2 4 ARCHAEOMETALLURGY OF COPPER AND SILVER ALLOYS IN THE OLD WORLD The production and processing of advanced materials, namely metals and alloys, began in the Old World about 8000 years ago and developed over many millennia, providing a lasting legacy for modern civilizations. Omid Oudbashi, Art University of Isfahan, Iran Russell Wanhill, Emmeloord, the Netherlands etals and alloys constitute an efforts to establish and promote scien- sometimes fail completely. Such exper- essential part of the develop- tific studies of (i) metallurgical process- iments enable a veritable appreciation ADVANCEDM MATERIALS & PROCESSESment | JULY/AUGUST of societies 2021 from Neolith- es and artifacts, from raw materials to of the empirically derived skills of an- ic times, and the earliest process metal- final production, and (ii) by-products, cient metalworkers. lurgy, melting and consolidation of na- tools, and equipment, e.g., slags, cruci- This article gives a brief overview tive metals, may be traced back to about bles, and furnaces (Fig. 1). of the production and processing of 6000 B.C. The importance of metals These studies use a wide range of ancient bronzes and silver in the Old technology in ancient societies is shown modern scientific methods and labora- World, and also mentions post-pro- by referring to the main periods of post- tory instruments to better understand cessing problems including corrosion Neolithic prehistory as the Copper, the complex processes involved, and and embrittlement, owing to long-term Bronze, and Iron Ages [1]. Approximate also the artifacts themselves and their burial before archaeological recovery. dates for the beginnings of these tech- eventual deterioration (especially cor- nologies in the Near East areas are: cop- rosion) over the millennia. This latter ANCIENT COPPER ALLOYS per (6000 B.C.), bronzes (3500 B.C.), and aspect is directly linked to conservation The first evidence of using na- iron (1500 B.C.). However, recent data and restoration techniques . tive copper to make small and decora- suggest that complex tin bronzes were The difficulties that had to be tive objects comes from the Near East smelted much earlier in the Balkans, overcome are well demonstrated by ex- and Caucasus and is dated to about around 4500 B.C.[2], but this technology perimental archaeometallurgy, i.e., py- 8000 B.C.[1]. Processing native copper was effectively lost after 4000 B.C. rometallurgical experiments to smelt by melting and casting began around Understanding process metallurgy metals from ores in ancient-style cru- 6000 B.C., and reduction of copper in the ancient world is a major remit of cibles and furnaces. Even with modern ores (smelting) to derive copper began archaeometallurgy. Over the last 50 to scientific knowledge these experiments around 4000 B.C. It is important to note 60 years there have been international may be only partially successful and that the ores were mined from copper sulfide deposits, where the weathered upper lay- ers consisted mostly of copper carbonates and ox- ide. These could be simply added to smelting cruci- bles and furnaces. How- ever, continued mining reached the sulfide depos- its, and these had to be oxidized (roasted) before Fig. 1 — Schematic of the main aspects of archaeometallurgical studies. Adapted from Bayley et al. [3] . smelting. 2 5 Early processing was done us- were often present in copper-bearing distinct possibility [8],, and the Andean ADVANCED MATERIALS & PROCESSES |JULY/AUGUST 2021 ing crucibles containing crushed ores ores. On the other hand, analysis of Ear- region study reinforces this [10]. The sec- and charcoal, with forced airflow pro- ly Bronze Age slags from Iran shows that ond hypothesis is disfavored by an ex- vided by bellows-powered blowpipes. speiss, an iron-arsenic alloy, was prob- tensive study and comparison of the Later on, crucible and hearth furnac- ably added to copper ore or during re- mechanical properties of arsenical and es using forced air via tuyères provid- melting to obtain arsenical bronzes [8,9]. tin bronzes [10]. There remains the possi- ed more controlled conditions. Figure 2 Also, although digressing here from the bility that smelting arsenical ores was is a schematic of a smelting furnace Old World, there is convincing evidence abandoned in Eurasia owing to health from the Near East Late Bronze Age that the Andes region arsenical bronz- concerns. However, arsenical bronz- (LBA: 1550‒1200 B.C). Besides initial es containing 0.5‒2 wt% arsenic were es were still being produced in the LBA metal production, such a furnace could intentionally produced from about (Fig. 3), 1000 years after tin bronzes be- be used to remelt additions of other 850 A.D. for cold-hammering into cul- came predominant. copper metal before tapping into clay turally desirable small implements and The majority of Near East tin or stone molds to cast ingots or arti- thin sheet materials [10]. bronzes have tin contents less than