Current Measurements from Multiple Pitting Sites in Engineered Aluminum Electrodes

Abstract

Corrosion of aluminum in a system of engineered copper particles was studied in order to elucidate the nature of the interactions between multiple pitting sites. Single copper particles were fabricated on isolated, thin-film aluminum electrodes to enable independent measurements of excess anodic and cathodic currents flowing during open circuit exposures to dilute aqueous chloride solutions. In-situ fluorescence microscopy allowed imaging of local changes in solution pH during the exposure. Under the experimental conditions explored here, the anodic dissolution of aluminum beneath the majority of particles was balanced by cathodic reactions on the same electrode. However, excess anodic currents were also observed on some single electrodes due to accelerated dissolution of aluminum beneath the copper particles. These excess anodic currents were balanced by cathodic currents distributed between the other electrodes. Fluctuations in the excess anodic currents were correlated with fluctuations in the open circuit potential and with localized increases in solution pH above each of the copper particles. The transient nature of this accelerated dissolution suggests that it may evolve under anodic control.Localized corrosion at particles in alloys is common, 1-16 but the details of how anodic and cathodic currents and solution pH evolve in a system containing multiple pitting sites is unknown. Interactions between pitting sites at particles in alloys 17, 18 and in macroscopic wire electrodes 19-21 have been investigated. One set of studies showed that when pitting is driven by polarization of particular electrodes, aggressive solution chemistries can damage the oxide on nearby electrodes and influence pit initiation at these remote sites. 19,20 Under free corrosion conditions, pitting sites on isolated Al wires displayed fluctuations in potential and current and required cathodic current from remote sites. 21 The complex microstructure found in typical alloys makes it difficult to predict where pits will develop, and whether balancing cathodic reactions at remote sites are needed for pit stabilization. 10,14,15 Model alloys have been employed to investigate several aspects of pitting such as Cu redistribution, 22 galvanic coupling, 23-26 and alkalization. 22, 27-29 The random particle size and distribution in aluminum alloys results in particle densities and anode to cathode area ratios that are not independently controlled. Furthermore, real-time, simultaneous measurements of currents at multiple pitting sites are extremely difficult.In an effort to elucidate the mechanistic details of the pitting process, pit propagation has been investigated using five individual pure Al thin film electrodes, each containing a single engineered Cu particle, as a model system for Cu-rich particles in Al-Cu alloys. This approach allows independent control of some of the variables thought to control pitting behavior in real alloys, in particular, the initial particle/matrix composition, cathode/anode area ratio, and particle distribution...

Finite Element Modeling of the Galvanic Corrosion of Aluminum at Engineered Copper Particles

“…Cu engineered particles on matrix thin films.-A detailed description of the bimetallic system can be found elsewhere. 30 It was composed of five 50 μm diameter Cu islands on five 150 μm square Al thin film electrodes, where the films were electron beam evaporated onto SiO 2 coated Si wafers and the dimensions were defined by standard photolithography. The spacing between electrodes was 50 μm (Fig.…”

“…The solutions were prepared using de-ionized water and mixed in a 45:1 ratio with 0.1 mM fluorescein in isopropyl alcohol for further fluorescence microscopy imaging. 30 The experimental results are fully described and analyzed in the companion paper. 30 The useful results for the construction of the model are the following: r two stages were defined: a passive stage with low activity and an active (propagation) stage.…”

“…30 The experimental results are fully described and analyzed in the companion paper. 30 The useful results for the construction of the model are the following: r two stages were defined: a passive stage with low activity and an active (propagation) stage. The onset times for the propagation stage were: 800s at 5 mM NaCl and 60s at 10 mM NaCl, as seen in Figure 2.…”

“…This assumption is justified based on the experimental results, which suggest that the current flowing between the Cu and Al on a single channel is very small. 30 Second, each electrode was considered to be circular, allowing us to move from a 3D to a 2D geometry. The simplification to the circular geometry is justified since the anodic reaction on the dominant anode takes place beneath the circular Cu island, and the oxygen reduction reaction takes place on the circular Cu islands on the dominant cathodes.…”

“…Although a millimetric bimetallic system can be used as a starting point for validation, smaller length scales would provide a more realistic representation of the behavior of a cathodic IMP in a matrix. In this work, five individual 150 μm square Al thin film electrodes with a 50 μm diameter Cu island at each of their centers 30 were used as a bimetallic system. This length scale offers a better representation of IMPs embedded in a matrix, than the previously studied millimetric bimetallic system.…”