Précédent Suivant

Reassessment of the prehistoric metallurgy at Arisman, Central Iran

p. 231-260

Résumés

The Arisman ancient metallurgical site is located in western Central Iran. This site hosts huge metallurgical remains from the late 4th to early 3rd millennium BCE, which attest to an extensive production of arsenical copper and silver at the same time. Despite the archaeometallurgical investigations that have so far been carried out at Arisman, some questions about the provenance of the ore, the technological procedures utilized and the possible connection between copper and silver production are still open. Thus, the authors have reviewed the previously published data and investigated and/or re-examined some of the metallurgical remains of the site including ore and slag pieces, as well as litharge fragments, through the use of ICP-MS, XRD, SEM, and optical microscopy. The results demonstrate a sharper picture for the metallurgical processes and a clearer provenance for the ore. It seems that the ore was provided from two polymetallic ore deposits of Baqoroq and Komjan in Central Iran. The polymetallic ore contained copper, arsenic, lead, and silver and was processed in two interconnected steps of smelting and cupellation, which produced arsenical copper and silver.

L’ancien site métallurgique d’Arisman est situé au centre-ouest de l’Iran. Ce site comporte d’énormes restes métallurgiques datant de la fin du IVe et du début du IIIe millénaires avant notre ère, attestant de l’existence d’une importante production de cuivre arsénié et d’argent. Malgré les investigations archéométallurgiques menées jusqu’à présent à Arisman, certaines questions concernant la provenance des minerais, les procédures techniques utilisées et les possibles liens entre production de cuivre et production d’argent restent ouvertes. De ce fait, les auteurs ont passé en revue les données précédemment publiées et ont étudié ou réétudié certains restes métallurgiques, comme du minerai, des scories ainsi que de la litharge au moyen d’une microscopie optique et d’analyses ICP‑MS, XRD et SEM. Les résultats permettent d’obtenir une image plus nette des processus métallurgiques et de la provenance du minerai. Il semblerait que ce minerai provienne des deux dépôts polymétallifères de Baqoroq et Komjan au centre de l’Iran. Le minerai polymétallique contenait du cuivre, de l’arsenic, du plomb et de l’argent ; il était traité en deux étapes interconnectées de réduction et de coupellation qui ont produit à la fois du cuivre arsénié et de l’argent.

Remerciements

This paper is based on research carried out in the framework of a three-month postdoctoral stay (in the autumn 2014) of the first author at the University of Heidelberg and the Curt Engelhorn Center (Mannheim) funded by the Gerda Henkel Foundation. Prof. Dr Ernst Pernicka and Prof. Dr Barbara Helwing have supervised this research. In this regard, the first author (N. Nezafati) would like to thank the Gerda Henkel Foundation for facilitating his stay and financing part of the analytical costs. His gratitude also goes to the staff of the Curt Engelhorn Center and to the Institute of Mineralogy of the University of Heidelberg – especially to Dr Steffen Kraus and Dr Roland Schwab for their kind assistance.


Texte intégral

Introduction

1The Arisman ancient metallurgical site, which extends over an area of 3 km2, is located 17.5 km north-east of Natanz and 7 km north of the foothills of the Karkas Mountains in western Central Iran (fig. 1). The site was first recognized in 1996 by Davood Hasanalian, a local scientist, and was excavated as part of the “Early mining and metallurgy on the western Central Plateau of Iran” project in five seasons (2000 to 2004) by a joint Iranian-German team (Chegini et al. 2000; Helwing 2008; Vatandoust, Parzinger, Helwing 2011).

Fig. 1 – a: Location of the study area on the geological subdivision map of central and western Central Iran; b: The general location map of the sites mentioned in the text (modified after Nezafati 2006).

Image

2The site of Arisman hosts archaeological remains dated from the late 5th to early 4th millennium BCE until the Islamic era (fig. 1, 2), but the Arisman I section of the site only yielded remains dated from the mid 4th to the beginning of the 3rd millennium BCE (which is equivalent to the periods Sialk III6‑7 – ca 3700‑3400 BCE – and Sialk IV – ca 3350‑2900 BCE – periods, the later also known as the proto‑Elamite period (Görsdorf 2011; Nokandeh 2010, pp. 75‑77; Pollard et al. 2013, pp. 44‑48, tab. 9). This time span covers the shift from a Chalcolithic occupation in increasingly complex and large villages to central places modeled on urban examples known from the lowlands of Susiana and Mesopotamia. This shift towards centralization encompasses major changes in the organization and scale of craft and trade (Helwing 2013a, 2013b; Helwing, this volume).

Fig. 2 – a: An overview of the Arisman ancient metallurgical site and the Karkas Mountains in the background (looking to the south); b: An overview of the Arisman I site; c: A section cut by a water channel in a slag heap (Arisman 1A); d: Smelting furnace with a possible place for the location of a crucible at the bottom of Arisman‑1A, trench A45; e: Slag cake from Arisman I; f: Part of a crucible with a handle-like portion in the lower part; g: Axe molds from Arisman I, trench C; h: A twin cupellation hearth from Arisman I (Nezafati 2000).

Image

3Extensive archaeometallurgical investigations were carried out on materials from Arisman I (e.g. Pernicka et al. 2011; Rehren, Boscher, Pernicka 2012). These provided first insights into the early steps of a copper and silver metallurgy that shifted from a systematic copper extraction by crucible smelting in the Late Chalcolithic to an industrial scale furnace smelting in the proto‑Elamite period. Over both periods, a sophisticated silver-cupellation technology was in use, apparently less affected by the eminent socioeconomic change that the proto‑Elamite occupation brought along. Copper and silver artifacts manufactured in Arisman and other highland sites fed into long distance trade networks, that connected the Iranian highland to the lowlands of Mesopotamia and Susiana, with heavy copper axes in the early and lighter flat axes in the later period as main shapes (Helwing 2011b). These trade relations also connected the Iranian highlands to the southern Caucasus, as is attested from imports of painted Sialk III ceramics in Azerbaijan (Helwing, Aliyev 2017, p. 13, fig. 6). Not surprisingly, contemporary 4th millennium BCE sites in the South Caucasian Mugan Steppe yielded metallurgical residues closely comparable to the Late Chalcolithic Arisman finds, including a furnace (Akhundov 2014, fig. 2), but also molds, crucibles and litharge (Stöllner 2016, p. 214, footnote 35).

4Despite the existing archaeometallurgical studies on the Arisman material, many questions about the provenance of the ore, the technological procedures used by early metallurgists and possible connections between the copper and silver productions remained open. For this reason, the samples stored in the rooms of the Curt Engelhorn Center for Archaometry (Mannheim), which were collected during the field campaigns between 1999 and 2004, were rechecked and partly sampled. These samples, which include ore and slag pieces, as well as litharge fragments, were re-examined and analyzed through different mineralogical and geochemical methods. The analytical data previously published by Pernicka et al. (2011) have also been integrated for the final interpretation.

Materials and methods

5First, the previous publications and data on Arisman and other contemporaneous similar ancient sites were carefully reviewed, and previously published data reassessed. This led to new hypotheses about the possible metallurgical procedures carried out on that site. Several samples were thus carefully restudied: altogether 40 samples including slag and ore pieces, as well as some fragments of litharge cores and cakes, were examined; 30 polished thin sections were cut from these samples and prepared at the Institute of Mineralogy (University of Heidelberg). These sections were studied with a polarizing microscope and SEM at the Curt Engelhorn Center for Archaeometry. At Zarazma Laboratories (Tehran, Iran), 57 elements from 36 samples were also analyzed with an ICP‑MS. The samples were first digested for the chemical analysis in a four-acid solution and (after dilution) were analyzed using a HP 4500 ICP‑MS system manufactured by GMI (detailed procedure of analysis is available under http://zarazma.com). The XRD analyses of 33 samples were performed at the material science laboratory of the German Mining Museum. The ICP‑MS costs were covered by the Gerda Henkel Foundation, while the microscopic and SEM investigations were supported by the Curt Engelhorn Center. The list of investigated samples is presented in tab. 1. During this research, we tried not to focus on one or two metallurgical processes only, but to observe the entire processes in order to bring out a broader picture of the whole procedure carried out on the site. Therefore ore, slag, and litharge samples were examined and investigated in parallel.

Tab. 1 – List of the samples investigated during this research together with XRD analysis results.

 Sample no.DescriptionContextPeriodXRD
1FG-012261Ore2001-AR-1B-35Sialk IIIQuarz, epidote, hematite, albite, eriochalcite, langite
2FG-030666Ore2003-AR02-B500 (B36, 11)Sialk IIIQuartz, hematite, clinochlore, microcline, albite
3FG-030667Ore2003-AR02-B500 (B36, 11)Sialk IIIQuartz, malachite, muscovite, calcite, dolomite
4FG-030663Ore2003-AR02-B0199 (B0596)Sialk III or IV (n.d., surface find) 
5FG-030666Ore2003-AR02-B500 (B36, 11)Sialk IIIQuartz, hematite, clinochlore, microcline, albite
6FG-030676Clay+Cu-Pb-oreAR02-C95 (C46, F41)Sialk IVGypsum, atacamite, quartz, goethite, albite, magnetite
7FG-030695Ore2003-AR02-C1079 (C46, F76)Sialk IV 
8FG-030685Ore2003-AR02-C917 (C46, F22)Sialk IV 
9FG-030689Ore2003-AR02-C1000 (C46, F31)Sialk IVDigenite, tennantite, mawbyite, quartz, kintoreite, clinomimetite, paratacamite
10FG-030125Iron oxide2003-AR-1B-45Sialk III or IV (n.d.) 
11FG-030126Alteration2003-AR-1C-55Sialk IV (n.d., surface find) 
12BQ-01OreBaqoroq mineBaqoroq mine 
13FG-030103OreKomjan mineKomjan mineCerussite, galena, fluorite, quartz, anglesite
14FG-10473 OreKomjan mineKomjan mineMalachite, quartz, goethite
15FG-10475 OreKomjan mineKomjan mineHemimorphite, quartz, malachite, calcite
16FG-10480 OreKomjan mineKomjan mineGalena, fluorite, laurionite, calcite, cuprite, quartz, pyromorphite
17FG-010505Brown slagAR-1A-surfaceSialk IV (n.d., surface find)Pyroxene, monticellite, paratacamite (zincian)
18FG-010506Brown slagAR-1A-surfaceSialk IV (n.d., surface find)Magnetite, augite, monticellite, gypsum
19FG-011995Brown slagAR-1A (Bochum)Sialk IV (n.d., surface find)Magnetite, augite, monticellite, quartz, omphacite
20FG-030119Brown slag2000-AR-1A-45Sialk IV (n.d., surface find)Augite, quartz, akermanite, hedenbergite, pharmacosiderite
21FG-040920Brown slag2004-ARTD, Trench DSialk III or IV (n.d.)Clinopyroxene, quartz, hedenbergite, monticellite
22FG-000116Green slagAR-1B-surfaceSialk III (n.d., surface find)Akermanite, quartz, cristobalite, clinopyroxene (alominous), diopside (aliminian)
23FG-030122Green slag + Brown slag2003-AR-1D-45Sialk III or IV (n.d.) 
24FG-000118Green slagAR-1D-surface-near qanatSialk III or IV (n.d., surface find)Diopside, magnatite, fayalite, hematite
25FG-012256Green slag2001-AR-1B-46Sialk IIIAugite, magnetite, monticellite, clinopyroxene (aluminous)
26FG-012260Green slag2001-AR-1D-45Sialk IVMagnetite, augite, paratacamite (zincian), fayalite
27FG-012262Green slag2001-AR-1B-46Sialk IIIMagnetite, clinopyroxene, cuprite, paratacamite
28FG-012275Green slag2001-AR-1D-45Sialk IVAugite, quartz, cristobalite, magnetite, clinopyroxene
29FG-030115Green slag2003-AR-1A1/5Sialk IVFassaite, cuprite, magnetite, olivine
30FG-030117Green slag2000-AR-1A-45Sialk IVAugite, magneite, clinopyroxene, diopside, cristobalite
31FG-030120Green slag2000-AR-1A-45Sialk IV 
32FG-030676Smelted Cu-rich material2003-AR02-C95 (C46, F41)Sialk IV 
33FG-012250Litharge2001-AR-1B-35Sialk IIILitharge, hydrocerussite, bilixite, asisite, calcite, cotunnite, quartz
34FG-012252Litharge2001-AR-1B-36Sialk IIICerussite, litharge, calcite, hydrocerrusite, laurionite
35FG-012253Litharge Surface, n.d.Hydrocerussite, cerussite, litharge, lead, laurionite, magnetoplumbite
36FG-012267Litharge2001-AR-1B-36Sialk IIILaurionite, litharge, hydrocerussite, paralaurionite, woodhouseite, mendipite, magentoplumbite
37FG-030682Litharge2003-AR02-C807 (C56, F22)Sialk IVLitharge, bilixite, hydrocerussite, lead oxide
38FG-030686Litharge2003-AR02-C926 (C46, F22)Sialk IVBilixite, litharge, hydrocerussite, laurionite, lanarkite
39FG-030697Litharge2003-AR02-D0006 (D82, F2)Sialk III or IVLitharge, massicot, cerussite, lead oxide
40FG-030698Litharge2003-AR02-D0058 (D82, F5)Sialk III or IVLitharge, hydrocerussite, massicot, cotunnite, protoenstatite, cerussite, lead oxide
41FG-993156LithargeAR1Surface, n.d.Litharge, hydrocerussite, calcite, bilixite, asisite, anglesite
42FG-030678Lead-litharge porous lump2003-AR02-C484 (C46, F37)Sialk IVFluorite, hydrocerussite, cristobalite, akermanite, laurionite, litharge
43FG-030701LithargeAR02-D0270 (D82, F7)Sialk III or IVHydrocerussite, litharge, cerussite, lanarkite, calcite, mattheddeite, leachillite

