Borosilicate glass alteration driven by magnesium carbonates

Abstract The alteration of simplified synthetic glass, representative of the French reference nuclear glass R7T7, in presence of hydromagnesite has been experimentally investigated and modeled. Magnesium in solution is known to potentially enhance glass alteration; nuclear glass clayed host rocks contain magnesium and can dissolve to maintain the concentration of magnesium in solution. For modeling purposes, it was suitable to study a simple system. Hydromagnesite was therefore chosen as a simple model mineral in order to check the influence of an Mg-rich mineral on glass alteration. Since the models use thermodynamic and kinetic parameters measured in pure water and pH-buffered solutions, changing the solution composition or adding minerals is a key step towards the validation of the modeling assumptions before using the model for predictive purposes. Experiments revealed that glass alteration is enhanced in presence of hydromagnesite. Modeling was performed using the GRAAL model implemented within the CHESS/HYTEC reactive transport code. Modeling proved useful both for explaining the mechanisms involved and quantifying the impact on glass alteration: Mg coming from hydromagnesite dissolution reacts with Si provided by the glass in order to form magnesium silicates. This reaction decreases the pH down to neutral conditions where magnesium silicates are more soluble than at the natural alkali pH imposed by glass or hydromagnesite dissolution. The driving force of the magnesium silicate precipitation is eventually the interdiffusion of alkali within the altered amorphous glass layer as this mechanism consumes protons. The model’s ability to describe the concentrations of elements in solution and formed solids whatever the glass/hydromagnesite ratio strongly supports the basic modeling hypothesis.