DFT calculations of the 'lone pair' effect - a tool for the chemist to predict molecular distortions?

摘要

The steric "lone pair" effect is a well-known phenomenon in chemistry. Not much is known about the chemical and structural criteria, however, on which it depends, and whether a distortion from the high symmetry structure actually occurs, and to which extent. The fast availability of reliable results by Density Functional Theory calculations provides a sound basis for attacking this problem. We analyse the energetic, steric and bonding properties of molecules AX{sub}3 (A: N to Bi; X=H, F to I). The application of a vibronic coupling approach of the PSJT type (D{sub}(3h) → C{sub}(3v) transition: A{sub}1{sup}' direct × α{sub}2{sup}" direct × A{sub}2{sup}" interaction) yields some valuable rules, from which two are mentioned. The vibronic coupling constant and the vibronic coupling energy, which predominantly determines the total D{sub}(3h) → C{sub}(3v) energy change, vary according to: F＞H＞Cl＞Br＞I(A: N to Bi); N＞P＞As＞Sb＞Bi(X: H, F) The dependence on A being only small or not present (X: Cl to I). Apparently the hardest molecules are the most susceptible to vibronic coupling, while a roughly inverse trend is observed for the extent of the angular distortion, where the influence of the force constant becomes distinct. Here the softest molecules such as Sb(Bi)Br{sub}3 exhibit the largest and NH{sub}3 the smallest deviations from D{sub}(3h). We further note that the vibronic coupling energy approximately equals the hardness difference η(C{sub}(3v)) -η(D{sub}(3h)), while the corresponding electro-negativity difference does not characterise the lone pair influence. The dependence on the coordination number is studied for AX{sub}(3+δ){sup}(δ-) entities (δ= -1, 1, 2, 3); one may broadly state, that - in agreement with structural data the tendency towards lone pair distortions decreases with increasing numbers of ligands.