fibula and snake bracelet from albania. a case study by om, sem-eds

SCIENTIFIC CULTURE, Vol. 2, No. 2, (2016), pp. 9-18 Copyright © 2016 SC

Open Access. Printed in Greece. All Rights Reserved.

DOI: 10.5281/zenodo.44895

FIBULA AND SNAKE BRACELET FROM ALBANIA. A CASE STUDY BY OM, SEM-EDS AND XRF.

Olta Çakaj1, Teuta Dilo1, Gert Schmidt2, Nikolla Civici3, Frederik Stamati4

1Faculty of Natural Science, University of Tirana, Albania 2Institut für Keramik, Glas- und Baustofftechnik, TU Bergakademie Freiberg, Germany

3Centre of Applied Nuclear Physics, Albania 4Centre of Albanological Studies, Tirana, Albania

Received: 20/11/2015 Accepted: 17/01/2016 Corresponding author: Olta Çakaj ([email protected])

ABSTRACT

Albania is situated west of the Balkan Peninsula while Shkodra and Durrës are respectively north-western and western cities. The objects of this study are a fibula (IX-VIII BC) found in Shkodra and a snake bracelet (III-I BC) found in Durrës. The purposes of this research paper are to study the chemical composition, pro-duction technology, corrosion products, microstructure and the possible minerals used for the objects’ man-ufacture. To fulfil these tasks the objects’ samples have been analysed with μ-XRF, OM, Vickers microhard-ness tester and SEM-EDS. Both objects resulted authentic copper-tin alloys with less than 2% of lead. They have approximate microhardness values (116.2HV and 113.6HV), with corrosion products from the first and second oxidized layer. The fibula and the snake bracelet has been casted, hot and cold worked to obtain the final shape. Both objects have lead and copper-iron-sulphur inclusions which suggest the possible use of copper-iron sulphite minerals.

KEYWORDS: antique bronze, microstructure, texture, OM, SEM-EDS, μ-XRF

10 Olta Çakaj et al

SCIENTIFIC CULTURE, Vol. 2, No 2, (2016), pp. 9-18

1. INTRODUCTION

This research was performed in order to study the alloy composition of two antique objects, their distinct production technology and the possible mineral used as raw material. This was archived through the examination of the samples’ chemical content, their microhardness, microstructure and its inclusions or secondary phases. Another section of the article covers the qualitative determination of the corrosion products which is essential to prove the objects’ authenticity (not fake replicas) and is helpful for the conservation process.

Albania has a favourable link position between Europe and Asia. Scodra (in Latin, today Shkodra) on the northwest of Albania was the capital of Ardi-aei kingdom since the III century BC. While Dyrra-chium (in Latin, today Durrës) was founded during the XIII - XI century BC and became one of the most important port cities of the western Balkan Peninsula (figure 1a). After the XI century BC the first iron ob-jects began to appear marking the initiation of the Iron Age but it took almost three centuries for this new chemical element to substitute partially bronze in working tools and weapons. During this time copper alloys were more preferable in the produc-tion of ornaments and jewellery. Later on the Hellen-istic Period was associated with the spread of Greek culture influences and trade exchanges during the expansion of the Empire founded by Alexander the Great. (Prendi 1977-1978; Boardman et al. 1982; Jacques 1995; Ceka 2000)

Most of the copper deposits from the Jurassic and Cretaceous periods are present in the volcanogenic massive sulphide compositions. They are located in Albania, Cyprus and a few in Turkey (Kure Asikoy, Murgul Maden, etc.) but are very rare in other coun-tries. (Economou-Eliopoulos et al. 2008; Hoxha 2014) Many chemical mineralogical studies have been per-formed through the years for the Albanian territo-ries. The central north (Rubik) and north (Mirditë) underground resulted rich in copper-sulphide min-erals associated with zinc such as pyrite (FeS2) - chal-copyrite (CuFeS2) - sphalerite [(Zn,Fe)S]. Later on because of the hydrothermal solutions changes and the partial oxygen pressure increments the quantity of chalcopyrite mineral had decreased and bornite (Cu5FeS4) had taken its place (figure 1b). (Çina 1975, 1976; Koçi 1977; Kati 1988) Two ancient and im-portant mines in the region are Rudna Glava of Ser-bia and Ai Bunar of Bulgaria. According to different studies their copper mineral deposits are hydro car-bonate such as azurite (Cu3(CO3)2(OH)2) and mala-chite (Cu2CO3(OH)2). (Jovanovic 1978; Gale 1990) The copper sources were so many and so important

