Tank bioleaching of a copper concentrate using the moderately thermophilic microorganisms Sulfobacillus thermosulfidoxidans KMM3 and Sulfobacillus acidophilus KMM26

2.1. Microorganisms

Following the method introduced by Ñancucheo and colleagues (Ñancucheo et al., 2016Ñancucheo, I., Rowe, O.F., Hedrich, S., Johnson, D.B. (2016). Solid and liquid media for isolating and cultivating acidophilic and acid-tolerant sulfate-reducing bacteria. FEMS Microbiol. Lett. 363 (10), fnw083. https://doi.org/10.1093/femsle/fnw083.), two strains of Sulfobacillus thermosulfidooxidans and Sulfobacillus acidophilus were isolated and identified from the samples received from Miduk Copper Mine in Kerman- Iran. A basal medium was prepared and inoculated by the samples containing the microbial community and while kept on orbital shaker at 150 rpm, inoculated for 7 days at 30 °C. Using a common laboratory practice (Johnson and Hallberg, 2007Johnson, D.B., Hallberg, K.B. (2007). Techniques for detecting and identifying acidophilic mineral-oxidizing microorganisms. Rawlings, D., Johnson D. (Eds) Biomining, Springer, Berlin, Heidelberg: Springer Berlin Heidelberg. pp. 237-261. https://doi.org/10.1007/978-3-540-34911-2_12.), pure cultures of the detected strains were then isolated from the solid medium (Johnson, 1995Johnson, D.B. (1995). Selective solid media for isolating and enumerating acidophilic bacteria. J. Microbiol. Methods 23 (2), 205-218. https://doi.org/10.1016/0167-7012(95)00015-d.).

The number of bacteria in the solution was counted daily under an optical microscope (Zeiss, Axioskop 40) using a blood cell counting chambers (Eftekhari et al., 2020aEftekhari, N., Kargar, M., Rokhbakhsh Zamin, F., Rastakhiz, N., Manafi, Z. (2020a). Bioremoval of iron ions from copper raffinate solution using biosynthetic jarosite seed promoted by Acidithiobacillus ferrooxidans. Rev. Metal. 56 (4), e182. https://doi.org/10.3989/revmetalm.182.; Eftekhari et al., 2020bEftekhari, N., Kargar, M., Rokhbakhsh Zamin, F., Rastakhiz, N., Manafi, Z. (2020b). The catalytic activity of biological seeds and Acidithiobacillus ferrooxidans on the process of ammonium jarosite. Journal of Microbial World 12 (4), 355-363. ). The initial inoculation containing 7×10⁷ cells/mL was added to the reactor in 10% (v/v). The bacterial counts reached from about 6×10⁶ to about 8×10⁸ cells/mL, from early days to the end of the experiment (day 7).

To extract genomic DNA, pure cultures were cultivated in broth medium and then, 5 ml of the bacterial cultures were centrifuged (15 min at 13000 rpm) and washed twice with Tris buffer (pH 8). According to modified protocol (Webster et al., 2003Webster, G., Newberry, C.J., Fry, J.C., Weightman, A.J. (2003). Assessment of bacterial community structure in the deep sub-seafloor biosphere by 16S rDNA-based techniques: a cautionary tale. J. Microbiol. Methods 55 (1), 155-164. https://doi.org/10.1016/s0167-7012(03)00140-4.), employed for DNA extraction. The fragment of 16SrDNA was amplified by PCR (polymerase chain reaction) using designed primers and then PCR was carried out (Webster et al., 2003Webster, G., Newberry, C.J., Fry, J.C., Weightman, A.J. (2003). Assessment of bacterial community structure in the deep sub-seafloor biosphere by 16S rDNA-based techniques: a cautionary tale. J. Microbiol. Methods 55 (1), 155-164. https://doi.org/10.1016/s0167-7012(03)00140-4.).

PCR products were sequenced after purification from the gel by Bioner South Korea. The desired sequences received the registration tracking code by blasting and registering on the NCBI site. Sequences derived from 16S rRNA gene were registered in the name of the researchers, in the World Gene Bank. The phylogenetic trees were constructed using neighbor-joining algorithm implemented in Molecular Evolutionary Genetics Analysis software (MEGA 6.0).

