The [formula omitted] NMR shifts are not a measure for the nakedness of the fluoride anion

Abstract

The solvent dependency of the 19 F NMR shifts of the fluoride anion in CH 3OH, H 2O, CH 3OCH 3, CHCl 3, CH 2Cl 2, CHF 3, CH 3CN, CH 3NO 2, (CH 3) 2SO, and CH 3COCH 3 solutions was studied at the MP2/6-31++G(d, p) and B3LYP/6-31++G(d, p) levels of theory and compared to the experimental values. It is shown that the free gaseous fluoride anion is most shielded. The stepwise build-up of a solvation sphere was modeled for the F −/ nH 2O system and results in a progressive deshielding of the F − nucleus with an increasing number of water ligands. The first solvation sphere of F − is comprised of six or seven monodentate water molecules. The F −···H bond distances increase from 1.42 Å in the monohydrate to 1.69–1.87 and 1.82 Å in the penta- and hexahydrates, respectively, and the transfer of negative charge from F − to the water ligands reaches its maximum for the tetrahydrate. The wide range of about 70 ppm observed for the chemical shift of F − in different solvents and the order of deshielding are confirmed by model calculations. Furthermore, it is shown that the deshielding observed for different solvents does not correlate with the calculated binding energies between F − and the corresponding solvent molecules, but parallels the increase in the calculated shielding anisotropy in the case of monodentate solvent F − adducts. Since the calculated shielding anisotropy can be taken as a qualitative measure for the paramagnetic shielding, the large solvent dependency of the F − shifts is best explained by the varying amounts of solvent induced paramagnetic shielding. The preferred structure of the F −·CH 3OH adduct involves hydrogen bridging through the hydroxyl and not the methyl group, and the minimum energy structures of F −·CH 3SOCH 3 and F −·CH 3COCH 3 exhibit bidentate solvent coordination. In solid fluorides, the chemical shift of F − spans more than 190 ppm and an increasing cation size results in increased deshielding. As previously shown, this deshielding is due to the electronic overlap effects. The MAS 19 F NMR spectra of solid N(CH 3) 4F and P(CH 3) 4F were also measured. The F − anion in the P(CH 3) 4 + salt is 19 ppm less shielded than in the N(CH 3) 4 + salt in accord with the increased cation size. However, in spite of its large size the deshielding caused by the N(CH 3) 4 + cation is only comparable to that of Rb + due to the methyl groups not providing as good an overlap as the smaller but softer Cs + cation. These results show that in both, the solid state and in solution, the chemical shift of F − is not a measure of its nakedness and that the fluoride anion is far from being naked.