in Cyprus since early times that its name derives from this chemical element. The copper ores of Cy-prus are hydro carbonate and sulphate such as mala-chite (Cu2CO3(OH)2), azurite (Cu3(CO3)2(OH)2) and chalcopyrite CuFeS2. (Constantinou 1982) While in Greece the regions rich in copper ores where is thought that objects production was supported by local copper deposits are Kamariza, Ilarion, Christi-ana, Agia Varvara, Sounion and Lavrion. According to the literature copper was obtained from native copper (Cu), cuprite (Cu2O), chalcopyrite (CuFeS2), malachite (Cu2CO3(OH)2), azurite (Cu3(CO3)2(OH)2), chalcocite (Cu2S), covellite (CuS), olivenite (Cu2AsO4OH), atacamite (Cu2Cl(OH)3) and tetrahe-drite ((Cu,Fe)12Sb4S13) (figure 1a). (Marinos & Pet-rascheck 1956)

Few studies have been performed about Albanian copper based objects this last years while other Bal-kanic countries have studied nearly all their culture heritage with physical and chemical analytical methods. Although Albania has a rich museum fund only a few objects have been studied so far. 122 sil-ver alloy coins from a hoard (III B.C.), minted by the Illyrian king Monounios and the ancient cities of Dyrrachium (Durrës) and Korkyra (Korfuz) was studied with EDXRF. One coin from Dyrrachium (Durrës) and one of Illyrian king Monounios from the III century B.C. resulted copper-silver alloys produced with a cold-struck process after casting. (Pistofidis et al., 2006; Civici et al., 2007) Another coin from Dyrrachium (Durrës) of the III-II century B.C. was determined to be a copper-tin-lead alloy produced in a similar way compared with the coins mentioned above. One nail from Dyrrah used in land (IV-III B.C.) and one from Saranda used in a ship (VI-IV B.C.) resulted pure copper which were hot worked after being casted. (Dilo et al., 2009) A necklace (VII-VI B.C.) and a belt application (VII-VI B.C) from Shkodra were casted copper-tin-lead al-loys while a button (VI-IV B.C) from the same region resulted a casted copper-tin composition. (Çakaj et al., 2014) Some weapons were studied as well such as six swords, three spearheads and a shield from the south-west and south-east of Albania. The shield from Apollonia (Fier, south-west Albania) was a copper-tin alloy produced throught hot working af-ter casting. While one sword from Erseka (south-east Albania), five swords of type Naue II and three speardheads from Korça (south-east Albania) result-ed copper-tin alloys. (Koui et al., 2006; Çakaj et al., 2009) Also some copper alloy sculptures from An-tigonea have been studied for restoration and con-servation reasons. (Budina and Stamati, 1989)

Fibula and snake bracelet from Albania. a case study by om, SEM-EDS and XRF 11


a) b)

Figure 1. a) Political map of the Balkan Peninsula and the Near East. It shows ancient copper mines in Serbia (Rudna Glava), Bulgaria (Ai Bunar), Lavrion (Greece); copper rich regions in Cyprus and in northern Albania (Mirditë). This

study objects were found in Scodra and in Dyrrachium. (http://www.worldatlas.com/webimage/countrys/europe/balkans.htm). b) The extension of FeS2-CuFeS2 and FeS2-

CuFeS2-(Zn,Fe)S ores in Albania. Regions rich in copper ores (Mirditë, Rubik), objects’ excavation locations (Scodra, Dyrrachium), excavation locations of former studied objects (Shkodra, Durrës, Saranda, Apollonia, Antigonea, Erseka,

Korça), the Albanian capital (Tirana). (Economou-Eliopoulos, Eliopoulos, and Chryssoulis 2008.)

2. MATERIALS AND METHODS

Sample from the fibula and the snake bracelet was removed using a small pincers and in both cases it didn’t exceed 6mm3 in volume (Figure 2a). The sampling process was carefully carried out in order to have almost no impact on the object integrity and to collect as much information as possible. The fibula (IX-VIII BC Iron Age, object no. 17960, code 015, file no. 485) dates back to the Iron Age and has a spiral string that forms two round discs 0.065m in diameter connected together (0.145m length from one side to the other) while each disc has a conical button on the centre (the two discs found separated in the excava-

tion). This fibula has no visible clip on the back side and its total length is approximately 2.3m. The snake bracelet (III-I BC. Hellenistic Period, object no. 60205, code 029, file no. 9786) comes from the Hellenistic Period, has a diameter of 0.047m, the height of spi-rals is 0.032m and the total length is 0.7m. The cross section of the fibula is 3.2∙10-6m2 and the one of the snake bracelet is 4∙10-6m2. Both objects have no visi-ble defects and the cross section is approximately the same along the whole length. (Prendi 1958, 2008) Figure 2a) shows the objects’ photos while 2b) shows their sketches.

a) b)

Figure 2. a) Photos of the fibula (left), the snake bracelet (right) along with the positions where each sample was taken, b) the corresponding objects’ sketches.

