| energy density functional | |
|---|---|
| Name | Energy density functional |
| Field | Quantum physics |
| Introduced | 1960s |
| Related | Density functional theory, Kohn–Sham equations, Exchange–correlation functional |
energy density functional
An energy density functional is a mapping from one or more local densities (such as particle density or spin density) to the total energy of a quantum many-body system. It provides an approach to compute ground-state properties of interacting electrons or nucleons by expressing energy in terms of spatially dependent fields rather than many-body wavefunctions. Energy density functionals underpin computational frameworks that balance accuracy and tractability in Quantum Physics and materials modelling.
In formal terms an energy density functional E[ρ, ...] assigns a scalar energy to a set of local densities ρ(r) and auxiliary fields (spin density, current density, pairing density). The concept stems from the Hohenberg–Kohn theorems which establish a one-to-one correspondence between external potentials and ground-state densities for interacting electrons, enabling the existence of a universal functional. Practical forms of the functional separate contributions such as kinetic energy, external potential energy, Hartree (classical Coulomb) energy, and the quantum mechanical exchange–correlation energy. The energy density is often expressed as an integral E = ∫ ε(ρ(r), ∇ρ(r), ...) d^3r, where ε is a local energy density per volume.
Energy density functionals are grounded in rigorous results from many-body theory and functional analysis. The Hohenberg–Kohn theorems and proof by Kohn and Sham motivate the use of density as the fundamental variable, while concepts from many-body Green's functions, reduced density matrix theory, and the variational principle inform construction and constraints. For nuclear systems, the Hartree–Fock and Hartree–Fock–Bogoliubov frameworks motivate energy density functionals that include pairing and isospin degrees of freedom. Theoretical tools such as the adiabatic connection and coupling-constant integration connect noninteracting reference systems to the full interacting problem, guiding derivation of exchange–correlation components.
The most widely used practical scheme is Kohn–Sham density functional theory (KS-DFT), which replaces the interacting many-electron problem by an auxiliary noninteracting system with identical ground-state density. KS-DFT introduces single-particle orbitals that reproduce ρ(r) and a set of self-consistent equations (Kohn–Sham equations) solved iteratively. Implementations appear across computational chemistry and materials science codes such as VASP, Quantum ESPRESSO, Gaussian, ABINIT, and CP2K. For nuclear applications, frameworks include the Skyrme interaction and Gogny force inspired energy density functionals encoded in codes like HFODD and Sky3D.
The exchange–correlation (XC) functional embodies many-body quantum effects beyond classical electrostatics and noninteracting kinetic energy. Exact XC is unknown; hence approximations are central. Common families include the local density approximation (LDA), derived from the uniform electron gas result; generalized gradient approximations (GGA) such as Perdew–Burke–Ernzerhof (PBE); meta-GGA functionals that depend on the kinetic energy density; and hybrid functionals (e.g., B3LYP, PBE0) that mix a fraction of exact Hartree–Fock exchange. For long-range correlation, methods like van der Waals density functional (vdW-DF) add nonlocal correlation terms. In nuclear EDFs, parameterizations such as Skyrme energy density functional and UNEDF project optimize parameters to reproduce nuclear masses and radii.
Numerical solution of energy density functional equations requires discretization choices and convergence strategies. Basis sets include plane waves, Gaussian-type orbitals, numerical atomic orbitals, and real-space grids; each choice affects efficiency and systematic errors. Self-consistent field (SCF) convergence employs mixing schemes (Pulay/DIIS), preconditioners, and level shifting. Treatment of periodic systems uses Brillouin zone sampling (k-point meshes) and pseudopotentials (norm-conserving, ultrasoft, or projector augmented-wave PAW) to reduce core-electron cost. Numerical integration of XC energy demands accurate quadrature grids, and for time-dependent extensions (TDDFT) one uses real- or frequency-domain propagators. High-performance computing and parallelization (MPI/OpenMP) are essential for large-scale simulations.
Energy density functional methods are central to predicting materials properties (band structures, formation energies, phonons) in solid-state physics, chemical reaction energetics and spectroscopy in computational chemistry, and global properties of nuclei in nuclear structure theory. Notable applications include screening of photovoltaic materials, catalysis studies with functionals like B3LYP and PBE, ab initio molecular dynamics (CPMD), and nuclear mass models for r-process nucleosynthesis used by collaborations such as UNEDF and Nuclear Energy Agency (NEA). DFT also feeds into multiscale modeling, linking to GW approximation for quasiparticles and Dynamical mean-field theory (DMFT) for strongly correlated systems.
Limitations arise from approximate XC functionals leading to delocalization errors, self-interaction, poor band gaps, and challenges with strongly correlated electrons. Active research areas include development of nonlocal and machine-learned functionals, incorporation of exact constraints (Perdew's constraint-based approach), embedding methods (DFT+DMFT, QM/MM), and time-dependent extensions for excited states. In nuclear physics, efforts focus on uncertainty quantification, Bayesian parameter estimation, and linking EDFs to ab initio Hamiltonians from chiral effective field theory. Cross-disciplinary initiatives involve software projects (e.g., Psi-k, Materials Project) and community benchmarks to validate and improve functionals.