of
ISSN (1897-3310)
Volume 15
Special Issue
4/2015
FOUNDRY ENGINEERING
23 – 28
ARCHIVES
Published quarterly as the organ of the Foundry Commission of the Polish Academy of Sciences
5/4
Copper and Arsenical Copper During
Eneolithic in Metallographic and Mechanical
Properties Examination
A. Garbacz-Klempka a*, J. Kozanaa, M. Piękośa, W. Cieślaka, M. Perek-Nowakb, Ł. Kowalskic, K. Adamczakc, J. Łośd
a
AGH-University of Science and Technology, Faculty of Foundry Engineering, Historical Layers Research Centre
Reymonta 23, 30-059 Kraków, Poland
b
AGH-University of Science and Technology, Faculty of Non-Ferrous Metals, Mickiewicza 30, 30-059 Kraków, Poland
c
Nicolaus Copernicus University, Institute of Archaeology, Szosa Bydgoska 44/48, 87-100 Toruń, Poland
d
Leon Wyczółkowski Bydgoszcz District Museum, Gdańska 4, 85-006 Bydgoszcz, Poland
*Corresponding author. E-mail address: agarbacz@agh.edu.pl
Received 20.11.2015; accepted in revised form 29.12.2015
Abstract
Arsenical copper has been used since 5th millennium cal.BC, later exchanged by application of Cu-Sn alloys in metallurgy. This work presents
the results of metallographic and mechanical properties studies performed on two flat axes connected with local Eneolithic societies (4500-3000
cal.BC). The axes are one of the oldest metal artifacts from Polish land. Originally they were made from Cu-As alloy, and their chemical
composition was established by X-ray fluorescence spectroscopy (XRF). Their microstructure was analysed using optical microscopy (OM) and
scanning electron microscopy conducted with energy-dispersive X-ray spectroscopy (SEM-EDS). The macrostructure analysis (OM) of the axes
was performed as well. On the basis of the results, the alloys used in the Eneolithic to cast the axes were reproduced in lab. In order to achieve the
characteristics of the alloys, their mechanical properties including ultimate tensile strenght (UTS), hardness (HB), microhardness (HV0,1) and
ductility were examined. The solidification process was studied by means of thermal analysis.
Keywords: Non-Destructive Testing, Copper, Arsenical Copper, Archaeometallurgy, Casting, Metallography, Eneolith
1. Introduction
The information on material structure and production technology
of the prehistoric metal artifacts can throw light on when they were
made, where they originate and their probable destination.
Microstructure research is relevant for documentation of the
prehistoric alloys, determining the condition of cast artifacts and
designing the program of their conservation and protection. The
elements that remain in the alloy as the consequence of copper
smelting are relevant for researching the alloy’s origin; the following
elements are common in such alloys: Fe, Co, Ni, Zn, As, Ag, Sn, Sb,
Pb and Bi. It means, that the ratio of concentration of these elements
to copper in the copper ore and concentration in the final product
makes their correlation possible [1].
Prehistoric artifacts were obtained by casting and forging.
Reforging of an already cast object could happen in order to change
its shape or improve its mechanical properties.
Arsenical copper has been used since 5th millenium cal.BC, later
exchanged by application of Cu-Sn alloys in metallurgy. Arsenic was
an element naturally introduced to alloys with no control in
concentration as it was incorporated in minerals, arsenides of copper:
Cu3AsS4 (enargite), Cu12As4S13 (tennantite), Cu3As (domeykite) or
Cu5.2-8As (algodonite) [2]. Today, an attempt is analysis and made
towards reproduction of those earliest Cu-As alloys basing on the
careful analysis of their composition. [1, 3-6]
ARCHIVES of FOUNDRY ENGINEERING Volume 15, Special Issue 4/2015
23
Arsenic is an impurity difficult to remove from liquid copper in
the metallurgical process [7-8]. It has detrimental effect on both
electrical and thermal conductivity of copper [9]. The addition of
arsenic can also lower plastic properties of copper thus worsening the
plastic properties of the material. But in some cases arsenic is
considered to be an advantageous alloying additive [10]. In copper it
improves the resistivity of alloys to corrosion [9]. Copper with the
content of 0.3-0.5% As is used nowadays for chemical equipment and
also in some appliances resistant to oxidation in increased
temperatures [11]. Mechanical properties of alloyed copper are
shaped respectively by heat treatment and plastic working.
