| Local density approximation | |
|---|---|
| Name | Local density approximation |
| Caption | Schematic electron density in an inhomogeneous system |
| Developer | Walter Kohn and collaborators |
| Introduced | 1960s |
| Field | Density functional theory |
| Related | Kohn–Sham equations, exchange–correlation functional |
Local density approximation
The Local density approximation (LDA) is an approach within density functional theory (DFT) that models the exchange–correlation energy of an interacting electron system as a functional of the local electron density. LDA matters in quantum physics and computational materials science because it provides a simple, computationally efficient approximation enabling practical calculations for atoms, molecules and solids. Its relative simplicity made it a foundation for many advances in electronic structure methods and high-throughput materials discovery.
LDA emerges from the recognition that the properties of an interacting many-electron system can be mapped to functionals of the one-particle electron density by the Hohenberg–Kohn theorem. In practice, LDA approximates the exchange–correlation energy at a point r by assuming the local environment behaves like a homogeneous electron gas with the same density. Historically, LDA traces to applications of the homogeneous electron gas studied by David Pines and Nozières and the quantum Monte Carlo work of Ceperley and Alder, and was formalized through the work of Walter Kohn and Lu Jeu Sham in the development of the Kohn–Sham equations. LDA's role is central to early implementations of DFT in packages developed at institutions such as Bell Labs and later in community codes like Quantum ESPRESSO and ABINIT.
Formally, the LDA exchange–correlation energy functional is written as E_xc^[LDA][ρ] = ∫ ρ(r) ε_xc^hom(ρ(r)) d^3r, where ε_xc^hom(ρ) is the per-particle exchange–correlation energy of the homogeneous electron gas at density ρ. The exchange part often uses the analytic result of Dirac or Slater determinant-based approximations, while the correlation part is frequently parameterized using results from quantum Monte Carlo calculations by Ceperley and Alder or fitting formulas such as those by Perdew–Zunger. LDA enters the self-consistent loop of the Kohn–Sham equations via the exchange–correlation potential v_xc(r) = δE_xc/δρ(r). Connections to many-body perturbation theory are made when comparing LDA to methods such as the GW approximation and dynamical mean field theory (DMFT).
LDA has been widely applied to compute ground-state properties: equilibrium lattice parameters, cohesive energies, band structures, and phonons for metals, semiconductors, and insulators. It underpinned early studies of materials in research at IBM Research, Argonne National Laboratory, Lawrence Berkeley National Laboratory, and universities such as University of California, Berkeley and University of Cambridge. In quantum chemistry, LDA informed hybrid schemes that incorporate exact exchange, leading to functionals such as B3LYP where localized approximations are combined with other corrections. LDA also serves as a starting point for more advanced many-body treatments (e.g., LDA+U to include on-site Coulomb repulsion, and LDA+DMFT) used for strongly correlated systems like transition-metal oxides and rare-earth compounds studied at facilities including Oak Ridge National Laboratory and the Max Planck Society.
Despite successes, LDA exhibits well-known biases: it typically overbinds (underestimates lattice constants), underestimates band gaps, and poorly describes van der Waals interactions and strongly correlated electrons. These limitations motivated systematic improvements: the generalized gradient approximation (PBE) includes density gradients; meta-GGA functionals add kinetic energy density; hybrid functionals (e.g., PBE0) incorporate exact exchange; and nonlocal van der Waals functionals address dispersion. Empirical corrections and parameterized schemes such as Perdew–Burke–Ernzerhof and Perdew–Zunger refine behavior. Benchmarks against quantum chemistry methods (e.g., coupled cluster theories) and experimental data drive assessment and development, with community efforts like the Materials Project and the NOMAD Repository cataloguing performance across diverse materials.
LDA's algebraic simplicity yields low computational cost and numerical stability, making it attractive for large-scale calculations and high-throughput screening on supercomputers at centers like National Energy Research Scientific Computing Center and NERSC. It is implemented in major electronic structure codes including VASP, Quantum ESPRESSO, ABINIT, GPAW, and CASTEP. Practical considerations include choice of pseudopotentials (norm-conserving, ultrasoft, or PAW), basis sets (plane waves, localized orbitals), k-point sampling, and convergence criteria. For production workflows, reproducibility and access to compute resources are critical; initiatives such as the Open Science Grid and cloud platforms have expanded availability but also highlighted disparities in access.
LDA's computational efficiency enabled democratization of electronic structure calculations, accelerating research in energy materials, catalysis, and quantum devices at academic groups worldwide. However, disparities persist: well-resourced institutions and corporations more readily access high-performance computing and proprietary software, amplifying inequities in scientific contribution and benefit. Community-driven, open-source projects (Quantum ESPRESSO, ABINIT) and data-sharing initiatives (Materials Project, Aflow) partially mitigate barriers, but equitable access also requires investment in training, infrastructure, and inclusive policies. Ethically directed research using LDA-derived predictions can inform socially beneficial technologies—clean energy, affordable electronics, and climate mitigation—when coupled with collaborative governance that centers underrepresented regions and communities in decision-making and capacity building.
Category:Density functional theory Category:Electronic structure methods