Dissolution kinetics of diopside as a function of solution saturation state: Macroscopic measurements and implications for modeling of geological storage of CO2

Measurements of the dissolution rate of diopside (r) were carried out as a function of the Gibbs free energy of the dissolution reaction (ΔGr) in a continuously stirred flow-through reactor at 90 °C and pH90 °C = 5.05. The overall relation between r and ΔGr was determined over a free energy range of −130.9 < ΔGr < −47.0 kJ mo1−1. The data define a highly non-linear, sigmoidal relation between r and ΔGr. At far-from-equilibrium conditions (ΔGr ⩽ −76.2 kJ mo1−1), a rate plateau is observed. In this free energy range, the rates of dissolution are constant, independent of [Ca], [Mg] and [Si] concentrations, and independent of ΔGr. A sharp decrease of the dissolution rate (∼1 order of magnitude) occurs in the transition ΔGr region defined by −76.2 < ΔGr ⩽ −61.5 kJ mo1−1. Dissolution closer to equilibrium (ΔGr > −61.5 kJ mo1−1) is characterised by a much weaker inverse dependence of the rates on ΔGr. Modeling the experimental r-ΔGr data with a simple classical transition state theory (TST) law as implemented in most available geochemical codes is found inappropriate. An evaluation of the consequences of the use of geochemical codes where the r-ΔGr relation is based on basic TST was carried out and applied to carbonation reactions of diopside, which, among other reactions with Ca- and Mg-bearing minerals, are considered as a promising process for the solid state sequestration of CO2 over long time spans. In order to take into account the actual experimental r-ΔGr relation in the geochemical code that we used, a new module has been developed. It reveals a dramatic overestimation of the carbonation rate when using a TST-based geochemical code. This points out that simulations of water-rock-CO2 interactions performed with classical geochemical codes should be evaluated with great caution.