1 INTRODUCTION

Objects exposed to corrosive environments for long periods of time, whether they are meteorites, archaeological objects, infrastructure components, or minerals exposed to weathering, frequently show degradation of a nature that cannot be duplicated in laboratory experiments, either in degree or sometimes even in kind. Some laboratory experiments have carefully exposed characterized specimens to known environments for decades (Romanoff 1957); however, for practical purposes the laboratory time frame is five years or less.

In this article, “long-term” will be applied to time frames sufficiently long that the areas where reactions take place may be regarded as macroscopic regions with local equilibrium, with the overall development of individual reactions controlled by mass transport considerations rather than local kinetics.

Central to this reasoning is the applicability of Pourbaix diagrams (Garrels and Christ 1965; Pourbaix 1966) to predict not only regions of immunity and passivity but also the identities of the species produced by corrosion. The use of diagrams is complemented by Ostwald's principle (Ostwald 1906) that when a component (e.g., a metal) reacts with an element in its environment (e.g., reduced sulfide species), the normal succession of reaction products is initially to produce those having the lowest sulfur-metal ratio and then gradually to produce products with higher sulfur-metal ratios, consistent with thermodynamic equilibria. There is an extensive literature on the application of thermodynamic stability diagrams to the prediction and interpretation of long-term corrosion products. A comprehensive review is beyond the scope of this p�per.

Sulfiding corrosion is a common consequence of biofouling, because sulfur compounds are common in the biosphere and ubiquitous in the industrialized world. Sulfate-reducing bacteria (SRB), a diverse group of anaerobic bacteria isolated from a variety of sulfur-containing environments (Postgate 1979; Pfennig et al. 1981), are capable of reducing sulfate and similar oxidized sulfur species to sulfides. SRB are routinely isolated from seawater where the concentration of sulfate is typically 25 mM (Postgate 1979). Even though seawater is generally aerobic (typical oxygen concentrations above the thermocline, the temperature change generally regarded as dividing near-surface from deep seawater, are in the range of 4–6 ppm, equivalent to .1–.15 mM), anaerobic micro-organisms survive in anaerobic microniches until conditions are suitable for their growth (Staffeldt and Kohler 1973; Costerton and Geesey 1986). If the aerobic respiration rate within a biofilm is greater than the oxygen diffusion rate, the metal-biofilm interface can become anaerobic and provide a niche for sulfide production by SRB (Little et al. 1990). Conversion of metals to sulfide minerals by SRB has been studied since the late 1800s (Daubree 1862; de Gouvernain 1875). Baas-Becking and Moore (1961) list sulfides produced by Desulfovibrio desulfuricans over a large pH range.

In the following sections, SRB sulfide production on pure metal surfaces will be reviewed. The metal interface under the biofilm and corrosion layers will be referred to as base metal to differentiate it from layers of minerals and metal ions that have been derivatized by corrosion reactions. Mineralogical data, thermodynamic stability diagrams (Pourbaix 1966; Wagman et al. 1982), and the simplexity principle for precipitation reactions (Goldschmidt 1953) will be used to rationalize corrosion product mineralogy in fresh and saline water and to demonstrate the action of SRB.