Study of a family of 40 hydroxylated β-cristobalite surfaces using empirical potential energy functions

We present a study of a family of 40 unique hydroxylated beta-cristobalite surfaces generated by cleaving the beta-cristobalite unit cell along crystallographic planes to include a combination of several low Miller index surfaces. The surface silicon atoms are quantified as percentages of Q(2) and Q(3) centers based on their polymeric state. We find that Q(3) centers are, on average, three times more abundant than Q(2) centers. To study the surface properties, we use two different empirical potential energy functions: the multibody potential proposed by Fueston and Garofalini (J. Phys. Chem. 1990, 94, 5351) and the newly developed CHARMM potential by Lopes et al. (J. Phys. Chem. B 2006, 110, 2782). Our results for the surface water interactions are in good agreement with previous ab initio theoretical studies by Yang et al. (Phys. Rev. B 2006, 73, 146102) for the (100) surface. We find that the most commonly studied family of {100} surfaces is unique and is the only surface with 100% abundance of Q(2) centers, whereas there are nine examples of surfaces with 100% Q(3) centers. The predominantly pure Q(3) surfaces show no hydrogen bonding with the neighboring surface hydroxyl groups and weakly adsorb water overlayers. This is markedly different from the {100} pure Q(2) surface that shows strong hydrogen bonding within the surface groups and with water. As compared to all the surfaces studied in this work, we find that the {100} surfaces are not true representations of the overall beta-cristobalite surfaces and their properties. We find that the two main factors that characterize the physical properties of silica surfaces are the polymeric state of the silicon atom and surface topography. Two types of pure Q(3) crystallographic planes have been identified and are labeled as Q(3A) and Q(3B) based on the differences in their topological features. Using the {111} and {011} surfaces as examples, we show that the Q(3A) surface adsorbs H2O that forms a stable monolayer, but the Q(3B) surface fails to form a stable H2O overlayer. Other crystallographic planes with different ratios of Q(2) to Q(3) centers are contrasted by the differences in the hydrogen-bonding network and their ability to form ordered H2O overlayers.