Predictions from a mathematical model of the propagation of a corrosion pit in copper are reported. The model uses equilibrium data for solid and aqueous species to calculate local chemical and electrochemical equilibria in small volume elements. Mass transport between elements under local internal equilibrium is calculated using aqueous diffusion coefficients with the constraint of electrical neutrality. Propagation of a corrosion pit is deemed possible when the fraction of the oxidised copper that forms solid corrosion products, at the copper metal, is insufficient to completely cover the underlying metal. The effect of pH and salt concentrations in the bulk water was studied by varying the composition of the water.

Results are presented in the form of E-log [Cτ] diagrams where E is the applied potential and [Cτ] is the total bulk concentration of chloride. The E-log [Cτ] diagram shows two separate areas where pitting is found to be possible. One region at low chloride concentration and high potential and one region at high chloride concentration and low potential. Increased sulphate concentration is found to be detrimental with respect to pitting corrosion, particularly in the high potential region. Increased carbonate concentration is found to be beneficial, particularly in the low potential region. Pitting corrosion of copper can be described as a case of galvanic corrosion where cuprous oxide at a pH similar to that of the bulk is the cathode material for oxygen reduction and copper metal at the local, lower pH in a corrosion pit may behave as the anode.