Résumé
Hydrothermal exchanges between fluids (seawater, hydrothermal and mantle-derived CO2-rich fluids) and mantle-dominated lithosphere at ridges represent a major means of mass exchange between the mantle and the Earth's envelopes. It results from/in a suite of strongly interconnected thermal, mineralogical, and (bio-) chemical processes within the seafloor, that contributes to the Earth's carbon budget through carbonation reactions (dissolution of silicates and precipitation of carbonates) and the production of complex carbon compounds. Recent experimental studies were realized to constrain the chemical parameters controlling these reactions (pH, T, kinetics, catalysis), one of the main applications being the development of techniques for geological storage of excess atmospheric CO2 into ultramafic ophiolitic massifs and in basaltic reservoirs. However, although these reactions will occur only if fluids can flow through the reacting rocks, the mechanisms controlling the hydrodynamic properties of the system during hydration and carbonation reactions remain poorly constrained. We present the results of three sets of percolation-reaction experiments during which CO2-depleted to CO2-saturated fluids were injected into olivine cores, dunitic samples and magmatic olivine with trace basalts. These experiments allowed us to explore the role of the initial mineralogy and structure of rocks, the composition of fluids (pCO2...) and flow rate and their effects on hydrodynamic properties and carbonation/hydration efficiency. The experiments produced broadly similar results, i.e. dissolution of olivine (+ or - mafic minerals and glass when present), precipitation of carbonates and of serpentine type minerals and a steady decrease in permeability with time. In details, differences in the compositions of the fluids and of the reaction products were observed from one set of experiments to the other as well as in the rate at which chemical and permeability changes occurred during the experiments: (i) during the earliest stages of the reactions, Ca-Mg-(Fe-)rich carbonates precipitate in low flow zones while Si-rich layer develop at olivine surface in higher flow zones, which provide a transient mechanism maintaining constant permeability; (ii) porosity changes are small and permeability decrease is triggered mainly by the precipitation of serpentine-type minerals at the expense of olivine, in low flow zones and/or downstream; (iii) carbonation efficiency at the scale of porous samples is controlled to the first order by feedback effects between flow rate, local pCO2 and the kinetics of (catalyzed) hydration reactions. These experiments show that reactions are controlled at pore scale, which results in the development of chemical microenvironments and of mineralogical, chemical and hydrodynamic heterogeneities at the scale of the samples. The spatial distribution of mixing, triggered by the variability of the velocity field in pores, controls the occurrence and the rate of reactions and must be taken into account to model the efficiency and the sustainability of carbonation in these environments.