This presentation addresses questions about the possibility of localized corrosion on copper in aqueous solutions containing (singly or in combination) chloride, sulfate, bicarbonate and bisulfide ions. We used new approaches involving multielectrode arrays (up to 50 electrodes) and statistical models to investigate possible surface roughening during localized corrosion, the likelihood of initiating pitting corrosion, and the chance of repassivating pits that did initiate.We employed a combination of different electrochemical methods, including potentiodynamic and potentiostatic polarizations to identify localized (vs active or transpassive) corrosion, to map out ranges of conditions under which pitting might be possible, and to obtain key parameters, such as passivity breakdown potentials (Eb) and repassivation potentials (Erp). Electrochemical impedance spectroscopy was also used to characterize the behaviour of Cu in SH⁻-containing solutions. The morphology and composition of various types of surface films were investigated using scanning electron microscopy and X-ray photoelectron spectroscopy. Focused ion beam cross-sectioning combined with scanning electron microscopy was used to evaluate the morphology and thickness of the various surface films on Cu, as well as the distribution of corrosion damage at the Cu/film interface. With these data, we were able to evaluate the pitting probability and possibility of repassivation for Cu in the presence of different surface films, such as Cu2O, CuO, Cu(OH)2, and Cu2S.We found that, under the conditions tested, the uniformity of general corrosion processes is determined by the grain size of the material due to the preferential dissolution of certain crystal planes over others. We were able to map out a range of solution compositions (types of dissolved salts, their concentrations, and the pH) and temperatures where passive film formation was either possible or not possible and then demonstrate that corrosion potentials (Ecorr), Eb, and Erp are all distributed parameters whose distribution width and average value vary with the exposure environment (though these may indeed be deterministic quantities that are sensitive to uncontrollable fluctuations in key factors such as surface roughness or the number and types of exposed grain boundary inclusions, for example).We have also investigated the film growth mechanisms for Cu in SH⁻-containing solutions, both under natural corrosion conditions and with the Cu polarized artificially, and shown that Cu2S films were generally porous and non-passivating, with the rate of film growth determined by a combination of SH⁻ transport through the growing film and the competition between SH⁻ and Cl⁻ for adsorption sites on the reacting Cu surface. Scanning electron microscopy (SEM) on corroded surfaces and cross-sections showed no evidence for pitting when only Cu2S films were present. Surface analyses demonstrated that, depending on the concentrations of SH⁻ and Cl⁻, the Cu2S films were composed of either one or two layers; a thin base layer and an outer crystalline deposit. Cyclic voltammetry on rotating disk electrodes showed that unless relatively high SH⁻ concentrations and mass-transport rates can be achieved, or a stronger oxidant than SH⁻ is present, classical pitting corrosion does not occur on Cu under truly anoxic conditions. In anoxic solution with low sulphide concentrations, the corrosion process is controlled by the supply of SH⁻ to the copper surface, resulting in the formation of porous Cu2S films on the Cu. A concern remains, however, that in solutions containing both bisulfide and trace amounts of oxygen (> micro-molar concentrations), some intergranular corrosion, enabled by the presence of oxy-sulfur compounds, might occur. This possibility remains under investigation.
Read full abstract
Dec 1, 2024
Jie Wu + 6
Open Access