###
The accuracy of *ab initio* molecular geometries for
systems containing second-row atoms

Sonia Coriani^{1}, Domenico Marchesan^{1},Jürgen Gauss^{2},
Christof Hättig^{3}, Trygve Helgaker^{4}, Poul Jørgensen^{5}

^{1}
Dipartimento di Scienze Chimiche, Università degli Studi di Trieste,
via Licio Giorgieri 1, I-34127 Trieste, Italy

^{2}
Institut für Physikalische Chemie, Universität Mainz, D-55099 Mainz, Germany

^{3}
Forschungszentrum Karlsruhe, Institute of Nanotechnology, P.O. Box 3640, D-76021 Karlsruhe, Germany

^{4}
Department of Chemistry, University of Oslo, P. O. Box 1033 Blindern, N-0315 Oslo, Norway

^{5}
Department of Chemistry, University of Århus, Langelandsgade 140, DK-8000 Århus C, Denmark

*J. Chem. Phys.* **123**, 184107 (2005).

(Received 1 August 2005; accepted 9 September 2005; published online 9 November 2005)

The performance of the standard hierarchy of ab initio models-that is,
Hartree-Fock theory, second-order Moller-Plesset theory,
coupled-cluster singles-and-doubles theory, and coupled-cluster
singles-doubles-approximate-triples theory-in combination with
correlation-consistent basis sets is investigated for equilibrium
geometries of molecules containing second-row elements. From an
analysis on a collection of 31 molecules (yielding statistical samples
of 41 bond distances and 13 bond angles), the statistical errors (mean
deviation, mean absolute deviation, standard deviation, and maximum
absolute deviation) are established at each level of theory. The
importance of core correlation is examined by comparing calculations in
the frozen-core approximation with calculations where all electrons are
correlated. (c) 2005 American Institute of Physics.

**View Article:**
PDF
(access restricted to domain theochem.rub.de)

Back to the list of Publications by the Quantum Chemistry (Hättig) Group