Review of the previous data

6In this study, only the samples from the Arisman I section of the ancient metallurgical site of Arisman were investigated. The surface of the site is covered by numerous ancient remains, including huge quantities of slag pieces (in various sizes and forms), pottery sherds, furnace and crucible remains, molds, lithic tools, and cupellation hearth pieces, all of which belong to the Sialk III6-III7b and Sialk IV periods (fig. 2).

7A major issue concerning the slag pieces collected on the site, which was recognized from the onset, is the coexistence of two distinct types of slag pieces respectively called “green slag” and “brown slag”. The grey to black copper-rich slag with distinct green stain (due to corrosion of copper) is referred to as green slag, while the rarer copper-poor grey slag with distinctive brown stains is called brown slag. These two types of slag pieces do not show any stratification and are totally mixed up. The green slag is more frequent than the brown one.

8Rehren, Boscher, Pernicka (2012) considered the quantity of these two types of slags to be equal and explained the brown slag as resulting from the production of metallurgical speiss, while the green slag was interpreted as some waste from the arsenical copper production on the site. They argued that both speiss and arsenical copper were produced side by side at the same site, but through the use of different ores and processes. They believe that arsenic was added to the copper ore (or copper metal) in the furnaces (or crucibles). In their opinion, this arsenic appeared in the form of speiss that has been produced from arsenopyrite in a separate step; not by using this mineral in its native form from an arsenic deposit.

9On the other hand, Pernicka et al. (2011) suggested that the processes of copper-ore smelting and silver production by cupellation were performed as two different unrelated, separate manufacturing operations (as regards the ore and the processes in use), which lead to the production of two distinct final products of arsenical copper and silver.

10A careful reassessment of the data published by Pernicka et al. (2011) and Rehren, Boscher, Pernicka (2012) makes their interpretations somewhat arguable for the following reasons:

  • No single piece of speiss or arsenopyrite-realgar-orpiment ore has yet been reported (or found) from Arisman, while several pieces of (copper-polymetallic) ore, metal prills, metal artifacts, and other metallurgical remains have been found on the surface and during the excavations. The only traces of speiss are the very tiny prills of the so-called speiss-prills (not larger than 300 microns in diameter), which have been reported in brown slag pieces by Pernicka et al. (2011) and Rehren, Boscher, Pernicka (2012). Such prills do not necessarily need to occur in a separate furnace; they may also be found in a furnace loaded with arsenic-copper bearing ore as suggested by Keesmann (1991).
  • No single piece of lead slag (a very usual waste material from the cupellation process of argentiferous galena) and/or galena has been found at Arisman.
  • Less than twenty percent of the total mass of the slag at Arisman is composed of the so-called “brown slag”. This brown slag does not have any stratigraphic or specific position compared to the green-black slag. On the other hand, both types of slag are totally mixed up.
  • The copper content of the slag pieces is highly variable and ranges from 0.01% to 31.7% (35.8% Cu2O). On the other hand, some slag pieces are very rich in copper and show contents as high as 31.7% Cu (although the average content is around 10%). This type of slag could well be copper ore in itself (a piece of copper ore found on the site contains 22% copper). This copper content is not only bound as matte, which was practically considered as waste by ancient metallurgists, but also as copper and arsenical copper prills that could be recovered from slags mechanically. So far, traces of sand slags that may result from the mechanical separation of copper from slag have been recovered from Arisman (Steiniger 2011; Helwing 2011a).
  • The presence of nickel in the slag and the finished artifacts is not as insignificant as previously considered. The pyroxenes of Arisman‑1C are rich in nickel and zinc. Some Ni‑Zn rich metallic phases (sulfides and arsenides) have been detected at the same samples. The nickel content of some analyzed copper-based artifacts is considerably high (up to 2.69%, Pernicka et al. 2011).
  • One sample (FG‑012269B‑Arisman‑1C) contains high concentrations of lead (12%), zinc (4%), and nickel (2%). Besides, in several slag pieces tiny globules of lead (from Arisman‑1B) and/or lead arsenides (Arisman‑1C) have been reported (Pernicka et al. 2011).
  • Antimony is a rather abundant element in many slag pieces in the form of antimony-bearing copper globules. The amount of Sb2O3 (4.9%), BaO (3.6%), As2O3 (0.79%), and Ag in a slag piece found near the qanat that crossed the site (FG‑000118) is high. In this sample silver is found next to copper alloy and copper antimonides (Pernicka et al. 2011).
  • No significant systematic chemical difference was detected between the green-black and the brown slag (Pernicka et al. 2011). The most significant differences are the generally higher contents of arsenic and lower contents of copper in the brown slag.
  • Sulfur and arsenic are present in almost all slag pieces, although the arsenic content is higher in brown slag (Pernicka et al. 2011). A considerable amount of brown slag contains almost no copper, while the As2O3 content of the non-brown slags reaches 5% at Arisman‑1C.
  • A piece of copper ore (FG‑030689) shows a paragenesis of copper-arsenic sulfide with impurities of zinc, nickel, lead and cobalt mostly in the form of chalcocite and very fine-grained nickel sulfide. This type of polymetallic ore could well be the typical ore used for producing metal on the site.
  • The analytical results obtained from litharge samples (Pernicka et al. 2011, tab. 9, 11) show considerable concentrations of arsenic, antimony, and copper therein, while these elements are rarely present in high quantity in the Nakhlak ore or even other typical silver bearing Cretaceous Mississippi Valley Type (MVT) deposits from western Asia. Based on tab. 11 (Pernicka et al. 2011), even the gold content is in a way considerably high for a litharge sample.
  • The litharge samples of Arisman can be divided into two groups according to their appearance and their trace element composition; group one has large irregular lumps showing higher amounts of arsenic (>200 ppm), antimony (500 ppm), and copper (3,000 ppm), while group two comprises smaller pieces with lower contents of arsenic (<45), antimony (75 ppm), and copper (1,000 ppm), also in one case 4.5% copper (measured using EDXRF).
  • The analyses of 28 copper-based artifacts (mostly pins and beads) from Arisman (Pernicka et al. 2011, tab. 15) show that most of the artifacts (25 out of 28) contain rather high concentrations of arsenic with highly variable contents (between 0.46 and 5.7%). If we consider that the metallurgical process was well-controlled, as suggested by Rehren, Boscher, Pernicka (2012; adding speiss to copper ore or metal), the arsenic content of the artifacts should be much more uniform than what we observe now. Interestingly enough, silver is generally present in high content in copper-based artifacts (9.9% at the highest). Nickel and antimony contents are considerably high in some cases (Ni: up to 2.69%, and Sb: up to 2.62%).

11All these elements challenge the interpretation presented by Pernicka et al. (2011) and Rehren, Boscher, Pernicka (2012). Therefore, previously published data need to be reassessed in the light of new analyses.

Mineralogy and geochemistry of the ore samples

12The ore pieces collected from archaeological sites are extremely valuable for provenance and technological studies since these ores have not undergone heat treatment or any other kind of manipulation. In this respect, they may be considered as intact, and illustrate the raw materials used by ancient metallurgists: not only can they give information on the mineralogy of the original raw material, but they can also provide a background for the mineralogy of the gangue material and the host rocks, both of which may also have functioned as metallurgical flux. As for the mineralogy of the ore, gangue, and the host rocks, it can provide very valuable data on the geological characteristics of the ancient mine that was exploited and help to localize this mine.

13Altogether eleven ore or ore-resembling pieces were found in the storeroom of the Arisman excavations in Mannheim from which nine samples were selected. These pieces, together with one sample from Baqoroq and four samples from the Komjan mines were studied through macroscopic and microscopic techniques as well as XRD, SEM, and ICP‑MS analyses.

14Judging by the macroscopic observations, all samples but one have a low ore content, which has strongly been oxidized (fig. 3). Their low content in ore has possibly raised the chance of survival of these pieces. Seven of them display visible green stains and are mainly composed of malachite impregnations. The ore microscopic investigations indicate that only very few primary sulfide minerals have remained that mainly include chalcopyrite and chalcocite. Malachite, hematite, and goethite are the major secondary minerals. Barite and some celestine are also present in some of the samples.

Fig. 3 – a: A piece of massive sulfide ore containing fahlore (FG‑030689); b: A polished section of the same sample; c: The backscattered microphotograph of part of the same sample containing tennantite, mawbyite, and quartz; d: A piece of quartz, Fe‑oxide‑malachite bearing ore (FG‑030663); e: A piece of oxide-sulfide ore in a carbonate host rock (FG‑030667); f, g: Pieces of oxidized ore (FG‑030685, FG‑030695) (N. Nezafati).