The examine procedure starts with µ-X ray fluorescence (µ-XRF) performed on 2-3 spots of the objects’ cleaned surface (without varnish) to deter-mine the alloy composition, taking account that the objects are chemically homogeneous. The equipment used was transportable µ-XRF spectrometer ARTAX Bruker (60μm spot diameter, detective capacity from Na to U) with Spectra ARTAX Version 7.2.5.0 and M-Quant-Calib (BRUKER) software. After this the samples (less than 12∙10-9m3) were prepared through mounting in epoxy resin, polishing with SiC

paper (30μm, 18μm, 5μm) and cloth with diamond paste (3μm, 1μm)associated with DP-Lubricant Blue. Vickers tester (attached on the Metalloplan Leitz mi-croscope) was used to measure the microhardness values of the samples. Kozo XJP304 optical micro-scope (OM) with reflected and polarized light along with Sony TCC-8.1 digital camera and View Version 7.3.1.7 image analysis software were used to observe the corrosion products and the microstructure. Scanning electron microscope-energy X ray disper-sive spectroscopy (SEM-EDS) XL30 ESEM-FEI with



EDAX Genesis Spectrum software was used to study the microstructure’s inclusions / secondary phases. In order to observe the microstructure the samples were etched with aqueous ferric chloride solution (H2O : HCl : FeCl3 = 12ml : 3ml : 1gr) for several se-conds. (Goldstein et al. 2003; MIT 2003; Potts & West 2008; Wayne 2009).

3. RESULTS AND DISCUSSIONS

For each of the objects were chosen two spots on the cleaned metal surface and one spot on the corro-

sion products. The metal spots on the fibula were chosen randomly while in the snake bracelet case, the first examine was performed on the snake head meantime the other on the bottom edge. Figure 3 shows the μ-XRF spectra of the first spot examined on each sample while table 1 lists the chemical ele-ments’ quantitative results with the corresponding standard deviations for all the analysis.

a) b)

Figure 3. μ-XRF spectra of the first metal spot examined on a) the fibula and b) the snake bracelet. The alloy results Cu and Sn with minor elements such as Pb, Fe, Ni, As and Cr.

Table I. µ-XRF results (elements percentage with the standard deviation) for a) the fibula and b) the snake

bracelet.

a) Fibula (IX-VIII BC)

b) Snake bracelet (III-I BC)

From the μ-XRF data obtained the fibula and the

snake bracelet resulted copper-tin alloys with lead content around 1.14% and 1.8% respectively. The tin percentage in the fibula’s alloy is around 6.64% and the one in the snake bracelet is around 9.25%. Other chemical elements such as Fe, Ni, As and Cr present less than 0.6% in the metallic alloys may originate

from the soil content where the objects were pre-served through the centuries or from the mineral used for the alloy production. According to the Cu-Sn phase diagram (usual casted condition) the alloy at low temperatures and low tin percentage is com-posed by α and δ eutectoid phases. The Cu31Sn8 δ phase is a crystalline solid intermetallic compound which tends to be very hard with no ductility. The composition of Cu and Sn in the δ phase is fixed (Cu:Sn=67.47%:32.53%) while the α phase can absorb more or less Sn according to the alloy consistence. Furthermore for bronzes with tin between 5-10% the adding of lead makes the alloy less viscous. For very low lead content such as this study objects the in-crease of the tin percentage causes the decrease of the material ductility or the elongation value (fibula 40%-50% and snake bracelet 20%-30%). (Scott 2012) The Cu percentage decreases in the outer surface while Sn percentage increases because of their dif-ferent diffusion coefficients and the formation of cor-rosion products of both elements (oxide and hydrox-ide). (Dillmannet al. 2007) On the corrosion spots oxygen is not detectable with μ-XRF and its percent-age is added by the software to the other present detectable elements (overestimation) in order to have a total of 100%. This can be observed in the μ-XRF results of the corrosion spots analysed on each sample. Figures 4, 5 show the corrosion products observed with polarized light OM for the fibula’s sample and the snake bracelet’s one respectively.

