Glass dissolution as a function of pH and its implications for understanding mechanisms and future experiments

Abstract Various rate equations for the dissolution of silicate glasses have been discussed in the literature. In this article, the published results from studies are discussed in which the dissolution rate data are collected under high flow conditions such that saturation with respect to alteration products is avoided. Additionally, the studies also covered broad ranges of temperature and pH. Starting with nuclear waste glass studies, a two-term rate expression is used to fit data with the result that the data point toward a three-term expression offered by Kohler et al. (2003). These rate expressions contain two or three pre-exponential or rate constants. However, it appears that a single rate constant, an intrinsic rate constant, is consistent with the data. Thus, a rate expression of the form R = k i exp - E a H + RT a H η H + exp - E a H 2 O RT + exp - E a OH - RT a OH η OH appears to be applicable to a broad range of glasses. Here, R is the rate of dissolution, mol/(m 2 ·s) or similar; E is the activation energy associated with the acid, water, or hydroxide activated reactions, kJ/mol; a is the activity of H + , H 2 O, or OH − , unitless; η is the order of the reaction with respect to H + , H 2 O, or OH − ; R is the gas constant, kJ/(mol·K); T is the temperature, Kelvin; and k i is the intrinsic rate constant, mol/(m 2 ·s) or similar. The contribution to the overall rate from the ‘water’ term is evident as a minor contribution in the middle pH range for some glass compositions and a major contributor for others. One nuclear waste glass (the Japanese P0798), a basalt glass (Kohler et al., 2003), and a glass with a mineral composition (Bourcier, 1998) exhibit this contribution as a relatively flat response to changes in pH in the range of 5–8. However, to distinguish between the possible rate laws, more experiments and more carefully constrained experimentation are needed. These may include experiments at pH values that differ by as little as 0.25. Lastly, experiments with glasses of different compositions are needed to determine the dependence of the intrinsic rate constant on the glass composition and structure, i.e. non-bridging oxygens, Si–O–Si and Si–O–X (X = a matrix-forming element, e.g. Al or Fe), and other glass structural properties, e.g. binding energies.