2. Materials for research, methodology
and experiments
2.1. Materials
Two Eneolithic flat axes are the object of this study. The first
of them was discovered about 1890 in Augustowo, gm. Krajenka,
pow. Złotów (refer to Fig. 1a; stored in the Bydgoszcz District
Museum under ACCN: MOB/A-374) [12-14]. The second one
was obtained about 1909 in Jezuicka Struga, gm. Rojewo, pow.
Inowrocław (Fig. 1b; stored in the Bydgoszcz District Museum
under ACCN: MOB/A-375) [14,15]. According to A. Szpunar the
axe from Augustowo is of Bytyń A type while the axe from
Jeziucka Struga is of Strzelin B type [14]. Having no
archaeological context, both axes should be treated as single
finds. Due to their typological features they may be placed
arbitrarily in the period of 4500-3000 cal.BC and must be
considered as imports.
Fig. 1. The Eneolithic flat axes: (a) Bytyń A type; (b) Strzelin B type
2.2. Experimental metod
The quantitative determination of the elemental composition was
performed by X-ray fluorescence spectrometry (XRF) with energy
dispersive X-ray fluorescence spectrometer SPECTRO Midex. The
objects were prepared by the removal of the corrosion products.
The macrostructure analysis was performed using a NIKON
SMZ 745Z stereoscopic microscope with a Nikon Digital Sight DsFi1
microscopic camera and a Nis-Elements BR picture analysis system.
The macrostructure of the artifacts was observed with respect to the
state of preservation and surface quality.
Their microstructure was analysed using a NIKON ECLIPSE
LV150 optical microscopy (OM) and scanning electron microscopy
(SEM, Hitachi S-3400N) eqiupped with energy-dispersive X-ray
spectrometer (EDS Thermo Noran). The macrostructure analysis
(OM) of the axes was performed as well. On the basis of the
24
quantitative analysis, the alloys used to cast the axes in the Eneolithic
period were reproduced in laboratory.
The foundry mixture was prepared from pure components (Cu,
As). The successive four melts were obtained in an induction furnace,
in a chamotte melting pot: (0) Cu, (1a) Cu+0.5% As,
(1b) Cu+0.6% As, (2) Cu+1.8% As. The pouring temperature fell
within the range 1200 – 1280oC. The oxygen concentration was
assessed based on the Leco system. In order to achieve the
characteristics of the lab alloys, their mechanical properties including
ultimate tensile strength (UTS), hardness (HB), microhardness
(HV0.1) and ductility were examined. The solidification process was
studied by means of thermal analysis.
The impact of arsenic on copper was analysed in four stages:
I – as cast (C)
II – after casting and heat treatment in the temperature of 950oC 1h
and 500oC 0.35h with water cooling and with 500oC 2h furnace
cooling (C+HT)
III – as cast and after plastic deformation by cold forging (C+F)
IV – as cast, after cold plastic deformation and heat treatment,
subsequently in the temperatures of 300oC, 400oC and 500oC for two
hours and slow cooling (C+F+HT).
At every stage the material hardness was controlled (HB
and HV0.1). The microstructure character was analysed using optical
microscopy (OM) and Scanning Electron Microscopy with X-ray
microanalysis (SEM-EDS).
3. Results description
3.1. Research of artefacts
The analysis of chemical composition showed that both copper
and arsenical copper were applied in manufacturing of the flat axes
(Tab.1). Table 1 presents the results of elemental composition taken
from quantitative analysis f the axes including the results obtained for
the flat axe from Augustowo published in 1924 by Kostrzewski [13].
Having a content of mercury (Hg) below a detection limit
(<0.001 wt%) together with a higher levels of cobalt (Co) and nickel
(Ni), the chemical profiles suggest that the axes were made of recast
copper ores [16]. The average iron (Fe) content in the objects did not
exceed 1 wt% which seems to be typical for the Eneolithic and the
Early Bronze Age artifacts [17].
The major contribution in the chemical profile of the axes was
made by copper (Cu) and in both cases it was completed by arsenic
(As), although the major secondary element contribution in Strzelin B
type profile was made by antimony (Sb).
The macroscopic observations confirmed the presence of the
casting defects on the surfaces of both axes. They were identified as
the gating system (Fig. 3b) and cast seams pointing to mould parting
line (Fig. 2a). Slag inclusions were present on the surface of the
Strzelin B type axe. Such structures allow us to assume that both axes
were cast in two-part closed moulds. Surface defects indicate the
reactions taking part at the metal and mould boundary, resulting from
the presence of moisture in mould and oxygen in metal. The
macroscopic observation pointed out the usage traces on the Bytyń A
type axe's blade which probably resulted from its usage in the
prehistory. The plastic forming of both axes was also recognized. For
improving the metal structure and refining their shapes the axes were
probably cold hammered just after being cast.
