(Corrosion Division Morris Cohen Graduate Student Award Address) Galvanic Corrosion of Copper-Coated Carbon Steel for Used Nuclear Fuel Containers

Abstract

Carbon steel vessels coated with ∼3 mm of Cu have been proposed for the permanent disposal of used nuclear fuel in a deep geological repository (DGR) in Canada. In the event that a container is emplaced in the DGR with an undetected defect in the Cu coating that exposes the steel substrate, galvanically accelerated corrosion of steel is, in principle, possible. To investigate this scenario, the progression of steel corrosion at the base of novel simulated through-coating defects was monitored electrochemically and imaged non-destructively using X-ray micro-computed tomography (micro-CT) as a function of time, O2 availability (including anoxic conditions), and coating method (cold spray deposition (with and without heat treatment) and electrodeposition). The corrosion products and surface damage were analyzed using Raman spectroscopy and scanning electron microscopy (SEM)/energy dispersive X-ray spectroscopy (EDX). These analyses showed how the corrosion damage to steel evolved over time and how it was affected by the method used to coat steel with Cu and the amount of O2 available. The results showed that steel exposed at the base of a through-Cu coating defect corroded and became covered by corrosion products, while there was no visible loss of Cu. The supply of O2 to the sample surface governed the corrosion rate, while the distribution of damage to steel at the base of the defect depended on the Cu coating method and the resulting quality of the Cu/steel interface. Cold spray Cu-coated steel specimens exhibited a radial spread of corrosion along the Cu/steel interface, while electrodeposited Cu/steel specimens experienced preferential interfacial corrosion in the direction in which the steel substrate was machined prior to the electrodeposition of Cu. An example of the progression of corrosion observed by micro-CT is shown in Figure 1. Less extensive corrosion along the Cu/steel interface was observed on samples that were shown by electron backscatter diffraction (EBSD) and adhesion tests to have a less stressed, more uniform, and more well-adhered interface. However, less extensive corrosion was typically at the expense of deeper penetration into the steel. In the absence of O2, the quality of the Cu/steel interface greatly affected both the overall amount of corrosion damage to steel and the distribution of damage. The corrosion rates were significantly lower under anoxic conditions than under oxic conditions, and tended to decrease over time. The influence of a wide range of cathode:anode area ratios and Cl− concentrations, and the availability of O2, was evaluated by monitoring the galvanic current passing between separate Cu and steel electrodes, connected through a zero-resistance ammeter (ZRA), and the galvanic potential of the couple. The galvanic corrosion of steel was most severe when it was exposed to air-sparged solution with a moderate [Cl−] as part of the couple with the largest Cu:steel area ratio. Figure 1