Thermodynamic probability analysis of the effects of Rb on the corrosion susceptibility of Cr-containing steels for nuclear materials canisters

Rubidium (Rb) generated from the β-decay of Kr-85 has been theorized to be corrosive toward steel, specifically in the storage of Kr-85 nuclear waste streams. In the present study, the phase equilibria of RbxCryOz oxides with Rb in dry oxygen and water are investigated to understand a possible pathway to unusual deterioration of the corrosion resistance of canister steels in the presence of Rb. It was found that, in dry oxygen environments, the accumulation of Rb (more than 0.01 mol) can completely consume the Cr in 1 mol of AISI 4130 steel by forming -Rb2CrO4 and Rb3CrO4 and prevent the formation of protective Cr2O3 scale. In aqueous environments, RbxCryOz oxides are metastable species. In order to investigate their role, the probability of forming various oxides is invoked in order to avoid the all-or-nothing approach to oxide formation typical of E-pH diagram, which only predicts the most stable species dissolved, ionized or solid ionized. Thus, the probability of forming RbxCryOz was considered and reported herein. It was found RbxCryOz can possess a larger than 7% probability of forming over Cr2O3 in Rb rich case and 15% in Cr rich case, indicating that it is expected to find small amount of RbxCryOz in the thermodynamically formed reaction products. Even though Cr2O3 is more stable than RbxCryOz oxides, the protective Cr2O3 scale is likely to have some vulnerability to Rb, leading to one possible route for the decline in the corrosion resistance of steel canisters in aqueous environments. Therefore, from a thermodynamic perspective, the current study supports the hypothesis that Rb can thermodynamically react with Cr in steels and can lead to formation of RbxCryOz at certain potentials and pH levels, showing the Rb influence of steel corrosion cannot be discounted. The paper considers experimental mixed potential and pH levels observed and relationship to thermodynamic probability. From this relative corrosion resistance can be assessed in a preliminary way in aqueous environments.