3 CONCLUSIONS

Micro-organisms within biofilms are capable of maintaining environments at biofilm/metal interfaces that are radically different from the bulk in terms of pH, dissolved oxygen, and concentrations of organic and inorganic species. As a consequence, micro-organisms within biofilms produce minerals and mineral replacement reactions not predicted by thermodynamic arguments based on the chemistry of the bulk medium.

Pourbaix diagrams, which correlate corrosion mineralogy with local redox and acidity conditions, can be used to demonstrate that certain minerals (in fact most sulfide minerals) form, in natural waters away from hydrothermal activity, only as a result of microbiological action. There remains the question of how thermodynamic stability diagrams can be applied to answer more detailed questions about archaeological objects. What was the original nature of the object? What was the original bulk composition of the specimen? Can the history of the specimen be determined? Of particular interest is the formation of minerals in natural environments over archaeological times that are unlikely to be found on forgeries.

The issue of using stability diagrams to determine whether corrosion product layers are the result of processes over archaeological times or whether they could be the product of relatively short-term processes has never been comprehensively addressed. One can use stability diagrams and the assumption of local equilibrium to construct trajectories on a stability diagram consistent with mineralogical observations (McNeil and Mohr 1993). One could in principle determine the time scale along the trajectory (Lichtner and Seth 1996), but so far only preliminary work has been done (Walton and McNeil 1994) on bronze disease. Another possible approach is to consider sulfur isotopic analysis. It is plausible that long-term microbiological sulfiding of buried artifacts would produce a sulfur isotope distribution different from that produced by short-term sulfiding corrosion, microbiological or otherwise, because the rate would characteristically be controlled by sulfur availability rather than sulfur conversion.

In studying corrosion of materials where corrosion processes can produce a variety of different mineral products, projections of four-dimensional Eh-pH-log anion-log cation diagrams taken normal to the Eh-pH plane (so each point in the diagram corresponds to a Pourbaix diagram) are useful. These are termed modified log activity diagrams (McNeil and Mohr 1992) and are particularly useful in systems where the metal ion has more than one valence state. No effort has been made to use them to interpret sulfiding corrosion. Combination of more extensive thermodynamic-diagram analysis with modeling of electrochemical kinetics (King et al. 1992; Walton and McNeil 1994) mechanisms will permit estimation of rates. This is a critical issue for the appliction of these scientific results to the broader world of archaeological corrosion, and it needs attention.

ACKNOWLEDGEMENTS

The authors wish to thank Dr. Lyndsie Selwyn of the Canadian Conservation Institute for careful review of this paper and for numerous valuable suggestions. This work was funded under Program Element 0601153N, Naval Research Laboratory (NRL). NRL Contribution No. NRL/JA/7333-98-0007.