Introduction
Bioleaching and bio-oxidation processes are based on the capability of certain micro-organisms to catalyse oxidation of ferrous iron (Fe2+) and/or reduced species of sulphur followed by a consequent dissolution of the sulphide minerals [1]. Such process has become attractive in the production of a wide range of metals such as nickel , copper , zinc and gold , given its lower operational and capital expenditures, when compared with traditional chemical methods [2].
In the gold industry, bio-oxidation is used as a pre-treatment of refractory gold ores, prior to the recovery of the noble metals (usually Au and Ag) by cyanidation . This occurs because the oxidation of the mineral sulphides exposes the gold present in these sulphides to the action of cyanide ions. Bio-oxidation alongside pressure oxidation and roasting are the main routes in the processing of refractory gold sulphide ores [3].
Various species of micro-organisms having such ability have already been identified. Amongst them, the mesophillic species Acidiothiobacillus ferrooxidans, which is the most commonly used species when it comes to the study of bioleaching of metallic sulphides [4]. However, aiming at improving dissolution kinetics as compared to what is achieved in systems containing mesophillic bacteria, the use of thermophiles emerged as a promising alternative [5]. In this direction, the moderate thermophilic species, Sulfobacillus thermosulfidooxidans, has been investigated when bacterial oxidation processes are concerned [6–8]
The presence of toxic elements such as Cu, Zn, As, Sb and Ag in ores and concentrates, often impairs bioleaching [9]. The process of adapting microorganisms to the solid, carried out through contact with increasing masses of the ore , leads to an improvement in the metal extraction as well as dissolution kinetics [10].
The current investigation addresses the bio-oxidation of a gold -bearing sulphide concentrate. This concentrate contains arsenic and antimony , which can affect the microbial growth and impair the oxidation of the sulphide minerals . In this sense, the performance of the species Acidiothiobacillus ferrooxidans and Sulfobacillus thermosulfidooxidans was compared, considering the antimony behaviour. After selecting the most suitable microorganism the oxidation of the concentrate in a 10% solids test was investigated followed by the cyanidation of the oxidised product.
Materials and Methods
Characterization of the Initial Solid Sample
In order to carry out bioleaching experiments, a refractory gold concentrate whose particle size was 100% below 37 µm (<400# Tyler) was selected. The quantification of the elements present in the initial sample, was carried out by acid digestion with hydrochloric, nitric and hydrofluoric acids followed by analysis in an ICP-OES (Varian 725-ES).
The characterisation of the concentrate was done by X-Ray diffraction (XRD ) (PanAnalytical, model Empirean). The XRD spectra were obtained at a 2θ scan range from 5º to 70º and step size of 0.02º/2θ per minute; at 40 kV and 20 mA. The analysis of the data was done through comparison to database standards (ICDD—International Committee of Diffraction Data) with the use of “Jade”, version 9.0 software.
Microorganisms
Pure strains of the species A. ferrooxidans and S. thermosulfidooxidans were kept in a shaker (New Brunswick Scientific) at 30 and 50 °C, respectively, under a stirring rate of 150 min−1 and a 5 cm orbit. The Norris medium containing 0.4 g/L of (NH4)2SO4; 0.4 g/L of KHPO4; and 0.8 g/L of MgSO4·7H2O was used in the bacterial growth. Yeast extract (0.1 g/L) was added to the growth medium in the tests with S. thermosulfidooxidans. These strains were adapted to the solid by mixing them with increasing contents of the concentrate. In order to inoculate the experiments, these cultures were filtered in cellulose membranes (Millipore-0.22 µm), which were transferred to the flasks.
Bio-Oxidation Experiments
Batch bio-oxidation experiments were carried out in shaking flasks. The effects of pulp density (2.0%, 4.0%, 6.0% w/v) and adaptation of the bacteria to the solid were assessed. The pH was kept at 1.80 in the experiments.
The experiments with A. ferrooxidans and S. thermosulfidooxidans were carried out in 250 mL Erlenmeyers flasks having a total volume of 150 mL of pulp made up of 10% (v/v) of Norris medium, 5 g/L Fe2+, the inoculum and also the sulphide concentrate under study. In experiments with the species S. thermosulfidooxidans, 0.1 g/L of yeast extract was added to the Norris medium. The flasks were maintained in a shaker (New Brunswick Scientific) at 30 °C (At. ferrooxidans ) and 50 °C (S. thermosulfidooxidans), under stirring at 150 min−1 (5.0 cm orbit). During the experiments, the evaporation losses were compensated with the addition of distilled water. The flasks were weighed at the beginning of the experiment and had their masses corrected during the assessments. For each one of bacterial strains investigated, a control test with a pulp density 4% (w/v) was also assessed.
The effect of 10% solids (w/v) in the bio-oxidation of the concentrate was also tested in an experiment with At. ferrooxidans , in accordance to the same experimental procedure described above. Apart from that, the rolling bottle procedure aiming at producing a higher amount of the bio-oxidized product to be used in the cyanidation tests was followed. Such procedure was carried out in a temperature -controlled room at 30 °C. A mechanical system was responsible for the agitation of the bottles.
