The Localised Corrosion Mechanism of Magnesium in Methanol Resolved Using SVET

Wednesday, 8 October 2014: 10:20
Expo Center, 1st Floor, Universal 11 (Moon Palace Resort)
H. N. McMurray, G. Williams, and J. E. Board (Swansea University)
It is well known that methanol is aggressively corrosive to light metals such as magnesium and that this can present a problem when methanol is present in cleaning solutions, fuel mixtures and coolant streams. However, the high electrical resistivity of methanol and (salt-free) methanol/water mixtures makes conventional electrochemical measurements involving polarisation difficult to perform. Consequently, whenever electrochemical studies of methanolic corrosion have been carried out they have typically been so in electrolyte mixtures comprising methanol, water and dissolved salt. Here we show that the scanning vibrating electrode technique (SVET) is capable of providing quantitative information on the intensity and localisation of electrochemical corrosion reactions as they occur on commercially pure magnesium immersed in pure methanol and salt-free methanol/water mixtures. As such it has been possible to show that, even in nominally anhydrous methanol the mechanism of corrosive attack is definitely electrochemical in nature and that the location of anode and cathode sites is often highly localized. 

Magnesium foil (2mm thick, 99.9% purity, temper as rolled) was obtained from Goodfellows Metals (nominal iron impurity level 280ppm). Methanol (ACS Reagent Grade, 99.8% purity) was obtained from Sigma Aldrich Ltd. The electrical conductivity of the as-received methanol was measured to be 1.170 mS/cm.  Fig. 1a shows pitting corrosion of magnesium, induced by immersion in methanol in contact with room air and Fig1b shows a corresponding SVET-derived current density map. The SVET data in Fig 1b demonstrate the extreme agression of methanolic attack, with local current densities of up to 500 amps per metre squared, they also show evidence of cathodic activation, in which initially anodic areas switch to cathodic activity. This last phenomena most probably occurs as a result of transition metal impurities (mainly iron) accumulating at the dissolving magnesium surface. In the current paper SVET  data are supported throughout with independent measurements of: hydrogen evolution rate and weight loss, all under conditions of known water activity and controlled oxygen partial pressure. Experimental arrangements are described which allow the continous, in-situ, measurement of evolved hydrogen using digital manometry and the continous, in-situ, measurement of magnesium weight loss using a digital balance and immersion bridge. By such means been possible to determine more exactly the kinetics methanolic attack and to show that both water and oxygen actually inhibit methanolic corrosion. In the case of water, the level of inhibition can be profound initially but tends to diminish over time when the original, oxide covered, magnesium surface breaks down and cathodic activation begins. 

The observed inhibitory influence of dissolved oxygen and water on the kinetics of corrosive attack in methanol are tentatively ascribed to the reinforcement of the original, air-formed, magnesium oxide surface film by methanol-insoluble magnesium hydroxide produced either through the direct reaction of magnesium with water or by the precipitation of magnesium cations through combination with hydroxide anions produced though oxygen reduction. In contrast, the principal product of methanolic attack is magnesium methoxide which is methanol-soluble and provides no protection or encouragement toward magnesium passivation. It should be noted however neither oxygen or water provide long-term protection but tend instead to lengthen an initial, slow, phase of corrosion which occurs prior to breakdown of the original oxide-covered magnesium surface and the onset of corrosion acceleration resulting from cathodic activation.