Image

15According to the geochemical analyses (tab. 2), copper (and in some cases iron) is the only significant trace element in most of the samples. Most ore samples contain quartz, calcite, dolomite and in some cases albite, microcline, epidote, barite, and celestine as accompanying gangue. Judging by the XRD and microscopic observations, the mineralization process has developed in a carbonate rock (limestone, dolomite, etc.), at least in part.

Tab. 2 – Bulk chemical analyses of ore samples measured by ICP‑MS (Zarazma Minerals Company).

Sample no.Cu%As%Pb%Ag (ppm)Zn%Fet%Bi (ppm)Sb (ppm)MnO%
FG-0122619.61<0.01190.026.6531.70.2
FG-03066715.60.01<1.30.11.880.39.90.03
FG-0306635.450.01<2.8<17.36.96.60.02
FG-0306660.92<<0.3<18.872.62.60.03
FG-03069520.20.010.018.80.0129.8156.44.90.01
FG-03068524.450.010.180.40.0130.681.41.70.06
FG-03068954.575.67.73178.61.132.463693.3<
FG-0301250.070.01<0.2<50.77<0.10.90.03
FG-0301260.070.010.030.2<18.452.94.80.04
BQ-0111.972.070.143.30.340.391.2741<
FG-0301030.03<78.9370.10.040.690.69.50.01
Sample no.BaO%S%Ni (ppm)Co (ppm)Mo (ppm)MgO%P2O5%Sr%
FG-012261<0.351519.71.74.510.17<
FG-0306670.010.035634.24.80.90.21<
FG-0306630.010.231780.23.70.480.09<
FG-0306660.030.65910.557.21.250.06<
FG-030695<0.1254.729.20.220.36<
FG-030685<0.131111.812.70.720.6<
FG-030689<16.943,668432.44860.040.23<
FG-0301250.060.222432.537.80.420.11<
FG-0301260.0111.962915.15.60.291.10.1
BQ-0130.5917137.90.10.160.8
FG-030103<12.1844.737.40.07<<

16An interesting ore piece demonstrates a massive sulphidic assemblage of minerals (FG‑030689; fig. 3a, 3b, 3c). Mineralogical investigations have shown that it is mainly composed of tennantite (Cu12As4S13) which is accompanied by digenite, mawbyite (PbFe3+2(AsO4)2(OH)2), clinomimetite (Pb5(AsO4)3Cl), kintoreite (PbFe3+3(PO4)2(OH, H2O)6), paratacamite (Cu2Cl(OH)3), and quartz (as veinlets). Mawbyite and clinomimetite, which are Pb‑As oxides, represent the major inclusions in tennantite ore; they measure up to 500 microns in size. SEM studies show that the content of silver reaches up to 0.5% in tennantite and Pb‑As minerals. According to the ICP-MS analysis, this type of ore, which can be considered as typical fahlore contains 54.57% copper, 5.6% arsenic, 7.73% lead, 1.13% Zn, 178 ppm silver, 3,668 ppm nickel, 693 ppm antimony, and little iron, cobalt, and molybdenum.

17The data obtained from the ore used by early metallurgists may give us some clues about the ore deposit exploited in ancient times. It seems that an argentiferous polymetallic ore (possibly from different sources) has been used for the production of copper and silver. This ore was a mixture of oxide, sulfide, and sulfosalt minerals with a gangue containing quartz, calcite, barite, and the remaining carbonaceous host rocks. The ore deposit(s) that provided such an ore was mainly composed of carbonate rocks (limestone and dolomite) in which quartz-calcite-barite-(fluorite) veins, veinlets and lenses hosted a sulfide mineralization that had later been partly oxidized due to weathering and supergene processes.

18One piece of ore from Baqoroq and four samples from the ancient mines at Komjan were investigated and analyzed (tab. 2). The piece from Baqoroq, which was oxidized ore collected from the superficial outcrops, contains 12.0% copper, 2.07% arsenic, 3.3 ppm silver and rather high anomalies for lead, zinc, and antimony. The analysis of a lead-rich sample from Komjan indicates 78.9% lead and 70.1 ppm silver. The other three samples of Komjan were analyzed by XRD. More detail about the ore of these two ancient mines is presented below.

Mineralogy and geochemistry of slag samples

19If morphological and macroscopic criteria are taken into account, the slag pieces of Arisman may be divided into three major types: green, brown, and glassy slag groups (fig. 4). The green slags are the most abundant, while the brown ones rank second and the glassy ones rank third in quantity. Morphological and macroscopic features (including color) help to discriminate them from each other. The so-called green slag is characterized by a green rust on the surface, but the dense matrix is grey to black with an occasional flow texture (fig. 4a, 4b, 4l), while the brown slag is more porous with a (brown) rusty surface (fig. 4c, 4d). Sixteen slag samples were investigated, five of which were brown, ten (8+2) were green and one contained interconnected green and brown slag (tab. 3, 4, 5). The major characteristics of the green and brown slag are discussed below.

Fig. 4 – a, b: Typical appearance of the so-called green slag pieces; c, d: Typical appearance of the so-called brown slag; e: Typical appearance of the glassy slag pieces; f: Pieces of brown slag with green (cupriferous) stain (transitional slag); g: A piece with both interconnected green (up) and brown (down) slags; h: Four pieces of the mixed coexisting green and brown slag (sample no. FG‑030122); i: Close-up view of a slag piece with adjoined brown (down) and green (up) slags stuck together as a single piece; j: Thin section of the sample shown in the pictures e and f; k: The microscopic view of part of the border between the green (up) and brown (down) slag; l: A piece of slag with a large copper prill surrounded by matte (sample no. FG‑012260); m: Microscopic photo of an unknown (to author) pattern in the copper piece of the sample no. FG‑012260, this pattern is not visible in SEM; n: Copper prills accumulated around a vug with some skeletal magnetite grains around; o: A globule of matte in the slag; p: Microscopic image of pyroxene laths in the slag (N. Nezafati).

Image

Tab. 3 – Bulk chemical analyses of major elements and oxides from the slag samples measured by ICP‑MS (Zarazma Minerals Company).

Sample no.DescriptionSiO2%Al%Fet%CaO%Na%K2O%
FG-010505Brown slag38.375.1618.5115.731.552.17
FG-010506Brown slag30.234.2624.7712.951.411.67
FG-011995Brown slag31.634.0626.3813.431.441.82
FG-030119Brown slag31.824.6615.5412.671.692.08
FG-040920Brown slag38.64.2820.0515.261.272.31
FG-000116Brown slag46.934.4110.3720.351.41.7
FG-030122Green slag+Brown slag11.171.6123.883.620.372.6
FG-000118Green slag38.632.7526.7712.110.630.82
FG-012256Green slag33.562.9327.7815.910.651.18
FG-012260Green slag27.832.925.6511.10.621.12
FG-012262Green slag29.572.3323.0813.270.630.94
FG-012275Green slag44.874.0617.8411.460.71.5
FG-030115Green slag34.995.8116.5213.570.9972.46
FG-030117Green slag37.73.320.899.830.831.31
FG-030120Green slag39.364.7715.7811.250.692.13
FG-030676Heated Cu-rich material26.753.8911.5611.680.931.16
Sample no.DescriptionMgO%MnO%P2O5%TiO2%BaO%Sr%
FG-010505Brown slag1.830.270.430.460.040.13
FG-010506Brown slag1.470.240.380.380.04<
FG-011995Brown slag1.520.250.420.390.040.1
FG-030119Brown slag2.170.190.420.40.040.12
FG-040920Brown slag1.830.480.50.440.060.16
FG-000116Brown slag2.960.140.460.370.040.25
FG-030122Green slag+Brown slag0.540.040.090.140.02<
FG-000118Green slag2.090.130.220.210.090.11
FG-012256Green slag1.730.30.310.360.710.21
FG-012260Green slag2.122.130.230.260.03<
FG-012262Green slag3.220.560.260.190.17<
FG-012275Green slag2.240.20.250.340.2<
FG-030115Green slag3.650.180.320.490.16<
FG-030117Green slag20.120.330.240.07<
FG-030120Green slag1.90.270.250.390.05<
FG-030676Heated Cu-rich material1.290.080.340.320.030.14

Tab. 4 – Bulk chemical analyses of metals of green slag samples measured by ICP‑MS (Zarazma Minerals Company).

Sample no.Cu%As%Pb%Ag (ppm)Zn%Fet%
FG-0001181.90.010.060.90.1226.77
FG-0122561.250.340.110.30.0227.78
FG-0122606.670.230.020.60.2525.65
FG-0122626.730.240.156.21.3523.08
FG-0122752.47 0.10.60.0317.84
FG-0301153.420.310.110.90.0116.52
FG-0301174.220.190.033.60.0220.89
FG-0301204.030.39<1.20.0215.78
FG-0306768.460.01<1.10.0111.56
Sample no.Bi (ppm)Sb (ppm)S%Ni (ppm)Co (ppm)Mo (ppm)
FG-0001182.123.40.084060.331.8
FG-0122560.93.60.232027.816.6
FG-0122600.5123.70.522691,391.716.3
FG-0122621.4167.60.16131276.970.4
FG-01227538.84.10.05353019.4
FG-0301150.83.40.4422132.122.9
FG-03011729.13.50.450389.2
FG-0301201.81.70.151634.231
FG-0306760.75.14.51341834.5

Tab. 5 – Bulk chemical analyses of metals of brown slag samples by ICP‑MS (Zarazma Minerals Company).

Sample no.Cu%As%Pb%Ag (ppm)Zn%Fet%
FG-0105050.221.930.0020.70.0218.51
FG-0105060.022.420.010.40.0324.77
FG-0119950.012.320.010.30.0326.38
FG-0301190.024.230.0120.0115.54
FG-0409200.041.820.013.20.0220.05
FG-0001160.070.750.030.90.0110.37
FG-0301220.4615.24<2.7<23.88
Sample no.Bi (ppm)Sb (ppm)S%Ni (ppm)Co (ppm)Mo (ppm)
FG-010505<0.11.80.341312.115.5
FG-010506<0.10.90.4189.912.3
FG-011995<0.10.90.2579.212.3
FG-0301190.11.10.651411.315.5
FG-0409200.21.30.271313.111.4
FG-000116<0.14.60.251610.76.6
FG-0301220.412.42.122528.937.2

Green slag

20Ten pieces of green slag were investigated in this study, two of which belong to the Sialk III period, while the others are dated to the Sialk IV period (or their age was not determined with confidence, tab. 1). The macroscopic examination of the green-slag pieces shows that they are composed of a dense material containing occasional pores, abundant globules of matte (5 microns to several centimeters, fig. 4o), and copper (5 microns to several centimeters, fig. 4n), as well as some remaining unmolten material from the original ore and its gangue and host rocks (fig. 4a, 4b, 4l). As a whole, the matrix of this slag demonstrates a heterogeneous texture with locally almost homogeneous parts.

21Mineralogical investigations (XRD and microscopic, tab. 1) indicate the presence of a rather wide series of minerals in this type of slag, which include quartz, cristobalite, pyroxene (in forms of clinopyroxene, diopside, augite, and fassaite), olivine (fayalite and monticellite), magnetite (mainly in skeletal form), hematite, paratacamite, cuprite, matte, and copper, arsenical copper, and occasional iron-arsenide prills.

