Abstract:Having access to accurate electron densities in chemical systems, especially for dynamical systems involving chemical reactions, ion transport, and other charge transfer processes, is crucial for numerous applications in materials chemistry. Traditional methods for computationally predicting electron density data for such systems include quantum mechanical (QM) techniques, such as density functional theory. However, poor scaling of these QM methods restricts their use to relatively small system sizes and short dynamic time scales. To overcome this limitation, we have developed a deep neural network machine learning formalism, which we call deep charge density prediction (DeepCDP), for predicting charge densities by only using atomic positions for molecules and condensed phase (periodic) systems. Our method uses the weighted smooth overlap of atomic positions to fingerprint environments on a grid-point basis and map it to electron density data generated from QM simulations. We trained models for bulk systems of copper, LiF, and silicon; for a molecular system, water; and for two-dimensional charged and uncharged systems, hydroxyl-functionalized graphane, with and without an added proton. We showed that DeepCDP achieves prediction R2 values greater than 0.99 and mean squared error values on the order of 10−5e2 Å−6 for most systems. DeepCDP scales linearly with system size, is highly parallelizable, and is capable of accurately predicting the excess charge in protonated hydroxyl-functionalized graphane. We demonstrate how DeepCDP can be used to accurately track the location of charges (protons) by computing electron densities at a few selected grid points in the materials, thus significantly reducing the computational cost. We also show that our models can be transferable, allowing prediction of electron densities for systems on which it has not been trained but that contain a subset of atomic species on which it has been trained. Our approach can be used to develop models that span different chemical systems and train them for the study of large-scale charge transport and chemical reactions.

Equivariant graph neural networks for fast electron density estimation of molecules, liquids, and solids

Equivariant graph neural networks for fast electron density estimation of molecules, liquids, and solids

GRAPH2MAT: UNIVERSAL GRAPH TO MATRIX CONVERSION FOR ELECTRON DENSITY PREDICTION

Learning Equivariant Non-Local Electron Density Functionals

Higher-Order Equivariant Neural Networks for Charge Density Prediction in Materials

High-Dimensional Neural Network Potential for Liquid Electrolyte Simulations.

Machine Learning Electron Density Prediction Using Weighted Smooth Overlap of Atomic Positions

Equivariant neural network for Green's functions of molecules and materials

Chemical Properties from Graph Neural Network-Predicted Electron Densities

Deep Neural Network Computes Electron Densities and Energies of a Large Set of Organic Molecules Faster than Density Functional Theory (DFT)

E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials

A Recipe for Charge Density Prediction

Neural Polarization: Toward Electron Density for Molecules by Extending Equivariant Networks

E3STO: Orbital Inspired SE(3)-Equivariant Molecular Representation for Electron Density Prediction

Toward Orbital-Free Density Functional Theory with Small Data Sets and Deep Learning

Transferable Machine-Learning Model of the Electron Density

Equivariant Neural Networks Utilizing Molecular Clusters for Accurate Molecular Crystal Lattice Energy Predictions

Highly Accurate Real-space Electron Densities with Neural Networks

Learning electron densities in the condensed phase

Neural network representation of electronic structure from $ab$ $initio$ molecular dynamics

Image Super-resolution Inspired Electron Density Prediction