Abstract:Electronic-structure calculations play a fundamental role in predicting important physical (optical, electrical, etc) properties of condensed matter. Based on density functional theory (DFT), Kohn-Sham (KS) equation replaces the interacting many-body (electrons) problem by an equivalent set of self-consistent single-particle equations. Different from well-adopted basis set approach, we present a real-space finite-difference method to discretize the KS equation where the Laplacian operator is represented by high-order differences. After spatial discretization, a derived nonlinear eigenvalue problem is solved by a self-consistent field (SCF) iteration scheme with the update of electron density and potentials. In spite of the simplicity and versatility of the finite-difference method, one has to perform large-scale calculations of both eigenvalue and electrostatic problems at each iteration step. For eigenvalue problems, we minimize the Rayleigh quotients by the conjugate gradient method; and find minimum eigenvalues of interest through subspace diagonalization without Gram-Schmidt procedure. For obtaining electrostatic potentials, we use Dirichlet boundary condition to truncate the computational domain after inserting electron density and Gaussian compensating charge density together into Poisson's equation. Then a preconditioned Krylov subspace solver is employed to solve the Poisson's equation. A three-dimensional quantum dot with few spin-polarized electrons (each electron has spin up) is simulated by the real-space DFT incorporating local spin density (LSD) approximation. Wave function, electron density, eigen-energy, and total energy are calculated and compared to published results. The real-space finite-difference method produces reliable simulation results with a high computational efficiency. This work is fundamentally important to quantum-mechanical ab-initio calculation.

Machine-learning Kohn-Sham potential from dynamics in time-dependent Kohn-Sham systems

Machine Learning Exchange-Correlation Potential in Time Dependent Density Functional Theory

Dynamic Learning of Correlation Potentials for a Time-Dependent Kohn-Sham System

By-passing the Kohn-Sham equations with machine learning

Self-Consistent Solution Of Kohn-Sham Equation By Real-Space Finite-Difference Method

Kohn-Sham accuracy from orbital-free density functional theory via $Δ$-machine learning

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

Development of a machine learning finite-range nonlocal density functional

Accelerating finite-temperature Kohn-Sham density functional theory with deep neural networks

Construct exchange-correlation functional via machine learning

Lattice dynamics simulation using machine learning interatomic potentials

Scalable learning of potentials to predict time-dependent Hartree-Fock dynamics

D4FT: A Deep Learning Approach to Kohn-Sham Density Functional Theory

Completing density functional theory by machine learning hidden messages from molecules

Machine Learning the Physical Nonlocal Exchange–Correlation Functional of Density-Functional Theory

Ab Initio Validation on the Connection Between Atomistic and Hydrodynamic Description to Unravel the Ion Dynamics of Warm Dense Matter

Subsystem Density Functional Theory Augmented by a Delta Learning Approach to Achieve Kohn–Sham Accuracy

Predicting The One-Particle Density Matrix With Machine Learning

Towards Efficient and Unified Treatment of Static and Dynamic Correlations in Generalized Kohn-Sham Density Functional Theory

Toward Efficient and Unified Treatment of Static and Dynamic Correlations in Generalized Kohn-Sham Density Functional Theory

Machine Learning Nonadiabatic Dynamics: Eliminating Phase Freedom of Nonadiabatic Couplings with the State-Intraction State-Averaged Spin-Restricted Ensemble-Referenced Kohn-Sham Approach