Significant advances have taken place in the treatment of refractory gold ores and concentrates during the past decade. In sulfide ores, pyrite is the most common host mineral for gold. Studies by Gasparrini (1983) showed that gold occurs in pyrite as native gold or electrum in all sizes and forms. Arsenopyrite is probably the second most common host for gold and forms similar associations with gold as pyrite. Gold may be in the form of coarse grains between grain boundaries, distributed in fine fractures, or may be totally encapsulated in the sulfide grains.
Arsenopyrite generally occurs with pyrite and gold may be associated with pyrite, arsenopyrite or both minerals in variable amounts. In some sulfide ores, fine grinding the concentrate may liberate the gold particles and make them amenable to cyanidation (Henley, 1975). In others, however, the gold occurs as either sub-microscopic inclusions or is chemically bound within pyrite or arsenopyrite (Wash, 1988). These types of ores are generally not amenable to cyanidation and require destruction of sulfide matrix by pretreatment methods (Jha, 1987). In addition to sulfides, carbonaceous material present in certain gold ores may also contribute to refractoriness.
Gold associations with orpiment and realgar are rather rare. However, gold has been observed to be associated with these minerals that form inclusions in, or coatings around pyrite (Mason and Nanna, 1989).
Gold-chalcopyrite associations also occur, but are not very common. In pyrite-chalcopyrite ores, gold is generally totally enclosed in pyrite or distributed at the pyrite-chalcopyrite grain boundaries (Gasparrini, 1983). In processing these types of ores, chalcopyrite concentrate and gold-bearing pyrite/arsenopyrite concentrate is produced (Tumilty et al., 1987).
Traditionally, gold-bearing pyrite/arsenopyrite flotation concentrates are roasted and the calcine is cyanided for gold recovery (Avraamides, 1989, Halverson, 1990). However, most arsenopyrite/ pyrite concentrates may remain partially refractory to cyanidation even after roasting (Robinson, 1988).
Bacterial oxidation of arsenopyrite/pyrite concentrates have also been investigated in pilot-plant-scale studies (Van Aswegen et al., 1991, Spencer and Budden, 1990) and the first commercial application of the bio-oxidation technology was reported at the Sao Bento plant in Brazil, as a pre-oxidation treatment to increase the capacity of the autoclave leach circuit (Suttill, 1990).
Pressure oxidation of the refractory sulfide/carbonaceous ores has gained world-wide acceptance. Compared with roasting, pressure oxidation yields higher gold extractions from ores or concentrates and enables better handling of environmentally sensitive impurities. Normally, gold is extracted from neutralized autoclave leach residue by cyanidation. Silver extraction, however, remains low due to silver jarosite formation. Atmospheric digestion of the oxidized solids slurry with lime at 90 to 95 °C prior to cyanidation enhances the silver recovery (Berezowsky and Weir, 1989).
The acidic thiourea process has been considered as an alternative to cyanidation for gold and silver leaching due to its faster kinetics. Furthermore, thiourea is considered non-toxic and environmentally safe. Reactions for gold and silver dissolution can be represented as follows:
2Au + ((NH2)2CSSC(NH2)2)²+ + 2NH2CSNH2 = 2Au(NH2CSNH2)2+……………………………
2Ag + ((NH2)2CSSC(NH2)2)²+ + 4NH2CSNH2 = 2Ag(NH2CSNH2)3+………………………………
As in the case of cyanide, thiourea leaching requires the presence of an oxidant such as ferric ion or hydrogen peroxide;
NH2CSNH2 + Fe³+ = 0.5 ((NH2)2CSSC(NH2)2)²+ + Fe²+…………………………………..
2NH2CSNH2 + H2O2 + 2H+ = ((NH2)2CSSC(NH2)2)²+ + 2H2O………………………….
Thus, ferric ion and sulfuric acid generated in bio-leaching or pressure oxidation systems provide very suitable conditions for thiourea leaching (Yen and Wyslouzil, 1986).
