Daniel Maynau

Size-consistent self-consistent configuration interaction from a complete active space

Abstract The size-consistent self-consistent (SC) 2 method is based on intermediate Hamiltonians and ensures size-extensivity of any configuration interaction (CI) by correcting its diagonal elements. In this work, an (SC) 2 dressing is proposed on a complete active space SDCI. This approach yields a more efficient code which can treat larger multireference problems. Tests are proposed on the potential energy curve of F 2 , the bond stretching of water and the inclusion of an Be atom in the H 2 molecule. Comparisons with approximate methods such as average quadratic coupled cluster (AQCC) are presented. AQCC appears as a good approximation to (SC) 2 .

Direct generation of local orbitals for multireference treatment and subsequent uses for the calculation of the correlation energy

We present a method that uses the one-particle density matrix to generate directly localized orbitals dedicated to multireference wave functions. On one hand, it is shown that the definition of local orbitals making possible physically justified truncations of the CAS (complete active space) is particularly adequate for the treatment of multireference problems. On the other hand, as it will be shown in the case of bond breaking, the control of the spatial location of the active orbitals may permit description of the desired physics with a smaller number of active orbitals than when starting from canonical molecular orbitals. The subsequent calculation of the dynamical correlation energy can be achieved with a lower computational effort either due to this reduction of the active space, or by truncation of the CAS to a shorter set of references. The ground- and excited-state energies are very close to the current complete active space self-consistent field ones and several examples of multireference singles and doubles calculations illustrate the interest of the procedure.

Can density functional methods be used for open-shell actinide molecules? Comparison with multiconfigurational spin-orbit studies

The geometries, electronic structures, and vibrational frequencies of two isoelectronic compounds PuO(2)(2+) and PuN(2) have been studied in detail at the density functional theory (DFT) and multiconfigurational ab initio levels of theory. Dynamic correlation was taken into account using second-order perturbation theory (CASPT2) and the variational difference-dedicated configuration interaction method for comparison with the results of the DFT study. Spin-orbit effects were included within the framework of an effective uncontracted spin-orbit configuration-interaction method which considers electron correlation effects and spin-orbit coupling on equal footing. The twelve lowest f-f electronic transitions are reported. The electronic ground state of both systems is found to be the Omega=4 component of (3)H(g). We thus disagree with an earlier assignment of the ground state of PuN(2) [E. F. Archibong and A. K. Ray, J. Mol. Struct: THEOCHEM 530, 165 (2000)]. Spin-orbit effects are small on both the geometry and vibrational frequencies of the ground states of PuO(2)(2+) and PuN(2), but they completely change the distribution of electronically excited states. A comparison of results obtained with the two classes of methods allows us to demonstrate that an unambiguous assignment of the electronic ground state and electronic spectra requires the use of multireference methods including spin-orbit coupling. Single-reference methods such as DFT provide a reasonable description of the electronic properties of ground states of these open-shell systems, and therefore also of their structural and vibrational properties. The experimental antisymmetric stretching frequency of matrix-isolated PuN(2) is reproduced well by both CASPT2 and DFT calculations; generalized gradient approximation formulations of DFT are more successful than hybrid versions in this respect. Ground-state properties of UO(2) (2+), UN(2), UO(2), PuO(2) (2+), and PuN(2) are compared and discussed.

Alkali dimers ground-state calculations using pseudopotentials

Abstract Ground-state calculations for alkali dimers XY (X, Y = Li, Na, K) are performed using a pseudopotential method. The Pseudopotential parameters are derived from atomic ab initio Hartree-Fock calculations. The influence of basis set size upon the results is underlined. Comparison with all-electron calculations using valence basis sets of the same quality shows the good performance of the present pseudopotential method.

A novel perturbation-based complete active space–self-consistent-field algorithm: Application to the direct calculation of localized orbitals

A complete active space–self-consistent-field (CAS–SCF) algorithm based on molecular orbitals that conserve their physical nature during the iterative process is proposed. The algorithm consists of an iterative procedure based on the imposition of the generalized Brillouin theorem to a complete active space-configuration interaction wave function. At convergence, the wave function is identical to the corresponding one obtained using canonical CAS–SCF orbitals, provided the nature of the active space is the same. If localized guess orbitals are used, the locality property is conserved by the final orbitals. Test calculations illustrate the interest of the proposed approach, that permits to control the nature of the active space.