ARCHIVES of FOUNDRY ENGINEERING Volume 15, Special Issue 4/2015
Table 1. The elemental composition (wt%) of the axes.
Axe type
Fe
Co
Ni
Cu
As
Se
Ag
Sn
Sb
Te
Pb
Bi
The Bytyń A (2015)
0.032
0.048
0.064
99.1
0.77
0.0032
< 0.020
< 0.051
< 0.051
0.00091 < 0.020
0.012
The Bytyń A (1924)
-
N/A
+
99.2
0.51
N/A
N/A
0.05
-
N/A
+
N/A
The Strzelin B
< 0.025
0.045
0.073
99.7
0.020
0.0072
0.0096
< 0.051
0.055
0.0010
< 0.020
0.015
3.2. Discussion
Due to the fact that Kostrzewski did not specify the applied
analytical method [13], the comparison of the chemical profiles of
the flat axe from Augustowo encountered some difficulties,
nevertheless, both profiles correspond quantitatively to each
other.
Fig. 4. Microstructure of prehistoric flat axe cast
and plastically deformed (forged) Strzelin B. (OM)
Fig. 2. The macrostructure of the Bytyń A type:
(a) the cast seam; (b) the blade usage traces
Fig. 5. Microstructure of prehistoric flat axe cast
and plastically deformed (forged) Bytyń A. (SEM)
Fig. 3. The macrostructure of the Strzelin B type:
(a) the cast defect; (b) the gating system.
The microstructure of the Strzelin B type axe, suggesting the
application of plastic deformation after casting, is presented in
Figure 4. In the microstructure of Strzelin B axe (0.02wt% As) copper
grains were observed, which underwent deformation. In the
microstructure of Bytyń A axe (0.77wt% As) the outlines of α solid
solution of arsenic crystallites were noticed, as well as numerous
impurities coming form ores (Fig. 5-6).
Additionally, marks of plastic deformation as well as corrosion
changes were observed (Fig. 6). Microhardness tests for both axes
yielded the the same value of 76HV0.1
Fig. 6. Microstructure of prehistoric flat axe cast
and plastically deformed (forged) Bytyń A. (SEM)
ARCHIVES of FOUNDRY ENGINEERING Volume 13, Issue 1/2013
25
3.3. Results and discussion
Based on the research of the prehistoric axes, the experiment
of recreating the model alloys with a variable arsenic content and the
evaluation of arsenic impact on the material properties. The chemical
content was controlled after the melt (Tab. 2)
It follows from the Cu-As phase diagram (Fig. 7), that the
maximum solubility of arsenic in copper equals 7.96 wt% in the
temperature of 685 oC and it decreases in step with decreasing the
temperature. Arsenic additives influence slightly but noticeably the
decrease of crystallization temperature, which is confirmed by both
the Cu-As phase diagram and by the cooling curve recorded for the
materials tested during the experiment (Fig. 8).
The conducted thermal analysis made it possible to register the
characteristic transformations taking place during the solidification of
the materials tested. In the case of pure copper solidifying at the
temperature of T1=1084 oC, it is worth pointing to the temperature of
T2=1066 oC, characteristic for the crystallisation of oxygen eutectic
(Fig. 9).
This eutectic disappears after adding arsenic, which is confirmed
by the decreased oxygen content in Cu-As in comparison to Cu
(Tab. 3).
Changes in crystallising conditions caused by the arsenic additive
determine the microstructure character. Already with 0.5% As content
(CuAs-1a) in as-cast condition (Stage I Casting) the dendritic
structure of solid solution is formed αCu(As) (Fig. 10-11).
Fig. 8. Thermal characteristics of copper with As additives: for Cu
T1=1084oC; T2=1066oC, for Cu+0.6%As T1=1079oC, and for
Cu+1.8%As T1=1070oC.
Fig. 9. Copper-oxygen phase diagram [19].
Arsenic locates itself at farther distances from the main dendrite
axes and in interdendritic spaces, which is confirmed by chemical
content analysis in microareas (Fig. 11). The structure of the other
samples, where the arsenic content was 0.6% (CuAs-1b) and 1.8%
(CuAs-2) is similar.
Mechanical properties: ultimate tensile strength Rm, hardness HB
and microhardenss HV0.1 of the as-cast material, independent
of the arsenic content, are similar (Figs 12-14). However, the arsenic
addition decreases the plasticity of the material.
Fig. 7. Copper-arsenic phase diagram [18]
Table 2. The elemental composition (wt%) of the alloys.