Element Metal spot 1 Metal spot 2 Corrosion spot

Cu (%) 91.79±2.66 90.95±4.28 28.5±0.8 Sn (%) 6.35±0.38 6.93±0.6 53.49±1.91 Pb (%) 0.95±0.08 1.34±0.16 5.57±0.33 Fe (%) 0.55±0.02 0.42±0.03 3.33±0.11 Ni (%) 0.21±0.01 0.18±0.02 0.32±0.02 As (%) 0.16±0.04 0.18±0.07 1.71±0.17 Ca (%) – – 7.07±2.84 Cr (%) – – –

Element Metal spot-

head Metal spot-

bottom

Corrosion spot

Cu (%) 89.11±4.3 88.26±3.32 55.77±1.78 Sn (%) 9.05±0.76 9.46±0.65 37.68±1.69 Pb (%) 1.55±0.12 2.06±0.12 6.04±0.25 Fe (%) 0.24±0.02 0.21±0.01 0.5±0.03 Ni (%) – – – As (%) – – – Ca (%) – – – Cr (%) 0.04±0.01 – –



Figure 4. Fibula’s corrosion products, photo obtained using OM with crossed nicols polarized light (left) and parallel nicols polarized light (right). Corrosion products introduced far inside the alloy. Present main phases that differ in

colours are CuO, Cu2O and Cu2CO3(OH)2.

Figure 5. Snake bracelet’s corrosion products, photo obtained using OM with crossed nicols polarized light (left) and parallel nicols polarized light (right).Corrosion products introduced far inside the alloy. Present main phases that differ

in colours are CuO, Cu2O and Cu2CO3(OH)2.

It is clear from figures 4, 5 that the corrosion products are introduced far inside the alloy. Differ-ent colours are visible on the polarized light OM photos (crossed nicols) for both samples such as bright and dark red mixed with brown areas over-lapped by bright and dark green. The observation of corrosion for both samples (figure 4, 5) leads to the conclusion that the objects are authentic (not fake replicas) because of these products’ introduction far inside the alloy which can’t be done artificially in a short time.

Based on the different corrosion colours from the OM crossed nicols polarized light photos the main phases present might be tenorite CuO (brown), cu-prite Cu2O (red) from the first corrosion layer and malachite Cu2CO3(OH)2(green) from the second one. (Pracejus 2008) The first corrosion layer is formed during the production process and the object’s us-age. It is mainly composed by tenorite CuO (dark brown with red internal reflections), cuprite Cu2O (bright/dark red) and SnO / SnO2 / Sn(OH)2 (grey colour shades). The second corrosion layer overlaps the first one and appears during the end of the usage period. The distinguish green colour comes from malachite Cu2CO3(OH)2 but also from nantokite CuCl and antlerite Cu3SO4(OH)4. When the percent-age of CO2 is high in contact with the object other phases are formed such as azurite Cu3(CO3)2(OH)2 (blue) and cerussite Pb2CO3(OH)2 (light grey). Last comes the third corrosion layer which is formed dur-ing the object’s burial and includes compounds from

the surrounding soil. The corrosion phases can be identified with XRD analysis or the main ones that differ in colours can be specified through observing the corrosion products with a microscope (polarized light) along with an optical mineralogical manual. (Vink 1986; Pracejus 2008; Sandu et al. 2008, 2014).

Table 2, 3 shows the EDS results of all examined areas and points for the fibula’s and snake bracelet’s sample.

Table 2. EDS results for the fibula (mean value of standard deviation ±0.3%).

Element Point A Area B Area C Area D

Cu (%) 08.00 91.81 65.36 06.62 Pb (%) 86.53 – – 53.81 O (%) 05.46 – – 21.67 Sn (%) – 08.19 – 10.83 S (%) – – 25.67 07.07

Fe (%) – – 08.97 –

Table 3. EDS results for the snake bracelet (mean value of standard deviation ±0.3%).

Element Area A Area B Point C

Cu (%) 61.84 91.24 79.21 Pb (%) 29.31 – – O (%) – – – Sn (%) 07.07 08.76 – S (%) 01.78 – 19.42

Fe (%) – – 01.37

Figures 6, 7 show the SEM images of with the

EDS examined areas or points along with the corre-sponding spectra for both samples.



