Electro-optical parameters of bond polarizability model for aluminosilicates - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Journal of Physical Chemistry A Année : 2006

Electro-optical parameters of bond polarizability model for aluminosilicates

Résumé

Electro-optical parameters (EOPs) of bond polarizability model (BPM) for aluminosilicate structures were derived from quantum-chemical DFT calculations of molecular models. The tensor of molecular polarizability and the derivatives of the tensor with respect to the bond length are well reproduced with the BPM, and the EOPs obtained are in a fair agreement with available experimental data. The parameters derived were found to be transferable to larger molecules. This finding suggests that the procedure used can be applied to systems with partially ionic chemical bonds. The transferability of the parameters to periodic systems was tested in molecular dynamics simulation of the polarized Raman spectra of -quartz. It appeared that the molecular Si-O bond EOPs failed to reproduce the intensity of peaks in the spectra. This limitation is due to large values of the longitudinal components of the bond polarizability and its derivative found in the molecular calculations as compared to those obtained from periodic DFT calculations of crystalline silica polymorphs by Umari et al. (Phys. Rev. B 2001, 63, 094305). It is supposed that the electric field of the solid is responsible for the difference of the parameters. Nevertheless, the EOPs obtained can be used as an initial set of parameters for calculations of polarizability related characteristics of relevant systems in the framework of BPM.

Dates et versions

hal-00267608 , version 1 (27-03-2008)

Identifiants

Citer

K.S. Smirnov, D. Bougeard, P. Tandon. Electro-optical parameters of bond polarizability model for aluminosilicates. Journal of Physical Chemistry A, 2006, 110, pp.4516-4523. ⟨10.1021/jp060151+⟩. ⟨hal-00267608⟩

Collections

CNRS UNIV-LILLE
16 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More