A study of the trace elements associated with gold from the Barberton District

Sixty eight samples of r»tive gold were analyzed for their silver, copper, iron and trace element sontenti. Forty of the samples were collected from different localities in the Barberton district; twenty two frodi various levels along the Zwartkopje ore shoot, Sheba Mine; and six from localities on the Witwatersrand and in Southern Rhodesia. The gold was extracted by either crushing the samples and then concentrating the gold using a superpanner, or by dissolving the rock in hydrofluoric acid. The particles of gold were thoroughly cleaned before analysis with hydrochloric, hydrofluoric and nitric acids. Great care was taken throughout the pre-analytical treatment to prevent contamination by extraneous material or by removal of trace elements from the gold. Three analytical methods were used to determine the composition of the gold. Atomic absorption spectroscopy was used to determine the copper and iron contents, the fire assay the silver content and emission spectroscopy the trace element composition. The concentrations of the trace elements are given semi-quantitatively as log intensity ratios. All other results are quantitative. Generally the composition of the gold samples is relatively simple. Silver, copper, iron and silicon were present in all the samples; magnesium was present in fifty-eight of them; aluminium in fifty-three; nickel in twenty-seven; lead in twenty-three; antinony in twelve; tin in eleven; bismuth in ten; cobalt in seven; mercury in six; zinc in four; titanium and molybdenum in thre*; beryllium and platinum in two; . .calcium/.. . calcium, manganese, palladium and vanadium in one. Of these trace elements, it is considered that beryllium, bismuth, C3balt, manganese, mercury> molybdenum, palladium, plot^r.jiTi, silver, t in and vanadium occur as alloy constituents in solid solution with .he gold. Antimony, copoer, iron, lead, nickel, titanium and zinc are also *hought to be present in the gold lattice, though they may also occur as mineral inclusions. Aluminium, calcium, magnesium and silicon are probably present as mineral inclusions. The results of the samples taken from the Zwartkopje shoot seem to indicate two different ty,_js of gold, occurring between the 20 and 26 and the 14 and ’ 8 levels, respectively. Gold sample? taken below the 2 0 level increased in fineness from 92^ to 950 with increasing depth, while those samples from above the 20 level, had fineness values increasing from 915 to 935 between the 14 and 18 levels. The copper content of the samples decreased from 700 p.p.m. to 400 p.p.m . between the i4 and 18 levels and increased from 160 to 5 00 p.p.m. between the 2 0 v M 26 levels. The iron content decreased from 500 to 300 p.p.m. between the )4 *nd 18 levels and from 1300 to 900 p.p.m . between the 20 and 26 iov«. s. The increase in copper content with increasing depth in ihe gold from the lower levels, correlat?s with increasing chalcopyr <te contwnx. witll incrersiig d»pth in th?se levojs. Trace elements occurring in the gold from the upper levels include aluminium, magnesium, mercury, n ick e l, while in the lower levels, slrninium, antimony, bismuth, magnesium and nickel ars presunt. . . . . T he / . • # The main fac/lors controlling the composition of the samples appear to be the temperature and pressure conditions prevailing at the time of deposition, the concentration of the elements in the ore fluids and the chemical aiKl mineralogical environment. The fineness values for the regional gold samples in the Barberton District show some regular distribution with the highest values around the Consort and Agnes Mines. The copper and iron contents show no systematic variation, nor do the trace elements. However, there does appear to be d dafinite association between trace elements from the same, or adjacent, groups in the Periodic Table. ?rom the few samples analyzed, it appears as if gold from :he Witwa'.ersrand mines, differs funda
Dissertation submitted for the degree of Master of Science in the Faculty of Soienoe, University of the Witwatersrand, Johannesburg. August, 1963.