Knife with a groove on both sides - Fe Alloy - Early medieval times - Switzerland

Marianne. Senn (EMPA, Dübendorf, Zurich, Switzerland) & Christian. Degrigny (HE-Arc CR, Neuchâtel, Neuchâtel, Switzerland)

Stratigraphic representation: none 

Fig. 4: Stratigraphic representation of the object in cross-section using the MiCorr application. This representation can be compared to Fig. 13.

Analyses performed:
Metallography (nital etched after etching with Oberhoffer’s reagent), Vickers hardness testing, LA-ICP-MS, SEM/EDX.

The remaining metal consists of two forged wrought iron bars, one of which includes the carburized tip (Table 1). They are separated by a welding seam (Figs. 8 and 9). The ferritic part is Cu-rich, whereas the carburised tip has a low and medium content of trace elements (Table 1). The metal contains elongated slag inclusions (Fig. 4) showing a structure of wüstite in a glassy matrix (Fig. 5 and Table 2). Most of the slag inclusions are arranged in rows, marking the welding seam (Figs. 4 and 8) and following the forging direction. Their chemical composition differs in the Mn content (Table 2): the latter is higher in the slag inclusions of the carburized tip. The high P content of the slag in the ferritic part must be noted since the metal in general has a medium P content (Tables 1 and 2). Etching with Oberhoffer’s reagent solution makes the P distribution visible (Fig. 9). Dark areas are depleted of P whereas P-rich zones, such as are found in the welding seam, appear in white. After nital etching, the very fine steel microstructure of the tip shows the transition from hypoeutectoid to eutectoid steel (ferrite component in white and pearlite component in black, partly bainite, Figs. 6, 7 and 8). The body of the knife is made of wrought iron with an annealed, irregular ferritic structure (Figs. 6 and 7). The average hardness of the wrought iron (HV1 130) is a little higher than expected, whereas the hardness of steel in the hypoeutectoid-eutectoid tip (HV1 360) is an indication of quench-hardening followed by tempering.


Elements V Cr Mn P Co Ni Cu As Ag Ni/Co C* mass%
Body (median of 2 similar analyses) mg/kg < < 7 400 60 20 1300 300 < 0.3 0/0.2
Tip (median of 7 similar analyses) mg/kg < 4 100 500 40 70 400 70 < 1.8 0.8
Detection limit mg/kg 0.7 2 0.4 68 0.4 3 2 0.8 0.4    
RSD1 % - 26 95 92 9 3 79 24      
RSD2 % - - 112 42 13 26 20 47      

*visually estimated

Table 1: Chemical composition of the metal. Method of analysis: LA-ICP-MS, Laboratory of Analytical Chemistry, Empa (for details see Devos et al. 2000).


Location Environment Na2O MgO Al2O3 SiO2 P2O5 K2O CaO TiO2 MnO FeO Total SiO2/Al2O3
Glass Pearlite, tip 0.7 < 9.8 72 < 6.2 3.3 0.9 1.7 8.9 104 7.3
n. d. Pearlite, tip < < 3.5 34 < 1.5 1.1 < 1.0 65 107 9.8
n. d. Ferrite (average of 4 similar analyses), body < 0.9 3.8 30 1.2 1.5 1.6 < < 64 104 7.8
n. d. Ferrite, body < < 1.5 14 < 0.7 0.6 < < 88 105 9.3

n. d. = structure not determined

Table 2: Chemical composition of the slag inclusions (mass%) at the tip (pearlite) and the body (ferrite) of the knife. Method of analysis: SEM/EDX, Laboratory of Analytical Chemistry, Empa.

The metal - corrosion products interface is irregular (Figs. 3, 4 and 10) and the average thickness of the corrosion crust is 200µm. In bright field, the corrosion appears grey, rather heterogeneous and heavily cracked. A thin light-grey layer (indicated by an arrow in Fig. 10) can be detected. Under polarised light the corrosion is more clearly stratified, the thin layer mentioned before is black and surrounded by light and orange-brown corrosion layers (Figs. 11 and 12). It contains less O (magnetite or hematite?) than the orange-red corrosion products (iron hydroxides?) (Table 3 and Fig. 13). The outer corrosion layer (covering the aforementioned thin black layer) contains external markers such as quartz grains and other rock fragments (Ca, Fig. 13). The shape of the blade is preserved in the corrosion crust (Fig. 13, arrows on the SEM image). The absence of P, an external marker, highlights where the limit of the original surface was located.



O Si P Ca Fe Total
Black layer (CP3i) 26 < < < 69 95
Dark-brown corrosion products (average of 3 similar analyses) (CP2i) 29 < < < 63 92
Red and brown corrosion products (CP1e) 32 1.4 1.1 0.7 56 92

Table 3: Chemical composition (mass %) of the corrosion layer (from Fig. 12). Method of analysis: SEM/EDX, Laboratory of Analytical Chemistry, Empa.

Corrected stratigraphic representation: none

The knife is forged from two wrought iron bars which have been welded together. The tip is carburized. The recrystallized structure of the ferrite is probably the consequence of tempering the tip. The metal compositions of both alloys differ from the one worked in the forges of Develier-Courtételle (Eschenlohr et al. 2007, 71). For this reason this well worked knife is identified as an importation to the early medieval village Develier-Courtételle. The limit of the original surface (limitos) is still preserved within the remaining corrosion layers. Chemically it can be located at the interface of the P-rich outer corrosion layer and the P-poor inner corrosion products. Visually it can be located by the presence of sediments in the outer corrosion layers and most likely by the hardness and coloration of the inner corrosion products (magnetite?). It is an example of a terrestrial corrosion crust.

References on object and sample

References object

1. Eschenlohr, L., Friedli, V., Robert-Charrue Linder, C., Senn, M. (2007) Develier-Courtételle. Un habitat mérovingien. Métallurgie du fer et mobilier métallique. Cahier d'archéologie jurassienne 14 (Porrentruy), 302.


References sample

2. Eschenlohr, L., Friedli, V., Robert-Charrue Linder, C., Senn, M. (2007) Develier-Courtételle. Un habitat mérovingien. Métallurgie du fer et mobilier métallique. Cahier d'archéologie jurassienne 14 (Porrentruy), 266.

References on analytic methods and interpretation