Self‐consistent intermediate Hamiltonians: A coupled cluster type formulation of the singles and doubles configuration interaction matrix dressing

This paper presents a new self‐consistent dressing of a singles and doubles configuration interaction matrix which insures size‐consistency, separability into closed‐shell subsystems if localized molecular orbitals (MOs) are used, and which includes all fourth order corrections. This method yields, among several schemes, a reformulation of the coupled cluster method, including fully the cluster operators of single and double excitations, and partially those of the triples (Bartlett’s algorithm named CCSDT‐1a). Further improvement can be easily included by adding exclusion principle violating corrections. Since it leads to a matrix diagonalization, the method behaves correctly in case of near degeneracies between the reference determinant and some doubles. Due to its flexibility this formulation offers the possibility of consistent combination with less expensive treatments for the study of very large systems.

A full-configuration benchmark for the N2 molecule

Abstract A full-configuration interaction (FCI) calculation has been performed for the nitrogen molecule using an ANO [4s3p1d] basis set. The FCI space for such a system contains about 9.68×109 symmetry-adapted Slater determinants. The FCI results are compared with several approximate methods, both of single- and multi-reference type, in order to test their accuracy.

Selected excitation for CAS-SDCI calculations.

A new selected‐configuration interaction method is proposed, based on the use of local orbitals. A corresponding code has been written, which is devoted to CI calculations of rather large systems (about 50–100 carbon‐like atoms). Taking advantage of the locality, and then of the fact that interactions vanish when the distance is large, the dimension of the CI space is largely reduced. The determinants that would be created by long range excitations are expected to have a small weight in the wave function and are therefore eliminated. This selected excitation CI space is particularly suited for large molecules. It is tested on large polyene chains and on a transition metal complex. For large enough systems, the CPU time saving is important and, what is more noticeable, calculations that were impossible to perform without selection are feasible in this approach.

Addressing Through-H Magnetic Interactions: A Comprehensive ab Initio Analysis of This Efficient Coupler.

The exchange coupling in a structuraly characterized Cu(II)2 complex is analyzed to highlight the role of H bonds in the generation of efficient magnetic interactions. The interest for complementary insights which are not accessible through DFT calculations (Desplanches, C. et al. J. Am. Chem. Soc. 2002, 124, 5197) has driven this state-of-the-art ab initio inspection. The wave function expansion based upon localized orbitals allows us to selectively turn on specific mechanisms and quantitatively evaluate their roles in the exchange interactions. Our singlet-triplet splitting calculations demonstrate the enhancement of the magnetic coupling through a concerted oxygen-to-metal charge transfer and electronic redistribution within the OH bond of the OH···O magnetic linker. This mechanism accounts for ∼35% of the total experimentally measured singlet-triplet energy difference. This analysis strongly suggests that H bonds might be particularly useful not only in the establishment of intermolecular contacts but also within the basic units of magnetic materials.

One-electron pseudopotential study of NanFn−1 clusters (2⩽n⩽29). I. Electronic and structural properties of the ground state

We introduce a one-electron pseudopotential model to study the structural and electronic properties of excess-electron alkali halide clusters. This model assumes total charge transfer between alkali and halide atoms. This ionic part of the system is described via repulsive and Coulomb potentials. The remaining electrons of the excess metal atoms are treated within an explicit quantal scheme via ion–electron pseudopotentials. Moreover, explicit core-polarization and core-electron correlation contributions are taken into account. This model is used to derive ground state structural, energetics, and electronic properties of one-excess electron NanFn−1 clusters in the range 2⩽n⩽29. We show that the structural characters are closely related with electron localization and we propose a classification into five types, two of them exhibiting rather strong localization namely F-centers and Na-tail structures, the others exhibiting a less bound electron localizing in a surface-state, an edge-state, or on an atom-dep...