The 13C nuclear magnetic shielding in benzene and ten monosubstituted benzenes was studied when these compounds were dissolved in cyclohexane, carbon tetrachloride, tetrachloroethylene, methylene bromide and methylene iodide. The results revealed that the observed changes of -C magnetic shielding are dependent on both the solute and solvent molecular properties, although the dependence on the solvent is much more significant. It was also shown that the solvent effects for aromatic carbons are independent of the m electron density distriiution in an aromatic ring. The observed -C deshielding was attriiuted mainly to overlap effects which take place during molecular collisions.
INTRODUCI'IONIt is well known that NMR chemical shifts are solvent dependent. Much research has been carried out in this field,14 which reflects the great importance of this problem in NMR spectroscopy as any routine measurement of chemical shifts includes an unknown solvent effect. It has been previously established that the NMR solvent effect (i.e. the change in nuclear magnetic shielding due to the solvent and in the absence of specific interactions'), usolv, consists of terms due to bulk magnetic susceptibility, a,, diamagnetic anisotropy of the solvent molecules, ua, the polar effect caused by the charge distribution in the neighboring solvent molecules, uE, and the so-called Van der Waals effect, a , , which is the only effect (apart from a,,) if the solute and solvent molecules are isotropic and non-polar : '~3 While ub, a, and uE are well defined and recognized, a,,, is usually used to describe the part of the solvent effect that cannot be explained by the previous terms. According to Rummens? the a, contribution is the effect of the dispersion or London forces, including higher order dispersion terms, and also of repulsive and overlap forces. The problem of the Van der Waals term is not so critical for proton chemical shifts, where a , is usually responsible for only a fairly small part of the solvent effect, and can be explained almost quantitatively by the dispersive attraction between molecules.6 Unfortunately, this is not true for the shielding of other nuclei, e.g. the cw term may easily dominate the solvent effect in 13C chemical shifts. However, a complicating feature of this problem is that the origin of a, may be different for various nuclei (e.g. 'H and 13C). There is increasing evidence that the repulsive forces effectively change the shielding constants of heavier nuclei (e.g. lZ9Xe, I9F and C) during intermolecular interaction^.^-'^ This may be the reason why the theoretical calculations of solvent effects based on the solvation model'4 were unable to predict even the proper direction of changes in the I3C magnetic shielding of methyl carbons (cf. the recent experimental results obtained by Tiffon and Ancian"). It seems that each a, effect requires an individual approach before its nature can be understood.Recently we reported the solvent effect on the I3C chemical shifts of some saturated hydrocarbons dissolved in carb...