Elements (wt%)
Fe
Co
Ni
Cu
As
Zn
Ag
Sn
Sb
Te
Pb
Bi
Cu-0
<0.025
0.04381 0.0596
99.62
0.00273 0.1155
0.0057
<0.051
<0.051
0.00100 <0.020
<0.0010
CuAs-1a
<0.025
0.0432
0.0616
99.04
0.5051
0.1407
0.1446
<0.051
<0.051
0.00092 <0.020
0.0093
CuAs-1b
<0.025
0.04830 0.0615
99.02
0.6438
0.1283
<0.020
<0.051
<0.051
0.00096 <0.020
<0.0010
CuAs-2
0.1095
0.0451
97.80
1.771
0.1515
0.0739
<0.051
<0.051
0.00095 <0.020
<0.0010
26
0.631
ARCHIVES of FOUNDRY ENGINEERING Volume 15, Special Issue 4/2015
Table. 3. Oxygen content in copper and alloying copper
Sample
Cu-0
CuAs-1b
CuAs-2
As content (wt%)
0.0
0.6
1.8
O2 content (ppm)
340
160
150
concluded that in these conditions arsenic strengthens the copper
structure.
During the III stage of work, the material cast earlier (C+F) was
plastically deformed by cold hammering. This process significantly
increased the material hardness in step with higher arsenic content
(Figs 13-14). During the IV stage, which was based on heat treatment
of the cast and forged alloy (C+F+HT) soaking at the temperature
range of 400-500 oC caused a distinct increase in hardness with the
addition of 0.5wt.%As (CuAs 1a) as well as a change in microstructure in comparison to the as-cast structure of the material (Figs.
15-16). This effect did not take place at the temperature of 300 oC.
Fig. 10. Microstructure of alloy Cu0.5As (OM)
2.00 wt.% As
1.66wt.% As
100 wt.% Cu
Fig. 13. Influence of As additive on copper hardness HB
of the as-cast structure and after heat treatment
Fig. 11. Microstructure and chemical composition (wt.%)
of Cu0.5As alloy (SEM-EDS)
Fig. 14. Microhardness of axes Bytyń A and Strzelin Band
the influence of As addition on the microhardness HV0.1
of plastically formed copper
Fig. 12. Influence of As additive on strength and deformation
of Cu and CuAs as-cast structure
At the II stage the influence of heat treatment on copper with
arsenic additives was tested (C+HT). Soaking at 500oC with
successive slow cooling in the furnace and quick water cooling,
caused, in each case, lowering of the material hardness. The soaking
temperature of 950oC with quick cooling causes hardness increase in
the sample with 0.6 wt% As (Fig. 13). Since the arsenic atomic radius
is smaller than the corresponding radius in copper, it should be
a
b
Fig. 15. Arsenic influence on the microstructure of Cu after Casting
and Forging (C+F) (a) and after Casting, Forging
and Heat Treatment 400oC (C+F+HT) (b)
ARCHIVES of FOUNDRY ENGINEERING Volume 15, Special Issue 4/2015
27
Acknowledgements
The financial support of the State Committee for Scientific
Research of Poland under the grant numbers 11.11.170.318 – 11.
References
[1]
a
b
Fig. 16. Arsenic influence on the microstructure of CuAs0.5
alloy after Casting and Forging (C+F) (a) and after Casting,
Forging and Heat Treatment 400oC (C+F+HT) (b)
[2]
[3]
4. Conclusions
The obtained research results allowed to determine the chemical
composition and microstructure of the prehistoric axes, which made it
possible to draw conclusions on utilizing casting technology and plastic
working when making them. Both axes were made from copper:
a relatively pure copper Strzelin B type and copper with arsenic Strzelin
A type. The arsenic content in one of them made designing
the experiment possible.
During the experiment the influence of arsenic on mechanical
properties of the material was assessed. Regarding the material
hardness, this influence is relevant, providing the soaking temperature is
950 oC and then it is water cooled. A significant hardness increase
caused by an arsenic addition also takes place with plastic working of an
already prepared cast. This effect disappears after the heat treatment at
500 oC and slow cooling, which results from copper recrystallization.
The microstructure and microhardness analysis (76HV0.1), conducted
for the prehistoric axe containing arsenic (Bytyń A), point to the process
of casting and forging the tool, which caused the increase of its
properties in comparison to the cast material, as well as the
recrystallization process, which lowered the properties with respect to
the forged metal. In pure copper no change in mechanical properties is
visible as the result of plastic working and heat treatment. That is why
plastic working and, possibly, heat treatment of the copper Strzelin A
axe influenced permanently the microstructure but not the material
properties.
