Direct configuration interaction and multiconfigurational self-consistent-field method for multiple active spaces with variable occupations. I. Method

Abstract

In order to reduce the number of ineffective configurations in a priori generated configuration spaces, a direct configuration interaction method has been developed which limits the electron occupations of orbital groups making up a total active space. A wave function is specified by first partitioning an active space into an unrestricted number of orbital groups and second by providing limiting values, in the form of minima and maxima, for the electron occupancies of each group. The configuration interaction problem corresponding to all possible determinants satisfying these conditions is solved in a fully direct manner by the use of Slater–Condon expressions in conjunction with single and double replacements. This configuration interaction approach, termed occupation restricted multiple active space-configuration interaction, has also been linked with orbital optimization programs to produce the occupation restricted multiple active space-self consistent field method.

Accurate ab initio potential energy curve of F2. I. Nonrelativistic full valence configuration interaction energies using the correlation energy extrapolation by intrinsic scaling method

“…In this context, the full configuration interaction ͑FCI͒ method 24,25 has as yet remained the benchmark approach because of its variational attributes. [26][27][28][29] However, while considerable strides have been made in CI methodology, [30][31][32][33][34][35][36][37][38] its excessively long configurational expansions still present the challenge of discovering the most efficiently converging reordering of its terms.…”

Quasi-degenerate second-order perturbation theory for occupation restricted multiple active space self-consistent field reference functions

“…These restrictions allow for just a single distribution (distribution 13) while all other distributions correspond to IECs (distributions [1][2][3][4][5][6][7][8][9][10][11][12][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. Not all of the distributions that correspond to IECs will contribute to the ORMAS-PT energy.…”

“…These α-groups and β-groups are then combined in a pair-wise procedure to find combinations that adhere to the modified electron occupation restrictions. For each valid combination, α-strings and β-strings are enumerated 24 from the α-group and β-group, respectively. These α-strings and β-strings are combined to form determinants; the α-strings and β-strings indicate which α and β molecular spin-orbitals are occupied.…”

“…It would therefore be beneficial to develop a general MRPT method that takes advantage of direct PT methodology and considers the IEC contributions to the MRPT properties. The present paper introduces an alternative MRPT method that employs the occupation restricted multiple active space (ORMAS) 24 wavefunction as the zero th order reference function. Similar to RASSCF and QCAS, the ORMAS method expands the wavefunction within a multi-electron basis that is typically much smaller than the corresponding CAS basis.…”

Model space diabatization for quantum photochemistry

“…Of special interest are methods applicable to electronic structure calculations that add dynamical correlation to multi-configurational selfconsistent field (MCSCF)-type wave functions [32][33][34][35][36] since currently these are the most reliable wave function approximations for describing potential energy surfaces of closely coupled electronic states. The present article is concerned with quasidegenerate perturbation theory (QDPT) methods in which the final step is the diagonalization of a lowdimensional effective Hamiltonian in a model space where the dimension is the number of adiabatic states of interest.…”