Figure 6. Photos obtained with SEM of the fibula’s sample along with the EDS area/point analysis and spectra. The bright inclusions are rich in Pb while the dark ones are Cu-Fe-S inclusions. The alloy is composed by Cu-Sn and the

corrosion area contains oxidised Cu, Pb, S and Sn.

Figure 7. Photos obtained with SEM of the snake bracelet’s sample along with the EDS area/point analysis and spectra. The bright inclusions are rich in Pb but also S while the dark ones are Fe-S inclusions. The alloy is composed by Cu-Sn.

In both samples the EDS examination on the alloy (area B in both objects) confirms the μ-XRF results taking account the standard deviation as well. Bright and dark inclusions are visible in the SEM photos of both alloys. After the EDS analysis all the bright in-clusions resulted rich in Pb while all the dark ones are mainly composed of Cu-Fe-S. Although only one EDS analyse of these inclusions for each sample is shown in this study.

If the analysed area/point is smaller than 5μm (the average dimension of the interaction volume) is probable that the detection of the alloy elements Cu and Sn comes from the surrounding area. This is the case of the examined point A on the fibula sample and area A / point C on the snake bracelet sample. Only one corrosion area was analysed on the fibula sample (area D) and among the chemical elements

(table 2) were also the ones detected with μ-XRF on the corrosion products (table 1).

The present of Cu-Fe-S inclusions in the alloy may be an indication of sulphide mineral usage in the production procedure. This mineral mainly con-sists of copper sulphide and copper-iron sulphide compounds that do not mix well with metallic cop-per and one of which is bornite Cu5FeS4. (Shugar 2003; Ashkenazi et al. 2012) A comparison between the Cu-Fe-S experimental percentages of the fibula’s area C with the elements’ weigh consistence in Cu5FeS4 was made. Area C is composed by 65.36%Cu - 25.67%S - 8.97%Fe (table 2) while the weight percentages of these elements in Cu5FeS4 are 63.32%Cu - 25.55%S - 11.13%Fe. This can suggest the possible use of bornite in order to produce the fibu-la’s alloy. (Figueiredo et al. 2011)



Table 4 shows the Vickers microhardness results along with the mean value for both samples. The fibula sample has a mean Vickers microhardness value of 116.2HV while the snake bracelet’s one is around 113.6HV (absolute error ±1HV). Adding ad-ditional chemical elements in copper (alloying) and submitting the metal to different mechanical / ther-mal processes can increase the Vickers microhard-ness value from 40-50HV (pure casted Cu) up to 220HV (fully worked Cu-Sn alloy). For the same per-centage of Sn the adding of Pb can slightly decrease the microhardness and on the other hand it increases after the order of the processes: casted → worked and annealed → cold worked. (Scott 1991)

Table 4. Measured and calculated mean values of the Vickers microhardness test (absolute error in measuring

the Vickers microhardness is ±1HV).

HV Fibula (IX-VIII BC) Snake bracelet (III-I BC)

Test 1 117 123 Test 2 109 121 Test 3 116 114 Test 4 121 115 Test 5 119 107 Test 6 115 109 Test 7 – 106

Mean value 116.2 113.6

Figure 8 a), b) and figure 9 a), b) show the fibula’s

and the snake bracelet’s respective microstructure using OM with reflected light and SEM.

a) b)

Figure 8. Fibula’s microstructure, photos obtained using: a) OM with reflected light etched for 1-2 seconds and b) SEM. The OM photo reveals twins (1), strain lines (2) and (3) elongated / orientated δ phase in the microstructure while the

both OM - SEM photos show elongated / orientated Pb inclusions (4).

a) b)

Figure 9. Snake bracelet’s microstructure, photos obtained using: a) OM with reflected light etched for 5 (above) and 7 (below) seconds and b) SEM. The OM photo reveals twins (1) and strain lines (2) in the microstructure while the SEM

photos show no texture.

