Is silica really an anomalous oxide? Surface acidity and aqueous hydrolysis revisited.

The single-site Solvation, Bond Strength, and Electrostatic (SBE) model accounts for the anomalous position of silica onthe surface acidity versus aqueous acidity correlation developed for metal oxides, by considering the solvation energy change in the protonation reaction implemented through the dielectric constant (1/epsilon(k)) and the electrostatic energy change through the Pauling bond strength to bond length ratio (s/r) of the oxide. I address here why inclusion of the solid's dielectric constant brings silica into the same correlation as other oxides like TiO2, Al2O3, and Fe2O3. The solvation and electrostatic contributions are interpreted in terms of classical concepts such as chemical hardness, polarizability, ionicity, electronegativity, and local charge densities. Silica is acidic (PZC < 7), not because of its small dielectric constant, its tetrahedral coordination, or its high bond strength alone. Surface acidity depends largely on high values of the s/r ratio. The dielectric constant of the solid affects acidity mainly by reflecting the nature of water-surface interactions. Solids with large values of epsilon(k) are interpreted as being less polarizable and more ionic so that water, a hard polar solvent, interacts favorably with such surfaces and scales similar to water-water interactions regardless of whether the metal-oxide bond is in the solid or in the aqueous state. For these oxides, pKa(s) = pKa(aq) +/- 1. Silica, with a small dielectric constant, is interpreted as being more polarizable and more covalent so that water-SiO2 interactions scale differently than for the more ionic oxides. Such an interpretation when combined with the Partial Charge Model for metal hydrolysis suggests that the surfaces of RuO2, W03, Sb2O5, and Ta2O5 should be acidic similar to silica. But, unlike silica, they would lie on the pKa correlation defined by the other oxides because of their larger dielectric constants. The mixed oxide, AlPO4, is predicted to behave like silica.