2.2. Copper concentrate

A copper concentrate from Miduk Copper mine was used in all experiments. The concentrate is produced by the Iranian Babak Copper Company (IBCCO) using a flotation method. Table 1 shows data of the chemical and mineralogical analyses of the concentrate. All elemental analyses were performed by atomic adsorption spectroscopy (by AAS, A gilent 200, Asturalia) and mineralogy was determined with X-ray diffraction (Philips X’pert-MPD system, CuKa, K ~ 0.154 nm, 40 kV, 40 mA).

2.3. Bioleaching tests

Three effective parameters including pulp density, pH and particles size (d₈₀) were comprehensively studied (Table 2). A 5-litre glass reactor (Sina-Shisheh, Iran) (Fig. 1) with the stirrer speed of 500 rpm was used for all bioleaching experiments (Anderson et al., 2002Anderson, C., Giralico, M., Post, T., Robinson, T., Tinkler, O. (2002). Selection and Sizing of Biooxidation Equipment and Circuits. In: Mineral processing plant design, practice and control. Mular, A.L., Halbe, D.N., Barret, D.J. (eds.), Society of Mining Engineers, Littleton.).

The contents of the 9K culture medium with some additive and modifications summarized in Table 3 (Anderson et al., 2002Anderson, C., Giralico, M., Post, T., Robinson, T., Tinkler, O. (2002). Selection and Sizing of Biooxidation Equipment and Circuits. In: Mineral processing plant design, practice and control. Mular, A.L., Halbe, D.N., Barret, D.J. (eds.), Society of Mining Engineers, Littleton.).

To start the experiments, 4500 ml of 9K medium was added to the reactor and inoculated by 450 ml bacterial suspension containing 10⁷ cells·mL^-1. The reactor was set at 50 °C (by coupling it with a water bath) which was aerated by 360 mL·min^-1. The temperature was controlled by a digital sensor to avoid sharp fluctuations. Redox potential, pH, and copper concentration were daily analyzed. After calculating the recoveries, a chemical control test without any bacteria was also carried out to compare its results with the test with highest Cu recovery.

2.4. Milling and grinding

To study the effect of grinding methods on the bioleaching efficiency, four different mills were used to decrease the particle size to smaller than 10 µm. Table 4 shows the mills properties. Scanning electron microscopy (SEM) studies were also carried out to investigate the particle shapes after grinding by each milling condition.

2.5. Aggregation and sedimentation

To study the agglomeration of the particles, a series of sedimentation tests have been operated at different pH values from 0 to 12 (totally 13 tests). For this purpose, the copper concentrate was mixed with the reagent at different pH values. Sulfuric acid and lime were used to set the pH. The pulp was transferred to 50 ml scaled cylinders and remained for 120 minutes. To complete the sedimentation studies, the point of zero charge was measured for the Cu concentrate sample (Particle Metrix, Germany).

3.1. Identification of bacteria

The two isolated bacterial strains were identified and named as Sulfobacillus thermosulfidooxidans KMM3 (MW175495) and Sulfobacillus acidophilus KMM26 (MW175507). Their phylogeny is shown in Fig. 2. After identification, the bacteria were stained and examined microscopically (Fig. 3). The two bacteria showed a similar morphology.

3.2. Optimization of bioleaching process

In total eight bioleaching tests along with one abiotic control test were conducted; the results are presented in Table 5. Generally, the test with a pulp density 5 %, pH 1 and a particle size (d₈₀) of 45 µm showed the highest copper recovery.

By decreasing the particle size, the specific surface area increases, thereby improving mass transfer and consequently enhanced bioleaching efficiency. It was expected that a greater surface area exposed to bacterial attack at smaller particle sizes (e.g., -10 µm) would result in metal dissolution increases. However, it was observed that decreasing particle size under 30 µm did not improve bioleaching rates, and resulted in lower recoveries. It was observed that at the particle sizes less than 30 µm, the solid particles were aggregated and thus, the bacteria could be possibly stocked in between the aggregates. Therefore, although viable cell counts were not significantly changed, their access to the minerals was limited and the overall recovery was reduced. This might be attributed to possible cell damage and deactivation and consequently, the reduced recoveries recorded by finer particles.