Oxidation of Pyrite and Arsenopyrite
Oxidation of pyrite has been the subject of numerous studies which have been summarized in two comprehensive reviews (Hiskey and Schlitt, 1982, Lowson, 1982). Basic studies involving pressure oxidation of pyrite, particularly under the conditions applicable to autoclave leaching of gold-sulfide ores have also been reported (Lowson, 1982). Although industrial conditions and leach parameters investigated in basic studies may differ considerably, these studies indicate that autoclave oxidation of pyrite is chemically controlled, showing a one-half order dependance at high temperatures and a first order dependance at low temperatures on the oxygen partial pressure. Ferrous sulphate and elemental sulphur are the primary reaction products at low acidities between 100-150°C. These are further oxidized to ferric sulphate and sulfuric acid, respectively, at higher temperatures and oxygen overpressures. Sulfuric acid production ceases when the acid concentration reaches 0.17M (pH=1.0) , and any further increase in acid is consumed by the ferrous ion produced by the oxidation of pyrite to ferrous ion and elemental sulfur:
FeS2 + 2O2 → FeSO4 + S°…………………………………………………(5)
2S° + 3O2 + 2H2O → 2H2SO4…………………………………………..(6)
4FeSO4 + O2 + 2H2SO4 → 2Fe2(SO4)3 + 2H2O…………………(7)
Consequently, the system becomes self-buffering at about a pH=1.0.
In constrast to pyrite, studies on the aqueous pressure oxidation of arsenopyrite are very limited. In a comprehensive study published recently, the oxidation of arsenopyrite was found to be faster than that of pyrite and extremely exothermic. A first order reaction with respect to oxygen partial pressure was also determined (Papangelakis and Demopoulos, 1990).
The following reactions were proposed to describe the reaction chemistry of arsenopyrite in a sulfuric acid leaching medium:
4FeAsS + 13O2 + 6H2O → 4H3AsO4 + 4FeSO4…………………………………………………
4FeAsS + 7O2 + 4H2SO4 → 4H3AsO4 + 4FeSO4 + 4S°………………………………………
4FeSO4 + O2 + 2H2SO4 → 2Fe2(SO4)3 + 2H2O……………………………………………….
Fe2(SO4)3 + 2H3AsO4 + 4H2O → 2FeAsO4·2H2O + 2H2SO4……………………………
Reactions  and  occur simultaneously as two parallel competing reactions with reaction  being favored at higher temperatures (>180 °C) and lower acidities. At high Fe/As ratios, ferric arsenic species precipitate  as ferric arsenate compounds (scorodite), which is considered relatively inert (Robins, 1987).
During the pressure oxidation of sulfide ores, hydrolysis products of ferric ion such as geothite (FeOOH), hematite (Fe2O3), or jarosite family compounds (MFe3(SO4)3(OH)6 where M is H3O+, Na+, NH4+, ½Pb²+, or Ag+) may precipitate depending on the temperature and acidity (Dutrizac, 1979). As the Eh-pH diagrams at high temperatures also indicate, the stability zones for both ferrous and ferric iron decrease with temperature giving way to the stability of Fe2O3 (Ferreira, 1975).
Five kilograms of gold and silver bearing drill-core concentrate was obtained from an operating mine in north-central Nevada. The chemical analysis of the concentrate is shown in Table 1. Deionized water, reagent grade sulfuric acid, and pure oxygen (99.5%) were used in all pressure oxidation experiments. Roasting experiments were carried out in a fireclay dish placed in a muffle furnace.
Preliminary cyanide and thiourea leach experiments were conducted using 50 g samples to determine the baseline extraction values. All cyanide leach experiments were conducted for 24 hours in a 0.5% sodium cyanide leach solution at a pH of approximately 11.5 and pulp density of 15% solids. The thiourea leach experiments were conducted for 2 hours at a 10% pulp density, and at initial reagent concentrations of 0.2 M thiourea, 0.1 M ferric ion, and 0.25 M sulfuric acid. Percent gold and silver extractions were determined by fire assay analysis of the leach residues. Thiourea or cyanide leach solutions were also analysed using atomic absorption spectrophotometry techniques.
Roasting experiments were carried out on 50 g samples of the drill-core concentrate at temperatures of 450, 550, 650, and 750°C. Thiourea and cyanide leaching of these samples were conducted and the percent extraction values determined as above to show the effectiveness of the pretreatment at varying temperatures.
The pressure oxidation treatment and the optimization of gold extraction were studied using a statistical experimental design. The pressure oxidation experiments were run in a Parr-2 liter autoclave with 50 g samples, in 450 milliliters of water (10% solids). Experimental conditions for the pretreatment were dictated by a three factor, two level fractional factorial design as illustrated in Table 2. The three factors considered, oxidation time, oxygen over-pressure, and temperature, were oscillated alternately between a high and low value. The high values correspond to 120 minutes, 120 psi, and 210°C, while the low values are 60 minutes, 60 psi, and 170°C . Three experimental runs were also made at mid-point values (90 minutes, 90 psi, and 190°C) to indicate any changing trends in bias errors.