22Judging by the geochemical results (tab. 3, 4), the composition of almost all samples is to some extent uniform and shows 26.75 to 44.87% (average: 35.81) SiO2, 2.33 to 5.81% (average: 3.6) Al, 15.78 to 27.78% (average: 21.79) Fe, 9.83 to 15.91% (average: 12.31) CaO, 1.27 to 1.69% (average: 1.46) Na, 0.82 to 2.46% (average: 1.43) K2O, 1.73 to 3.65% (average: 2.36) MgO, 0.12 to 2.13% (average: 0.48) MnO, 0.22 to 0.33% (average: 0.27) P2O5, 0.19 to 0.49% (average: 0.31) TiO2, and 0.03 to 0.71% (average: 0.18) BaO, 0.05 to 0.52% (average: 0.25) S, 1.25 to 6.73% (average: 3.84) Cu, 0.01 to 0.39% (average: 0.24) As, 0.02 to 0.15% (average: 0.08) Pb, 0.3 to 6.2 ppm (average: 1.78) Ag, 0.01 to 1.35% (average: 0.2) Zn, 0.5 to 38.8 ppm (average: 9.42) Bi, 1.7 to 168 ppm (average: 41.38) Sb. This demonstrates a high quantity of copper and iron in the samples with in some cases considerable contents of arsenic, lead and zinc. The content of sulfur reaches 0.52%. Silver, antimony, bismuth, nickel and cobalt are present in a fairly high quantity in a few samples. In the large copper prill of sample FG‑012260 several inclusions of silver (up to 30 microns) are visible.

23The mineralogy of the green slag and especially the presence of cristobalite demonstrate that this type of slag was exposed to temperatures reaching between 1,200 and 1,400oC. Cristobalite is a high-temperature polymorph of quartz. The temperature required to stabilize cristobalite is 1,470oC in the pure silica system, and the temperature needed to melt SiO2‑Al2O3 material is ca 1,600C. However, the presence of other fluxing elements such as Ca, Na, and K may have decreased the temperature by approximately 100‑200oC (Kierczak, Pietranik 2011). Cristobalite forms favorably in the hottest zone of the firing vessel at a temperature range of 1,200 to 1,400oC (Baumgarten, Eldar 1984; Hauptmann 2007).

Brown slag

24Mineralogical investigations (XRD and microscopic) also indicate the presence of a rather wide series of minerals in this type of slag, which include akermanite (a melilite), quartz, cristobalite, pyroxene (in the form of clinopyroxene, diopside, augite, omphacite, and hedenbergite), monticellite, magnetite (mainly in skeletal form), gypsum, pharmacosiderite (KFe4(AsO4)2(OH)4.(6-7)H2O), iron-arsenide (speiss) prills, and a few copper and matte prills.

25Judging by the geochemical results (tab. 3, 5), the composition of almost all samples is to some extent uniform (tab. 35) and shows 30.23 to 46.93% (average: 36.26) SiO2, 4.06 to 5.16% (average: 4.47) Al, 10.37 to 26.38% (average: 19.27) Fe, 12.67 to 20.35% (average: 15.06) CaO, 0.37 to 0.99% (average: 0.70) Na, 1.67 to 2.31% (average: 1.95) K2O, 1.47 to 2.96% (average: 1.96) MgO, 0.14 to 0.48% (average: 0.26) MnO, 0.38 to 0.5% (average: 0.43) P2O5, 0.37 to 0.46% (average: 0.41) TiO2, and 0.04 to 0.06% (average: 0.043) BaO, 0.25 to 0.65% (0.36) S, 0.01 to 0.22% (average: 0.06) Cu, 0.75 to 4.23% (average: 2.24) As, 0.25 to 0.65% (average: 0.36) S, 0.002 to 0.03% (average: 0.01) Pb, 0.3 to 3.2 ppm (average: 1.25) Ag, 0.01 to 0.03% (average: 0.02) Zn, <0.1 to 0.2 ppm Bi, 0.9 to 4.6 ppm (average 1.76) Sb. Ni, Co, and Mo amounts are not significant. This demonstrates a high quantity of arsenic and iron in the samples with considerable contents of copper and sulfur (0.65%) in some cases.

Comparison of the green and brown slags

26The mineralogical and geochemical comparison of the green and brown slags shows that they differ by their external stain color, as well as by their lower quantity of copper for the brown slags and the lower contents of arsenic for the green slags. Other minor differences between green and brown slags are the following:

  • The quantity of Fe, K2O, Na, MgO, MnO, BaO, Pb, Ag, Zn, Bi, and Sb is slightly higher in the green slag, while SiO2, Al, CaO, and S are present in slightly higher quantities in the brown slag.
  • (Arsenical) copper and matte prills are more abundant in the green slags, while the iron-arsenides (speiss) prills are abundant in the brown slags but scarce in the green slags.
  • Pyroxene is the most abundant silicate mineral in both green and brown slags (fig. 4p). Omphacite and hedenbergite, two types of pyroxene, were observed in brown slag, while fassaite is present in green slag. Akermanite is present in much higher quantity in brown slag. Cristobalite was observed more frequently in green slag.

27Although most of the green and brown slags are very distinct from each other because of their appearance (color) and geochemistry, two interesting samples bear the characteristics of both. Sample FG‑010505 has a brown-slag appearance with some green cupriferous stains (fig. 4f). Moreover, the geochemical results of this sample indicate 0.22% of copper, which is a rather high copper content in comparison with other brown slags.

28The most interesting is sample FG‑030122, which comes from a sample bag containing four similar pieces (fig. 4g, 4h, 4i, 4j, 4k). This sample is composed of two interconnected green and brown slags. The bulk analysis of this slag indicates 15.2% of arsenic and 0.46% of copper. The occurrence of both green and brown slags beside each other as one single piece would suggest that both green and brown slags were produced in the same furnace (or crucible).

29One sample taken from a copper-rich material seemed to be heated ore (FG‑030676). The bulk analysis of this sample indicates 8.46% of copper and 4.51% of sulfur. Other trace elements in this sample are low. Two rather large grains of scheelite (a calcium tungstate mineral) were also observed.

Mineralogy and geochemistry of the litharge fragments

30Eleven litharge fragments have been studied. Three belong to the Sialk III period, three to Sialk III or Sialk IV periods, five to the Sialk IV period, and two come from the surface of the Arisman I area. According to macroscopic examinations (fig. 5), most of the litharge fragments are composed of heavy concave pieces with a pale yellow to cream and buff exterior. Several of these fragments were cut with a laboratory saw in order to examine their internal parts. Most of the fragments show a layered structure, which is dark in the lower and light in the upper part. In one case a small piece of slag (with a 0.7 cm length) was observed in the lower part of a fragment. Another sample (FG‑030686) was mainly composed of massive lead and lead oxide (instead of litharge), with a lower amount of other minerals. This sample shows the highest quantity of silver among the measured samples. One interesting sample (FG‑030678) demonstrates a porous lead and lead oxide with some carbonate and silicates. The lead content of this sample is too high to consider it as a lead slag, nevertheless two slag-related high-temperature minerals, namely akermanite (melilite) and cristobalite are observed in it. This sample shows the highest zinc content among the measured samples.

Fig. 5 – a: Section of a litharge fragment with a small piece of slag in the lower part; b, c: Porous litharge-lead minerals (FG‑030678); d: A piece from a litharge fragment mainly composed of lead with a copper prill in the lower middle part (FG‑030686); e: Thin section of the sample in the picture d; f, g: EDX mapping by SEM (element distribution images) showing inclusions of copper and silver in the lead matrix of sample FG‑030698 (N. Nezafati).

Image

31Mineralogical investigations (XRD and microscopic, tab. 1) show a rather uniform mineralogy for most of the litharge fragments, which include litharge, cerussite, hydrocerussite, lead (+As), calcite, pyroxene, plagioclase, quartz, bilixite (Pb2Cl(O, OH)2), asisite (Pb7SiO8Cl2), cotunnite (PbCl2), laurionite (PbCl(OH)), paralaurionite, woodhouseite ((Ba,Ca,Pb,Sr)(Al,Fe)3(S,As,P,O4)(SO4)(OH)6), mendipite (Pb3O2Cl2), magnetoplumbite (Pb(Fe3+,Mn3+)12O19), fluorite, cristobalite, akermanite, lanarkite (Pb2(SO4)O), massicot (PbO), protoenstatite (MgSiO3), anglesite, as well as some lead (+As) prills.

32The presence of minerals like barite, fluorite, as well as lead halide, is very interesting because the same or similar minerals have also been detected in the ore samples from the Baqoroq and Komjan ancient mines and also in some of the green and brown slags.

33Judging by the geochemical results of the major and trace elements (tab. 6, 7), the composition of most samples is almost uniform (tab. 6, 7) and shows 0.48 to 8.77% (average: 4.39) SiO2, 0.54 to 1.38% (average: 0.73) Al, 0.06 to 0.82% (average: 0.36) Fe, 4.76 to 10.78% (average: 7.71) CaO, 0.14 to 2.09% (average: 1.08) MgO, 0.01 to 0.26% (average: 0.15) P2O5, 203 to 745 ppm (average: 463) Cu, 11.2 to 1,700 ppm (average: 258) As, 0.01 to 3.18% (average: 0.59) S, 4 to 76.7 ppm (average: 28) Sb, 51.6 to 68.8% (average: 61.4) Pb, 19.8 to 204 ppm (average: 90.5) Ag.

Tab. 6 – Bulk chemical analyses of major elements and oxides of the litharge fragments measured by ICP‑MS (Zarazma Minerals Company).

Sample no.SiO2%Al%Fet%CaO%Na%K2O%
FG-0122504.420.570.318.560.130.25
FG-0122523.741.210.587.250.240.24
FG-0122674.540.730.348.860.090.14
FG-0306823.340.540.36.270.290.12
FG-0306860.480.0710.064.760.120.07
FG-0306975.410.640.245.460.240.22
FG-0306988.771.380.8210.780.360.55
FG-9931564.410.690.259.780.070.11
FG-03067819.261.240.8720.680.650.61
Sample no.MgO%MnO%P2O5%S%TiO2%BaO%
FG-0122501.160.040.170.120.050.01
FG-0122520.830.060.260.120.070.13
FG-0122672.090.040.110.140.060.02
FG-0306821.050.040.080.340.050.01
FG-0306860.14<0.010.390<
FG-0306970.740.030.20.10.030.01
FG-0306980.990.040.253.180.090.01
FG-9931561.70.040.120.360.050.01
FG-0306783.730.050.160.430.090.21

Tab. 7 – Bulk chemical analyses of metals in litharge fragments measured by ICP‑MS (Zarazma Minerals Company).