facts, for example, vessels, tools, and Returning to Eurasia, two more about 12 wt%, typically ranging from ornaments. important questions arise. Why did tin 5‒10 wt% from about 3000 B.C. [4]. The Large “oxhide-shaped” copper in- bronzes become the main type, large- earliest EBA alloys have lower tin con- gots were widely used in Eurasia as ly replacing arsenical bronzes after tents, 1‒3 wt%; and there are occa- trade items in the LBA [4], and these could 2500 B.C., and why did antimony bronz- sional exceptions, the high-tin alloys be remelted with additions of tin or tin es almost disappear after 2000 B.C. [7]? already mentioned. Hence most of the oxide (cassiterite), and possibly other Possible answers have been given, but materials and artifacts would have had alloying metals, to produce bronze in- there is no consensus. Firstly, antimony homogeneous single-phase microstruc- gots or cast artifacts including vessels, bronzes may have been supplanted be- tures after working and annealing, very tools, weapons, and ornaments. cause their lesser hardness, and hence different from the inhomogeneous as- Near East Early Bronze Age (EBA: lesser strength, made them unsuitable cast structures (Fig. 4). 3300‒2100 B.C.) ingots were proba- for tools or weapons. This could have bly forged by cold-working rather than resulted in a lack of demand and trade ANCIENT SILVER ALLOYS hot-working [5], and with intermittent an- in favor of tin bronzes, though this is not Owing to native silver’s scarcity, nealing, depending on the metalsmith’s (yet) known [7]. there is limited evidence of its direct use experience with the materials and the The more intriguing question is for artifacts, a few of which have been required artifacts. This practice con- the predominance of tin bronzes over dated to 4300‒4000 B.C. [11]. Silver was tinued well into the Iron Age, beyond arsenical bronzes, beginning in the later more abundant as a minor component 1500 B.C. Hot-working would have grad- EBA. There are three basic hypotheses: in the ores of other metals, especial- ually developed as an alternative, ex- (i) tin bronzes were intentional alloys ly lead [12]. Beginning before 3000 B.C., cept for high-tin bronzes, because “hot but arsenical bronzes were not; (ii) tin lead obtained from smelting argentif- shortness” (brittle cracking at high tem- bronzes had superior mechanical prop- erous lead ores was further processed peratures) would become increasingly erties; (iii) smelting arsenic-containing likely with tin contents above 8 wt% [6]. ores resulted in poisonous fumes that by cupellation to extract the silver. This became recognized as a health haz- process became the primary source of ANCIENT BRONZES ard. The first hypothesis has been dis- ancient silver and silver artifacts, al- The history of ancient bronzes cussed already: intentional alloying to though some artifacts were obtained is complex, spanning a “classic” peri- obtain Eurasian arsenical bronzes is a from direct smelting of silver ores [12]. od of more than 2000 years in Eurasia (3300‒1200 B.C.). Many issues are still unresolved, despite extensive studies since the early 20th century. Perhaps the most important question is whether the presence of alloying elements in copper was always accidental or became inten- tional. Considering the three main types of bronzes, antimony bronze, arsenical bronze, and tin bronze, the evidence of intentional alloying for tin bronzes is in- controvertible. However, deliberate al- loying with antimony and arsenic can be questioned[5,7], since these elements Fig. 2 — Schematic copper smelting furnace, Crete, Late Bronze Age. Adapted from Tylecote [4] . 2 6 Cupellation was a multistage pro- groove and was discarded. More bul- the litharge cones discarded, and the cess employing three separate hearths. lion was added until sufficient sil- rods re-dipped. Eventually this second Figure 5 is a schematic of a first stage ver-enriched lead was obtained for the stage left a silver globule on the hearth. hearth for enriching smelted lead bul- second stage. Then the enriched lead In the third stage, a number of globules lion. This was remelted to a high tem- was transferred to a second hearth and were melted and further refined in an- perature using wood fuel. Bellows- again oxidized, but here the litharge other hearth to obtain ingots, the re- powered tuyères oxidized the lead to was removed by dipping iron rods into maining PbO being absorbed by pores litharge (PbO), which melts at 880°C, it (before 1000 B.C., wooden poles) to in the cupel wall. hence the need for a high tempera- form layered litharge cones on the rods. Cupellation is very effective in pro- ture. The litharge drained via a surface These rods were repeatedly removed, ducing silver above 95% purity. It usu- ally contains minor-to-trace amounts of (a) (b) copper, gold, bismuth, and lead (gener- ally below 1 wt% for each), and traces of antimony, arsenic, tellurium, zinc, and nickel.