This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| CFOUR | |
|---|---|
| Name | CFOUR |
| Author | John F. Stanton; J. Gauss; J. F. Stanton Group |
| Developer | CFOUR developers; A. Tajti; J. F. Stanton; J. Gauss |
| Released | 1990s |
| Latest release | ongoing development |
| Programming language | Fortran; C; Python (wrappers) |
| Operating system | Linux; Unix; macOS |
| Platform | x86_64; POWER; Intel; AMD |
| Genre | quantum chemistry software |
| License | proprietary; academic licenses |
CFOUR is a quantum chemistry software package specialized in high-accuracy, coupled-cluster electronic structure methods and response theory for molecular properties. It provides implementations of coupled-cluster singles and doubles with perturbative triples and higher-order correlation treatments, aimed at spectroscopy and benchmark computations. The program is widely used by researchers in computational chemistry, chemical physics, and molecular spectroscopy across academic and national laboratory settings.
CFOUR traces its lineage to programs and developments by pioneers such as J. F. Stanton, R. J. Bartlett, G. E. Scuseria, and T. J. Lee and grew alongside initiatives at institutions like Northwestern University, Iowa State University, and Argonne National Laboratory. The codebase incorporated algorithms from seminal works by Per-Olov Löwdin, John Pople, and Martin Head-Gordon and was influenced by the design philosophies of packages such as Gaussian (software), MOLPRO, GAMESS (US), and ACES II. Throughout the 1990s and 2000s, contributions from authors affiliated with University of Florida, University of Georgia, and University of Texas expanded its coupled-cluster and response modules. Community workshops and conferences including Gordon Research Conference on Computational Chemistry, American Chemical Society symposia, and meetings of the Molecular Quantum Mechanics Group fostered collaboration and dissemination. National initiatives like those at Lawrence Berkeley National Laboratory and Pacific Northwest National Laboratory provided computational environments for large-scale testing and benchmarking.
CFOUR offers capabilities for high-accuracy electronic structure calculations used in spectroscopic prediction and method development. Core functionalities include coupled-cluster methods (CCSD, CCSD(T), CCSDT, CCSDTQ), analytic gradients, and linear- and non-linear-response theory for properties such as polarizabilities, optical rotation, and vibrational frequencies. The package supports harmonic and anharmonic vibrational treatments, zero-point vibrational corrections, and rovibrational analysis for small and medium-sized molecules, interoperating with basis sets such as those from Dunning (chemistry), Pople basis sets, and correlation-consistent families. Interfaces and utilities enable use with external tools like Molpro, Psi4, ORCA (quantum chemistry), NWChem, and visualization via Avogadro (software). Parallel execution on clusters managed by resource managers such as SLURM, PBS (software), and LSF is supported.
The package implements a hierarchy of coupled-cluster theories and related many-body methods for ground and excited states. Methods include CCSD, CCSD(T), CCSDT, CCSDT(Q), and equation-of-motion coupled-cluster (EOM-CC) for ionization potentials, electron affinities, and excitation energies, drawing on formalisms developed by Stanton–Gauss theory contributors and related to techniques from Monkhorst–Pack sampling adaptations for molecular point groups. Response implementations cover linear-response (LR-CC), quadratic response, and frequency-dependent properties following approaches by J. F. Stanton, J. Gauss, and contemporaries such as A. Szabo and N. S. Ostlund. Perturbative treatments and explicitly correlated corrections (F12 variants) reflect methods advanced by K. R. Ranasinghe and contributors in the F12 community. Vibrational-configuration interaction and second-order vibrational perturbation theory (VPT2) tools reflect developments by J. M. Bowman and S. Carter.
CFOUR is predominantly implemented in modern Fortran with performance-critical kernels in Fortran and C and higher-level orchestration via scripts and Python wrappers. The codebase emphasizes modularity with separate modules for integral transformation, coupled-cluster amplitude solvers, and response routines. Integral evaluation and density-fitting support use libraries and standards such as Libint, BLAS, LAPACK, and parallelization through MPI. Memory and disk management strategies follow patterns used in high-performance codes at centers like Oak Ridge National Laboratory and Los Alamos National Laboratory. The program supports point-group symmetry handling for subgroups like C2v, D2h, and C1 and uses allocator schemes optimized for modern CPU caches such as those in Intel Xeon and AMD Epyc processors.
Development is coordinated by a core team of academic and national-lab developers with contribution models akin to projects at Argonne National Laboratory and Pacific Northwest National Laboratory. Distribution typically occurs under proprietary or academic licensing terms; binaries and source access are granted to academic groups, governmental organizations, and commercial entities under negotiated agreements, similar to licensing practices of Molpro and Gaussian (software). User support, documentation, and training are provided through mailing lists, workshops at venues like American Chemical Society meetings, and tutorials at research centers such as Max Planck Institute for Chemical Physics of Solids and university computational chemistry centers.
CFOUR is optimized for accuracy-driven benchmarks and has been employed in comparison studies against MRCC (software), DALTON (program) , Molpro, Psi4, and ORCA (quantum chemistry). Benchmark suites include energy and spectroscopic comparisons on small molecules like H2O, NH3, CO2, and N2O, and larger test sets such as the W4 and G2 sets. Performance scales with available parallel resources and benefits from dense linear algebra optimizations using Intel Math Kernel Library and tuned BLAS/LAPACK stacks. Cross-comparisons often reference standards set by groups at NIST and publications in journals like Journal of Chemical Physics and Journal of Physical Chemistry A.
Researchers employ CFOUR for high-accuracy studies in thermochemistry, spectroscopy, and method validation. Notable application areas include accurate prediction of molecular geometries, potential energy surfaces, and vibrational spectra for small to medium molecules relevant to atmospheric science and astrochemistry (e.g., work connected to NASA and European Space Agency projects). Studies in photochemistry, transition-state energetics, and nonadiabatic coupling have cited CFOUR computations in concert with experimental work at facilities like Brookhaven National Laboratory and Lawrence Livermore National Laboratory. The package is also used in benchmarking novel electronic structure methods developed at institutions such as Harvard University, Massachusetts Institute of Technology, California Institute of Technology, Stanford University, University of Cambridge, ETH Zurich, Max Planck Society, and Imperial College London.
Category:Quantum chemistry software