Sample no.Cu (ppm)As (ppm)Pb%Ag (ppm)Zn%Bi (ppm)Sb (ppm)MnO%
FG-01225034459.563.29178.2<0.423.50.04
FG-01225274510061.3224.70.020.476.70.06
FG-01226748613.959.619.8<0.416.60.04
FG-03068245111.263.5134.2<0.34.20.04
FG-03068630215.267.23203.9<14<
FG-03069720310068.7727.10.010.631.10.03
FG-0306985201,70051.55134.3<0.64.10.04
FG-99315666068.955.78101.6<0.345.10.04
FG-03067866574.229.3819.81.830.447.10.05
Sample no.BaO%S%Ni (ppm)Co (ppm)Mo (ppm)MgO%Sr%
FG-0122500.010.12116.80.91.160.14
FG-0122520.130.12219.430.830.46
FG-0122670.020.14167.51.62.090.14
FG-0306820.010.34189.42.11.05<
FG-030686<0.39713.53.20.14<
FG-0306970.010.1106.31.80.74<
FG-0306980.013.18205.76.20.990.14
FG-9931560.010.36145.621.70.14
FG-0306780.210.4398.124.73.730.14

34The SEM (point-analysis) investigations indicate that some lead prills, as well as the sample mainly composed of lead, contain up to several percent of arsenic, while silver reaches 0.5%. Some inclusions of copper sulfides are also observed in some of the litharge samples (fig. 5f, 5g). An EDXRF analysis by Pernicka et al. (2011) shows 4.5% of copper in a litharge fragment (FG‑030166), while the other analyses indicate 0.3 (0.65)% of copper and up to 500 (690) ppm of antimony.

35From the geochemical analysis (tab. 6, 7), it appears that lead (up to 68.8%), SiO2 (up to 8.77%), CaO (up to 10.78), and MgO (up to 2.09%) are the major components of most of the litharge fragments. Nevertheless, sulfur, copper, arsenic, and antimony, as well as P2O5, present considerable anomalies in some cases. High quantities of arsenic, antimony, and copper are not typical of the argentiferous lead ore of Nakhlak, which isotopically matches very well with the litharge samples of Arisman.

36The silver concentration in the nine litharge samples of Arisman ranges between 19.8 and 204 ppm (average: 90.5), four of which display a silver content higher than 100 ppm, while the silver content of the other five range between 19.8 and 34.2 ppm. This is much higher than the silver content of the litharge fragments reported from the contemporaneous ancient sites of Fatmalı-Kalecik in Anatolia (Hess et al. 1998) and Habuba Kabira in Syria (Kohlmeyer 1994). The gold measured by the point-analysis of SEM shows a concentration up to 0.88% in these samples, which is considerable.

37From the above, it seems that the cupellation hearths of Arisman were made of a material composed of silica and carbonate (CaO+MgO), and possibly salt, in which argentiferous lead containing some sulfur, arsenic and antimony was worked. The presence of salt in the building material of litharge was concluded from several chlorine-bearing minerals identified by XRD.

Geochemistry of analyzed artifacts

38No geochemical analyses were performed on the metal artifacts of Arisman in the frame of the present study, but the results of the analyses published by Pernicka et al. (2011) have been used and reassessed in order to draw final, comprehensive conclusions.

39Pernicka et al. (2011) performed an EDXRF analysis on 28 artifacts from Arisman (tab. 15), 25 of which were copper-based and three lead-based artifacts. These metal finds belonged to different categories, which include beads (4 pieces), pins (9), spirals (3), awls (3), chisel (1), some metal fragments (7), and one pendant.

40The results of the analyses are the following: <0.05 to 15.4% Fe, 72 to 100% Cu, 0.059 to 5.7% As, <0.01 to 2.69% Ni, <0.005 to 2.62% Sb, <0.01 to 8.4% Pb, and <0.005 to 9.9% Ag. This shows that all the copper-based artifacts contain arsenic with variable contents (in most cases high amounts of arsenic). Silver, iron, lead, nickel, and antimony contents are also considerably high in some artifacts. Among the objects, a pendant contains the highest amounts of iron and silver with 72% copper and 1.6% lead. Interestingly, a litharge sample has 4.5% copper. Zn, Sn, and Au were sought out, but they appeared to be below the detection limit of ca 0.01% for Zn and Au, and 0.005% for Sn in all samples (Pernicka et al. 2011).

41The content of trace elements in the artifacts does not seem to depend on the type or function of the objects. Moreover, the variable contents of arsenic and some other elements like Fe, Pb, Ni, Co, Sb, Bi, and Ag could result from the uncontrolled procedure of smelting, especially for arsenic. This may also indicate that the composition of the final metal product was much dependent on the composition of the ore.

Lead isotope investigations

42In this section, the lead isotope data from Arisman and Tappeh Sialk previously published by Pernicka et al. (2011), Nezafati, Pernicka (2012), Nezafati, Pernicka, Malek Shahmirzadi (2008), Schreiner (2002) and Schreiner, Heimann, Pernicka (2003) will be reviewed and reassessed (tab. 8). Through the comparison of the lead isotope ratios of the copper prills, litharge pieces, and slag pieces of Sialk and Arisman, together with ore samples from the mines of the Karkas Mountains and the Anarak area, we can draw the following conclusions (fig. 6):

  • The litharge pieces of Arisman and Tappeh Sialk – which belong to the same time-period – have matching lead isotope signatures; they are also in a good isotopic correspondence with some of the ore deposits of the Anarak area. The lead ore of the Nakhlak ancient mine appears as the best match, but the Baqoroq and Chah Mileh ores also match very well with the archaeological pieces. The area marked with a red line in fig. 6 illustrates this narrow lead isotope range.
  • The brown and green slag samples are in a good correlation isotopically, except for one brown slag sample that matches with the Sialk III period copper prills of Arisman.
  • Interestingly, the lead isotope ratios of green and brown slag pieces of Arisman are in a very good correspondence with the same ratios of litharge pieces of Arisman and Tappeh Sialk.
  • The Pb‑isotope ratios of the slag from Arisman and Tappeh Sialk are similar, although some differences can be observed.
  • There is a very good match of lead isotope ratios between the ore from a few deposits in the Anarak area and in the Karkas Mountains and the Pb‑ratios of Arisman slag.
  • The copper prills collected from different-period occupation levels at Arisman I (Sialk III and Sialk IV periods), show different lead isotope signatures. Interestingly, they also match with isotopically different ores. Only one copper prill from the Sialk IV period at Arisman matches well with the slag pieces of Arisman, as well as with the ore from Baqoroq, Chah Mileh, Khuni, Darhand, and Hollabad, while the three copper prills from Sialk III period do not match with the aforementioned slag and ores but only with the ore from Komjan.
  • The ore from the mines of Baqoroq, Chah Mileh, Khuni, Ghebleh, Chah Gorbeh in the Anarak area and the ores from Hollabad and Darhand show very similar lead isotope signatures. Interestingly, the mentioned deposits of the Anarak area contain polymetallic ore with arsenic, lead, copper, silver, and in some cases (e.g. Khuni) gold. The Darhand deposit represents a native and oxide copper ore where no ancient mining has been observed. The Hollabad deposit exhibits some ancient mining relics, but hematite (with rare visible copper) is the main ore exploited from it (Nezafati 2000).
  • In this report, only the slag and ores that were in a good correspondence with the available data from Arisman and Sialk have been considered, therefore the ores that do not match very well with the Arisman and Sialk slag pieces, such as the ore from Veshnaveh, have not been mentioned. The Veshnaveh ore does not match very well with the slag of Arisman and Sialk from geochemical point of view either (for instance this ore contains almost no arsenic; Pernicka et al. 2011; Nezafati, Stöllner 2017).

Tab. 8 – Lead isotope analysis results of ore, slag, litharge, and artifacts from Arisman and Tappeh Sialk.

Sample no.LocationMaterial208Pb/206Pb207Pb/206Pb206Pb/204PbReference
TS-2D-693SialkSlag2.096930.8571718.299Schreiner 2002
TS-2D-694SialkSlag2.08320.8453518.566Schreiner 2002
TS-2D-695SialkSlag2.083190.845118.555Schreiner 2002
TS-2D-696SialkSlag2.079420.8455718.563Schreiner 2002
TS-2D-698SialkSlag2.081350.8457918.567Schreiner 2002
TS-2D-715SialkSlag2.083240.845118.568Schreiner 2002
TS-2D-716SialkSlag2.080670.846118.546Schreiner 2002
TS-3D-699SialkSlag2.090230.848118.501Schreiner 2002
TS-3D-700SialkSlag2.080230.8437118.581Schreiner 2002
TS-3D-701SialkSlag2.084130.8559618.485Schreiner 2002
TS-3D-702SialkSlag2.080340.8451718.504Schreiner 2002
TS-3D-703SialkSlag2.080730.8452718.539Schreiner 2002
TS-3D-704SialkSlag2.084590.8474418.517Schreiner 2002
TS-3D-717SialkSlag2.084520.8479118.521Schreiner 2002
TS-3D-718SialkSlag2.082120.8457218.702Schreiner 2002
FG-012267ArismanLitharge2.08720.8447118.5237Pernicka et al. 2011
FG-012250ArismanLitharge2.08710.8447418.5215Pernicka et al. 2011
FG-012251ArismanLitharge2.08720.8447518.5235Pernicka et al. 2011
FG-012252BArismanLitharge2.08700.8446618.5229Pernicka et al. 2011
FG-012252AArismanLitharge2.08700.8446818.5235Pernicka et al. 2011
FG-012253ArismanLitharge2.08720.8448218.5206Pernicka et al. 2011
FG-012233BArismanBrown slag2.08540.8443418.553Pernicka et al. 2011
FG-012269-BArismanBrown slag2.09140.8458818.500Pernicka et al. 2011
FG-010506ArismanBrown slag2.12060.8729917.961Pernicka et al. 2011
FG-010505ArismanCopper slag2.08560.8453118.563Pernicka et al. 2011
FG-000119ArismanCopper slag2.09250.8514918.655Pernicka et al. 2011
FG-010476CArismanCopper slag2.08270.8438318.799Pernicka et al. 2011
FG-010484ArismanCopper slag2.08160.8425818.615Pernicka et al. 2011
FG-012256-AArismanCopper slag2.08610.8429818.567Pernicka et al. 2011
FG-012270AArismanCopper slag2.09290.8442018.652Pernicka et al. 2011
FG-012270BArismanCopper slag, porous2.08760.8426118.590Pernicka et al. 2011
FG-010504AArismanCopper slag2.09280.8451618.568Pernicka et al. 2011
FG-010504ArismanCopper slag2.09220.8516218.492Pernicka et al. 2011
FG-012275ArismanCopper slag2.09360.8459418.534Pernicka et al. 2011
FG-000118ArismanCopper slag2.09640.8477718.503Pernicka et al. 2011
FG-010501AArismanArisman II-Slag2.08330.8412918.620Pernicka et al. 2011
FG-010501ArismanArisman II-Slag2.08540.8481618.606Pernicka et al. 2011
FG-012257ArismanCopper prill2.12160.8723717.979Pernicka et al. 2011
FG-012244ArismanCopper prill2.09680.8525318.379Pernicka et al. 2011
FG-012247ArismanCopper prill2.11390.8665918.108Pernicka et al. 2011
FG-012246ArismanCopper prill2.10770.8619718.190Pernicka et al. 2011
FG-012245ArismanCopper prill2.09460.8470818.519Pernicka et al. 2011
FG-012250ArismanCopper prill2.08710.8447418.522Pernicka et al. 2011
FG-010473KomjanOre2.11560.8672018.079Pernicka et al. 2011
FG-010502HollabadOre2.09220.8516218.492Pernicka et al. 2011
FG-012033DarhandOre2.08290.8419018.599Pernicka et al. 2011
FG-010486DarhandOre2.08400.8437118.599Pernicka et al. 2011
FG-012140BaghoroghOre2.08590.8446318.533Pernicka et al. 2011
FG-012147Chah MilehOre2.08980.8436618.579Pernicka et al. 2011
FG-012147Chah MilehOre2.09080.8436218.583Pernicka et al. 2011
FG-012147Chah MilehOre2.09080.8436018.581Pernicka et al. 2011
FG-012153KhuniOre2.08970.8476918.399Pernicka et al. 2011
FG-012153KhuniOre2.09210.8480118.409Pernicka et al. 2011
FG-012153KhuniOre2.09140.8478418.41Pernicka et al. 2011
FG-012194GheblehOre2.08960.8448818.517Pernicka et al. 2011
FG-012215Chah GorbehOre2.08830.8429518.593Pernicka et al. 2011
SRP-TBSialk III, IVLitharge2.08650.8444818.513Nezafati et al. 2008
SRP-Reg.22Sialk IVLitharge2.08680.8445818.518Nezafati et al. 2008

