Condensed Matter Modelling and Density Functional Theory

Condensed matter modelling seeks to predict the structure, electronic behaviour and functional properties of solids and interfaces by solving the equations of quantum mechanics. Central to this effort is density functional theory (DFT), which replaces the many-electron wavefunction by the ground-state electron density and maps an interacting system to an auxiliary non-interacting one. In practice, the Kohn–Sham formulation of DFT introduces single-particle orbitals subject to an effective potential comprising the external (nuclear) term, the classical Coulomb (Hartree) term and an exchange–correlation contribution that captures all many-body quantum effects. Despite its foundation as an exact theory, the form of the exchange–correlation functional must be approximated, leading to a hierarchy of methods—from the local density approximation (LDA) and generalized gradient approximations (GGAs) to hybrid and range-separated functionals. With advances in algorithms and high-performance computing, DFT now underpins the prediction of lattice constants, band structures, magnetic orderings and response functions across metals, semiconductors, oxides and low-dimensional materials. By balancing accuracy and cost, DFT has become the workhorse for materials discovery, guiding experiments and enabling the design of novel compounds with tailored optical, electronic and mechanical properties.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

A new all-electron plane-wave framework has been developed to treat both core and valence electrons on the same footing by employing analytic norm-conserving Coulomb potentials and very high energy cut-offs. Benchmark applications to diamond, silicon and molecular clusters demonstrate sub-millielectron-volt accuracy in band gaps and forces, providing a reference for pseudopotential and localized-basis methods. A broad roadmap for electronic structure software has charted the path to exascale computing, reviewing over a dozen major codes and outlining strategies to exploit massive parallelism, heterogeneous accelerators and mixed-precision arithmetic. This survey identifies key bottlenecks—such as efficient sparse linear algebra and sustainable code design—and recommends community-wide standards for portability and reproducibility. In parallel, breakthroughs in linear-scaling ab initio molecular dynamics have enabled simulations of over one hundred million atoms by combining sparse matrix techniques, non-orthogonal local submatrix methods and mixed-precision GPU execution. Sustained petascale performance has been achieved, bringing realistic modelling of complex materials and biological systems within reach of first-principles techniques.

Condensed Matter Modelling and Density Functional Theory publication trend

The graph below shows the total number of articles in condensed matter modelling and density functional theory across all publications each year (not limited to Nature Index journals).

Technical terms

Exchange–correlation functional: A term in DFT that encapsulates all quantum mechanical many-electron effects beyond classical Coulomb interactions and single-particle kinetic energy.

Pseudopotential: An effective potential replacing the atomic core electrons and nuclear attraction, allowing one to focus computational effort on valence states with smoother wavefunctions.

Plane-wave basis: A set of periodic sinusoidal functions used to expand electronic wavefunctions in crystalline materials, offering systematic convergence and easy treatment of periodic boundary conditions.

All-electron method: An approach that retains explicit treatment of both core and valence electrons, avoiding pseudopotentials at the cost of handling rapidly varying core states.

Ab initio molecular dynamics (AIMD): A simulation method that couples classical nuclear motion with instantaneous electronic structure calculations, enabling finite-temperature dynamics from first principles.

Exascale computing: Next-generation supercomputing platforms capable of 10^18 floating-point operations per second, empowering large-scale electronic structure and dynamics simulations.

References

1. Density Functional Methods.
2. All-Electron Plane-Wave Electronic Structure Calculations. Journal of Chemical Theory and Computation (2023).
3. Roadmap on electronic structure codes in the exascale era. Modelling and Simulation in Materials Science and Engineering (2023).
4. Towards electronic structure-based ab-initio molecular dynamics simulations with hundreds of millions of atoms. Parallel Computing (2022).