Pitting corrosion initiation is a poorly understood phenomenon. Understanding of the factors which lead to the initiation of a pit at a specific time and location has been limited by restrictions in temporal and special resolution, as well as environmental restrictions for high resolution techniques. With new advances in liquid cell transmission electron microscopy (LCTEM), an electrochemical cell specifically designed for corrosion studies directly in the TEM has been developed. The cell, consisting of a thin-film of flowing electrolyte in a TEM sample holder, allows in-situ observations at the nanoscale. In this work, the corrosion behavior of pure aluminum as observed in the LCTEM is compared and contrasted to more conventional laboratory scale benchtop observations. The cell has been instrumented such that traditional three electrode electrochemistry experiments can be conducted directly in the microscope. The electrochemistry of localized corrosion behavior of aluminum in LCTEM experiments in saline solutions of 0.001 M NaCl will be shown in this presentation. There are some observable but explainable differences in in situ TEM experiments and more conventional benchtop experiments of aluminum foils. Similar to more conventional electrochemical studies, metastable pitting behavior and pit initiation were observed at potentials active to the zero current potential.

An interesting observation is that micron-scale pits were initiated under the electron beam, with pits tens-to-hundreds of nanometers in scale forming in areas away from the beam. The electron beam induced micron-scale pits initiate as dense clusters of hundred-nanometer scale pits separated by hundreds of nanometers in the irradiated region. The extent of irradiation induced localized corrosion is a function of both electrochemical polarization and beam current. The application of noble potentials mitigates the irradiation induced pitting morphology. The beam energy does not have a substantial impact on the rate of beam-related diffraction contrast loss (perforation of the thin foil), suggesting that the mechanism of irradiation damage is a radiolysis reaction in the electrolyte, generated by a secondary electron cascade, rather than a high-energy displacement damage to the metal caused by the primary electron beam. Figure 1 shows the morphologies of localized attack both under irradiated and unirradiated conditions.

This study has also shown that the geometry of the microfluidic cell can have significant effects on the electrochemical behavior of the aluminum films contained in the cells. Polarization experiments were performed on an aluminum foil in the microfluidic cell, a similar aluminum film tested in a benchtop experiment. The polarization curves for each of these conditions are shown in Figure 2. There is a significant shift in the zero-current potential between thin films of Al in the cell and out of the cell, although the breakdown potentials are essentially the same. This shift in the open circuit potential can be ascribed to the narrow electrolytic channel in the TEM cell effectively acting as an artificial crevice, polarizing the metal in the noble direction and effectively acting as a pre-formed pit. A second consideration is that the solution is not deaerated. The aluminum and other components in the cell deaerated the solution in the cell and the spacer chip creates an occluded environment in the solution channel across the sample, thus further shifting the corrosion potential in the noble direction. Variations in OCP between samples in the microfluidic cell are within a +/-100 mV range and show no systematic shift due to beam conditions. The nominal pitting potentials are similar for benchtop and TEM measurements although the open circuit potential differ by several hundred millivolts.

The results of this investigation have provided some new insights regarding pit initiation on electrochemically passive surfaces exposed to chloride solutions, and shown that microfluidic cells directly incorporated into transmission electron microscopic stages can be useful tools in studying corrosion at the nanoscience level. This research has also shown that the presence of a focused electron beam can have significant effects on the onset and morphology of localized corrosion damage in passive metals.

The authors acknowledge NSF (DMR-1309509) for funding this work, as well as Frances Ross (IBM- T.J. Watson Research) for thoughtful discussions, and Ray Dove (RPI) for support on the electron microscope.