Towards a Physical Description for the Origin of Enhanced Catalytic Activity of Corroding Magnesium Surfaces
Introduction
The corrosion of magnesium (Mg) and its alloys remains a practical barrier to their wider industrial application. As such, unambiguously understanding Mg corrosion is increasingly topical given the demand for product weight reduction across several industries, including automotive, consumer electronics, and aerospace. There has been both mystery and debate regarding the origin of the so-called “negative difference effect” (NDE), which involves an increase in the rate of hydrogen production with increasing applied anodic potential, for over a century [1]. This phenomenon is readily observed on Mg, where the principal cathodic reaction upon Mg is water reduction (2H2O + 2e− → H2 + 2OH−). The work of Petty et al. in the 1950s [2] was amongst the first to suggest that Mg may be corroded as univalent Mg+ ions. In the corrosion field, a popular theory that hydrogen evolution on corroding Mg surfaces arises from a secondary chemical reaction between the univalent Mg+ ion and water to form excess hydrogen gas has been developed in recent decades [3], [4], [5], [6], [7]. It is prudent to note that the study of Petty et al. did not include any experimental data, nor has any researcher isolated or confirmed the existence of univalent Mg+ ions by spectroscopic [8], [9] or any other means to date. Alternative theories put forth to explain the NDE effect include, Mg particle undermining [10], [11], [12], [13], [14], formation and dissolution of magnesium hydride (MgH2) [15], [16], and chemical equilibrium [17].
In a recent study by Frankel et al. [18], a mechanism to account for hydrogen evolution on Mg during anodic polarization was described in terms of enhanced catalytic (i.e. cathodic) activity on the corroding Mg surface. For example, if the ability to support water reduction is enhanced during corrosion, then the cathodic reaction may still appreciably persist at potentials anodic relative to the corrosion potential of Mg. Such a mechanism would indeed account for the so-called “negative difference effect”; however, a physical basis for the origin of such a mechanism has yet to be presented. In this vein, Williams and co-workers [19], [20] have revealed via the scanning vibrating electrode technique (SVET) that corroded regions of the Mg surface are subsequently ‘cathodically activated’ and take the role of a cathode during the continued corrosion of Mg. The SVET studies are important because they demonstrate that cathodic sites on anodically polarized Mg are spatially associated with prior corrosion. It has been postulated that the sites of cathodic activity may be sites of relatively noble impurity metals [19], [21], based on the foundational findings of McNulty and Hanawalt [22]. Given further consideration, it may be reasonable to assume accumulation of metal impurities that are more noble than Mg (such as Fe) as they would remain cathodically polarized at potentials where Mg is anodically polarized. As such, preferential anodic corrosion of Mg could reasonably occur, akin to de-alloying, albeit that the Mg is the significant major constituent of the alloy.
To date enrichment of other metals at the surface of Mg has not been definitively shown, likely due to the combination of it not being regarded as a significant issue and that it was simply difficult to discern. For example, with respect to the latter issue, conventional scanning electron microscopy (SEM) based energy dispersive spectroscopy (EDS) is not sensitive to the near surface, and other techniques like Auger electron spectroscopy (AES) or X-ray photoelectron spectroscopy (XPS) may lack the required signal to noise ratio or spatial resolution to detect minor enrichment. It is noted that typical impurity levels in Mg are at the ppm amount (such that the sum of impurities may be ∼200 ppm), and even a doubling or tripling of the impurity level would not be significant in terms of signal to noise ratio unless the experiment was carefully executed. However, recent work has shown that focused ion beam (FIB) prepared cross-section lamellae can be studied using transmission electron microscopy (TEM) to reveal the morphology and local composition of the surface of Mg-alloys following exposure to an electrolyte [23], [24], [25], [26], and it was more recently revealed that localized enrichment of metallic Al could be detected in an Mg-Al alloy using this technique [27]. As such, the FIB preparation method for TEM investigation was used to survey the morphology, structure, and chemistry of the darkened surface of pure Mg following anodic polarization. The aim was to provide a physical description of the surface and reveal whether or not noble metal enrichment is present.
Section snippets
Experimental
Square samples (10 mm × 10 mm) were prepared from 2 mm thick polycrystalline cold-rolled Mg sheet (99.99% nominal purity with 50 ppm Fe). The samples were annealed for 350 °C for 0.5 h in air and quenched in water to obtain a recrystallized grain structure. Each working electrode was prepared by attaching a coated copper (Cu) wire to the rear of the Mg sample and then cold mounted in epoxy resin. The working surface was mechanically abraded using silicon carbide (SiC) paper up to a 4000 grit surface
Composition and Structure of Corroded Surface
An image of the macroscopic visual appearance of the Mg electrode surface after conditioning at the open circuit-potential for 24 h is shown before [Figure 1(a)], during [Figure 1(b)] and after [Figure 1(c)] anodic polarization at -1.0 VSCE for 0.5 h. Anodic polarization was observed to initiate and sustain dark regions of localized corrosion on the Mg surface, upon which hydrogen gas evolved. Away from the edges, these dark corroded regions were radial-like in appearance, which is similar to
Conclusions
- 1.
Transmission electron microscopy (TEM) of focused ion beam (FIB) prepared cross-section lamellae was used to successfully characterize the structure and composition of the dark coloured film formed on an area of corroded Mg where hydrogen gas evolved.
- 2.
The film was found to be a bi-layered structure consisting of columnar crystalline mixed MgO-Mg(OH)2 outer layer on top of a crystalline MgO inner layer. EDS determined that Fe-rich particles were embedded in the columnar outer layer. The
Acknowledgements
This work was financially supported by General Motors of Canada Limited, Natural Sciences and Engineering Research Council of Canada (NSERC) and Initiative for Automotive Innovation (ORF Research Excellence). The authors acknowledge the staff at the Canadian Centre for Electron Microscopy (CCEM) at McMaster University for their technical support with the FIB milling and TEM measurements. The CCEM is a National Facility supported by NSERC and McMaster University. Thanks also to CanmetMATERIALS
References (37)
- et al.
Corros. Sci.
(1997) - et al.
Corros. Sci.
(1997) - et al.
Corros. Sci.
(2012) - et al.
Corros. Sci.
(2010) Electrochim. Acta
(1992)- et al.
Corros. Sci.
(2013) - et al.
Electrochem. Commun.
(2013) - et al.
Electrochem. Commun.
(2013) - et al.
Corros. Sci.
(2012) - et al.
Corros. Sci.
(2013)