Noble metals can be separated from base metals by liquidation of the metal with lead, then cupellation of the lead, however, this does not separate silver from gold. To remove silver from gold a method of parting has to be used. This was by either acid parting, or the salt cementation process, and later by sulphur parting. For full discussion of the history of parting see Rampage and Craddock 2000.
Acid parting
Although weak organic acids could improve the colour of gold containing copper, these would not remove silver, so that silver-rich gold would remain a pale colour. The removal of silver required the use of strong acids (hydrochloric and nitric in particular). Acid parting only became possible when strong mineral acids became generally available. The distillation is thought to have been first used during the eleventh or twelfth centuries in Europe, although it is some evidence that they were used by the 10th century in the Islamic world.
The method is not mentioned by Theophilus, but various 16th century European metallurgical treatises and handbooks describe the method in some detail; they suggest that it was not a cost effective alternative to salt parting. However, it seems likely that it became more widely used by the end of the 16th century and during the 17th century.
Antimony Parting
See sulphur and sulphide parting.
Salt Cementation Parting Until the discovery of strong mineral acids salt cementation parting was the main method used to refine gold. The literature would suggest that there were a number of different variations on the process, but the essentials were the use of common salt cement and the presence of alumino-silicate ceramics preferably with an appreciable iron oxide content (in the form of an earthenware container, and possibly the addition of crushed brick or earthenware powder to the salt cement).
The impure gold was beaten into thin sheets, or was granulated to produce fine grains, or was naturally in the form as an ore. The gold was interleaved with layers of ‘cement’ formed of either salt or a mixture of slightly moistened ground up alumino-silicate material (burnt clay – old pot or tile) and salt (NaCl) in a sealed clay vessel. On heating the salt would react with the alumino-silicates of the burnt clay together with the water or urine, used to moisten the cement, to liberate volatile hydrogen chloride or chlorine gas. This gas, in turn, reacts with the silver at the surface of gold to form volatile silver chloride (AgCl), which was absorbed by the alumino-silicate in the ‘cement’.
The process had to be carried out at as high as a temperature possible to speed the diffusion process. But the temperature was limited by either the melting points of either the alloy, or of the active reagents if no carrier medium was used. If the gold alloy melted, or the temperature rose very close to its melting point, the foil would ball up. This would reduce the surface area available for attack and increase the diffusion distances, thus slowing the reaction. If no carrier material (brick or pottery dust) was used, the gold would sink through a simple salt cement if the temperature rose too close to the melting point of salt (804°C).
At the end of the process the gold would be melted into small ingots that could be tested for purity and if found to be of low purity these could be forged into foils so that the parting process could be repeated. The spent cement would be smelted with lead, which would be cupelled to recover the silver.
Various versions of the process have been described in the ancient literature, with the addition of other salts such green vitriol (hydrated ferrous sulphate), saltpetre (potassium nitrate) in addition to common salt. However, the presence of large amounts of nitrates in the mixture would result in the loss of gold as the combination of hydrochloric and nitric acid will dissolve the gold as well as the silver.
Although widely used until the 18th century, the process was largely forgotten and the distinctive debris from the process is rarely identified, as the sherds do not have the vitrified internal surfaces typical of used crucibles. However, parting vessels do have a distinctive purple tinge due to the presence of silver chloride.
See Craddock 1995 216-219 and Bayley 2001, Rampage & Craddock 2000 for more information.