (MEAM), and machine-learning-based (ML-IAP) interatomic potentials. A systematic quantitative comparison and discussion of the results is reported.
View Article and Find Full Text PDFAnalyzing Atomic Interactions in Molecules as Learned by Neural Networks.
J Chem Theory Comput
January 2025
BIFOLD─Berlin Institute for the Foundations of Learning and Data, 10587 Berlin, Germany.
While machine learning (ML) models have been able to achieve unprecedented accuracies across various prediction tasks in quantum chemistry, it is now apparent that accuracy on a test set alone is not a guarantee for robust chemical modeling such as stable molecular dynamics (MD). To go beyond accuracy, we use explainable artificial intelligence (XAI) techniques to develop a general analysis framework for atomic interactions and apply it to the SchNet and PaiNN neural network models. We compare these interactions with a set of fundamental chemical principles to understand how well the models have learned the underlying physicochemical concepts from the data.
View Article and Find Full Text PDFHessian QM9: A quantum chemistry database of molecular Hessians in implicit solvents.
Sci Data
January 2025
IBM Research, Hursley, SO21 2JN, UK.
A significant challenge in computational chemistry is developing approximations that accelerate ab initio methods while preserving accuracy. Machine learning interatomic potentials (MLIPs) have emerged as a promising solution for constructing atomistic potentials that can be transferred across different molecular and crystalline systems. Most MLIPs are trained only on energies and forces in vacuum, while an improved description of the potential energy surface could be achieved by including the curvature of the potential energy surface.
View Article and Find Full Text PDFExperimental and Theoretical Force Constants as Meaningful Indicator for Interatomic Bonding Characteristics and the Specific Case of Elemental Antimony.
Adv Mater
January 2025
Felix-Bloch-Institut für Festkörperphysik, Universität Leipzig, Linnéstraße 5, 04103, Leipzig, Germany.
Stable Sb exhibits a rhombohedral structure, often referred to as distorted primitive cubic, with each Sb atom having three short and three longer first neighbor bonds. However, this crystal structure can also be interpreted as being layered, putting emphasis on only three short first neighbor bonds. Therefore, temperature-dependent extended X-ray absorption fine structure (EXAFS) spectroscopy is carried out at the Sb K-edge in order to obtain more detailed information on local structural and vibrational properties.
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!