Comparison of ab initioand density functional methods for vibrational analysis of TeCl4

Vibrational analysis of tellurium tetrachloride, TeCl4, was performed with Hartree–Fock (HF), MP2, and generalized gradient approximation density functional theory (DFT) methods supplemented with polarized double‐zeta split valence (DZVP) basis sets and relativistic effective core potentials (RECP) of Hay and Wadt. The molecular geometry is best reproduced at the HF and MP2/RECP+DZVP [polarized Hay and Wadt RECP for Te and 6–31G(d) basis set for Cl] levels of theory. The DFT methods gave rise to poorer results, especially those using Becke's 1988 exchange functional. Generally, the vibrational frequencies calculated by the MP2 and B3‐type DFT methods with the all electron and RECP+DZVP basis sets as well as at the HF/RECP level were in satisfactory accord with the experimental data. The agreement was good enough to assist the assignment of the measured vibrational spectra. The best agreement with the experimental vibrational frequencies was achieved with the scaled HF/RECP force field. Consistent results were obtained for the unobserved A2(ν4) fundamental, where the results of the best methods were within 4 cm−1. The best force fields were obtained with the following methods: Becke3–Lee–Yang–Parr and Becke3–Perdew/all electron basis, MP2 and Becke3‐Perdew/RECP+DZVP, and HF/RECP. The methods using RECPs are advantageous for large‐scale computations. The RECP basis set effectively compensates the errors of the HF method for TeCl4; however, it provides poor results with correlated methods. © 1998 John Wiley & Sons, Inc. J Comput Chem 19: 308–318, 1998