Stochastic studies of passivity breakdown

Abstract

Passivity breakdown triggers the initiation of various types of localized corrosion, which exhibits a wide scatter of initiation time and/or site distribution even at a controlled condition. Examples of the statistical and probabilistic analysis of the scattered data of localized corrosion such as pitting corrosion, crevice corrosion and stress corrosion cracking are presented. It is emphasized that a common probability distribution law is an exponential distribution which is simply derived from a Poisson stochastic process. A birth and death stochastic model had been found to be most suitable for analyzing the distribution of induction time of pit generation as well as pitting potential. From the analysis of the distribution of induction time, the rate of pit generation and its repassivation could be easily decided as a function of applied potential. Pitting potential is defined as a potential at which the rate of pit generation is equal to the rate of pit repassivation. This model was justified by a Monte Carlo simulation which assumes independent stochastic events of pit generation and its repassivation, the probability of which is controlled by a Poisson process. The effect of the parameters of the model on the distribution of induction time as well as pitting potential could be evaluated by the simulation. Electrochemical noise is also simulated by the same birth and death stochastic model.