Graph-neural-network predictions of solid-state NMR parameters from spherical tensor decomposition

Abstract

Nuclear magnetic resonance (NMR) is a powerful spectroscopic technique that is sensitive to the local atomic structure of matter. Computational predictions of NMR parameters can help to interpret experimental data and validate structural models, and machine learning (ML) has emerged as an efficient route to making such predictions. Here, we systematically study graph-neural-network approaches to representing and learning tensor quantities for solid-state NMR -- specifically, the anisotropic magnetic shielding and the electric field gradient. We assess how the numerical accuracy of different ML models translates into prediction quality for experimentally relevant NMR properties: chemical shifts, quadrupolar coupling constants, tensor orientations, and even static 1D spectra. We apply these ML models to a structurally diverse dataset of amorphous SiO${}_2$ configurations, spanning a wide range of density and local order, to larger configurations beyond the reach of traditional first-principles methods, and to the dynamics of the $α\unicode{x2013}β$ inversion in cristobalite. Our work marks a step toward streamlining ML-driven NMR predictions for both static and dynamic behavior of complex materials, and toward bridging the gap between first-principles modeling and real-world experimental data.

13 pages, 7 figures

References in corpus (25)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• All-electron magnetic response with pseudopotentials: NMR chemical shifts
• PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges
• Universal Fragment Descriptors for Predicting Electronic Properties of Inorganic Crystals
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• SchNetPack: A Deep Learning Toolbox For Atomistic Systems
• Chemical Shifts in Molecular Solids by Machine Learning
• A Transferable Machine-Learning Model of the Electron Density
• Machine Learning for Quantum Mechanical Properties of Atoms in Molecules
• Accurate molecular polarizabilities with coupled-cluster theory and machine learning
• Evidence for supercritical behavior of high-pressure liquid hydrogen
• First principles theory of the EPR g-tensor in solids: defects in quartz
• Learning the electronic density of states in condensed matter
• New fitting scheme to obtain effective potential from Car-Parrinello molecular dynamics simulations: Application to silica
• MatterSim: A Deep Learning Atomistic Model Across Elements, Temperatures and Pressures
• Thermodynamics and dielectric response of ${\text{BaTiO}}_3$ by data-driven modeling
• NMR Studies on the Temperature-Dependent Dynamics of Confined Water
• Modelling atomic and nanoscale structure in the silicon-oxygen system through active machine learning
• Predicting electronic structures at any length scale with machine learning
• Accurate computational prediction of core-electron binding energies in carbon-based materials: A machine-learning model combining density-functional theory and $\mathbf{GW}$
• Structure and dynamics of Oxide Melts and Glasses : a view from multinuclear and high temperature NMR
• Crystal structure identification with 3D convolutional neural networks with application to high-pressure phase transitions in SiO${}_2$