Fig. 6 – a: Lead isotope plot of ore, slag and litharge fragments of Arisman, Tappeh Sialk, and some ancient mines of the Karkas and Anarak areas; b: Comparison of the lead isotope signatures of the samples from Nakhlak (ore and slag), Sialk (litharge), Arisman (litharge and Cu‑prill), and Baqoroq ore; note that the scale of the diagram is greatly expanded, all 2‑a (95% confidence level) errors are less than 0.05% for the 207Pb/206Pb and 208Pb/206Pb ratios; diagram drawn based on the data presented in tab. 8. (N. Nezafati).

Image

Tentative sources for the ancient ore

43During the Arisman project the first author (occasionally together with other colleagues) surveyed the Karkas and Anarak areas where he visited and studied several ore deposits and ancient mines. In this research, the two ancient mines of Baqoroq and Komjan were chosen and re-examined because of their favorable (geochemical and mineralogical) characteristics. In the following section, these two ancient mines will be described briefly.

The ancient mine of Baqoroq

44The Baqoroq deposit is located about 170 km south-east of Arisman and 4 km north-west of the Nakhlak mine. Arisman could be accessed via a flat plain from Baqoroq (the access road between these two sites is not mountainous). The mineralization developed in Upper Cretaceous rock units, which appear as a series of hills surrounded and partly covered by the sand dunes of the Dasht‑e Kavir (fig. 7a, 7b).

Fig. 7 – a: Location map of the Baqoroq ancient mine; b: Geological map of the Baqoroq-Nakhlak mines (modified after Anarak 1:250,000 geologic map, Geological Survey of Iran); c: Part of (possible) ancient diggings at Baqoroq; d, e: mineralization at Baqoroq; f: Geological and location map of the Komjan ancient mine (modified after Kashan 1:250,000 geologic map, Geological Survey of Iran); g: A conical depression ancient digging at Komjan (N. Nezafati).

Image

45The mine has been mentioned and described by Ladame (1945), Bariand (1962), Bazin, Hübner (1969), Romanko et al. (1984), Pernicka et al. (2011) and Jazi, Karimpour, Malekzadeh Shafaroudi (2016). It has been exploited between 1935 and the outbreak of the Second World War. The maximum annual production was reported to have been about 1000 t of sorted ore with a 5 to 12% copper grade.

46The remains of mining are rather vast and composed of multilevel underground works. Several old pits including two 75m‑shafts, a large open pit, an adit, together with the remains of a few buildings and a copper smelter (including a furnace, roasting furnaces, and huge amounts of large conical slags) are preserved and visible (fig. 7c, 7d).

47The host rock is composed of a partly brecciated limestone containing arenaceous and pelitic (calcareous conglomerate and sandstone) subunits from the Upper Cretaceous age (similar to the host rocks of the Nakhlak mine). The main part of the primary rich mineralization on the surface has been exploited by the previous mining activities and the remaining mineralization is very sparse and mainly of secondary character. The mineralization has an open space filling and replacement nature and has mainly occurred in E‑W and NE‑SW oriented calcite‑barite and barite veins (fig. 7d, 7e).

48Unfortunately, only a few mineralogical and analytical records from Baqoroq are available and even these are mainly based on the oxidized surface samples and do not present all the trace elements of the ore. Anyhow, drawing on one ICP-MS analysis (this study) and several other results published by Romanko et al. (1984), Pernicka et al. (2011), and Jazi, Karimpour, Malekzadeh Shafaroudi (2016), we may say that the ore of Baqoroq is composed of chalcocite, chalcopyrite, fahlore (tennantite), arsenopyrite, magnetite, covellite, cuprite, chrysocolla, malachite, and azurite. Barite, calcite, dolomite and some quartz are the main gangue minerals. The oxidized ore shows higher amounts of copper (up to 33.8%), arsenic (up to 2.07%), silver (up to 303 ppm), and rather high anomalies of lead (up to 0.1%), zinc (up to 1.8%), and antimony (up to 9,932 ppm).

49The mineralogy of the ore, gangue, and the host rocks, the trace elemental pattern, as well as the lead isotope signature of the Baqoroq mine matches (very) well with ore that may have been used in Arisman I.

The ancient mine of Komjan

50The polymetallic (Cu, Pb, Zn, and Ag) ancient mine of Komjan is located about 33 km south-west from the ancient site of Arisman (fig. 7f). The mine is situated in the Paleozoic-Triassic core of the Karkas anticline (Stöcklin, Ruttner, Nabavi 1964). In this area the Neoproterozoic‑Early Paleozoic dolomite, which contains cherty nodules and bands, hosts the mineralization. Several travertine outcrops are observed around the area (Nezafati 2000).

51Ancient mining activities are present in two major forms: (1) at least 22 conical depressions measuring between 2 and 20 m in diameter and surrounded by waste dumps of mining – these depressions, which very much resemble the ancient mining remains of the Deh Hosein ancient mine (Nezafati 2006), are scattered over an area of 1.5 km2 (fig. 7g) –; (2) horizontal tunnels less than 5 m long, which seem to belong to the Islamic period activities. An Arabic inscription and a nearby rather large old metallurgical site with Islamic era pottery sherds (at Karvand farm) seem to be related to this type of mining works (Nezafati 2000). The analysis of a piece of slag from the nearby metallurgical site by Nezafati (2000) shows high amounts of Cu, Pb, Zn, Ag, and As.

52Although some mining remains from the Islamic period are present in the Komjan ancient mine, its main mining activities (the conical depressions) may be related to the prehistoric era. Unfortunately, as at Baqoroq, very few mineralogical investigations and geochemical analyses have been performed at Komjan. Only a few results are available from Nezafati (2000) and Pernicka et al. (2011), which have been used here. Four ore samples were analyzed (using ICP and XRD) during this study.

53According to Nezafati (2000) and this study, the ore minerals of Komjan are composed of bornite, chalcopyrite, covellite, cuprite, pyrite, malachite, smithsonite, calamine, galena, magnetite, pyromorphite (Pb5(PO4)3Cl), laurionite (PbCl(OH)), valleriite ((Fe2+,Cu)4(Mg,Al)3S4(OH,O)6), hematite, limonite, cerussite, and siderite. Barite, fluorite, quartz, and calcite compose the major gangue minerals. The chemical analyses (ICP; Nezafati 2000; and this study) from the ore that was collected from the surface shows higher amounts of copper (up to 50%), lead (up to 79%), arsenic (up to 1.3%), zinc (up to 6.4%), and silver (up to 23 ppm). The presence of lead chlorophosphate and lead halide minerals such as pyromorphite and laurionite in the Komjan mine is significant, especially because of the presence of similar minerals in the litharge fragments of Arisman.

Discussion and conclusions

54Drawing on the investigations that were so far performed on the Arisman metallurgical relics, we may reach the following conclusions:

  • An argentiferous polymetallic ore has been used for the production of copper and silver. This ore was a mixture of oxide, sulfide, and sulfosalts of copper, lead, iron, arsenic, and silver with subordinate amounts of antimony, nickel, cobalt, and bismuth (a mixture of fahlore and other minerals). The gangue was composed of quartz, calcite, barite, fluorite, and the remains of a carbonaceous host rocks. The ore deposit that provided this ore was mainly composed of carbonate rocks (limestone and dolomite) in which quartz-calcite-barite-(fluorite) veins, veinlets and lenses hosted a sulfide mineralization that had later been partly oxidized due to weathering and supergene processes.
  • The so-called green and brown slags are both waste materials resulting from the same production processes, not from two different metallurgical operations. This conclusion is supported by the presence of several slag pieces bearing the characteristics of both the green and the brown slags, as well as some interconnected green-brown slag pieces. The close proximity of the geochemical pattern of the two slag-types, the good match of their lead-isotopic signatures, the lower quantity of the brown slag compared to the green slag; all these elements appear as a rather convincing evidence.
  • A rather good trace elemental match is observed between the ore and the artifacts.
  • The same type of ore was used during smelting and cupellation. This is suggested by the considerable quantity of arsenic and some other metals in the litharge fragments, a significant amount of silver in the finished artifacts, and the very good lead isotopic match of the litharge and the slag pieces. Moreover, slag and litharge both show close genetic connections.
  • The presence of several chlorine-bearing minerals in the litharge samples is interesting because some similar chlorine minerals have also been detected in the ore from the Baqoroq and Komjan ancient mines. Nevertheless, the addition of salt in the cupellation process is also possible; this question requires more investigations.
  • The whole metallurgical activity was designed to produce arsenical copper at the same time as silver from one ore type and through two processes namely; (1) smelting for producing arsenical copper, argentiferous lead (final or semi-final products), and slag (waste); as well as (2) cupellation, which yielded silver and litharge.
  • Judging by the mineralogical evidence, the temperature reached during the smelting process was between 1,200 and 1,400oC.
  • Metallurgical activities including copper production and silver extraction by cupellation started in the Sialk III period and continued in the Sialk IV period, when the furnace-bound copper production flourished.
  • As concerns the provenance of the ore used at Arisman, the two ancient mines of Baqoroq and Komjan appear to be very good matches judging by the mineralogical, geochemical, and isotopic investigations so far carried out. The polymetallic sulpharsenide oxidic ore of the Baqoroq and Komjan deposits used by ancient metallurgists contain copper, arsenic, lead, and silver in a carbonate-siliceous host rock that Pb‑isotopically also matches well with the Arisman metallurgical remains. According to the lead isotopic and archaeological evidence, the Komjan ancient mine could have been the first mine exploited by the ancient miners of Arisman (Sialk III period), while the Baqoroq mine was exploited later during the Sialk IV period. Although the Nakhlak ore isotopically matches well with the litharge and the slag of Arisman and Sialk, it rarely contains copper, arsenic, and other elements that are vividly present in both the slag and litharge pieces of Arisman and Sialk. Moreover, the ores from Komjan, Baqoroq, Chah Mileh, Khuni, Ghebleh, and Chah Gorbeh not only match well, isotopically speaking, with the Arisman and Sialk ancient metallurgical remains, but they also correspond with them from a geochemical point of view. At any rate, the ore from Nakhlak should not be disregarded as possible source of silver. It could be used as a complementary additive during smelting.