In all tests, the pH of the pulp was controlled daily by using 1 mol/L H2SO4 or 6 mol/L NaOH solutions. That was accomplished using a pHmeter (Digimed), equipped with a glass membrane electrode. Electrode calibration was carried out in pH 4.0 and 7.0 buffers. The reduction potential of the solution was also monitored, through a Ehmeter (Digimed). All Eh measurements refer to the Ag/AgCl (297 mV) couple. In addition to that, samples of the solution were collected regularly and the metal concentrations were determined in an ICP-OES.
The cyanidation tests were carried out with the bio-oxidized material produced in the 6L rolling bottle tests. The cyanidation was carried out in 1000 mL flasks in an orbital shaker (Ika KS, 260 Basic) at 175 min−1 for 72 h with excess of KCN to ensure gold dissolution. The losses resulting from evaporation were compensated for with distilled water. The work volume was 300 mL and the pH of the pulp (33% (w/v) solids) was maintained at 10.5–11.0, by addition of drops of NaOH, 6 mol/L.
Results and Discussion
Characterization of the Initial Sample
DRX analysis of the original concentrate (a) and the product of bio-oxidation at 10% solids (b)
Bio-Oxidation Experiments
Eh profile at different pulp densities in experiments with adapted (a) and non-adapted (b) bacteria. Experimental conditions: pH 1.8, 30 °C (Af) or 50 °C (St), 5 g/L Fe2+. Captions: Af: At. ferrooxidans ; St: S. thermosulfidooxidans
In the case of adapted microorganisms, (Fig. 2a), it can be observed that the Eh values recorded in biotic conditions for both species were high above (>500 mV) the values observed in the abiotic experiments. In the face of this fact, the catalytic effect of the microorganisms on the oxidation of the sulphides in relation to chemical leaching by the dissolved oxygen is confirmed. Moreover, a faster increase of the Eh fostered by A. ferrooxidans as compared to what was observed for S. thermosulfidooxidans was verified in all conditions tested. This fact results in a more efficient Fe+2 oxidation by At. ferrooxidans and consequently in the dissolution of the sulphide minerals . In accordance with Figs. 2a and b, a slower adaptation of S. thermosulfidooxidans to 6% (w/v), of the solids, resulted in Eh values close to the values observed in the abiotic tests, which suggests growth inhibition under such condition.
Pyrite bio-oxidation is difficult to be assessed through concentration of ions in solution, for the solubilised species (Fe and S) precipitate in the system when jarosite is formed i.e. the oxidation products are not soluble. The same occurs with antimony , which is also not soluble in acid media . Being so, the aqueous concentration of arsenic was the only option in the assessment of the level of bio oxidation of the sulphides present in the concentrate under study.
Arsenic extraction (%) at different pulp densities in experiments with adapted (a) and non-adapted (b) bacteria. Experimental conditions: pH 1.8, 30 °C (Af) or 50 °C (St), 5 g/L Fe2+. Captions: Af: At. ferrooxidans ; St: S. thermosulfidooxidans
Conversely, the profiles of arsenic extractions for the species At. ferrooxidans under “without adaptation” and “with adaptation” conditions were similar, reaching maximum extractions near 70%, as stated. Such results as well as the Eh profile did not suggest deleterious effects of the solid on the growth of At. ferrooxidans up to 6% (w/v) solids. It is worth mentioning that the reduction in the arsenic extraction values observed at the end of the experiments with the species A. ferrooxidans did not imply in a reduction in the efficiency of the process, once it is a result of the formation of arsenic -bearing jarosite [12–14].
Profiles of Eh, Fetot concentrations and arsenic extractions in the bio-oxidation tests carried out with 10% solids (w/v) and At. ferrooxidans . Experimental conditions: pH 1.8, 30 °C, 2.5 g/L Fe2+
According to Fig. 4, the Eh profile presented a sharp rise up to the third day of tests in the presence of microorganisms; suggesting that the bio-oxidation of the ferrous iron started in no time, as observed in other tests (Fig. 2). From the third day on, the Eh levelled off throughout the experiment, reaching values around 500 mV. However, the abiotic condition presented Eh values practically constant all through the experiment—below 400 mV, indicating low or no microbial activity. By comparing Figs. 2 and 4, one can infer that the pulp density had little effect on the values of the redox potential; thus corroborating what was observed by Zhang et al. [15] in their study of pyrite bioleaching using At. ferrooxidans .
The initial bacterial concentration was in the order of 107cells/mL. Up to the second day of test, the cell concentrations presented similar results for all experimental conditions investigated, however at the end of the experiment carried out with 10% (w/v) of solids, values around 109 cells/mL were attained. Conversely, the pulp density affected the yield of the sulphide oxidized significantly, because the arsenic extraction halved in relation to data presented in Fig. 3, i.e.−40% for 10% pulp density. It is well worth mentioning that a great amount of solid was present in the system and a longer test time is believed to have led to better results. Similar results were obtained by Deng et al. [16] in studies into refractory ores containing mainly pyrite and arsenopyrite as gold bearers. For a pulp density of up to 10% (w/v), the authors reported that the arsenic extractions decreased with a rise in the pulp density i.e. 90%, 78% and 68% for pulp densities of 5%, 7% and 10%, respectively. On studying the bio-oxidation of a gold refractory concentrate, with a mixed mesophillic bacterial culture (At. ferrooxidans , At. thiooxidans and L. ferrooxidans), Ciftci and Akcil [17] also reported that the iron and arsenic extractions were reduced as the pulp density rose.
