1Archaeological iron objects from anoxic and carbonated sites are often in a better preserved state than excavated ironwork from other sites. This is attributed to the formation of the corrosion product siderite under specific environmental burial conditions. Research shows that the presence of chlorides accelerates the corrosion of iron, but also seems to increase the rate of siderite precipitation. Furthermore, the oxidation of siderite upon excavation leads to the formation of stable iron corrosion products if stored dry (Matthiesen, Hilbert and Gregory, 2003:185). The aim of this Master project is to gain understanding of the siderite formation process and its effect on the corrosion crust and its chloride content.
2Archaeological iron objects recovered from the site of Nydam Mose in Denmark have significant differences in condition. The better preserved iron objects are thought to have been directly incorporated into the reducing environment, whilst the more poorly preserved objects were buried in an initial oxidising environment that gradually changed to a reducing one. The corrosion product siderite only occurs on iron artefacts from anoxic and carbonated environments, indicating that the state of preservation is dependent on the precipitation rate and the density of the siderite layer formed. The initial corrosion products on the more poorly preserved objects would predominantly have been composed of iron oxides and iron hydroxide oxides, reduced upon incorporation into the anoxic and carbonated environment.
3The processes of formation of siderite and the reasons for its stability in the burial environment have become a field of study. Short-term laboratory tests suggest that an increase in chloride content of the electrolyte will cause a more rapid formation of a denser siderite layer on iron coupons (Jiang and Nešić, 2009:5). Given that siderite oxidizes to stable products suggest, in my opinion, that the corrosion process and precipitation of siderite may have an effect on the chlorides within the pores of the corrosion layers initially formed around the artefact during the first period of burial.
4A series of short-term laboratory experiments have been designed in collaboration with researchers specialised in mineral growth and dissolution from the Department of Geosciences (University of Oslo). The experiments were performed in a glass cell reactor, and tests were undertaken on both iron coupons and archaeological finds from both oxidising terrestrial and marine environments.
5During preliminary experiments clean and beforehand corroded iron coupons were kept for 4 days in electrolytes with different NaCl concentrations in a synthetic anoxic environment achieved by the bubbling of CO2 through it. Otherwise the parameters were set to optimize for siderite growth. The objective of these experiments was, among others, to establish at which NaCl concentration siderite growth were more efficient.
6Results from the first preliminary test showed a rapid siderite crystal growth at 74 °C and pH 6,4 ± 0,1 when the electrolyte was saturated with Fe2+ ions, the CO2 partial pressure was at ca 0,54 bar, total pressure ca 1,0 atm and the NaCl concentration was ca 20 % (w/v). The siderite crystal growth was clearly denser at the surface of the clean coupon than on the surface of the already formed corrosion products (fig. 1).
7The bands at 1402/1409, 859/864 and 734/737 cm-1 present in the FTIR-spectra indicates the formation of siderite (Matthiesen, Hilbert and Gregory, 2003:185). These spectra also show a pronounced difference in purity of the two samples (fig. 2).
8Upon completion of the preliminary experiments, archaeological samples will be kept in the most optimal electrolyte for siderite growth at different lengths of time. Comparative analysis of both treated and untreated cross-sections will be performed by SEM-EDS, thus providing data of the elemental distribution within the corrosion products. Furthermore, FTIR analysis will provide complementary analysis of the corrosion products formed. The performed analysis will then hopefully provide insight in siderite formation and its effect upon the chloride content of a given corrosion layer.
Fig. 1 Siderite crystals enlarged with scanning electron microscopy
Siderite crystal growth on iron coupons after 4 days in the glass cell reactor: Top left: clean coupon 1000x, 15 kV, Bottom left: corroded coupon 1000x, 15kV, Top right: clean coupon 300x, 15kV, Bottom right: corroded coupon 300x, 15kV
Credits: Ruben With
Fig. 2 Fourier transform infrared spectra of the newly precipitated corrosion products
FTIR-spectra of the precipitated product on both a corroded and a clean iron coupon, top and bottom respectively.
Credits: Ruben With
9During the period between the conference and time of printing the experiments were completed with encouraging results. The newly formed corrosion products apparently separated the chlorine from the metal surface, seemingly without affecting the corrosion crust of the object. However, more statistically valid research is needed to verify the obtained results.