First principles quantum mechanical calculation of the electric dipole polarizabilities of the borate,carbonate, nitrate and related ions

Electric dipole polarizabilities have been calculated from first principles of quantum mechanics for the BO ₃ ^3? , CO ₃ ^2? , NO ₃ ^? series and for NO ₂ ^? and LiNO₃(g). Calculated trends in average polarizability and polarizability anisotropy in the BO ₃ ^3? -NO ₃ ^? series are in agreement with experiment and can be qualitatively interpreted in terms of the varying energies of the a₁′, a₂″ and e′ symmetry unoccupied MO's of the oxyanions. Embedding a CO ₃ ^2? ion in a D_3h symmetry array of divalent cations reduces both the average polarizability and its anisotropy, particularly when diffuse s and p functions are included in the calculation. Calculations on the gas phase LiNO₃ molecule and on the free NO ₃ ^? ion in the distorted geometry found in LiNO₃(g) allow us to separate polarizability contributions internal to the NO ₃ ^? and Li⁺ ions from those which arise from the Li⁺-NO ₃ ^? interaction. The Li⁺-NO ₃ ^? interaction term so obtained is much smaller than the NO ₃ ^? contribution but is in turn larger than the Li⁺ contribution, suggesting that the inclusion of this interaction term is essential for obtaining accurate results for ion pairs. Although static polarizabilities are in reasonable agreement with experiment for NO ₃ ^? the wavelength dispersion of the polarizability is underestimated by about a factor of two, apparently as a result of inadequacies in the quantum mechanical method. Calculated values are also presented for ¹⁴N NMR shieldings in the nitrogen oxyanions but these are in only qualitative agreement with the experimental values. Similarly, calculated values of magnetic susceptibility are in only qualitative agreement with experiment although trends along the BO ₃ ^3? -NO ₃ ^? series are properly reproduced.