Configuration interaction

Configuration interaction
Name	Configuration interaction
Classification	Post-Hartree–Fock method
Founders	Douglas Hartree, Vladimir Fock, Clemens Roothaan
Related	Coupled cluster, Møller–Plesset perturbation theory, Density functional theory
Software	Gaussian (software), GAMESS (US), NWChem, Psi4

Configuration interaction. It is a post-Hartree–Fock wave function-based technique used in computational quantum chemistry to solve the non-relativistic Schrödinger equation for molecules. The method systematically improves upon the single-configuration approximation of Hartree–Fock theory by expressing the total electronic wave function as a linear combination of Slater determinants, which represent different electron configurations. This approach provides a conceptually straightforward route to incorporating electron correlation, a critical effect for accurate predictions of molecular properties, reaction energies, and spectroscopic data.

Overview and theoretical foundation

The fundamental concept originates from the variational principle, which states that the expectation value of the Hamiltonian for any trial wave function is an upper bound to the true ground state energy. In Hartree–Fock theory, the wave function is approximated as a single Slater determinant constructed from a set of optimized one-electron functions, the molecular orbitals. This mean-field approach neglects the instantaneous, correlated motion of electrons. To recover this correlation energy, the wave function is expanded in a basis of configuration state functions, which are symmetry-adapted linear combinations of Slater determinants. The expansion includes the Hartree–Fock reference determinant plus excited determinants where electrons are promoted from occupied to virtual orbitals. The coefficients of these determinants are determined by diagonalizing the Hamiltonian matrix in this many-electron basis set, a process formally equivalent to a full configuration interaction calculation within the given orbital basis.

Mathematical formulation

For an N-electron system, the total wave function is expressed as a linear combination: Ψ_CI = C_0 Φ_0 + Σ_{i,a} C_i^a Φ_i^a + Σ_{i>j, a>b} C_{ij}^{ab} Φ_{ij}^{ab} + ..., where Φ_0 is the Hartree–Fock reference determinant. The terms Φ_i^a represent singly excited configurations, Φ_{ij}^{ab represent doubly excited configurations, and so forth. The indices i, j denote occupied orbitals from the Hartree–Fock solution, while a, b denote unoccupied (virtual) orbitals. The expansion coefficients, C, and the corresponding energy are obtained by solving the secular equation, H C = E S C, where H is the Hamiltonian matrix and S is the overlap matrix between configurations. The matrix elements between determinants are evaluated using Slater's rules and the Slater–Condon rules, which express them in terms of one- and two-electron integrals over the molecular orbitals.

Types of configuration interaction

The method is categorized by the level of excitation included in the expansion. Full configuration interaction includes all possible excitations within a given one-electron basis set, providing the exact solution to the Schrödinger equation for that basis, but is computationally prohibitive for all but the smallest systems. Configuration interaction with singles and doubles is a common truncation, capturing a significant portion of correlation energy. Configuration interaction with singles, doubles, and triples offers higher accuracy at greatly increased cost. Specialized variants include Multi-reference configuration interaction, which is essential for describing systems with significant static correlation or near-degeneracy, such as in transition metal complexes or during bond dissociation. The Complete active space self-consistent field method is often used to generate a reference for subsequent multi-reference CI calculations.

Computational considerations and limitations

The primary computational challenge is the factorial growth in the number of Slater determinants with system size and basis set, a manifestation of the combinatorial explosion. Even CISD scales formally as N^6, where N is the number of basis functions. This limits practical applications to relatively small molecules. The method is also not size-consistent when truncated; for example, the energy of two infinitely separated hydrogen molecules calculated with CISD is not twice the energy of an isolated H₂ molecule. This deficiency is addressed by more advanced methods like Coupled cluster theory. Furthermore, the accuracy is intrinsically limited by the choice of the one-electron basis set; reaching the complete basis set limit requires extrapolation techniques.

Applications in quantum chemistry

Despite its cost, it remains a valuable benchmark tool for developing and calibrating more approximate methods like Density functional theory. It is extensively used for calculating accurate potential energy surfaces for small polyatomic molecules, which are critical for understanding reaction dynamics and molecular spectroscopy. The method provides reliable predictions for electronic excitation energies through configuration interaction singles and higher-level variants, making it useful for simulating UV/Vis and photoelectron spectra. It is also employed for precise computation of molecular properties such as dipole moments, polarizabilities, and hyperpolarizabilities.

Relationship to other methods

It is historically and conceptually linked to several other ab initio techniques. Møller–Plesset perturbation theory can be viewed as an approximation to it, where correlation is treated as a perturbation to the Hartree–Fock Hamiltonian. Coupled cluster theory, often considered the gold standard for single-reference correlation, can be derived from an exponential ansatz of the excitation operators and typically provides superior accuracy and size-consistency compared to truncated CI. Density functional theory takes a fundamentally different approach by focusing on the electron density rather than the wave function. For multi-reference problems, methods like Complete active space perturbation theory combine the advantages of CASSCF and perturbation theory, offering a more scalable alternative to full multi-reference CI.

Category:Computational chemistry Category:Quantum chemistry Category:Ab initio quantum chemistry methods