Copper is an important candidate material considered as corrosion barrier for used nuclear fuel containers (UFC) (1). The current Canadian UFC design involves a Cu-coated carbon steel vessel. In the Deep Geological Repository (DGR), the Cu coating will be exposed to a continuous flux of g-radiation emitted from the fission products trapped in the spent fuel matrix and also to water trapped in the gap between a clay buffer material and the UFC. The water volume per unit surface area will be small and under stagnant condition (2). In the presence of g-radiation, condensed water on copper surface decomposes and forms redox active species such as H2O2 (3). Predicting the corrosion rate of copper in DGR environments, and at timescales, that are well beyond those that can be easily tested in a laboratory is very challenging. Existing copper corrosion models ignore the interfacial solution reactions and cannot explain many of the observed relationships between corrosion results and solution parameters. This makes it very difficult to apply them to the Cu corrosion in the DGR environments that can vary as corrosion progresses. The work presented here was performed to develop a high-fidelity model that can predict the long-term evolution of the Cu corrosion rate as a function of solution parameters.In this work systematic studies are being performed to decouple the effect of solution parameters on copper corrosion. We focus on establishing the relation between the overall corrosion rate and the rate of dissolution and oxide formation by monitoring the solution and surface changes. Corrosion in solutions containing different anions was investigated but only the results from the tests performed in 5 mM sulphuric solution is presented here. The effect of water volume to surface area ratio, and solution pH were studied by exposing Cu coupons to the test solution in the presence and absence of g-radiation and monitoring the surface and solution as a function of time.The changes in pH and the amount of copper dissolved in the solution was measured during corrosion time. Surface and cross section analyses were performed to investigate the morphology and the composition of oxide formed on the surface of coupons corroded for different times. This work focuses on establishing the relation between the overall corrosion rate and the rate of dissolution and oxide formation.Our study shows that copper corrosion progresses through different stages in which different elementary processes determine the overall corrosion rate. At all tested conditions, the time-dependent behaviour of the corrosion shows three characteristic kinetic stages. During each stage, the same trend of copper dissolution, pH changes, and oxide evolution is followed. The result from one of studied solution depths (2.5 cm) in the absence of g-radiation is presented in figure 1. The rate of corrosion in stage I, has a linear trend and is limited by the dissolution of copper cations and is independent of solution depth. The second stage starts when quasi-equilibrium is established near the surface and this condition assist Ostwald ripening of oxides formed on the surface. In stage III, the dissolved copper species precipitate from saturated bulk solution and form Cu(I) and/or Cu(II) oxides depending on the stability of different oxides. Any solution paramter including the initial and evolving parameters may affect the duration of each kintetic stage. The results of radiation experiments are in a good agreement with our purposed mechanism. The radiation products may affect the equilibrium reactions in the second stage of corrosion.This study clearly demonstrates that in a small stagnant water volume, the concentration of metal cations released due to corrosion can be high even for a corrosion resistant material like Cu. The solution reactions and transport of dissolved metal cations occur at rates that can strongly couple with electrochemical metal oxidation and precipitation of oxide deposits. In the presence of such strong systemic feedback, the effects of different solution parameters on the overall corrosion rate cannot be evaluated based on linear chemical dynamics.
Read full abstract