According to the fibula’s OM reflected light pho-to the presence of twins and some strain lines (figure 8a) suggests that the alloy might have been firstly casted then cold worked and at the end annealed (hot worked). Another possible production process starts with casting then annealing and last cold / hot working. Adding Pb in Cu alloys increases the fluid-ity during casting but it’s not suitable in high per-centage for bronzes that will be hammered because lead can decrease the alloy ductility. This happens because Pb is not soluble in Cu, it usually segregates in grain boundaries and during hammering the propagation of cracks can occur along Cu-Pb inter-faces. It is clear from the OM and SEM figure 8 a), b) that the Pb and Cu-Fe-S inclusions are elongated and orientated in one direction evidencing a texture mi-crostructure caused by probable hammering. Taking

into account that lead has a significantly lower melt-ing point compared to copper, the most probable production technique might be the one where the alloy after casting has been hot worked in order to obtain the object’s final shape. (Mortazavi et al. 2011; Ashkenazi et al. 2012)

On the snake bracelet’s microstructure (figure 9a) from the OM photos with reflected light are clearly visible twins and strain lines. According to the litera-ture the alloy has been initially casted, then it might has followed these possible processes: 1) cold worked, annealed and again cold worked at the end or 2) hot and cold worked to obtain the final shape. On the SEM photo (figure 9b) of the snake bracelet’s sample no texture is visible, meaning that the Pb and Cu-Fe-S inclusions have no orientation and are not elongated in one specific direction.



Based on the literature after the alloy is casted the microstructure is composed by dendrites (like fern-like growth) which can lead to bent ones after cold working or to hexagonal grains after the alloy has been annealed (hot worked). If the bent dendrites microstructure is annealed (hot worked) a twinned one will be obtained. On the other hand the twinned microstructure can also derive from hot working the hexagonal grains’ alloy. The increase of the alloy strength from cold working is higher compared with the one from hot working but in both cases the duc-tility of the material decreases. Further cold working can bend the twins and create stain lines. (Scott 2012)

4. CONCLUSIONS

Different analytical methods were used in this study in order to have the fibula’s and snake brace-let’s characterization as complete as possible.

From the μ-XRF examination both alloys are α + δ phases Cu-Sn with the addition of low percentage of Pb. The fibula is composed by 6.64% Sn and 1.14% Pb meanwhile the snake bracelet has a higher con-sistence of the two elements respectively 9.25% Sn and 1.8% Pb. Soil chemical elements are present in both samples such as Fe, Ni, As, Ca, Cr.

The EDS results of the alloy area confirm the μ-XRF analysis. The SEM-EDS analysis show the pres-ence of Pb and Cu-Fe-S inclusions. This result and based on the geological - mineralogical studies of

Albanian copper ores brings to the assumption that local sulphite minerals might have been used as raw material.

According to the OM photos with polarized light the first and second corrosion layer main phases are present in both samples. Based on these products’ insertion far inside the alloy both objects result au-thentically antique ones.

The SEM and OM with reflected light photos show the presence of twins, strain lines, orientated elongated Pb and Cu-Fe-S inclusions, δ phase and pores in the fibula’s microstructure. The presence of twins and strain lines is very clear on the OM photos with reflected light of the snake bracelet’s sample. This might be explained if the objects after casting have been hot worked probably by hammering and then cold worked where the final shape has derived. The Pb and Cu-Fe-S inclusions in the snake bracelet’s sample (SEM photo) result not elongated and orien-tated. The Vickers microhardness value is 116.2HV for the fibula and 113.6HV for the snake bracelet.

These objects are evidence of unique and distinct handicraftsmen skills in metal processing since the IX-VIII BC in Albania. By using simple tools such as hammers and furnaces the handicraftsmen were able to produce beautiful wire-like ornamental objects with considerable length and no visible defects.

ACKNOWLEDGEMENTS

The authors are very grateful to the Archaeological Department, Historical Museum of Shkodra and the In-stitute of Cultural Monuments in Tirana for providing the objects for this study; the Centre of Applied Nu-clear Physics, the Faculty of Natural Sciences, UT, Albania and the Institut für Keramik, Glas- und Baustofftechnik, TU Bergakademie Freiberg, Germany for their support with analytical equipment and fund-ing. The authors would like to thank DAAD for financially sustaining the analysis performed in Germany and project coordinator Prof. Dr. Thomas Bier. Last but not least the authors are grateful for the unremitting collaboration of Mr Zamir Tafilica (archaeologist) and Mrs Edlira Çaushi (restorer).

REFERENCES

Ashkenazi, D., Iddan, N. and Tal, O. (2012) Archaeometallurgy characterization of Hellenistic metal objects. The construction of the bronze objects from Rishon Le-Zion (Israel). Archaeometry 3, Vol. 54, 528-548.

Boardman, J., Edwards, I. E. S., Hammond, N. G. L. and Sollberger, E. (1982) The Cambridge ancient history: The Prehistory of the Balkans; and the Middle East and the Aegean world, tenth to eighth centuries BC. UK, Cambridge University Press, 209-231.