55The Sialk III period at Arisman may have had different raw material (ore) sources for the production of metals. The ore used in the Sialk III period may have come from local sources such as the Komjan deposit, while the ore of the Sialk IV period may have come from sources like Baqoroq and possibly from even more distant deposits of the Anarak area.

56The data produced during our research and the revision of previously published data offer a different scenario (from what Pernicka et al. 2011 and Rehren, Boscher, Pernicka 2012 have proposed) for the ancient metallurgy at Arisman. There seems to be a strong technical and manufactural link between the copper production and the cupellation processes: the two processes were not separately carried out, but thoroughly interrelated. Smelting was probably carried out first and yielded two products; (1) arsenical copper, and (2) lead (or lead arsenide) containing silver and/or gold. It is even possible that the production of the latter was the first priority for the ancient metallurgists of Arisman, while arsenical copper was considered as a kind of by-product. Otherwise, they could have recovered a lot of arsenical copper mechanically from the existing slag, a process which has already been reported from Arisman (Steiniger 2011; Helwing 2011a).

57Smelting served as a double operation for producing arsenical copper, on the one hand, and a silver(-gold)-rich phase (possibly lead or lead arsenide), on the other hand. Through the production of arsenical copper during smelting, the rest of the metallic charge was diluted from copper, arsenic and sulfur, all elements that heavily impede the progress of a cupellation process. In this regard, arsenic and sulfur were enriched in the slag phase, while copper was partly enriched as arsenical copper in the form of matte and metal prills in the slag phase. At the same time the lead-bearing phase of the charge was melted by absorbing most of the silver (and gold) of the ore. Since lead and lead sulfides have a lower melting point than usual arsenic and copper-bearing minerals, they could melt faster and consequently absorb more silver and gold than arsenic and copper: they would leave the system faster than the others. The lead obtained from this stage could have been cupelled for gaining silver and gold. As a case in point, at least one lead ingot has been found at Arisman (with 6,000 ppm silver, by EDXRF; Pernicka et al. 2011).

58Drawing on these data and explanations and inspired by Keesmann’s publication (1991 and 1993), we propose a model (fig. 8) that could explain the coexistence of two types of slag, namely green (black) slag and brown slag at Arisman.

Fig. 8 – Schematic model for the furnace load of Arisman I involving complex argentiferous sulpharsenide ore where a good segregation between five major phases, according to their respective specific gravities is achieved; 1: glassy slag; 2: speiss-rich brown slag; 3: black (green) matte Cu‑rich slag, 4: copper metal; 5: argentiferous lead metal (based on the macroscopic and microscopic observations as well as the models presented by Keesmann 1991, 1993).

Image

59The smelting of a mixed oxidic and sulphidic (sulpharsenide) polymetallic ore was a one-stage procedure (mixed- or co-smelting operation) resulting in two products. This smelting was performed in furnaces (or crucibles), where the main body of the furnace charge would undergo a high temperature (1,200 to 1,400oC), highly reducing conditions (due to the presence of numerous metallic iron and iron-arsenides), while the marginal part of the charge would experience oxidizing conditions (delafossite in contact with ceramic material). As a whole, the furnace load goes through heterogeneous conditions as regards temperature, redox conditions, and material. The ore was rich in copper, arsenic, lead, silver, and maybe even gold (in some cases also in iron, nickel, zinc and antimony) and may well have been mainly composed of fahlore. Due to poor temperature control, the quantity of resulting slag in the furnace load was high, and this slag has functioned as a trap (or filter) for many elements including arsenic, sulfur, iron, part of copper, etc., while some elements like lead, silver, and part of the copper would go through. In fact, during the smelting process under the above-mentioned conditions, a good segregation between four major phases, which followed their respective gravity specificities would occur; (1) silicate slag (on the top), (2) an iron-arsenide-rich slag phase, (3) a matte-rich slag phase, and (4) copper (?) + lead metal (at the bottom). In such a process, most of the arsenic, iron, sulfur and part of the copper were trapped in the slag phases. This is probably the reason why most of the slag pieces are rich in copper and arsenides. The main yield of such a process was a mixture of copper and silver (gold) bearing lead. The metallic copper and lead could have later easily been separated by a simple melting due to their highly different melting points. It should be borne in mind that such copper still contains considerable amounts of arsenic (arsenical copper) and some other elements present in the ore (e.g. antimony, nickel, iron and some silver). It is also possible that most of the copper is stuck in the slag and the final metallic yield is only metallic lead enriched in silver. The (arsenical) copper prills were later recovered by mechanical separation. The lead that is now highly enriched in silver and gold still contains some of the undesired metals, which may impede the cupellation process. Therefore – and this conclusion is also based on the results of Pernicka et al. 2011 – the cupellation process was also performed in two steps. At first, the yield from such metals was more diluted; in the second step, silver and gold were extracted at a higher rate. It should be borne in mind that the whole process may occur in a furnace (but not in a crucible). Such furnaces have already been recovered from the slagheaps of Arisman (Steiniger 2011).

60Concerning the provenance of the ore used at Arisman, the polymetallic deposit of Baqoroq area appears as an ideal candidate, since it contains chalcocite, fahlore, cuprite, galena, and barite (thus a polymetallic sulpharsenide oxidic ore) with copper, arsenic, lead, and possibly silver just beside the famous Nakhlak deposit. This deposit is characterized by an Upper Cretaceous host rock of limestone, conglomerate and sandstone. The lead isotope signature of the Baqoroq deposit is also very close to the ore from Nakhlak. Nevertheless, the Komjan polymetallic deposit in Natanz area together with some other polymetallic or lead deposits of the Anarak area (including Nakhlak) have also probably contributed to the ore provision of metallurgical activities of Arisman during Sialk III and/or Sialk IV periods.

61It should also be noted that at Tappeh Hesar (also known as Tepe Hissar), where Thornton, Rehren, Pigott (2009) have mentioned speiss-slag with some inclusions of lead and As‑Sb‑rich copper, a considerable amount of litharge fragments has been recovered (Nezafati, Pernicka 2012). This could be an indication that similar processes were carried out both at Tappeh Hesar and Arisman, especially if we take the chronological similarities of the two sites into account.

Wider implications

62Reassessing the archaeometallurgy of Arisman from an analytical perspective has provided welcome new insights into the complex processes behind the development of early metallurgy on the Iranian Plateau by contributing to a more robust reconstruction of the technical procedure of copper and silver working, and the possible interrelation of the two metals. This development bridges the transformation from village communities to proto‑Elamite sites that are formed on urban models of social organization, bringing with it a shift from small to industrial scale metal production. In Arisman, new ore sources were exploited during this later phase, with an apparent shift to Baqoroq as main supplier. This entailed an increased need for long-distance transportation, possibly facilitated by the now available domesticated donkey as a beast of burden (Benecke 2011; Potts 2011).

63Although long-distance ore supply is only attested since the proto‑Elamite period for Arisman, the early urban system piggybacks on long-distance relations existing long before, not only with the alluvial lowlands but also with the southern Caucasus where a number of similarities with the Iranian sites indicate a shared development. Closely parallel are the early heavy copper tools, mostly shafthole axes, known in the Caucasus and also produced in Iran at the same time (Helwing 2017). Further evidence for a shared tradition are the recently found metallurgical remains from Alkhantepe, and the imported Sialk III ceramics from MPS 16 in the Mil Steppe. These strong relations seem to vanish in the later 4th millennium BCE, when the developments in both regions apparently move apart: the southern Caucasus metallurgical evidence is scarce for the post‑Leilatepe period, while the Iranian highlands see a surge of production in the new proto‑Elamite centers like Arisman or Sialk. The revival of the Caucasian copper production happens only in the 3rd millennium BCE with the full flourish of the Kura-Araxes tradition at a moment when the proto‑Elamite sites are just coming to an end.

Recommendations for further research

  • A thorough survey of the Komjan and Baqoroq ore deposits together with additional lead isotopic and trace elemental analysis of the ores could be decisive for getting a better grasp of the provenance of the ores used at Arisman.
  • The measurement of the gold content of the samples may reveal that gold was also produced on the site.
  • The lead isotopic analysis of some of the finished artifacts, and a few more ore samples from Arisman with consideration of their archaeological stratigraphic level would give better information on the use of the ores, and their provenance, through time.
  • Similar investigations on the metallurgical remains of Tappeh Sialk would help to reconstruct the raw material supply and trade of the Early Bronze Age in Central Iran.

Bibliographie

Des DOI sont automatiquement ajoutés aux références bibliographiques par Bilbo, l’outil d’annotation bibliographique d’OpenEdition. Ces références bibliographiques peuvent être téléchargées dans les formats APA, Chicago et MLA.

Akhundov 2014: T.I. Akhundov, “At the beginning of Caucasian metallurgy”, in G. Narimanishvili (dir.), International Conference “Problems of Early Metal Age archaeology of Caucasus and Anatolia” (Tbilisi, November 19-23, 2014), Tbilisi, 2014, pp. 11‑16.

Bariand 1962: P. Bariand, Contribution à la minéralogie de l’Iran, PhD, Université de Paris, Faculté des sciences (ser. A, no. 980), 1962, pp. 2‑64 (unpublished).

10.3406/bulmi.1963.5609 :

Baumgarten, Eldar 1984: Y. Baumgarten, I. Eldar, “Neve Noy: a Chalcolitic site near Beer-Sheba”, Qadmoniot 17, 1984, pp. 51‑56.

Bazin, Hübner 1969: D. Bazin, H. Hübner, Copper deposits in Iran, Tehran, Geological survey of Iran, 1969.

Benecke 2011: N. Benecke, “Faunal remains of Arisman”, in A. Vatandoust, H. Parzinger, B. Helwing (dir.), Early mining and metallurgy on the western Central Iranian Plateau. Report on the first five years of research of the Joint Iranian-German Research Project, published in Archäologie in Iran und Turan 9, 2011, pp. 376‑382.

Chegini et al. 2000: N.N. Chegini, M. Momenzadeh, H. Parzinger, E. Pernicka, T. Stöllner, R. Vatandoust, G. Weisgerber, in collaboration with N. Boroffka, A. Chaichi, D. Hasanalian, Z. Hezarkhani, M. Mir-Eskandari, N. Nezafati, “Preliminary report on archaeometallurgical investigations around the prehistoric site of Arisman near Kashan, western Central Iran”, Archäologische Mitteilungen aus Iran und Turan. AMIT 32, 2000, pp. 281‑318.

Ghirshman 1951: R. Ghirshman, L’Iran des origines à l’Islam, Paris, Payot, 1951 (English translation: Iran From the Earliest Times to the Islamic conquest, England, Penguin Books, 1978, 368 p. and Persian translation by Mohammad Moiin, Tehran, Elmi va Farhangi Publication, 1957).

10.3917/albmi.ghirs.1976.01 :

Görsdorf 2011: J. Görsdorf, “14C datings of Arisman”, in N.N. Chegini, B. Helwing, H. Parzinger, T. Stöllner, M. Momenzadeh, A. Vatandoust (dir.), Early mining and metallurgy on the Central Iranian Plateau. Report on the first five years of research of the Joint Iranian-German Research Project, published in Archäologie in Iran und Turan 9, 2011, pp. 370‑373.

