Abstract:
Pitting Corrosion is usually considered to consist of two stages, an initiation process, followed by the growth of pits. This project was designed to study the initiation stage of the pitting of anodised aluminium.
Electrochemical techniques were used, and most measurements were made under potentiostatic conditions. Supplementary studies using potentiodynamic and galvanostatic techniques were also made. A novel electrode design, using a double-anodising procedure, permitted closely-controlled experiments without the problem of crevices. Scanning electron microscopy provided some information about the form of pitting at times shortly after initiation.
Pitting was studied in acidic chloride solutions, and the acidity of these media was characterised by indicator methods based on the Hammett acidity function HO. It was found that the salt effects could be explained by a simple hydration model.
The rate of pit initiation is inversely proportional to τ, the induction time for pitting, and the latter was used to follow the kinetics of the process. The irreproducibility of results was attributed to a substantial random component in the initiation process, and it was found that the results obtained by repeating an experiment many times fitted a log-Normal distribution. This indicates that mG, the geometric mean of the τ values, should be used to follow the behaviour of the system, and thus a set of results large enough for statistical analysis is required.
The effect of a number of experimental variables was studied. mG increases linearly with anodised film thickness, and also increases with the degree of sealing. mG decreases when the area is increased, but less than proportionately. The roughness of the pre-anodised surface affects mG, the rougher the surface the lower mG. mG decreases linearly with HO as the acidity increases. mG shows a first-order dependence on the chloride activity at low activities, but the order becomes zero at concentrations greater than 0.5 mol.l-1 chloride. mG is largely independent of the applied potential between -0.6 and +1.0 V(SHE) at high chloride concentrations and high acidity. Electrode rotation does not affect the rate of pit initiation.
Bromide causes pitting behaviour similar to that seen with chloride, although the rate of breakdown is somewhat slower. Fluoride also causes pitting but the attack is very rapid. All three halides tested show Arrhenius-law behaviour, and the apparent activation energies are Ea(Cl-)=69 kJ.mol-1, Ea(Br-)=56 kJ.mol-1, and Ea(F-)=17 kJ.mol-1. There were indications that CrO4- ion inhibits the propagation of pits, without affecting the initiation process. Alloys show a lower resistance to pit initiation than pure aluminium, and alloying elements seem to introduce cathodic sites into the anodised film.
A two-step mechanism for pit initiation is proposed. The rate-determining step is the dissolution of the surface by the film by H+ and Cl- ions. This is followed by some relatively rapid failure at localised points, which penetrates the rest of the film. A stress-cracking process, or penetration down a pore, are suggested for this second step.