Sacrificial knife (tumi) - Cu-As Alloy - Peru - Cu Alloy

Christian. Degrigny (HE-Arc CR, Neuchâtel, Neuchâtel, Switzerland) & Gerber. Alice (HE-Arc CR, Neuchâtel, Neuchâtel, Switzerland) & Valentin. Boissonnas (HE-Arc CR, Neuchâtel, Neuchâtel, Switzerland)

Complementary information

The original archaeological context is unknown, so there is very little information about the culture of origin and no exact dating. The object entered through a purchase and donation to the museum.

The schematic representation below gives an overview of the corrosion layers encountered on the object from a first visual macroscopic observation

Fig. 9 MiCorr schematic representation of the corrosion

X-ray radiography and tomography [1], XRF[2] and SEM-EDS [3].

[1] Conditions of the analyses are unknown.

[2] With a portable X-ray fluorescence spectrometer NITON XL3t 950 Air GOLDD+ analyser from Thermo Fischer®

[3] With a Jeol JSM-6400 device

The metal is an arsenical copper alloy with more or less 3% As (Table 1).

A XRF analysis has been conducted on one area, on the front side of the partly-cleaned artifact (Fig. 11).


Element Cu As Cl S P Ca BAL
% 75.04 2.33 11.47 0.192 0.170 0.119 10.34

Table 1 Chemical composition of the partly-cleaned metal surface. Method of analysis: XRF. Mining mode Cu/Zn, acquisition time 120s (filters: M30/Lo30/H30/Li30). BAL = non-analysed elements (probably O and C).


X-ray radiography gives technological indications, such as the way incrustations are put in place, and highlights the thickest and least corroded parts (Fig. 12)

The X-ray tomography shows us that the metal, at least in the handle decorated with a character, is relatively porous (Fig. 13)

Complementary information

The metallographic structure was not observed, but it is likely that the object was cast and then reworked cold.
Depending on the area, there should be cored grains with remaining dendritic structures and, if the metal was annealed, grains with twin lines.

Results from the XRF analysis of the partly-cleaned metal surface (Table 1) shows an elevated concentration of the element chlore, indicating a possible high amount of copper chlorides.

Depending on the appearance and morphology of each corrosion product, it is possible to assume that on this object are present copper carbonates such as Malachite (CP3) and Azurite (CP4), copper oxides such as Cuprite (CP5 and CP6), copper chlorides such as Nantokite (CP7) and copper hydroxychlorides such as Atacamite and Paratacamite (CP2). Some corrosion products can not really be identified (CP1).

SEM-EDS analyses also show the presence of phosphorus P in many compounds. This is an element that suggests that the object was probably buried in a grave, or near any other consequent amount of decomposing organic materials.

Complementary information

The surface of the object is corroded in a very heterogeneous way. Some zones correspond to Type I corrosion models according to Robbiola, and others to Type II with the formation of pustules. Parts of the object that have been cold worked and / or deformed are those with the most corroded surfaces, sometimes with complete lost of the limitos. The edge of the blade, for example, was hammered after casting and was bent during its use: it is one of the most corroded areas.

Following comes a new stratigraphic representation of the corrosion layers, corrected and improved thanks to the analysis carried out. The limit of the original surface  was identified as still present and is located at the interface between CP5 and CP6.

In this case the metal is an arsenical copper alloy. The object was cast and then cold-worked, probably hammered. The metal is heterogeneously corroded with Robbiola types I and II and pustules. In the analyzes carried out, the chlore element appeared, indicating a so-called "active" corrosion of the metal. The limit of the original surface is at the CP5 and CP6 interface.

Scott, D. Copper and Bronze in Art: Corrosion, Colorants, Conservation. Getty Conservation Institute, Los Angeles, 2002.

Scott, D. Metallography and Microstructure of Ancient and Historic Metals. Getty Conservation Institute, Los Angeles, 1991.

Robbiola, L. Caracterisation de l'altération de bronzes archéologiques enfouis à partir d'un corpus d'objets de l'âge du bronze. Mécanismes de corrosion. Université Pierre et Marie Curie - Paris VI, 1990.

Gutknecht, 2018 : Gutknecht Naïma. La corrosion active sur les alliages cuivreux archéologiques - Evaluation de la stabilisation par biopassivation fongique. Mémoire de Master, Haute École Arc Necuhâtel, Conservation-Restauration, non-publié, 2018.