Hauptmann 2007: A. Hauptmann, The Archaeometallurgy of copper, Berlin/Heidelberg, Springer, 2007.

10.1007/978-3-540-72238-0 :

Helwing 2008: B. Helwing, “Early metallurgy in Highland Iran, on the basis of the joint Iranian-German excavations at Arisman”, in L.R. Weeks (dir.), The 2007 Early Iranian metallurgy workshop at the University of Nottingham, published in Iran 46, 2008, pp. 335‑345.

Helwing 2011a: B. Helwing, “The small finds from Arisman”, in A. Vatandoust, H. Parzinger, B. Helwing (dir.), Early mining and metallurgy on the Central Iranian Plateau. Report on the first five years of research of the Joint Iranian-German Research Project, published in Archäologie in Iran und Turan 9, 2011, pp. 254‑328.

Helwing 2011b: B. Helwing, “Conclusions: the Arisman copper production in a wider context”, in A. Vatandoust, H. Parzinger, B. Helwing (dir.), Early mining and metallurgy on the western Central Iranian Plateau. Report on the first five years of research of the Joint Iranian-German Research Project, published in Archäologie in Iran und Turan 9, 2011, pp. 523‑531.

Helwing 2013a: B. Helwing, “Some thoughts on the mode of culture change in the fourth-millennium BC Iranian highlands”, in C.A. Petrie (dir.), Ancient Iran and its neighbours. Local developments and long-range interactions in the fourth millennium BC, Institute of Persian Studies Archaeological Monographs Series 3, Oxford/Oakville, Oxbow Books, 2013, pp. 93‑105.

10.2307/j.ctvh1dn46 :

Helwing 2013b: B. Helwing, “Early metallurgy in Iran: an innovative region as seen from the inside”, in S. Burmeister, S. Hansen, M. Kunst, N. Müller-Scheessel (dir.), Metal matters. Innovative technologies and social change in Prehistory and Antiquity, Rahden, Leidorf, 2013, pp. 105‑135.

Helwing 2017: B. Helwing, “Networks of craft production and material distribution in the Late Chalcolithic: metallurgical evidence from Iran and the southern Caucasus”, in E. Rova, M. Tonussi (dir.), At the northern frontier of Near Eastern archaeology: recent research on Caucasia and Anatolia in the Bronze Age. Proceedings of the International Humboldt-Kolleg (Venice, January 9‑January 12 2013), Publications of the Georgian-Italian Shida Kartli archaeological project 2/Subartu 38, Turnhout, Brepols, 2017, pp. 51‑78.

Helwing, Aliyev 2017: B. Helwing, T. Aliyev, “Excavations in the Mil Plain sites, 2012‑2014”, in B. Helwing, T. Aliyev, B. Lyonnet, F. Guliyev, S. Hansen, G. Mirtskhulava (dir.), The Kura Projects. New research on the later prehistory of the southern Caucasus, Archäologie in Iran und Turan 16, Berlin, Dietrich Reimer, 2017, pp. 11‑42.

Hess et al. 1998: K. Hess, A. Hauptmann, H.T. Wright, R. Whallon, “Evidence of fourth millennium BC silver production at Fatmali-Kalecik”, in T. Rehren, A. Hauptmann, J.D. Muhly (dir.), Metallurgica Antiqua. In Honour of Hans-Gert Bachmann and Robert Maddin, published in Der Anschnitt. Beiheft 8, 1998, pp. 57‑67.

Jazi, Karimpour, Malekzadeh Shafaroudi 2016: M.A. Jazi, M.H. Karimpour, A. Malekzadeh Shafaroudi, “Mineralization, geochemistry, fluid inclusion and sulfur stable isotope studies in the carbonate hosted Baqoroq Cu‑Zn‑As deposit (NE Anarak)”, Iranian Journal of Economic Geology 7/2, 2016, pp. 179‑202 (in Persian).

Keesmann 1991: I. Keesmann, “Rio Tinto. Die Technik der Silbergewinnung zu Beginn des Mittelalters”, in P. Benoit (dir.), Argent, plomb et cuivre dans l’histoire, Lyon/Villeurbanne, CNRS, 1991, pp. 1‑13.

Keesmann 1993: I. Keesmann, “Naturwissenschaftliche Untersuchungen zur antiken Kupferund Silberverhüttung in Südwestspanien”, in H. Steuerand, U. Zimmermann (dir.), Montanarchäologie in Europa. Berichte zum Internationalen Kolloquium “Frühe Erzgewinnung und Verhüttung in Europa” (4. bis 7. October 1990), Archäologie und Geschichte 4, Sigmaringen, 1993, pp. 105‑122.

Kierczak, Pietranik 2011: J. Kierczak, A. Pietranik, “Mineralogy and composition of historical Cu slags from the Rudawy Janowickie Mountains, southwestern Poland”, The Canadian Mineralogist 49/5, 2011, pp. 1029‑1044.

Kohlmeyer 1994: K. Kohlmeyer, “Zur frühen Geschichte von Silber und Blei”, in R.‑B. Wartke (dir.), Handwerk und Technologie im Alten Orient, Mainz, Philipp von Zabern, 1994, pp. 41‑48.

Ladame 1945: G. Ladame, “Les ressources métallifères de l’Iran”, Schweiz. miner. petrogr. Mitt. 25/1, 1945, pp. 165‑298.

Nezafati 2000: N. Nezafati, Study on the metallic minerals of the Natanz area, Central Iran, MSc thesis, Research Institute for Earth Sciences, Geological Survey of Iran, Tehran, 2000 (unpublished, in Persian with an English abstract).

Nezafati 2006: N. Nezafati, Au‑Sn‑W‑Cu‑Mineralization in the Astaneh-Sarband Area, West Central Iran; including a comparison of the ores with ancient bronze artifacts from western Asia, PhD thesis, University of Tübingen, 2006 (published, available on http://hdl.handle.net/10900/48972).

Nezafati, Pernicka 2012: N. Nezafati, E. Pernicka, “Early silver production in Iran”, Iranian Archaeology 3, 2012, pp. 37‑45.

Nezafati, Pernicka, Malek Shahmirzadi 2008: N. Nezafati, E. Pernicka, S. Malek Shahmirzadi, “Evidence on the ancient mining and metallurgy at Tappeh Sialk (Central Iran)”, in Ü. Yalcin, H. Özbal, A.G. Paşamehmetoğlu (dir.), Ancient Mining in Turkey and the eastern Mediterranean, Ankara, Atilim University, 2008, pp. 329‑350.

Nezafati, Stöllner 2017: N. Nezafati, T. Stöllner, “Economic geology, mining archaeological, and archaeometric investigations on the Veshnaveh ancient copper mine, Central Iran”, Metalla 23/2, 2017, pp. 67‑90.

Nokandeh 2010: J. Nokandeh, Neue Untersuchungen zur Sialk III-Periode im zentraliranischen Hochland: auf der Grundlage der Ergebnisse des “Sialk Reconsideration Project”, Berlin, Dissertation.de/Verlag im Internet, 2010.

Pernicka et al. 2011: E. Pernicka, K. Adam, M. Böhme, Z. Hezarkhani, N. Nezafati, M. Schreiner, B. Winterholler, M. Momenzadeh, A. Vatandoust, “Archaeometallurgical researches at Arisman in Central Iran”, in A. Vatandoust, H. Parzinger, B. Helwing (dir.), Early mining and metallurgy on the Central Iranian Plateau. Report on the first five years of research of the Joint Iranian-German Research Project, published in Archäologie in Iran und Turan 9, 2011, pp. 633‑705.

Pollard et al. 2013: A.M. Pollard, H. Fazeli Nashli, H. Davoudi, S. Sarlak, B. Helwing, F. Saeidi Anaraki, “A new radiocarbon chronology for the North Central Plateau of Iran from the Late Neolithic to the Iron Age”, Archäologische Mitteilungen aus Iran und Turan 45, 2013, pp. 27‑50.

Potts 2011: D.T. Potts, “Equus asinus in highland Iran. Evidence old and new”, in N.J. Conard, P. Drechsler, A. Morales (dir.), Between sand and sea. The archaeology and human ecology of southwestern Asia. Festschrift in honor of Hans-Peter Uerpmann, Tübingen, Tübinger Monographien zur Urgeschichte, 2011, pp. 167‑176.

Rehren, Boscher, Pernicka 2012: T. Rehren, L. Boscher, E. Pernicka, “Large scale smelting of speiss and arsenical copper at Early Bronze Age Arisman, Iran”, Journal of Archaeological Science 39, 2012, pp. 1717‑1727.

Romanko et al. 1984: E. Romanko, Y. Kokorin, B. Krivyakin, M. Susov, L. Morozov, M. Sharkovski, Outline of metallogeny of Anarak area (Central Iran), Moscow, Technoexport/USSR Ministry of Geology, 1984.

Schreiner 2002: M. Schreiner, Mineralogical and geochemical investigations into prehistoric smelting slags from Tepe Sialk, Central Iran, MSc thesis, TU Bergakademie Freiberg, Germany, 2002 (unpublished).

Schreiner, Heimann, Pernicka 2003: M. Schreiner, R.B. Heimann, E. Pernicka, “Mineralogical and geochemical investigations into prehistoric smelting slags from Tepe Sialk, Central Iran”, in S. Malek Shahmirzadi (dir.), The silversmiths of Sialk, published in Sialk Reconsideration Project Monograph 2, 2003, pp. 13‑24.

Steiniger 2011: D. Steiniger, “Excavations in the slagheaps in Arisman”, in A. Vatandoust, H. Parzinger, B. Helwing (dir.), Early mining and metallurgy on the Central Iranian Plateau. Report on the first five years of research of the Joint Iranian-German Research Project, published in Archäologie in Iran und Turan 9, 2011, pp. 69‑100.

Stöcklin, Ruttner, Nabavi 1964: J. Stöcklin, A. Ruttner, M. Nabavi, New data on the Lower Paleozoic and pre-Cambrian of North Iran. Report No. 1, Tehran, Geological Survey of Iran, 1964.

Stöllner 2016: T. Stöllner, “The beginnings of social inequality: consumer and producer perspectives from Transcaucasia in the 4th and the 3rd millennia BC”, in M. Bartelheim, B. Horejs, R. Krauß (dir.), Von Baden bis Troia. Ressourcennutzung, Metallurgie und Wissenstransfer. Eine Jubiläumsschrift für Ernst Pernicka, Rahden, Leidorf, 2016, pp. 209‑234.

Thornton, Rehren, Pigott 2009: C.P. Thornton, T. Rehren, V.C. Pigott, “The production of speiss (iron arsenide) during the Early Bronze Age in Iran”, Journal of Archaeological Science 36, 2009, pp. 308‑316.

10.1016/j.jas.2008.09.017 :

Vatandoust, Parzinger, Helwing 2011: A. Vatandoust, H. Parzinger, B. Helwing (dir.), Early mining and metallurgy on the Central Iranian Plateau. Report on the first five years of research of the Joint Iranian-German Research Project, published in Archäologie in Iran und Turan 9, 2011.

Précédent Suivant

Le texte seul est utilisable sous licence Creative Commons - Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International - CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.