Molecular Dynamics Simulations of SrTiO₃ with Oxygen Vacancies using Neural Network Potentials

2025/06/11 by Kazutaka Nishiguchi, Nishiguchi, Kazutaka, Ryota Yamamoto +11
Materials Science · Biochemistry, Genetics and Molecular Biology · #Machine Learning in Materials Science #Electronic and Structural Properties of Oxides #Advanced Electron Microscopy Techniques and Applications

paper · pdf · doi:10.48550/arxiv.2506.09372

Abstract

A precise analysis of point defects in solids requires accurate molecular dynamics (MD) simulations of large-scale systems. However, ab initio MD simulations based on density functional theory (DFT) incur high computational cost, while classical MD simulations lack accuracy. We perform MD simulations using a neural network potential (NNP) model (NNP-MD) to predict the physical quantities of both pristine SrTiO₃ and SrTiO₃ in the presence of oxygen vacancies (V_O). To verify the accuracy of the NNP models trained on different data sets, their NNP-MD predictions are compared with the results obtained from DFT calculations. The predictions of the total energy show good agreement with the DFT results for all these NNP models, and the NNP models can also predict the formation energy once SrTiO₃:V_O data are included in the training data sets. Even for larger supercell sizes that are difficult to calculate using first-principles calculations, the formation energies evaluated from the NNP-MD simulations well reproduce the extrapolated DFT values. This study offer important knowledge for constructing accurate NNP models to describe point-defect systems including SrTiO₃:V_O.

Citations

• Generator of Neural Network Potential for Molecular Dynamics: Constructing Robust and Accurate Potentials with Active Learning for Nanosecond-scale Simulations
• Two-step growth mechanism of the solid electrolyte interphase in argyrodyte/Li-metal contacts
• MLIP-3: Active learning on atomic environments with Moment Tensor Potentials
• Hyperactive Learning (HAL) for Data-Driven Interatomic Potentials
• The MLIP package: Moment Tensor Potentials with MPI and Active Learning
• DP-GEN: A concurrent learning platform for the generation of reliable deep learning based potential energy models
• Role of the exchange-correlation functional on the structural, electronic and optical properties of cubic and tetragonal SrTiO₃ including many-body effects
• On-the-Fly Active Learning of Interpretable Bayesian Force Fields for\n Atomistic Rare Events
• Less is more: Sampling chemical space with active learning
• The atomic simulation environment—a Python library for working with atoms
• From ultrasoft pseudopotentials to the projector augmented-wave method
• Electron-energy-loss spectra and the structural stability of nickel oxide: An LSDA+U study
• Generalized Gradient Approximation Made Simple
• Efficient iterative schemes forab initiototal-energy calculations using a plane-wave basis set
• Projector augmented-wave method
• Ab initiomolecular-dynamics simulation of the liquid-metal–amorphous-semiconductor transition in germanium
• The energy and elastic dipole tensor of defects in ionic crystals calculated by the supercell method

Related

• The Small Unit Cell Reconstructions of SrTiO3 (111)
• First-principles modeling of oxygen interaction with SrTiO3(001)\n surface: Comparative density-functional LCAO and plane-wave study
• First-principles study of cubic perovskites SrMO₃ (M = Ti, V, Zr and Nb)
• A dataset of DFT-computed oxygen vacancy formation energies of metal oxides
• Atomic-Scale Investigation of an Asymmetric SrTiO₃ Grain Boundary