Also, the influence of arsenic on copper deoxidation was
confirmed, which impacts both the microstructure and eliminates the
surface defects of the cast. It can be also observed in analysis of
prehistoric axes: Bytyń A with 0,77wt% As exhibits much less surface
defects than Strzelin B with low As content. Arsenic copper is also
characterized by better fluidity what improves a casting process.
Based on the results of studying the real casting of the flat axes,
an attempt towards characterization of the used alloys with regard to
their microstructure and mechanical properties was undertaken. The
analyses involved three different states of material: as-cast, forged
(plastic working) and heat-treated states. A change in mechanical
properties was observed.
It is likely that prehistoric smiths in the Eneolithic were conscious
of mechanical properties limitations of the alloys they used for
manufacturing metal objects. The forming technique (cold or hot
forging) was a deliberate treatment used for improving properties
of the cast objects.
The obtained results contribute to a better understanding of
metallurgy, casting and forging techniques during the Eneolithic on
Polish land.
28
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
Pernicka, E. (2014) Provenance Determination of Archaeological Metal Objects. In B.W. Roberts & Ch.P. Thornton
(Eds.), Archaeometallurgy in Global Perspective. Methods
and Syntheses (pp. 239-268).
Bolewski, A., Manecki, A. (1993). Mineralogia szczegółowa.
Warszawa: PAE.
Scott, D.A. (1991) Metallography and Microstructure of
Ancient and Historic Metals. The Geety Conservation
Institute, The J. Paul Getty Museum.
Kienlin, T.L., Bischoff, E., Opielka, H. (2006) Copper and
Bronze during the Eneolithic and Early Bronze Age:
a Metallographic Examination of Axes from the Northalpine
Region, Archaeometry (48)3, 453–468.
Lechtman, H. & Klein, S. (1999) The Production of Copper–
Arsenic Alloys (Arsenic Bronze) by Cosmelting: Modern
Experiment, Ancient Practice. Journal of Archaeological
Science 26, 497–526.
Lechtman, H. (1996) Arsenic Bronze: Dirty Copper or
Chosen Alloy? A View from the Americas. Journal of Field
Archaeology, 23(4) Winter, 477-514.
Kucharski, M. 2003. Pirometalurgia miedzi. AGH, Kraków.
Czekaj, E., Fajkiel, A., Glądała, K., Turzyński, J. (2006)
Application of Modern Technology to Remove Arsenic from
Copper under the Local Conditions of KGHM Polska Miedź
S.A. Archives of Foundry 6(18) (1/2), 261-266.
Rdzawski, Z. (2009) Alloyed Copper. Gliwice: Wyd.
Politechn. Śląskiej.
Sobczak, J., (1997) The Principales of Alloys Synthesis.
Kraków: Instytut Odlewnictwa
Metals, Technical Directory (2012), Wrocław: Metale
Agencja Promocyjna
Yearbook of the Historical Society for the Noteć District in
Bromberg (1891), Bromberg. (In Deutsch)
Kostrzewski, J. (1924). Research on Early and Middle Bronze
Age settlement on Polish land. Archaeological Review, 2(2),
161-218. (In Polish)
Szpunar, A. (1978). The axes in Poland I (flat axes, flanged
axes, chisels) (Die Beile in Polen I (Flachbeile, Randleistebeile, Randleistemeißel)). Prähistorische Bronzefunde IX: 16,
München: C.H. Beck’sche Verlagsbuchhandlung.
Inventory Book of Noteć District Historical Society in
Bydgoszcz (In Polish).
Pernicka, E., Begemann, F., Schmitt-Strecker, S., Todorova,
H. & Kuleff, I. (1997). Prehistoric copper in Bulgaria,
Eurasia Antiqua 3, 41-180.
Cook, S.R.B. & Aschenbrenner, S. (1975). The Occurence of
Metallic Iron in Ancient. Copper. Journal of Field
Archaeology 2(3), 251-266.
Subramanian, P.R. & Laughlin, D.E. (1988) The As-Cu
(Arsenic-Copper) System. Bulletin of Alloy Phase Diagrams
9(5): 605-617, 625-626.
Neumann, J.R., Zhong, T. & Chang, Y.A.(1984) The Cu-O
(Copper-Oxygen) System. Bulletin of Alloy Phase Diagrams
5(2), 136-140.
ARCHIVES of FOUNDRY ENGINEERING Volume 15, Special Issue 4/2015