Plug-In Specimens for Measurement of the Corrosion Rate of Mg Alloys ZHIMING SHI, 1 ARVIND PRASAD, 1 and ANDREJ ATRENS 1,2 1.—Division of Materials, The University of Queensland, Brisbane, QLD 4072, Australia. 2.—e-mail: Andrejs.atrens@uq.edu.au Magnesium alloy corrosion is often nonlinear. The corrosion rate accelerates to steady state after an initial period of low corrosion. Plug-in specimens permit simultaneous measurement of the corrosion rate using hydrogen evo- lution, P H , and Tafel extrapolation of cathodic polarization curves, P i . More- over, weight loss allows independent verification. P H is consistently greater than P i . The data, for short exposure periods up to 10 days, are consistent with the unipositive Mg + ion being a short-lived intermediate. Tafel extrapolation needs to be used with caution for estimation of Mg corrosion, as the measured corrosion rate can have a significant contribution from crevice corrosion. Furthermore, measurements made at short immersion times may not reflect the steady-state corrosion rate, and the corrosion reaction at the Mg surface may be decoupled from the electrochemical measurement. INTRODUCTION Corrosion limits more widespread use of Mg alloys in aircraft and automobiles. In contrast, complete corrosion within a finite period allows Mg to be considered for use in biodegradable medical implants. Our research on Mg corrosion, 1–8 Mg flammability, 9 and stress corrosion cracking 10,11 has provided many key insights. This paper summarizes recent advances at the University of Queens- land 12–15 regarding measurement of the Mg corro- sion rate and mechanism. The weight loss rate, designated P W (mm year 1 ), provides a benchmark. This is a simple method, al- though experimental skill is required in the case of Mg, particularly to ensure removal of all corrosion products. The evaluated corrosion rate is less than the actual corrosion rate if some corrosion products remain on the surface. Some literature results appear to be erroneous due to this effect. 12 Our work has found 13 that chromic acid solution removes no Mg metal, and removes only the corrosion products. One drawback of this approach is that the weight loss rate is an average over the experiment duration and does not easily allow for measurement of the time variation of the corrosion rate. The Mg corrosion mechanism 1–8 is summarized in Fig. 1. Corrosion mainly occurs at breaks in a partly protective surface film. The anodic partial reaction occurs in two steps, (1) and (2). In the second step, some fraction, k, of unipositive Mg + ions undergoes electrochemical anodic oxidation to the equilibrium Mg ++ ion. The complement fraction, 1  k, reacts chemically by reaction (3) to produce the stable Mg ++ ion and H 2 . The two anodic partial reactions (1) and (2) are balanced by the cathodic partial reaction (4). Mg ! Mg þ þ e anodic partial reaction ð Þ (1) k Mg þ ! k Mg þþ þ ke anodic partial reaction ð Þ (2) 1  k ð Þ Mg þ þ 1  k ð Þ H þ ! 1  k ð Þ Mg þþ þ 1  k ð Þ 1 = 2 H 2 chemical reaction ð Þ ð3Þ 1 þ k ð Þ H þ þ 1 þ k ð Þ e ! 1 þ k ð Þ 1 = 2 H 2 cathodic partial reaction ð Þ ð4Þ Equations 1–4 are summed to give the following overall reaction: Mg þ 2H þ ! Mg þþ þ H 2 (5) The overall corrosion reaction indicates that measurement of the evolved hydrogen 1–8 can be JOM, Vol. 64, No. 6, 2012 DOI: 10.1007/s11837-012-0335-z Ó 2012 TMS (Published online June 3, 2012) 657