Defect formation and anode/cathode relationships in underfilm corrosion of heterogeneous aluminium alloys
Synthesis and Properties of Functional Materials
Final Report Abstract
The aim of the project was to study the influence of intermetallic constituent phases contained in heterogeneous aluminum alloys on the failure of organic coatings, especially with respect to the presence of local anodes and cathodes on the surface. The role of galvanic coupling on the corrosion initiation and propagation, the occurrence of local anodes and cathodes, the influence of corrosion products and the interdependency between corrosion and coating deterioration were to be studied. Methods to be applied were electrochemical techniques, especially electrochemical impedance spectroscopy, Confocal Laser Scanning Microscopy using also ion-sensitive fluorescent dyes incorporated into the coating, quartz crystal microbalance measurements for the study of coating properties, localized electrochemical impedance spectroscopy to delineate the usually complex impedance response seen in underfilm corrosion, and (surface) analytical techniques to study the role of corrosion products and copper replating. These studies were originally planned to be performed using a well-known, good commercial coating system. However, this coating system is ill defined, because especially the polyamide curing component shows a range of molecular masses and the supplier also does not publish the exact composition. Therefore the reviewers of the proposal demanded the use of well-defined model systems and the application of further techniques for the study of the metal-coating interface. Therefore Confocal Raman microscopy studies were envisaged to learn more about chemical changes at the interface. A number of different coatings was investigated for applicability in this study. All of them were based on variants of the diglycidyl ether of bisphenol A. The effect of different curing agents and solvent addition was studied. Characterisation was performed by recording of electrochemical impedance spectra as a function of immersion time. The best defined curing agents, diethylene triamine (DETA, M = 103,17 g/mol) and triethylene tetraamine (TETA, M = 146.2 g/mol), showed insufficient corrosion protection and had strong issues with phase separations. Polyetheramine based curing agents led to excellent corrosion protection, but showed non-wetting behavior for polar surfaces like quartz. Studies with the electrochemical quartz crystal microbalance technique (EQCM) therefore relied on the use of a commercial resin formulation. The EQCM analysis of coated quartz resonators showed very interesting results. These resonators have an Au electrode that does not corrode, so any change in the signal can be fully attributed to the influence of the coating. Already the presence of the cured, dry polymer film caused not only the expected decrease in the resonance frequency, but also an increase in the damping of the resonator. The frequency response observed during immersion in slightly acidic 0.5 M NaCl was compared with the capacitance changes measured in parallel by electrochemical impedance spectroscopy. At early immersion times, there was a good agreement between the amount of water calculated from electrochemical impedance data and from EQCM assuming a purely gravimetric response. Later the water uptake slowed down, but the measured frequency continued to decrease, leading to a poor agreement between EQCM and EIS data. This was explained by a change in the viscoelastic behavior of the polymer coating, leading to an increased damping and a non-gravimetric contribution to the frequency decrease, and the chemical deterioration of the polymer coating. In the CLSM using fluorescent light chemical information can be obtained in addition to the information from reflected light. The fluorescent light comes from the coating and is therefore ideal to monitor the surface of the delaminated coating at the blister side. The fluorescence of the dye is quenched at acidic pH values as they are typical for the occluded solutions in coating blisters on AA2024-T3. One therefore would expect a partial quenching of the fluorescence response at the blister side. However, the dye embedded in the outer part of the coating was as active at the blister side as at uncorroded places of the substrate. The fluorescent images showed some very bright spots that could be agglomerates of dye in the coating, and features like fluorescent streaks, rings, and other defects located at the same x-y-coordinates as a corrosion side within the blister, and proved to be valuable for the detection of corrosion defects. The microelectrochemical setup enables the measurement of the electrochemical properties of the coated alloys with local resolution. Electrochemical impedance spectra at different locations on coated alloy samples were performed. However, until now no progress beyond the state of the art has been accomplished applying this technique. Overall a large number of studies have been performed in order to identify the best coating for this project, but it turned out that no single coating seems to satisfy the requirements for all types of measurements and the aim of finding generally valid trends in corrosion failure of coated AA2024-T3 alloys. The equipment for the different types of measurements has been set up, and some interesting results have been obtained. Further studies would be needed in order to achieve the progress beyond the state of the art anticipated at the time of writing the proposal. In future work for different samples regular local electrochemical measurements using a fully automated protocol (for the measurement of a specific specimen) should be performed in order to systematically map the development of the coating properties. Locations identified as sites of potential breakdown then should be studied with CLSM in both fluorescent and topography mode at high resolution. Once local coating breakdown is taking place, studies of the interfacial properties with Raman microscopy to learn more about the nature of the corrosion products and the changes in the interaction between coating and substrate are envisaged. The statistical nature of the distribution of intermetallic compound phases makes the correlation between intermetallic compound phases and sites of corrosion initiation difficult. Here the use of microstructured synthetic alloys consisting of islands of Cu, Al2CuMg or Al2Cu embedded in an Al matrix might be used as samples to facilitate the correlation of corrosion site and original microstructure. The electrochemical quartz crystal microbalance technique can be expanded to include measurements at odd multiples of the fundamental resonance frequency in order to learn more about the property changes of the coatings. Possible applications of such results lie in defining the requirements for the development of improved coatings and alloys with improved microstructure. For both issues the mechanistic details of coating degradation and the influence of the microstructure would be very valuable.