Budina, Dh. and Stamati, F. (1989) Disa objekte bronzi nga Antigoneja. (Some bronze objects from Antigo-nea.) Monumentet, 2.

Ceka, N. (2000) Përpara se të shkruhej historia (Before history was written). Albania, SHBLU Publ, 36-37. Çakaj, O., Dilo, T., Civici, N. and Stamati, F. (2009) Përbërja elementore dhe mikrostruktura e një shpate dhe

një mburoje antike prej bronzi. (Elements composition and microstructure of a bronze antique sword and shield.) Buletini i Shkencave Natyrore, 8, 5-14.

Fibula And Snake Bracelet From Albania. A Case Study By OM, SEM-EDS And XRF. 17


Çakaj, O., Duka, E., Tafilica, Z., Stamati, F., Civici, N. and Dilo, T. (2014) Characterisation of copper alloy antique ornamental accessories found in northern Albania. Open Journal of Archaeometry, Vol. 2, 129-133.

Constantinou, G. (1982) Geological features and ancient exploitation of the cupriferous sulphide ore bodies of Cyprus. Proceedings of the Symposium Early metallurgy in Cyprus 4000-500 B.C., Cyprus, GreeceEconomou-Eliopoulos, M., Eliopoulos, D. G. and Chryssoulis, S. (2008) A comparison of high Au massive sulphide ores hosted in ophiolite complexes of the Balkan Peninsula with modern analogues: Genetic significance. Ore Geology Reviews, 33, 81-100.

Civici, N., Gjongecaj, Sh., Stamati, F., Dilo, T., Pavlidou, E., Polychroniadis E. K. and Smit, Z. (2007) Compo-sitional study of III-rd century BC silver coins from Kreshpani hoard (Albania) using EDXRF spec-trometry. Beam interaction with Materials & Atoms, Nuclear Instruments and Methods in Physics Re-search, B 258, 414-420.

Çina, A. (1975) Gërshetimet ndërmjet mineraleve dhe shkalla e çlirimit të tyre në xeherorët sulfurorë të bakër-zinkut të disa vendburimeve të vendit tone. (Interlacing between minerals and their release rate in sulphur ores of copper-zinc deposits of our country.) Përmbledhje studimesh, 2, 107-120.

Çina, A. (1976) Mbi zonalitetin vertical të pjesës veriore të vendburimit kolçedan të bakrit në Rubik. (Study on vertical zonality of northern kolçedan copper deposit in Rubik.) Përmbledhje studimesh, 3, 107-116.

Dilo, T., Civici, N., Stamati, F. and Çakaj, O. (2009) Archaeometallurgical characterization of some ancient copper and bronze artefacts from Albania. American Institute of Physics, 985-990.

Gale, N. H. (1990) Metals and Metallurgy in the Chalcolithic Period. Proceedings of the Symposium Chalcolithic Cyprus, Cyprus, Greece.

Hoxha, L. (2014) Compiling of lithologo stratigraphic-tectonic sections, increase substantially resources of Cu-Fe, Cu-Fe, Zn, Au, Ag sulphides hosted by volcanogenic massive sulphide deposits (VMS), in Albanian ophiolites. Proceedings of the Geological Society of America, USA.

Jacques, E. E. (1995) The Albanians - an ethnic history from prehistoric times to the present. USA, Mc Farland & Company, Inc, 2-28.

Jovanovic, B. (1978) The oldest copper metallurgy in the Balkans. A study of the diffusion of copper from Asia Minor to Southeastern Europe. Expedition, 9-17.

Koçi, M. (1977) Ndryshimet anësore përreth trupave xeherorë të bakrit në bazë të studimeve mineralogo-petrografike. (Lateral changes of copper ore bodies under mineralogy-petrography studies.) Përm-bledhje studimesh, 2, 37-47.

Koui, M., Papandreopoulos, P., Andreopoulou-Mangou, E., Papazoglou-Manioudaki, L., Priftaj-Veveçka, A. and Stamati, F. (2006) Study of Bronze Age copper-based swords of type Naue II and spearheads from Greece and Albania. Mediterranean Archaeology and Archaeometry 1, Vol. 6, 49-59.

Kati, P. (1988) Disa veçori kimiko-mineralogjike të mineraleve kryesore të pasurimit dytësor të bakrit në vendburimet e Mirditës. (Some chemical-mineralogical characteristics of the main minerals on sec-ondary enrichment Mirdita copper deposits.) Buletini i shkencave gjeologjike, 3, 67-74.