Figure 4 shows the SEM images for Cu concentrates before and after grinding. The EDS studies indicated that large particles after grinding contained higher amounts of iron and sulfur than fine particles, which contained more copper and less iron and sulfur (data not shown). This obviously shows that chalcopyrite is enriched in smaller particles compared to pyrite and other minerals. The samples also contained low concentrations of Al and Si, which indicated that the concentrate contained lower amounts of silicates as gangue.

Smaller particle size distribution (Fig. 4b) using grinding by ceramic media can also be one of the reasons for low recovery as finer particles become more agglomerated under the bioleaching conditions.

Figure 5 shows pH and oxidation-reduction potential (ORP) changes in different bioleaching tests. Among the selected parameters, acidity (pH) had lowest effect on the dissolution rate. Via the production of S⁰, chalcopyrite dissolution could be considered as an acid consuming reaction (Smart et al., 2000Smart, R.S.C., Jasieniak, M., Prince, K.E., Skinner, W.M. (2000). SIMS studies of oxidation mechanisms and polysulfide formation in reacted sulfide surfaces. Miner. Eng. 13 (8-9), 857-870. https://doi.org/10.1016/s0892-6875(00)00074-1.), in contrast to an expected complete oxidation of the sulfur in the chalcopyrite to sulfuric acid according to Eq. (1) (Tanne and Schippers, 2019Tanne, C.K., Schippers, A. (2019). Electrochemical investigation of chalcopyrite (bio)leaching residues. Hydrometallurgy 187, 8-17. https://doi.org/10.1016/j.hydromet.2019.04.022.):

This finding indicates incomplete sulfur oxidation due to inactive bacteria towards the end of the experiments and probably accumulation of elemental sulfur occurring during abiotic chalcopyrite oxidation according to Eq. (2) (Tanne and Schippers, 2019Tanne, C.K., Schippers, A. (2019). Electrochemical investigation of chalcopyrite (bio)leaching residues. Hydrometallurgy 187, 8-17. https://doi.org/10.1016/j.hydromet.2019.04.022.).

Eftekhari and Kargar reported that increasing the pH causes iron precipitation in form of iron hydroxide or jarosite which can cover the minerals surface and prevent reagent access to surface (Eftekhari and Kargar, 2018Eftekhari, N., Kargar, M. (2018). Assessment of optimal iron concentration in the precipitation of jarosite and the activity of Acidithiobacillus ferrooxidans. Modares. Journal of Biotechnology 9 (4), 525-529.; Eftekhari et al., 2020bEftekhari, N., Kargar, M., Rokhbakhsh Zamin, F., Rastakhiz, N., Manafi, Z. (2020b). The catalytic activity of biological seeds and Acidithiobacillus ferrooxidans on the process of ammonium jarosite. Journal of Microbial World 12 (4), 355-363. ; Eftekhari et al., 2020cEftekhari, N., Kargar, M., Rokhbakhsh Zamin, F., Rastakhiz, N., Manafi, Z. (2020c). A review on various aspects of jarosite and its utilization potentials. Ann. Chim. - Sci. Mat. 44 (1), 43-52. https://doi.org/10.18280/acsm.440106.).

The microscopic images (by Leica DMLP optical microscope) of the concentrate before and after grinding, and after bioleaching showed pyrite, chalcopyrite and covellite as the main minerals with various sizes down to d₈₀10 µm (Fig. 6). The microscopic observation of un-milled concentrate (a), milled concentrate (b) and milled concentrate after bioleaching process (c) are presented accordingly.

Microscopic examinations clearly show agglomeration of fine particles during the leaching process (Fig. 6). As this figure shows, very fine grinding results in the production of very fine particles (Fig. 6b) with much more surface area available for bioleaching. However, in an oxidizing leaching environment, their agglomeration prevents to achieve high copper recoveries because of not leached mineral particles (Fig. 6c).

