Bondi’s van der Waals radii

The S-S bond a appears to be almost normal, but there is also some shortening of the O-S distance b by comparison with the van der Waals contact distance. We have already accepted 1.70 A as the van der Waals radius of sulfur. According to Bondi,105 a mean value acceptable for oxygen seems to be 1.50 A. The S-0 contact distance would then be 3.20 A, which is much larger than the measured b distance, even if we make substantial allowance for the uncertainty in the determination of van der Waals radii. For instance, even if we accept errors of 0.2 A for both radii, we find a minimum S-0 contact distance of 2.80 A, still considerably larger than the measured value. 

The definition of the cavity (shape and size) is an intricate and delicate question that may have a considerable influence on the results (even qualitatively). In the original Onsager s theory, the molecular cavity was defined as a sphere and the volume was taken equal to the partial molecular volume of the solute in the solution. In practice, this volume can be assumed to be equal to the average volume in the pure liquid. Experimental values are then easily deduced from the experimental density of the liquid at 20°C when this quantity is available. Obviously, in SCRF applications, it became rapidly necessary to achieve a theoretical definition of the cavity applicable to any molecular structure. In former works carried out by our group [28,62], it was shown that a simple linear relationship exists between the experimental volume derived from the liquid density (Onsager s recipe) and the van der Waals volume, i.e., the volume enclosed by a set of overlapping atomic spheres with Bondi radii [63], Roughly, this relationship is... 

---