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.

A Recipe for Charge Density Prediction

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

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

Deep Charge: A Deep Learning Model of Electron Density from One-Shot Density Functional Theory Calculation

Deepks: A Comprehensive Data-Driven Approach Toward Chemically Accurate Density Functional Theory

Charge-Optimized Electrostatic Interaction Atom-Centered Neural Network Algorithm

A representation-independent electronic charge density database for crystalline materials

Transferable Machine-Learning Model of the Electron Density

Machine learning for accuracy in density functional approximations

Completing density functional theory by machine learning hidden messages from molecules

Predicting The One-Particle Density Matrix With Machine Learning

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

A Transferable Machine-Learning Model of the Electron Density

Machine Learning Approaches toward Orbital-free Density Functional Theory: Simultaneous Training on the Kinetic Energy Density Functional and Its Functional Derivative

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

Efficient Sampling for Machine Learning Electron Density and Its Response in Real Space

DeepAtomicCharge: a new graph convolutional network-based architecture for accurate prediction of atomic charges

Improving Density Functional Prediction of Molecular Thermochemical Properties with a Machine-Learning-Corrected Generalized Gradient Approximation.

Graph Nets for Partial Charge Prediction

Transferable and extensible machine learning derived atomic charges for modeling hybrid nanoporous materials