Application of Measurement Models for Interpretation of Impedance Spectra for Corrosion

摘要

Experimental identification of errors has allowed enhanced interpretation of a wide variety of spectroscopy techniques, including measurements for electrochemical systems for which detailed process models are available. The physicochemical models available for corrosion systems, however, are typically not sufficiently detailed to represent adequately impedance data to within the experimentally determined noise level of the measurement. Even in the absence of detailed physicochemical models, enhanced interpretation of corrosion processes is possible through the use of measurement models which can account quantitatively for differences in surface reactivities. In this paper, the necessary steps for the quantification of corrosion processes using the measurement model approach of Agarwal et al. are described. The stationary stochastic error contribution to impedance spectra is identified from replicated measurements. Presence of a bias error contribution, caused, for example, by instrumental artifacts and non-stationary behavior, is identified from application of the Kramers-Kronig transforms. The zero and high frequency limits obtained through the measurement model are used to determine the polarization resistance in a way that accounts for the experimentally determined error structure. The results of this procedure, weighted by an appropriate statistical analysis, can be used to monitor electrochemical systems as functions of time or process conditions. The approach described herein is validated for model systems, such as the reduction of ferricyanide on platinum for which accurate process models are available. The corrosion examples presented here involve the transient growth of corrosion-product films on copper in synthetic seawater and on cast iron in Evian water.