Figure 4 also shows that the iron concentration in solution did not stabilise during the test; assigning values in the order of 5.4 g/L. Under abiotic conditions, the final iron concentrations were around 5.0 g/L, which means that the values for abiotic and non-abiotic conditions did not differ much, suggesting a negligible formation of jarosite . A significant content of this solid would result in a smaller total iron concentration in the biotic experiments, due to a higher Eh (Fig. 3) and would be detected by XRD , but Fig. 1b does not suggest that. In addition, a smaller total iron concentration in the biotic experiments Moreover, these results justify the high bacterial population observed, since the ferrous iron is a substrate for the At. ferrooxidans growth. On the other hand, the final antinomy extractions were below 0.5% in the biotic and abiotic tests with 10% of solids, which were quite inexpressive when compared to the extraction of arsenic . Higher extractions, in the control tests and biotic extractions alike were obtained with a low pulp density (~2% solids), for under this condition there was a higher dissolution of the concentrate; however the extraction was not greater than 6%.
SEM-EDS analysis of the bio-oxidation product obtained in the experiment regarding the influence of 2% solids on the bio-oxidation of the concentrate. Experiments with S. thermosulfidooxidans (a), highlighting arsenopyrite (1 and 2), pyrite (3 and 4) and gudmundite (5 and 6) grains. Experiments with A. ferrooxidans (b) highlighting stibnite (12), silicate (13) and pyrite (14) grains
In the product of bio-oxidation with At. Ferrooxidans , pyrite and arsenopyrite grains were either not conspicuous or were reduced in size in relation to the initial sample (Fig. 5b—14), as also observed with S. thermosulfidooxidans. This fact indicates higher extraction values for arsenic in inoculated conditions (Fig. 3). However, the antimony sulphides were more frequent, which explains the low antimony extractions observed in the current study; as already discussed. By means of XRD analysis, Torma and Gabra [18], in their studies into stibnite oxidation (Sb2S3) by At. ferrooxidans , attributed the low solubility of this element to the formation of insoluble antimony oxides ((SbO)2SO4 e (SbO2)2SO4) In the current investigation, the formation of such phases was not detected, despite the fact that the extraction results suggest low antimony solubility in acid media . Besides being abundant (Fig. 5), the antimony sulphide grain presented morphology similar to that found in the original sample i.e. massive structures unlike the pyrite grains, which revealed to have been attacked (as indicated by the cracks). This fact suggests non-reactivity of the antimony sulphides in bio-oxidation systems, i.e. according to SEM-EDS analysis, the formation of new phases bearing antimony in the bio-oxidation experiments were not observed.
In order to conclude the study, the concentrate was bio oxidised in a 10-litre-bottle and then submitted to cyanidation . In the bio-oxidation experiments, the conditions observed were: pH equal to 1.8; pulp density equal to 10% and initial Fe2+ concentration equal to 5.0 g/L. In accordance to what was indicated in the shaking flask experiments (Fig. 4), a longer reaction time in the rolling bottle experiment was necessary as to improve the results of the bio-oxidation assessed by means of arsenic extraction . In the face of this fact, experiments with a 40 day reaction time were carried out.
The original concentrate and the finishing product of bio-oxidation in the rolling bottle test were submitted to cyanidation . Cyanidation of the sample without pre-oxidation treatment attained a gold extraction of 21%, confirming the refractory nature of the concentrate under investigation. The results of the cyanidation of the solid products obtained after 40 days of bio-oxidation showed a significant rise in gold extraction , attaining 85%. Similar results were observed by Ubaldini et al. [19] in an integrated process (using a mixed At. ferrooxidans and At. thiooxidans strains) in the treatment of pyrrhotite followed by cyanide leaching in the recovery of gold . In the aforementioned investigation, the gold recovery from the solid product was 84%, which is consistent with the findings presented herein.
Conclusions
Considering the Eh, and also the arsenic extractions, it was possible to infer that strains of At. ferrooxidans present higher capability of adaptation to the concentrate sample. On the other hand, the species S. thermosulfidooxidans proved more sensitive to the solid and the time required for its adaptation is longer. Therefore, the use of mesophillic bacteria in the bio-oxidation of the sample investigated is recommended. In such case, it was possible to confirm the exposure of the gold particles to cyanide, as the gold extraction was high—(85%) after bio-oxidizing the concentrate at 30 °C. Under the conditions applied, antimony sulphides were not oxidised, implying that the recovery of the element requires further hydrometallurgical techniques.
Acknowledgements
The authors should like to thank the institutions FINEP, FAPEMIG, CNPq and CAPES, also UFOP, for support throughout the development of the current research.