A more general method for ranking the enrichment of alloying elements in passivation films

The corrosion of alloys is often characterized by an enrichment of one or other of the alloying elements within the surface oxide or even within the underlying metallic phase. For some three decades the measurement of such surface enrichment has been one of the most important contributions made by XPS to corrosion science. X-ray photoelectron spectroscopy permits the composition of the oxide film to be distinguished from that of the metallic surface underlying the oxide so that both can be determined independently. In the mid-1970s these measurements were crucial in showing that the selective enrichment of chromium in the electrochemical passivation of stainless steel occurred by selective dissolution of iron, and not by selective oxidation of chromium. Measurements of this type are important also in active corrosion, where there is no stable oxide film, as in studies of dealloying phenomena. The literature now contains numerous cases, many published by one of us (K.A.) based on studies of novel alloys produced in the Institute for Materials Research at Sendai. Typically, measurements are made for a series of binary compositions A(x)B(y) ranging from A-rich to B-rich alloys, in media that reflect the intended use of the alloy. The results are normally produced in the form of plots of cation composition A(n+)/(A(n+) + Bm+) against bulk composition A/(A + B) or, in the case of dealloying, as A(surface)/(A(surface) + B-surface) against the bulk composition. Although graphical representation is useful in giving the full picture of the alloy's behaviour, it is not so useful in cataloguing or indexing performance. In this paper we suggest a means to give a rank or performance index as a single number that will characterize the behaviour of the alloy over a wide range of composition for a given medium or exposure condition. The rank number does not imply that any particular mechanism of enrichment is in operation; in various cases of corrosion it might occur by selective precipitation (FeOOH on steels, CuO on cupronickels), by selective dissolution (dezincification of brass) or by preferential incorporation of ions in a passivating oxide or oxyhydroxide (Cr3+ on stainless steels). Although the rank has no mechanistic implications, it is useful within a known class of mechanistic behaviour for indicating the magnitude of the effect and thus for enabling an XPS measurement to be indicative of wanted, or unwanted, behaviour. For example, the rank number could be useful in indicating, perhaps from a single alloy composition, whether a degree of passivation is likely to be achieved for that mix of metallic elements. Copyright (C) 2004 John Wiley Sons, Ltd.