Pulp density was the most important factor. The rate of copper extraction decreases with increasing the pulp density, probably due to shear stresses resulted from higher percentage of solids, and consequently increases the toxicity of metal ions (Wang et al., 2014Wang, Y., Zeng, W., Qiu, G., Chen, X., Zhou, H. (2014). A moderately thermophilic mixed microbial culture for bioleaching of chalcopyrite concentrate at high pulp density. Appl. Environ. Microbiol. 80 (2), 741-750. https://doi.org/10.1128/aem.02907-13.). It has shown that grinding the concentrate could increase the copper recovery to more than 10 % (Manafi, 2021Manafi, Z. (2021). Evaluation and improve the performance of moderate thermophiles in increasing of copper extraction from chalcopyrite concentrate. PhD thesis.). At the same pulp density, when the particle size gets smaller, the easier and more effective access to the surface of the mineral and metal content increases the reaction rate. On the other hand, a significant decrease in particle size increased the apparent viscosity of the pulp, and reduced the oxygen transmission rate which could be harmful for the bacterial activity (Derjaguin and Landau, 1993Derjaguin, B., Landau, L. (1993). Theory of the stability of strongly charged lyophobic sols and of the adhesion of strongly charged particles in solutions of electrolytes. Prog. Surf. Sci. 43 (1-4), 30-59. https://doi.org/10.1016/0079-6816(93)90013-L.; Anderson, et al., 2002Anderson, C., Giralico, M., Post, T., Robinson, T., Tinkler, O. (2002). Selection and Sizing of Biooxidation Equipment and Circuits. In: Mineral processing plant design, practice and control. Mular, A.L., Halbe, D.N., Barret, D.J. (eds.), Society of Mining Engineers, Littleton.).

3.3. Effect of grinding on the bioleaching efficiency

In tank bioleaching tests, non-agglomerated particles are critical to achieve maximum contact surface. When the particles are close enough, the van der Waals attraction can interact between particles causing agglomeration of particles. Further, electrostatic equilibrium (DLVO theory) and steric and electrostatic force cause agglomeration. Agglomerated particles could significantly decrease the metal extraction efficiency. It is evident that type of fragmentation and milling in this series of tests was quite effective. Sphere particles have the lowest surface to volume ratio. If the grinding is such that instead becoming orb, particles can be created in angular shape, the surface energy of the angular particles increases. Therefore, choosing the right method for grinding has great importance on the extraction recovery (Marshall, 1949Marshall, C.E. (1949). Theory of the stability of lyophobic colloids. The interaction of sol particles having an electric double layer. Elsevier, pp. 413-414. https://doi.org/10.1002/pol.1949.120040321.; Biggs and Healy, 1994Biggs, S., Healy, T.W. (1994). Electrosteric stabilization of colloidal zirconia with low-molecular-weight polyacrylic-acid. An atomic-force microscopy study. J. Chem. Soc., Faraday Trans. 90 (22), 3415-3421. https://doi.org/10.1039/FT9949003415.).

Therefore, in order to create spherical particles with minimal surface changes due to regrinding, different grinding intermediates (grinding media) were used to regrind the concentrate. Although the surface area of the particles obtained using ceramic grinding media was more uniform and also relatively larger, the higher recoveries were obtained using steel metal media. One of the possibilities is perhaps the drastic structural changes of the concentrate particles when using steel balls, which could be considered as a kind of mechanical activation. In this case, the defected crystalline concentrate particles dissolve more rapidly in the leaching environment due to these drastic structural changes. Another purpose of using ceramic pellets was to prevent the possible formation of reactive oxygen species that are toxic to microorganisms. However, better results obtained from steel pellets show that, firstly, the probability of the formation of active radicals at this level of crushing (d₈₀ 8 µm) is very low. Secondly, the results of solid leaching studies (SEM and mineralogical photomicrographs) reinforced the conclusion that the main reason for not extracting 100 % of copper is the agglomeration of very fine particles during the leaching process. Therefore, it seems that although softening is quite useful in increasing recovery, arrangements must be made so that the mineral particles have maximum dispersion during the bioleaching period. This, of course, needs to be examined more closely to determine what happens to the particle surface during the leaching period. Accordingly, it may be possible to overcome this problem by surfactants or by precisely controlling the pH of the environment or by adjusting other operating parameters.

