Bracelet bus-na 98 BZ 11 - Bronze - Iron Age - Switzerland

Bracelet bus-na 98 BZ 11

Christian. Degrigny (HE-Arc CR, Neuchâtel, Neuchâtel, Switzerland) & Naima. Gutknecht (HE-Arc CR, Neuchâtel, Neuchâtel, Switzerland) & Valentina. Valbi (Laboratoire Métallurgie et Culture LMC-IRAMAT-CNRS-UTBM, Belfort, Franche-Comté, France)

Complementary information

This bracelet is part of a corpus of four bracelets (bracelets bus-na 98 BZ 11 / bus-na 98 BZ 23 / bus-na 98 BZ 38 / bus-na 98 BZ 67) found on the same site.

The schematic representation below gives an overview of the corrosion structure encountered on the bracelet from a first visual macroscopic observation.

Strata Type of stratum Principal characteristics
S1 Soil light brown, thin, non-compact, powdery
CP1 Corroded product green, thin, discontinuous, compact, soft
CP2 Corroded product blue, thick, discontinuous, non-compact, very soft
CP3 Corroded product red, thin, continuous, compact, very soft
M1 Metal orange, metallic, soft

Table 1: Description of the principal characteristics of the strata as observed under binocular and described according to Bertholon's method.

The transition between CP3 and the metal M1 is irregular and rough. There are cracks through S1, CP1 and CP2 that generate the flaking of CP1 and CP2.

Fig. 5: Stratigraphic representation of the corrosion structure of the bracelet observed macroscopically under binocular microscope using the MiCorr application with reference to Fig. 4. The characteristics of the strata are only accessible by clicking on the drawing that redirects you to the search tool by stratigraphy representation, credit MiCorr_HE-Arc CR, N.Gutknecht.

Complementary information

None.

Analyses performed:

Non-invasive approach

- XRF with handheld portable X-ray fluorescence spectrometer (NITON XL3t 950 Air GOLDD+, Thermo Fischer®). General Metal mode, acquisition time 60s (filters: Li20/Lo20/M20).

Invasive approach (on the sample)

- Optical microscopy: the sample is polished, then it is observed on a numerical microscope KEYENCE VHX-7000 in bright and dark field.

- Metallography: the polished sample is etched with alcoholic ferric chloride and observed by optical microscopy in bright field.

- SEM-EDS: the sample is coated with a carbon layer and analyses are performed on a SEM-FEG JEOL 7001-F equipped with a silicon-drift EDS Oxford detector (Aztec analysis software) with an accelerating voltage of 20 kV and probe current at about 9 nA. The relative error is considered of about 10% for content range <1mass%, and of 2% for content range of >1mass%. 

- µ-Raman spectroscopy: it is performed on a HORIBA Labram Xplora spectrometer equipped with a 532 nm laser with 1800 grating, the laser power employed is between 0.04 and 0.55 mW with acquisition time varying between 1 and 5 minutes.

The XRF analysis of the bracelet was carried out before sampling on four areas (Fig. 2). All strata (soil, corrosion products, and metal) are analyzed at the same time. The metal is presumably a copper-tin alloy with some lead (and perhaps Bi), while the other elements detected (Fe, Si, P, Al) are from the burial environment. 

Elements (mass %) 1 2 3 4
Cu 40 28 33 27
Sn 39 44 39 43
Pb 2 2 2 3
Bi 1 1 1 1
Fe 6 13 8 14
Si 6 5 9 5
P 3 5 2 5
Al 2 >2 3 2

Table 2: Chemical composition of the surface of the bracelet at four representative points shown in Fig.2. Method of analysis: XRF. The results are rounded up to the nearest whole number.

EDX analysis (Table 3) of the residual metal on cross-section indicates that it is a low tin bronze (7-8 mass% Sn) with a low percentage of Pb (0.5 mass%). 

The metal etched with alcoholic ferric chloride reveals a main equiaxe grain structure (Fig. 7) with several twinned grains (Figs. 8 and 9), indicating that the object underwent an annealing procedure. It is possible to observe on the thinner extremities of the sample the deformation of the grains (shown for the right extremity on Fig. 8) and the presence of slip lines (Fig. 8) caused by cold mechanical working of the metal after annealing. 

Small Pb inclusions (2-3 μm, Figs. 8-9) are homogeneously distributed on the whole surface of the sample.

Intergranular corrosion is observed (Fig. 9), as well as a thick transgranular fracture.

Elements mass %
Cu 91
Sn 7.8
Pb 0.5
As <0.5
Ni <0.5
Al <0.5
Si <0.5

Table 3: Chemical composition (mass %) of the alloy over a general area of analysis obtained by SEM-EDX.

Complementary information

The interface roughness between the metal and the CPs, corresponding to the stratum identified as corroded metal (CM1) was measured through optical microscopy observation and the following parameters were determined (in µm): Rp=56, Rv = 68, Rt = 124, Ra = 41.

The observation of the sample in cross-section in dark field allowed identification of an external discontinuous greenish/yellowish stratum (CP1), a blue stratum (CP2), a red discontinuous thin stratum (CP3) at the interface with the metal, and a corroded metal area (CM1)(Fig.2). This is confirmed in Fig. 10. 

The EDX elemental analysis (Table 4) and mapping (Fig. 11) of the visually identified CPs by binocular and cross-sectional observations show that CP1 is Cu depleted, O, Sn and Pb enriched and polluted with Ca, Fe and P, while CP2 is also Cu depleted and O and Sn-rich but with fewer external polluting elements and CP3 is Cu rich. The results are in agreement with those of the preliminary XRF analysis (table 2), with the exception of aluminum and bismuth which are apparently contained only in S1. Indeed, the XRF analysis indicates tin and Pb enrichments on the four areas investigated of the metal surface. 

