Benchmarking of Model Core Potentials:  Application to the Halogen Complexes of Group 4 Metals

Stephen A. Decker and Mariusz Klobukowski*
Department of Chemistry, University of Alberta, Edmonton, Alberta, CanadaT6G 2G2, and Computational Research on Materials Institute, Department of Chemistry, University of Memphis, Memphis, Tennessee 38152-6060
J. Chem. Inf. Comput. Sci., 2001, 41 (1), pp 1–7
DOI: 10.1021/ci0000474
Publication Date (Web): December 14, 2000
Copyright © 2001 American Chemical Society

 University of Alberta.

,

 University of Memphis.

,
*

 Corresponding author. E-mail:  Mariusz.Klobukowski@UAlberta.CA. Phone:  (780) 492-2568, 492-3170.

Abstract

The reliability of the model core potential (MCP) method was probed in a systematic RHF and MP2 study of the geometries of the group 4 metal halogen complexes (MX4; M = Ti, Zr, Hf and X = F, Cl, Br, I). The computed bond lengths were compared with experimental values, as well as those predicted using effective core potentials. Provided that electrons from the outermost core shell of the metal atom are treated explicitly in the calculation, both the MCP and ECP methods predict M−X bond lengths within 0.02−0.03 Å of experiment. The reaction energies for a simple set of halogen substitution reactions of the MX4 complexes leading to the mixed halogen complex, MX2Y2, were also studied. Although no experimental values are available for these reactions, comparison was made with the values computed using effective core potentials. The predictability of the different pseudopotential techniques and the importance of the metal atom valence basis set contraction scheme and polarization space are discussed.

Tools

History

  • Published In Issue January 22, 2001
  • Received May 18, 2000

Recommend & Share

Related Content

Other ACS content by these authors: