Corrosion mechanisms of materials buried in soil remain insufficiently understood due to complex, porous, heterogeneous nature of the medium and environmental changes over long period, producing thick, intricate corrosion products. Short-term, simplified laboratory experiments cannot reliably replicate such corrosion layers. Archaeological iron artifacts, having corroded naturally over centuries or millennia, offer a unique perspective into corrosion phenomena. However, excavation can alter physicochemical properties of corrosion layers through environmental exposure, and invasive analytical techniques must be avoided given the irreplaceable nature of such artifacts.
This thesis addresses these challenges through an integrated approach to determine corrosion rates and elucidate corrosion initiation and progression of ferrous materials at the archaeological site Bois de Châtel (Avenches, Switzerland). Four sample categories were studied: pure iron, wrought iron representative of buried archaeological nails, archived archaeological nails, and freshly excavated nails from Bois de Châtel. Wrought iron samples were produced by a blacksmith using the Roman-era direct reduction method.
To bridge laboratory and real-world conditions, natural corrosion experiments were conducted by burying samples at the site, and through artificial corrosion experiments in laboratory using electrolytes produced from sampled soils. Repeated burial, excavation, and bimodal neutron and X-ray imaging cycles enabled novel temporal monitoring of corrosion progression within the same sample. Corrosion rates from computed tomography were compared against electrochemical methods and oxygen consumption measurements. A COMSOL corrosion model supported interpretation and extended insights beyond measurable conditions. Cross-sections were validated via scanning electron microscopy, Raman spectroscopy, and energy dispersive X-ray analysis.
Across all methods, wrought iron consistently exhibited lower corrosion rates than pure iron, attributed to surface oxide layers from its production process acting as a physical barrier against electrolyte and oxygen ingress. Discrepancies in absolute values between methods reflect the different physical and chemical phenomena each method captures and differing experimental conditions. This highlights the sensitivity of iron corrosion kinetics to environmental exposure.
When ferrous material is buried, adhered soil triggers corrosion beneath its surface. Newly formed corrosion compounds replace the consumed metal's space. During wet-dry cycles, soil pores are filled with electrolyte and oxygen. Once corrosion products exceed solubility limits, they precipitate in soil pores, embedding soil particles to form a transformed medium layer. The dense corrosion product layer advances inward toward the metal core. The transformed medium layer progresses outward, simultaneously reducing the soil layer.
The initial corrosion product, ferrihydrite, transfor
