Abstract:
A long-time goal of the aluminium smelting industry has been to develop a material that could be deployed as a non-consumable anode. Anodes based on a metallic monolith have several potential advantages due to the good electrical and mechanical properties of metals. The principle challenge for the development of metallic anodes is to prevent their corrosion in molten cryolite. This may be achieved by using alloys that form self-passivating oxide films. The film should physically separate the metal monolith from the electrolyte. All oxides have a degree of solubility in cryolite, so the film must also form in situ at a rate approximately equal to its rate of dissolution. Alloys that form a thin aluminium oxide film are of interest because the inevitable dissolution of the oxide will not contaminate the aluminium product. The key to developing metallic anodes lies in understanding the mechanisms of film formation and dissolution that occur during service. This study focused on identifying the films that formed on candidate alloys under anodic potential in cryolite melts, the effectiveness of these films at preventing corrosion of the alloy substrate, and the mechanisms by which the substrates corroded. Three aluminium bronze alloys and a nickel aluminde were tested, both with and without pre-formed alumina films. All these materials corroded to varying degrees. The corrosion behaviour was determined by the oxygen activity in the surface region, which was a function of current density. In this regard, the effect of current density during electrolysis is analogous to that of oxygen partial pressure during high temperature oxidation in a gaseous environment. Pre-treatment of the nickel aluminide resulted in a thin alumina film, with poor surface coverage. This film did not prevent oxidation of the Ni substrate, and the corrosion behaviour was not markedly different between nickel aluminide anodes without pre-treatment. All anodes formed thick NiO scales, which were electrically insulating, non-passivating, and mechanically inadequate. The NiO scale propagated through a sequence of growth and cracking, which allowed the electrolyte to make contact with exposed metal. Growth of this scale resulted in an unacceptably rapid rise in the cell voltage, and poor electrochemical performance. Pre-treatment of the aluminium bronze alloys produced alumina scales that were thick, porous, and poorly adherent. These scales did not form a physical barrier between the Al-depleted metal substrate and the electrolyte. As a result, the Cu-metal behind the alumina scale was readily oxidised. Untreated aluminium bronze alloys formed an Al-based oxide surface film when the oxygen activity at the anode was relatively low; at operating current densities of 0.5A.cm-2 of less. Low current densities provided low oxygen diffusion rates, which could be matched by the temperature dependent surface diffusion of aluminium. Under these conditions, the surface concentration of aluminium was sufficient for preferential formation of Al-based oxides. The particular oxide that formed varied by alloy; Cu-Al formed Al2O3 or CuA1O2, and the alloys containing Fe and/or Ni formed spinels based on (Ni,Fe)AhO4. These films were not sufficiently passivating, and the underlying Cu-metal was again readily oxidised. The principle aluminium bronze corrosion product was Cu2O. Growth of the Cu2O was an electrochemical process and was therefore proportional to current density. The anode electrical resistance did not increase with the scale thickness due to the low resistivity of Cu2 O. The Cu2 O propagated into the metal, driven by oxygen migration. Dynamic concentration gradients were established for oxygen and aluminium within the surface region, and resulted in internally oxidised bands of alumina beneath the Cu2O scale. These bands also failed to halt the substrate corrosion. The untreated aluminium bronze anodes showed that alumina can form in situ, but also that this process alone is insufficient to protect the metal. Future anode success may be built upon a dense pre-formed alumina coating in combination with· the demonstrated Al migration and in situ alumina formation.