Experimentally determined dissolution kinetics of SON68 glass at 90 °C over a silica saturation interval: Evidence against a linear rate law

Flow-through dissolution experiments were carried out on the SON68 glass—the inactive analog of the French high-level waste glass—at 90 °C and pH = 9 over a silica saturation interval [0–150 ppm Si(aq)]. Two types of specimens were subjected to dissolution studies: powders (249–150, 149–75 or 74–45 μm diameter size fractions) or monoliths (∼1 cm × 1 cm × 4 mm). At each Si concentration interval individual coupon and glass powder experiments were run at the same flow-to-surface area (q/S) conditions. Under dilute solution conditions, the dissolution rate becomes independent of q/S and define the forward rate of reaction of 0.30 g/(m2 d). Dissolution rates of powders normalized on either a geometric or a B.E.T. surface area basis were compared to rates determined on monoliths. In every case, the geometric surface area normalized rates matched those of the monoliths, conclusively showing that B.E.T. normalized rates are too low by a factor of ∼3×. In Si-free solutions changes in flow rates resulted in differences in Al released to solution (∼50 to 3140 ppb), correlating inversely with dissolution rates. Si was added to input solutions and kept constant (“direct”), but in select experiments Si concentrations were above target concentrations, and then dropped to their target values (“indirect”). Both “direct” and “indirect” Si addition experiments yielded identical dissolution rates within experimental uncertainty. Progressive addition of Si to the input solution caused a decrease in rates from 0.30 to 0.0020 g/(m2 d), or a factor of ∼150×, but the decrease is strongly non-linear with respect to Si. These data are inconsistent with simple models relating glass dissolution to the chemical affinity of reaction.