Clay host rock and engineered barrier systems are the key elements in the concept adopted by several countries to isolate the high level nuclear waste from the biosphere. Mainly composed by illite, mixed-layer illite–smectite (I/S) and montmorillonite, these clays are characterized by their properties of high retention and low permeability for released radionuclides. After closure of the repository in deep geological formation, groundwater in equilibrium with the host rock, forming the pore water, will saturate the engineered barrier system and come in contact with the nuclear glass waste package. The leaching of the different silicate minerals as well as the uptake of dissolved silicic acid by these phases is an important factor in the dissolution of the nuclear glass waste. To understand how the Si interacts with clay materials, this study aims at evaluating the solubility as well as the dynamics of solid/solution exchange reactions, which control the dissolved silicic acid concentration in solution in contact with Callovo-Oxfordian claystone. The results were compared with the dissolution of illite, bentonite and quartz at 25, 50 and 90 °C in pore water. The experiments were conducted in batch system and in controlled atmosphere conditions and were followed in continue until the equilibrium between the solid and solution is reached. The results present a stabilization of the pH-values at 8.2 after 211 days for all the samples. The nature of solids and the temperature were not the determining factors on the pH-value but the chemical composition of the pore water and the working atmosphere (here, nitrogen). Dissolution and precipitation rates were calculated from the concentration of Si released from solids and the activity of 32Si-radiotracer added in solution, respectively. Dissolution rates were in the range of (8.7 ± 0.4) × 10 −12–(6.8 ± 0.3) × 10 −11 mol Si/m 2/s for quartz, (1.6 ± 0.1) × 10 −13–(6.4 ± 0.3) × 10 −13 mol Si/m 2/s for Callovo-Oxfordian claystone, (2.4 ± 0.1) × 10 −13–(9.4 ± 0.5) × 10 −13 mol Si/m 2/s for illite du Puy and (1.2 ± 0.6) × 10 −12–(9.1 ± 0.5) × 10 −12 mol Si/m 2/s for bentonite. Precipitation rates were in the range of (8.4 ± 0.4) × 10 −12–(2.0 ± 0.1) × 10 −11 mol Si/m 2/s for quartz, (2.0 ± 0.1) × 10 −13–(2.5 ± 0.1) × 10 −12 for Callovo-Oxfordian claystone, (2.4 ± 0.1) × 10 −12–(5.2 ± 0.3) × 10 −12 mol Si/m 2/s for illite and (1.9 ± 0.1) × 10 −13–(6.9 ± 0.3) × 10 −13 mol Si/m 2/s for MX80 bentonite. The plot of dissolution rates versus precipitation rates gave a slope near one indicating a dynamic process of dissolution/precipitation. Through the experiment with the addition of a spike of 32Si-radiotracer in solution, we have showed that the activity of 32Si decreases in contact with clay containing at least 6% of Fe and Al, as for example illite, which seems to be a necessary condition. Finally, in the condition of nuclear glass waste disposal, the Callovo-Oxfordian claystone would be the phase controlling the solubility of dissolved Si from the nuclear waste and clay with a solubility value of 4.8 × 10 −4 mol/L at 90 °C in pore water of Bure site.