| CASTEP | |
|---|---|
| Name | CASTEP |
| Developer | Cambridge University (initial), Accelrys/Biovia/commercial partners, academic contributors |
| Released | 1990s |
| Programming language | Fortran, C (components) |
| Operating system | Unix-like, Linux, macOS |
| Platform | x86, x86-64, HPC clusters |
| Genre | Electronic structure software, Density functional theory |
| License | Commercial and academic licensing; mixed-source historically |
CASTEP
CASTEP is a plane-wave pseudopotential electronic structure code used to calculate the properties of solids, surfaces and molecules from first principles. It implements density functional theory (DFT) and related methods to predict electronic, vibrational and spectroscopic properties, and is widely used in computational condensed matter physics and materials science for research, engineering and policy-relevant work on energy, sustainability and equitable technology deployment.
CASTEP directly implements the quantum mechanical many-electron problem within practical approximations to solve for ground-state and excited-state properties. By numerically solving the Kohn–Sham equation of DFT using a plane-wave basis and pseudopotential techniques, CASTEP connects microscopic quantum mechanics to experimentally observable quantities such as band structures, phonons and optical spectra. Its role in quantum physics is as a production-grade tool enabling tests of fundamental approximations (e.g., exchange–correlation functionals) and providing data that inform experimental design and public-interest topics like low-carbon materials and equitable access to clean technologies.
CASTEP is founded on density functional theory and the Kohn–Sham formalism, employing popular exchange–correlation functional approximations such as the Local density approximation (LDA) and Generalized gradient approximation (GGA) (e.g., PBE). It uses a plane-wave basis set combined with norm-conserving or ultrasoft pseudopotential and projector-augmented wave (PAW) approaches to remove core electrons from explicit treatment, improving efficiency for solids. CASTEP also implements perturbative and linear-response extensions to compute vibrational properties via density-functional perturbation theory (DFPT), and time-dependent extensions for optical response approximations related to time-dependent density functional theory (TDDFT). The code supports many-body corrections such as GW approximation and hybrid functionals (e.g., HSE06) for improved band gaps, linking to broader quantum many-body theory.
CASTEP provides workflows for electronic structure, geometry optimization, molecular dynamics, phonons, and spectroscopy. Notable capabilities include: - Self-consistent field (SCF) ground-state DFT with plane-wave convergence controls and k-point sampling (Monkhorst–Pack). - Geometry optimization and transition-state searches using algorithms related to conjugate gradient and quasi-Newton methods. - Phonon calculations via finite-displacement and density-functional perturbation theory to obtain vibrational spectra and thermal properties. - Electronic excitation and optical properties via linear response, GW, and hybrid functionals. - Simulations of surfaces, defects and interfaces important for materials science problems such as battery electrodes, catalysts and photovoltaic materials. These features make CASTEP relevant for practitioners in academia, national laboratories such as Culham Centre for Fusion Energy and industrial research groups focused on sustainable materials.
CASTEP is implemented primarily in Fortran with performance-critical components optimized for high-performance computing (HPC) architectures. It uses Fast Fourier Transforms (FFTW, vendor libraries) for plane-wave/real-space transformations, iterative diagonalization (e.g., Davidson or preconditioned conjugate-gradient solvers), and parallelization across k-points, bands and plane-waves using MPI and hybrid MPI+OpenMP models. Memory and I/O strategies address large-scale supercell and defect calculations. Performance scaling has been validated on academic and commercial clusters, and CASTEP has been compared with other codes such as VASP, Quantum ESPRESSO, ABINIT and WIEN2k in benchmark studies. Algorithmic developments in CASTEP have included efficient handling of ultrasoft pseudopotentials, improved preconditioners, and scalable DFPT kernels for phonons.
CASTEP is widely applied to compute electronic band structures, density of states, defect energetics, adsorption on surfaces, and vibrational spectra—central problems in condensed matter physics and materials design. Typical application areas include: - Battery and energy materials: predicting intercalation voltages and diffusion barriers for lithium-ion battery electrodes. - Catalysis and surface science: adsorption energetics on metal and oxide surfaces relevant to green chemistry and decarbonization. - Photovoltaics and optoelectronics: assessing band alignments and defect-induced nonradiative recombination. - Quantum materials: exploring correlated behavior and topological properties when combined with beyond-DFT methods. Researchers from universities such as University of Cambridge, University of Oxford, Imperial College London and institutions like Rutherford Appleton Laboratory have used CASTEP to produce peer-reviewed studies that inform material selection, standards, and equitable access to technologies.
CASTEP originated from developers at University of Cambridge and has evolved through collaborations with commercial vendors and academic contributors. Historically distributed under academic licenses and commercial arrangements via software companies (e.g., Accelrys/BIOVIA), its development model mixes proprietary and academic-release practices. The user community includes academic groups, national labs and industry, with workshops and tutorials often presented at conferences such as the American Physical Society meetings and specialized symposia on electronic structure. Community efforts emphasize reproducibility, open data, and equitable participation; users increasingly integrate CASTEP outputs with open databases like the Materials Project and NOMAD repository to accelerate inclusive materials discovery.
Category:Density functional theory software Category:Computational physics Category:Materials science software