NEW HORIZONS — Fingerprinting gold

Historians of gold have always pondered the origin of the metal used by Venetians in the 12th century for their ducat coins and by Benvenuto, the Renaissance goldsmith, for his famed gold salt cellar.

We may get some answers from technology developed in South Africa by The Anglo American Research Laboratories (AARL) to fingerprint samples of gold as an effective means of determining their source. The fingerprinting was designed, not to aid historians, but to assist police in identifying gold stolen from South African mines. As much as 300,000 oz. may be stolen each year.

AARL has compiled the Gold Bullion Databank, which contains profiles of gold from South African mines, as well as a growing number from mines around the world.

A single profile comprises 15 samples taken from several days’ production at a mine or shaft. These samples are revised with fresh samples every two years. The profiles reveal minor and trace element impurities that vary from mine to mine and, sometimes, from shaft to shaft. Initially, a microscopic sample of gold is removed by laser and analyzed with a mass spectrometer to reveal the presence or absence of 58 elements and their 151 isotopes.

The resulting profile is then compared with those in the databank. In a test case in 1995, a consignment of 4,000 oz. that allegedly originated in Mozambique was proved to have come from 11 different South African mines.

Since 1995, the origins of nearly 10,000 oz. gold in 24 separate cases of theft have been identified.

Over the next few years, it is hoped that the Gold Bullion Database will expand to embrace profiles of most of the world’s gold mines. Indeed, insurance companies might insist on a profile of each mine or open pit as a protection against theft. To date, the database contains tentative profiles from mines in Mozambique, Zimbabwe, Tanzania, Mali and Zambia.

This information is providing the first evidence of the other benefits of fingerprinting — tracking the historic sources of gold. Metal working in southern Africa can be traced back at least 2,000 years, and the gold may have gone not just north to the Mediterranean, but across the Indian Ocean to India or China. Thus, important clues to intercontinental trading may eventually be revealed.

Seventy-six artifacts recovered by the University of Capetown from three archeological sites dating back to the 10th century provided profiles that showed gold from two of the sites had similar origins, somewhere along the Limpopo River. The gold used at the third community, farther north in Botswana, has a different profile.

One of the towns along the Limpopo River, Mapungubwe, in South Africa, is thought to have dominated Indian Ocean trade from the 10th to 13th centuries. A breakthrough would be matching gold artifacts in Asia with their African sources.

While such breakthroughs may be years away, the Anglo American laboratory is now testing 18th-century gold coins made in London to see what similarities they share with profiles of modern West African sources such as Ghana, which, as the Gold Coast, was an important source of the yellow metal.

— The preceding is an excerpt from “Gold News,” published by the Washington, D.C.-based Gold Institute.