Anodic Activation of Aluminum by Trace Element Tin
MetadataShow full item record
Anodic activation of commercial and model aluminum alloys in chloride solution became of practical importance in connection with filiform corrosion of painted aluminum sheet in architectural application and aluminum components of brazed heat exchangers. Activation in chloride solution manifests itself in the form of a significant negative shift in the pitting potential relative to pure aluminum and a significant increase in the anodic current output at potentials where aluminum is normally expected to be passive. This activation is caused by trace elements in Group IIIA-VA in aluminum alloy, mainly Pb, Bi, Ga, In, and Sn, with low melting points. For alloys heat treated at 600°C, which is a relevant temperature, e.g., for brazing, segregation of Pb, a common trace element in most commercial alloys, as a nanofilm between the thermal oxide and the metal substrate was identified as the main cause of anodic activation. This nanofilm was independent of Pb concentration in the alloy and could only form by heat treatment at 600°C. However, certain recycled commercial alloys were activated by annealing at temperatures significantly lower than 600°C, e.g., 300°C, at which Pb did not segregate to the surface because of higher melting point and lower mobility. The effect of other trace elements with lower melting points than Pb, such as Sn, was suggested to be taken into consideration. The purpose of this thesis is to investigate surface segregation of tin, present at small concentrations in pure aluminum, by heat treatment and chemical etching and resulting activation of the alloy surface. The study focuses on the effect of Sn concentration in the alloy, heat treatment temperature and time, the method of cooling after heat treatment, oxidation during heat treatment and cooling, the nature of segregation and/or enrichment at the surface as a result of heat treatment (thermal segregation) and by dealloying during subsequent anodic polarization in chloride solution (anodic segregation), the role of chloride ions in the solution, and the electrochemical mechanism of activation. Model binary alloys, containing 30, 100, 500, and 1000 ppm Sn, denoted as AlSn30, AlSn100, AlSn500, and AlSn1000, respectively, were prepared by casting from pure components. The samples were prepared by scalping, homogenization, cold rolling and metallographic polishing of the cast alloys. The specimens were then heat treated in an air circulating furnace at various temperatures between 100 and 600°C for 1 h. After annealing, the samples were quenched in water or in 96% ethanol solution, or they were cooled in air for comparison. Most samples were characterized electrochemically in this condition. The AlSn1000 specimens, which formed a thick oxide layer on the surface as a result of this treatment, were immersed in a standardized hot chromic-phosphoric acid stripping bath (ASTM G1-90) to dissolve the oxide without damaging the underlying metallic surface and to obtain information about the properties of the substrate underneath the oxide. Some AlSn30 samples were mechanically polished after heat treatment and re-annealed at selected temperatures. Selected AlSn30 samples were also etched in 10 wt% NaOH for 10 s at 60°C and then de-smutted for 1 min in concentrated nitric acid at 25°C before electrochemical characterization. The highest degree of thermal segregation of Sn to the metal-oxide interface occurred by annealing of the model binary AlSn alloys at 300°C. Optimal segregation of Sn at this temperature resulted from a balance between high mobility of Sn in the liquid state and low solubility of Sn in aluminum. Segregation of Sn was reduced at lower and higher temperatures as a result of decreased mobility of Sn and increased solubility of Sn in Al, respectively. Thermal segregation of Sn at 300°C occurred in the form of nanoparticles and enrichment along certain crystallographic planes at the metal-oxide interface. The latter form was responsible for activation of aluminum alloy in chloride solution and independent of the bulk Sn concentration. Anodic segregation of Sn also occurred in chloride solution with increasing Sn concentration in the bulk, and these segregations also contributed to the activation of the AlSn alloy surface. Without heat treatment, tin concentration had an insignificant effect on the electrochemical behavior up to 500 ppm. However, alloy containing 1000 ppm Sn was significantly activated by lowering the pitting potential to -1.38 VSCE in chloride solution. Anodic activation of AlSn alloys in chloride solution, obtained by annealing at 300°C, was however temporary. Corrosion was significantly reduced to near passivity as the segregated Sn was etched away from the surface, to a level that was independent of Sn concentration. This was caused by segregation of most bulk Sn to the surface during heat treatment, reducing the bulk Sn concentration to the solid solubility level of Sn with aluminum at 300°C. Corrosion of samples annealed at 300°C occurred in the form of uniform etching from a macroscopic viewpoint, in the active potential region. Microscopically, corrosion was in the form of needle-shaped attack, undermining the oxide and following certain crystallographic planes by a mechanism similar to filiform corrosion. This directional activation of the surface was attributed to enrichment of Sn along preferred crystallographic planes near the surface. Annealing at 600°C caused increasing activation with increasing Sn concentration. Activation occurred as a result of Sn enrichment at the metal surface by dealloying of aluminum during corrosion (anodic segregation). Corrosion was localized in the form of grain boundary corrosion for Sn concentration less than and equal to 500 ppm and pitting following the triple grain boundaries for 1000 ppm. Both annealing time and cooling rate had insignificant effect on the polarization behavior of AlSn30 alloys in chloride solution, because the amount and state of tin segregation at the metal-oxide interface was not affected. In contrast, mechanical polishing removed any segregated tin layer and nearly eliminated activation of aluminum. Alkaline etching increased activation of AlSn30 alloys by causing anodic segregation of Sn. A thick oxide layer was formed on 300°C-annealed AlSn500 and AlSn1000, as well as 600°C-annealed AlSn1000 samples during water quenching, demonstrating that Sn at these levels activated Al in chloride-free water/steam at a high temperature. For samples treated at 300°C, activation of alloys AlSn500 and AlSn1000 was caused mainly by Sn that was segregated during heat treatment. For AlSn1000 samples annealed at 600°C, activation was caused by Sn segregating in situ by dealloying during water quenching. Activation during water quenching was related to the combined presence of segregated liquid Sn and super-heated steam at the surface. This thick oxide layer could act as a barrier layer and introduce partial passivation during polarization in chloride solution. The activation mechanism of Al by Sn in chloride solution and during water quenching was attributed to the amalgamation theory, which is originally based on the activation of aluminum by Hg at room temperature and also attributed to the activation of aluminum by low melting point elements Ga and In in the literature. The present work showed that Sn is a more effective activator than indium under identical conditions, although Sn has a higher melting point. The observed difference between Sn and In is difficult to explain in view of the close proximity of the two elements in the periodic table, especially the fact that the presence of In does not give rise to voluminous oxidation of Al during quenching. Solubility of aluminum in the liquid film forming at the metal-oxide interface is suggested as an important factor determining the rate of aluminum oxidation according to the amalgamation theory. The fact that the solubility of Al in liquid Sn is significantly higher than that in liquid In may explain the activating capability of Sn. Another possible factor is the melting point depression caused by the size of segregated film and particles of these elements activating aluminum, which is envisaged as an important factor for maintaining the fluidized state of the film down to room temperature. It is argued that Sn can undergo a more significant depression of its melting (or solidification) point with reduction in size compared to In.