Corrosion Mechanisms and Protection of WE43 Mg Alloys in "Biologically Relevant" Environments

Corrosion mechanisms of Mg alloys (mostly the MgAl-Zn AZ series) have been extensively documented in the past [1-3]. Only very recently, Mg alloys (especially the Mg-Y WE43) gained interest in the medical community as biodegradable implants of different type (from pins, small plates to stents). The corrosion mechanisms are quite different in complex biological media then in simple chloride and sulfate solutions. First studies [4-5] are estimating corrosion rates which is an important parameter when degradation needs to be optimized but insufficient for valuable prediction of the implant lifetime. In the present work, specific corrosion aspects related to biodegradable Mg implants are investigated. The most obvious one is the difference between uniform and localized attack taking place on the surface. By Electrochemical Impedance Spectroscopy (EIS), it is possible to assess the stability of an Mg alloy surface in different body fluids. Figure 1, illustrates a further important question we are facing when working with “body fluids. Measured polarization resistances by EIS in Simulated Body Fluid (SBF: pH 7.4 buffered S9 formulation) and Artificial Blood Plasma (AP: pH 8.4, non buffered) are totally different. The two solutions are relatively similar in ionic content except for the higher buffering capacity of SBF and a higher carbonate content of the AP. Corrosion rates are differing by order of magnitude (Rp of 10 in AP and 10 for SBF). Based on this observation a detailed study of 3 important parameters has been conducted: i) buffering ability, ii) carbonate content and iii) influence of proteins. Carbonate is interesting because it can form stable corrosion products with Mg. Due to the uncertainty in the real environment and numerous SBF formulations, it is important to have a fundamental understanding of the role on the corrosion processes of small electrolyte changes. A second important aspect is temporary surface protection necessary even for a degradable implant. Here again, the buffering ability of the electrolyte is crucial. Figure 2 shows two completely different barrier surface oxide that have been produced [6]: an 800nm Yttrium rich by thermal oxidation at 550°C (Figure 2a) and an 350400nm Mg rich by anodizing in 0.05M Na3PO4 (pH 13) (Figure 2b). The Yttrium rich oxide that seemed a promising candidate for temporary protection has a certain corrosion resistance in SBF, but dissolves in AP. During these investigations and especially for coating evaluation, the limitation of EIS was evidenced. The method is suitable for the corrosion resistance or uniform corrosion rate determination. For localized attack in otherwise stable coating occurred, results where of little use. On Figure 3a, it is observed that the polarization resistance drops (transition zone 1 to zone 2), but only a very small part of the surface is attacked, figure 3b. An additional hydrogen gas evolution (correlated to the cathodic reaction rate) measurements sensing uniform dissolution is added to the EIS characterization. Initial slow uniform dissolution is the desired mechanism for biodegradable implants; localized attack has to be avoided. a) b)