Describing the Effect of Alkalization in Localized Corrosion of Aluminum Alloys by Combining Finite Element Modeling, Model Al/Cu Bimetallic Systems and EQCM (Electrochemical Quartz Crystal Microbalance) Measurements. N. Murer 1 , N. A. Missert 2 , R. G. Buchheit 1 1 Fontana Corrosion Center, Ohio State University, 2041 College Road, Columbus, OH 43210 2 Sandia National Laboratories, 1515 Eubank SE, Albuquerque, NM 87123 In neutral, mildly aggressive, unbuffered media, local alkalization, originating from cathodic O 2 reduction on intermetallic particles (IMP), has been shown to occur during localized corrosion of Al alloys at the IMP/matrix interface 1 . The driving force for localized corrosion is believed to be a microgalvanic coupling between the cathodic particle and the surrounding anodic matrix. An accurate experimental proof of this microgalvanic coupling and a quantitative study of the effect of pH on local dissolution in Al alloys is experimentally difficult to achieve due to the geometrical and microstructural complexity of Al alloys and its passive nature 2 . Still, local electrochemical techniques and model alloys can be used to validate finite element models which can give a quantitative prediction of the pH evolution at a local scale 1,2 . Simulated pH profiles, obtained with a model based on the assumption of a cathodic O 2 diffusion control showed good agreement with the experimental pH profiles 1 . In this work, a bimetallic model system, composed of 5 engineered 50 µm diameter Cu islands deposited on five separated Al interconnects, has been designed as shown in Figure 1. This system, following the results shown in [3], allows the measurement of the net currents originating from each Al/Cu couples under free corrosion conditions in 5 and 10 mM NaCl. Based on the results obtained in [1] and [3], we want to use these coupling net currents to find the conditions for which the output of the model is physically consistent. The validated model is then used to describe the pH evolution with time at a local scale and its effect on the anodic dissolution of the Al matrix. EQCM is used to provide the relationship between pH and the Al anodic dissolution rate, so that the actual effect of pH on the corrosion current density can be described. The finite element modeling software that was used was COMSOL TM 3.4. The governing equation was the Nernst-Planck equation combined with the electroneutrality equation. The current distribution is due to the motion of charged species. The electrochemical dissolution of Al takes place at the anodic surface. At the cathodic surface, the four-electron reduction of O 2 occurs. EQCM experiments were done on 100 nm Al thin films deposited on the quartz crystals. A 10 nm copper layer was deposited on the Au-coated crystal before Al deposition. The solutions at different pH were prepared by using NaOH to adjust pH. The net current monitored for channel 1 is anodic and the other four net currents are cathodic. We consider that channel 1 is behaving as a single anode coupled with the four other cathodic channels. The geometry used in the model is consequently simplified. The current vs time curves exhibit an initiation/incubation passive regime and a propagation active regime during which pitting is occurring beneath the Cu islands. Using the experimental polarization curves of Al and Cu as boundary conditions of the model, we can reproduce the stationary experimental current distributions in the passive regime. This approach contains limitations, already tackled in [2]. In a similar way, if we assume a cathodic O 2 diffusion control of the coupling, the simulated current distribution is in agreement with the experimental current values in the propagation regime as already found in [1]. Using a time-dependent resolution, we show that the maximum pH on the anode at the end of the initiation step is 9.2. The EQCM data allows us to show that the simulated current density lies around 0.04 A.m -2 , which is ten times less than the simulated maximum anodic current density during the propagation step in a steady regime, assuming O 2 diffusion control. The role of alkalization on the onset of corrosion propagation seems to be very limited. This finite element model was used to describe the current and pH evolution occurring at the surface of a model bimetallic system composed of engineered Cu islands deposited on pure Al. After validation of the model, it was shown that under free corrosion conditions in near-neutral 5-10mM NaCl, the effect of alkalization on anodic Al cannot explain the transition from the passive state of corrosion to the active state. ACKNOWLEDGMENTS N. Murer would like to acknowledge the Fontana Corrosion Center and the Ohio State University for the financial support of this project. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the U.S. Dept. of Energy under Contract DE-AC04-94AL85000. Part of this work was carried out under the auspices of the U.S. Dept. of Energy, Division of Materials Sciences, Office of Science, under Contract DE-AC02-98CH1088. REFERENCES 1. J. O. Park, C. H. Paik, Y. H. Huang, R. C. Alkire, J. Electrochem. Soc., 146, 517 (1999). 2. N. Murer, R. Oltra, B. Vuillemin, O. Néel, J.Electrochem. Soc, submitted. 3. N. Missert, J. C. Barbour, R. G. Copeland, J. E. Mikkalson, JOM, 53, 7 (2001). Figure 1 : Engineered Cu Islands Deposited on Al Matrix.