Marinos, G. P. and Petrascheck, W. F. (1956) Lavrion, geological and geophysical research. IGME Athens 1, Vol. IV, 149-206.

Pistofidis, N., Vourlias, G., Pavlidou, E., Dilo, T., Civici, N., Stamati, F., Gjongegaj, S., Prifti, I., Bilani, O., Stergioudis, G. and Polychroniadis, E. K. (2006) On the comparative study of three silver coins of the III-rd century B.C. minted in Korkyra, Dyrrachion and by the Illyrian king Monounios. Applied Physics A Material Science & Processing, 83, 637-642.

Prendi, F. (1958) Materiale të kulturës ilire të zbulueme në Shqipërinë e veriut (Materials of Illyrian culture discov-ered in northern Albania). Albania, BUSHT, 110-132.

Prendi, F. (2008) Studime Arkeologjike (Archaeological studies). Kosova, ARK-KOS, 453-474. Goldstein, J. I., Newbury, D. E., Echlin, P., Joy, D. C., Lyman, C. E., Lifshin, E., Sawyer, L. and Michael, J. R.

(2003) Scanning Electron Microscopy and X-Ray Microanalysis. USA, Kluwer Academic/Plenum Pub-lishers.

Massachusetts Institute of Technology (2003) The Metallographic Examination of Archaeological Artifacts, Laboratory Manual. Summer Institute in Materials Science and Material Culture, USA. Available at: http://ocw.mit.edu/courses/materials-science-and-engineering/3-094-materials-in-human-experience-spring-2004/laboratories/manual_suppl.pdf

Prendi, F. (1977-1978) Epoka e bronzit në Shqipëri. (Bronze Age in Albania.). Iliria, VII-VIII, 5-58.



Potts, P. and West, M. (2008) Portable X-ray Fluorescence Spectrometry Capabilities for In Situ Analysis. UK, The Royal Society of Chemistry.

Wayne, R. (2009) Light and Video Microscopy. USA, Elsevier Inc. Scott, D. A. (2012) Ancient Metals: Microstructure and Metallurgy, Volume I, Copper and Copper Alloys. USA, D.

A. Scott. Dillmann, P., Béranger, G., Piccardo, P. and Matthiesen, H. (ed.) (2007) Corrosion of metallic heritage artefacts.

Investigation, conservation and prediction for long-term behaviour. England, Woodhead Publishing, 240-243.

Pracejus, B. (ed.) (2008) The ore minerals under the microscope, an optical guide. Hungary, Elsevier. Vink, B. W. (1986) Stability relations of malachite and azurite. Botswana. Mineralogical Magazine, Vol. 50, 41-

47. Sandu, I., Ursulescu, N., Sandu, I. G., Bounegru, O., Sandu, I. C. A. and Alexandru, A. (2008) Pedological

stratification effect of corrosion and contamination products on Byzantine bronze artefacts. Corro-sion Engineering, Science and Technology 3, Vol. 43, 256-266.

Sandu, I., Mircea, O., Sandu, A. V., Vasilache, V. and Sandu, I. G. (2014) Study of the Liesegang chemical ef-fects in antique bronze artefacts during their stay within an archaeological site. Rev. Chim. (Bucha-rest) 3, Vol. 65, 311-319.

Shugar, A. N. (2003) Reconstructing the chalcolithic metallurgical process at Abu Matar, Israel. Proceedings of the International Conference Archaeometallurgy in Europe, Vol.1, Milan, Italy, 449-458.

Figueiredo, E., Valério, P., Araújo, M. F., Silva, R. J. C. and Soares, A. M. M. (2011) Inclusions and metal composition of ancient copper-based artefacts: a diachronic view by micro-EDXRF and SEM-EDS. X-Ray Spectrom., Vol. 40, 325-332.

Scott, D. A. (1991) Metallography and microstructure of ancient and historical metals. Singapore, The J. Paul Getty Trust.

Mortazavi, M., Salehikakhki, A., Golozar, M. A. and Tala’i, H. (2011) Preliminary metallurgical investigation of copper-based artefacts at Tepe Sagzabad in Qazvin Plain, Iran (1500-800 B.C.). Iranian Journal of Archaeological Studies 2, Vol. 1, 49-59.

fibula and snake bracelet from albania. a case study by om, sem-eds

Documents