Copper Corrosion Dynamics under Deep Geologic Repository Conditions
Abstract
Canada’s plan for the permanent disposal of spent nuclear fuel involves a multiple-barrier system contained within a deep geologic repository (DGR). The key engineered barrier is the used fuel container (UFC), which is a copper-coated carbon steel vessel. The environments to which the copper coating will be exposed in the DGR are very different from those commonly encountered in copper corrosion studies. The copper (Cu) will be in contact with thin stagnant layers of water, and the corrosion system will be exposed to a continuous flux of γ-radiation. In the presence of γ-radiation, water and humid air decompose to redox-active species such as H2O2 andHNO3, which make the corrosion system more complex. Therefore, predicting the corrosion rate of Cu in the DGR and over long timescales is challenging.
Corrosion involves many electrochemical and chemical reactions that are coupled with interfacial transfer and solution transport of metal cations. The initial soluble corrosion products will accumulate with time, and the reactions and transport of the initial and intermediate corrosion products will provide routes for developing strong systemic feedback, which can induce autocatalytic reaction cycles. Hence, the overall corrosion rate is not expected to have a simple dependence on solution parameters, as would be predicted using linear chemical dynamics. This makes it difficult to apply existing corrosion models to predict the corrosion rate of copper in the environments anticipated in the DGR. Consequently, in order to develop a corrosion model that can be used to predict the long-term integrity of the UFC with confidence, it is critical to identify and decouple the key elementary processes controlling the overall rate.
In this thesis, systematic studies were conducted to decouple the effects of solution parameters on the Cu corrosion dynamics. The parameters investigated were pH, oxidant concentration, type of anion (e.g., SO42- and NO3-), and solution layer thickness (dsol), in the presence and absence of γ-radiation. A combination of coupon exposure tests and electrochemical measurements, along with post-test analyses, were performed to extract the rate parameters required for a predictive model. In this work, the bulk solution properties (dissolved copper and pH change) and the average chemical behaviour in the interfacial region (morphology/composition of the corroded surface) were simultaneously measured in order to investigate the copper corrosion dynamics.
This work has demonstrated that Cu corrosion progresses through different dynamic stages, and the key rate-determining steps have been identified for each stage. Stage 1 involves the oxidation of Cu0(m) (bound to the bulk metal) to solvated Cu2+(solv) in the interfacial region, followed by transport of Cu2+(solv) from the interfacial region to the bulk solution. When the bulk concentration of Cu2+(solv) approaches the solubility limit of Cu(OH)2, transport of Cu2+(solv) is significantly impeded, and hydrolysis occurs, accelerating the formation and precipitation of a Cu(OH)2 hydrogel. In Stage 2, the reduction of CuII species (Cu2+(solv) and Cu(OH)2) to CuI species (Cu+(solv) and Cu(OH)) in the hydrogel can be coupled with the oxidation of Cu0. The CuI species thus produced in the hydrogel layer then precipitate and grow as Cu2O crystals. Cuprous oxide crystal growth occurs via a redox-assisted Ostwald ripening process via coupling of the redox reactions of the redox active species (H2O2 and NO3-/NO2- redox pair) with the Cu0, CuI and CuII species. In Stage 3, the growth of thermodynamically more stable cupric-hydroxide-anion (CuII-OH-X) salt crystals occurs at the expense of Cu2O crystals.
The corrosion mechanism proposed in this thesis can successfully explain the effect of solution parameters on the overall corrosion rate. The solution parameters affect the rates of progression through the individual stages and the overall corrosion rates within the individual stages. From these time-dependent studies, the metal oxidation rate parameters required for model development can be extracted. This study is a step towards developing a high-fidelity corrosion model.