The problem of the origin of Cu:Sn bronze has intrigued historians for many years. Bronze artifacts have been dated to 5000 BCE on the Iranian Plateau.  It is thought that the earliest bronzes were arsenical in nature. The presence of arsenic in copper metal or copper ore is not uncommon.

Copper can be found as the native metal but the smelting of copper ore appears to date back to ca 5000 BCE in southeastern Europe in what is now Serbia.

Most commonly today, the word bronze refers to a range of copper alloys comprising various proportions of copper (major, e.g., 88 %) and tin (minor, e.g., 12 %).  As the tin content increases, the resulting alloy changes properties and may have a unique purpose and name. For instance, a ratio of ca 2:1 :: Cu:Sn is called speculum and was prized for it’s ability to take a high polish for mirror applications.

Further down the composition range are varieties of pewter which are alloys comprised substantially of tin and a few percent of copper and antimony for hardening.  Many specalized compositions of pewter have been developed. Britanium or Britannia metal is an alloy comprised of 93 %Sn, 5 % Sb, and 2 % Cu. This alloy serves as the base metal Oscar Award Statue upon which gold is plated.  Pewters composed of Sn:Pb were commonly used as well.

Tin is not found in the metallic state in nature. It is oxophilic and occurs primarily as the tin (IV) oxide mineral, cassiterite. Tin ore was mined in Cornwall, England, for instance, for many centuries before recorded history.  Today, most of the worlds tin comes from Asia, South America, and Australia.

The jump to “engineered” bronze was a step change that involved the reduction of a tin mineral either in situ with copper or in isolation to produce discrete tin. It is thought that polymetallic copper ores were smelted, producing Cu:Sn bronze directly. Eventually, tin ore was identified as a source of smeltable metallic tin.  Why anyone would think to apply reduction conditions to a mineral as seemingly featureless and uninteresting as cassiterite is an intriguing question.

Below is a photo of the result of my first attempt at smelting a cassiterite simulant (SnO2, Aldrich). The SnO2 was treated with carbon black at 900 C for 4 hours in a covered porcelain crucible in a muffle furnace.  After a  failed attempt with a large excess of carbon, the ratio was reversed and heated for a longer period.  For the illustrated sample, the mass ratio of SnO2 to carbon black was ~2:1. All of the carbon black was consumed, leaving a white mass of needles on the granular solids.  Using a USB microscope I searched for evidence of reduction to the metallic state and found numerous examples of sub-millimeter sized pieces of metal.  The yield of metallic tin is estimated at < 1 %.

The purpose of this exercise (for me) is to try gain a better sense of what problems people might have faced smelting tin in antiquity.  Using basic principles, I strongly heated the SnO2 under reducing conditions until the carbon was consumed.  What I did not expect was the large amount of white crystalline material produced. It’s composition is as yet unknown to me.

Next I will make some charcoal or even wood shavings as a reductant for authenticity sake. Who knows, maybe some carbon monoxide generation might be helpful. The muffle furnace does not simulate a reverberatory furnace very well. It could be that gases from a reducing flame are important.

Smelting of Cassiterite Simulant