  CP1  CP2 CP3
Cu 16.2 28.4 57.6
Sn 33.6 34.7 18.2
O 32.4 30.2 18.1
Pb 1.3 0.8 0.7
Si 1.6 1.9 0.9
P 5.0 2.4 1.8
Fe 6.5 < 0.5 < 0.5
Ca 2.8 0.9 < 0.5
Cl < 0.5 < 0.5 0.8

Table 4: Chemical composition (mass%) of the corrosion layers over a general area of analysis in cross-section obtained by SEM-EDX.

Analyses with µ-Raman were performed on the three identified strata (Fig. 11). The R01 (Fig. 12) point of analysis was performed on the red CP3 layer and the obtained spectrum presented the typical main peaks (145, 218, 632 cm-1) of the Raman RUFF reference spectrum of cuprite Cu2O (Lafuente et al. 2015). The R02 point of analysis was performed on the blue layer and the spectrum obtained has a broad peak at 560 cm-1 that can be attributed to nanocrystals of cassiterite Sn2O thanks to comparison with the work of Ospitali et al. 2012. The R03 analysis point performed on the green/yellow external phase was too disturbed by fluorescence signal and an identification of the phase was impossible. The blue color of CP2 is probably caused by residual Cu ions, while the greenish/yellowish color of CP1 is given by the presence of Fe ions.

Complementary information

None.

Fig. 13: Stratigraphic representation of the sample of bracelet observed in cross-section under dark field using the MiCorr application. The characteristics of the strata are only accessible by clicking on the drawing that redirects you to the search tool by stratigraphy representation. This representation was build according to Fig. 10, credit MiCorr_LMC-CNRS, V.Valbi.

Three CPs were identified with both binocular and cross-sectional observations, but some differences between the two methods of observation should be noted. The outer sediment layer (S1), easily identified under binocular, was barely visible as a thin dark yellow layer in the cross-section and could have been interpreted as part of CP1 if observed only under cross-section.

On the contrary, the presence of a CM1 stratum was only revealed through cross-section observation. Moreover, information obtained under binocular examination such as brightness, compactness, cohesion, and adherence are not accessible during cross-sectional observation. Conversely, the physico-chemical characteristics obtained by cross-section examination are not easily accessible during the binocular examination. The differences observed underline the necessity and complementarity of these two approaches.

The bracelet is a tin bronze with Pb inclusions. The metallographic observation revealed that the metal was annealed with final cold-working

The characterization of the corrosion products showed a typical corrosion structure for an archeological bronze. The object presents the phenomenon of decuprification with Sn enrichment. This is a common phenomenon observed on bronze archaeological objects buried in moderately aggressive natural conditions (such as oxygenated sandy soils) and accompanied by the formation of cuprite at the interface with the metal. Cuprite is the first compound often to be formed during the corrosion of bronzes (Robbiola et al. 1998, Scott 2006). The local enrichment in Fe and P in the CP1 was also previously observed on archaeological objects (Robbiola et al. 1998, Papadopoulou et al. 2016) and can be attributed to the diffusion of these elements from the burial soil. 

The limit of the original surface according to Bertholon's method is probably located on the upper interface between S1 and CP1. CP1 presents hammering traces which are typical markers of the original surface. This surface is partially lost due to the flaking of the layers CP1 and CP2. The conservation of the original surface allows to identify the corrosion form as a Type I according to Robbiola et al. 1998 classification.

This bracelet is part of a corpus of four bracelets (bracelets bus-na 98 BZ 11 / bus-na 98 BZ 23 / bus-na 98 BZ 38 / bus-na 98 BZ 67) found on the same site. These artefacts are all drafts of bracelets but correspond to different stages of advancement: two bracelets are "as cast" and present no further working (bracelet bus-na 98 BZ 23 and bus-na 98 BZ 38) while two bracelets present signs of annealing and cold-working (bracelet bus-na 98 BZ 11 and bus-na 98 BZ 67).

References on object and sample

1. MiCorr_Bracelet bus-na 98 BZ 23
2. MiCorr_Bracelet bus-na 98 BZ 38
3. MiCorr_Bracelet bus-na 98 BZ 67

References on analytical methods and interpretation

4. Lafuente, B., Downs, R. T., Yang, H., Stone, N. (2015) The power of databases: the RRUFF project. In: Highlights in Mineralogical Crystallography, T. Armbruster and R. M. Danisi, eds. Berlin, Germany, W. De Gruyter, 1-30.
5. Ospitali, F., Chiavari, C., Martini, C., Bernardi, E., Passarini, F., Robbiola, L. (2012) The characterization of Sn-based corrosion products in ancient bronzes: a Raman approach. Journal of Raman Spectrpscopy, 43 (11), 1596-1603.
6. Papadopoulou, O., Vassiliou, P., Grassini, S., Angelini, E. and Gouda, V. (2016) Soil-induced corrosion of ancient Roman brass – A case study. Materials and Corrosion, 67, No. 2.
7. Scott, D. (2006) Metallography and microstructure of ancient and historic metals. J Paul Getty Museum Publications.
8. Robbiola L., Blengino M., Fiaud C., (1998) Morphology and mechanisms of formation of natural patinas on archaeological Cu–Sn alloys.