| Abstract: | SUMMARIES. Desiccation has long been used to store chloride-contaminated archaeological iron but there are no precise guidelines on the degree of desiccation required to prevent corrosion occurring. Akaganéite (ß-FeOOH), ferrous chloride tetrahydrate (FeCl2 • 4H2O) and ferrous chloride dihydrate (FeCl2 • 2H2O) have been recorded on archaeological iron. Iron corrodes in the presence of FeCl2 •4H2O and ß-FeOOH but not in the presence of FeCl2 •2H2O. The rate of desiccation of FeCl2 •4H2O at various levels of relative humidity (RH) was determined by experiment and found to be an exponential relationship. The point at which FeCl2 •2H2O first becomes a stable hydrate was established. Rates of corrosion for iron mixed with FeCl2•4H2O and with ß-FeOOH were examined for a range of RH. The hygroscopicity of ß-FeOOH and the RH at which it ceases to cause iron to corrode were established. Corrosion of iron in contact with FeCl2 •4H7O and ß-FeOOH speeds up as RH rises and is appreciable at 25% RH and above. On the basis of these results, recommendations are made that 12% should be the maximum allowable RH for long-term storage of archaeological iron from chloride-bearing soils. Low RH requirements raise problems for long-term monitoring of storage microclimates.
CONCLUSION. The results show that FeCl2•2H2O does not corrode iron in contact "with it. In contrast, iron in contact with FeCl2•4H2O corrodes over the stability range of this corrosion product. The rate of corrosion increases with increasing RH. ß-FeOOH corrodes iron in contact with it from 15% RH upwards, with increasing RH producing a faster corrosion rate. Although iron in contact with ferrous chloride could be safely stored at 19% RH without any corrosion occurring, most archaeological iron objects have some ß—FeOOH present on their surface, due to post-excavation corrosion. Since the iron/ß-FeOOH corro¬sion model appears to have ceased to operate at 12% RH, all chloride-contaminated archaeological iron should be stored at or below this RH to prevent corrosion. If RH exceeds this value, then over the range 15% to 20% RH ß-FeOOH will determine the corrosion rate of chloride-contaminated iron. Above 20% RH FeCl2•4H2O will form and will also contribute to corrosion of iron. At 25% RH and above both the FeCl2•4H2O/iron and ß-FeOOH/iron corrosion reactions progress at an appreciable rate and speed up, at least initially, as RH rises. Even if RH cannot be kept at a level where corrosion ceases, it is advantageous to keep it as low as possible to reduce the rate of corrosion. While it is essential to store iron objects with a metal core in this way, totally mineralized objects can be stored in high humidity environments, as there is no more iron left to oxidize and they should be stable [41]. The problem facing the conservator is to successfully identify such objects. Radiography remains the best guide. Comparing the rate of corrosion for FeCl2•4H2O/ iron powder at 25% RH, with that of ß-FeOOH/iron powder at the same RH (Figures 6 and 9), indicates that β-FeOOH corrodes iron significantly faster than FeCl2•4H2O at this low humidity. Corrosion of iron at low humidity has implications for chloride-contamin¬ated iron stored in sealed plastic boxes whose internal microclimate is controlled by a desiccant. Small rises in their internal RH are difficult to detect, but may be sufficient to support the iron corrosion model studied here. Synergistic effects of FeCl2•4H2O/ß-FeOOH/iron mixtures have not been considered in this paper. It may be that the corrosion rate is greater than the sum of the individual corrosion rates when both compounds are present. Work on this is currently being prepared for publication along with investigations into the effect of fluctuating RH on corrosion rate and deliquescence of FeCl2•4H2O. To gain quick results, the experiments reported here represent worst case scenarios. Experimental design produces large surface areas of metal in contact with quantities of FeCl2•4H2O and ß-FeOOH that far exceed what might be expected on the surface of archaeological objects. This emphasizes the aggressive tendencies of the reactions examined. In reality, damage may be more localized and corrosion rates slower, although ultimately cleavage at the metal/corrosion layer interface will occur. |
|---|