Upon completion of the oxidation pretreatment, the pulp was filtered and the filtrates were analyzed for ferric ion, ferrous ion, total iron, Eh, and pH. The residues were subsequently leached in a thiourea solution for two hours and filtered. Percent gold and silver extractions were determined by fire assay analysis of the leach residues.
Using a statistical computer software, mathematical models were established using multiple linear regression techniques to predict gold and silver recoveries within the pressure oxidation conditions studied. Then, using these models, contour plots were derived to help elucidate the optimal conditions (Figures 1-6).
The conditions of experiment 7 were repeated in experiment 12 (Table 3). However, after the pressure oxidation pretreatment was completed, a lime-conversion step was performed on the residue to decompose any argento-jarosite which may have precipitated during oxidation. The residue was then leached in a 0.5% sodium cyanide solution for 24 hours and the gold and silver extractions were determined.
Results and Discussion
Table 3 shows the results of the baseline leach tests for the concentrate. These results show that gold extraction with either thiourea or cyanide leaching were similar whereas silver extractions varied.
Table 4 gives the thiourea and cyanide leach results for the roasted drill-core concentrate samples. Maximum gold and silver recoveries for the roasted thiourea leach samples were in the range of 85-90% and 60-70% respectively, while those for the roasted cyanide leach samples were 70-75% and 60-70%. The lower extraction values obtained at the higher roasting temperatures (650-750°C) are probably due to the direct formation of magnetite and hematite during the roasting operation, thus occluding the gold and silver in the cinders (Cheng and Lowson, 1987, Robinson, 1988).
Table 5 gives the results of the thiourea leach tests on pressure oxidized concentrate samples. It is apparent that gold extractions varied in the range of 70-92% while silver extractions were much lower (10-40%).
The following mathematical models were established to predict the gold and silver extraction by thiourea leaching:
where XI, X2, X3 represent oxidation time, oxygen over-pressure, and oxidation temperature, respectively.
Using these models, the contour plots of Figures 1 to 6 were derived showing predicted gold and silver recoveries from this concentrate as a function of two parameters studied within the experimental range.
Figure 1 shows that percent gold extraction increases with increasing temperature between 170 to 210°C, and also increases with an increase in oxygen pressure up to a maximum, in the range of 90 to 100 psi. However, above this range, extraction decreases probably due to the apparent rapid precipitation of iron in the form of geothite and jarosites.
Figure 2 shows the gold extraction increases as temperature is increased at constant oxygen pressures. It is also apparent that at a given pressure, an increase in leaching time results in an increase in extraction.
Figure 3 shows the simultaneous effects of oxidation time and oxygen over-pressure on percent gold extraction for a given temperature. At 170°C (Figure 3a), an increase in time corresponds to an increase in extraction. Also, an increase in oxygen over pressure up to approximately 90 to 100 psi produces an increase in gold extraction. However, as Figure 3c indicates, an increase in time in the range of approximately 70 to 90 minutes results in no corresponding increase in extraction. As Figures 1 through 3 show, optimal extraction values correspond to approximately 90 psi oxygen over-pressure, 210°C, and 60 minutes oxidation time.
Figures 4 through 6 show the contour maps for the percentage silver extraction with thiourea from the concentrate following pressure oxidation. It is evident that increasing temperature or oxygen pressure results in lower silver recoveries. This is probably due to the precipitation of iron and silver in the form of argento-jarosite which hinders silver recovery in the subsequent leaching operation.
Table 5 also shows the results of experiment 12, where a lime-boil treatment was applied to the autoclave leach residue. A 98% gold extraction and 95.7% silver extraction was realized as compared to 92.5% and 7.1% gold and silver extraction, respectively, under the best pressure leach conditions without a lime boil. X-ray diffraction analysis of the pressure leach residues indicated the presence of goethite and plumbo-jarosite, whereas presence of silver-jarosite could not be confirmed due to overlapping diffraction peaks.
Baseline leach experiments have established the refractory nature of the concentrate. Roasting of this concentrate at 450-550°C for one hour prior to leaching has increased gold and silver extraction values to approximately 89% and 70% respectively by thiourea leaching, and 76% and 74% by cyanidation. Acidic pressure oxidation conditions for optimal gold extraction by thiourea leaching correspond to an oxidation temperature of 210°C at an oxygen over-pressure of approximately 90 psi for 60 minutes. However, because the silver extraction under these conditions by either cyanide or thiourea leaching is quite low, a jarosite conversion step must be performed on the autoclave residues prior to leaching, for complete silver recovery. After the lime-boil, the conditions which would have made thiourea leaching marginally attractive (due to its faster kinetics) no longer exist.