3.4. Effect of milling method

As discussed in section 2.4, bioleaching efficiency can be influenced by the milling method.

The particle size distributions for Cu concentrate before and after ultra-fine grinding are demonstrated in Fig. 7. The d₈₀ for sample before grinding was 47.6 µm. 50 % and 10 % of particles were finer than 24.5 and 4.5 µm, respectively. The d₈₀ for ball mill with ceramic balls after 840 min grinding reach to ~10 µm and with steel balls reach to ~8 µm.

Figure 8 shows the copper recoveries for un-milled concentrate and concentrate ground with different mills. According to Fig. 8, with 49 % the highest Cu recovery was achieved for original Cu concentrate without any grinding process.

3.5. Effect of particles size

The Cu recoveries for copper concentrate at different d₈₀s are depicted in Fig. 9. All samples were milled with a ball mill using ceramic or steel balls as grinding media. The copper recovery dropped by decreasing the particle sizes to d₈₀ 10 µm, even using high-energy planetary mill with highly mechanical activation effect. By decreasing the particle size, the agglomeration increases and this phenomenon cause decreasing Cu recovery. While using concentrate with d₈₀ 20, 25 or 30 µm had a negligible effect on Cu recovery. Therefore, it is suggested to avoid over-grinding for bioleaching process. In the other words, decrease the particles size cause increase the energy consumption without any significant improvement in bioleaching efficiency.

The agglomeration phenomenon is detectable from the solid residuals achieved from different experiments. In fact, by reducing size to d₈₀ 10 µm the agglomeration increased. Numbers of small particles were observable around each large particle. The obtained results showed that the recovery increased when steel balls used which concluded that the grinding method effect on some physiochemical properties of particles such as roundness (Fig. 9). The particle roundness was calculated by using following equation (Gilgannon et al., 2017Gilgannon, J., Fusseis, F., Menegon, L., Regenauer-Lieb, K., Buckman, J. (2017). Hierarchical creep cavity formation in an ultramylonite and implications for phase mixing. Solid Earth. 8 (6), 1193-1209. https://doi.org/10.5194/se-8-1193-2017.):

Based on above equation and SEM images (Fig. 4), particles which were ground by steel balls had roundness ~44 while particles milled by ceramic balls had roundness ~34. This means that grinding with steel balls increased the free space and this improved the bioleaching efficiency. Optical photomicrographs of the minerals indicated that there was not any chalcocite in bioleaching residuals, which shows complete decomposition of the chalcopyrite, but the chalcopyrite content also decreased from 24.97 % to 13.01 %. On the other hand, the covellite content increased from 8.79 to 30.6 %.

3.6. Aggregation of the fined particles

The behaviour of the concentrate sedimentation at different pH values was visually observed in the bakers. The lowest stability occurred at tests with pH values of 3, 5, and 11, and the highest stability was at pH 0, 7 and 12. Interestingly, a little change in pH caused a significant change of sedimentation. For instance, although pulp had a high stability at pH 7, it had a low stability at pH 6 or 8. The point of zero charge (ZPC) is one of the most important factors that control the sedimentation. Figure 10 depicts the ZPC for the concentrate in comparison with the pure samples of chalcopyrite, chalcocite and covellite. By increasing the pH from 0 to 5 the zeta potential dropped from 47 to 5 mV. In addition, increasing the pH up to 7 -12 cause a ZPC decrease to lower than -40 mV. Generally, zeta potential of the concentrate has a decreasing-increasing-decreasing trend. A comparison between sedimentation tests and zeta potential results indicated that the highest stabilities happened at the highest and lowest zeta potential ranges (pH 0, 7 and 12). It should be noted that setting pH at 0 or 12 is better because at pH 7 few changes could result in significant changes in stability, which became important factor when regrinding the concentrate under 25 µm size distributions.

Thus, in general it can be concluded that at lower pH values the particles are more stable. However, changes in pH in the environment due to acid consumption can change this brittle stability. Of course, more important than these are the very large surface changes (“passivation” or by-product layers) in the mineral particles in the concentrate during bioleaching, which can completely invalidate these estimates.