Column Bioleaching of Nickel from Sulfidic Samples with Different Nickel and Magnesium Content

Abstract Nickel is a valuable metal that is becoming more prevalent in the industry. Column bioleaching was used in this study to extract nickel from magnesium-bearing sulfide minerals. Two different sulfidic samples with different nickel and magnesium content were utilized to investigate the performance of column bioleaching. It was discovered that mesophilic cultures’ adaptation is delayed by increasing magnesium contents. Bioleaching outperformed leaching in terms of recovery by 80% compared to 50% in sample 1 and 70% compared to 40% in sample 2. Jarosite is precipitated in samples with a high magnesium content due to the high pH and oxidation level, which lowers bioleaching effectiveness. The pretreatment method using acid washing before the start of bioleaching treatment can reduce the amount of magnesium in samples, which increases the Ni recovery in both samples. SEM analysis was performed on each bioleaching residue. The result showed that high amounts of magnesium in the second sample could be a factor in the precipitation of jarosite. Finally, it can be concluded that the pretreatment method is a feasible Bio-heap operation. RESEARCH HIGHLIGHTS Column bioleaching of nickel was tested by mix mesophile culture in two different samples. Bacteria were found to be less adaptable to magnesium-rich minerals. More than 95% of nickel was extracted from both samples. Acid washing was identified as the best pretreatment method for each sample. Graphical abstract


Introduction
Nickel, an essential metal in various industries, belongs to the transition group of the periodic table along with iron and cobalt (Everhart 1971).Its applications include stainless steel (Patnaik et al. 2020), non-ferrous alloys (Rajesh et al. 2021), nickel-based catalysts (Gao et al. 2021), turbine blades (Shao et al. 2021), foundry, PCBs (Rezaee et al. 2022), and batteries (Mystrioti et al. 2018;Saneie et al. 2022).Most of the world's nickel resources are found in laterite or magmatic sulfide mineral deposits (Mudd and Jowitt 2014).It should be mentioned that while stainless steel production consumes the majority of nickel, the usage of nickel in electric vehicle batteries is growing, resulting in a considerable increase in nickel demand in the near future (Henckens and Worrell 2020).Sulfide ores have historically provided most of the nickel production, with laterite ores contributing only a minor contribution (Meshram et al. 2019).However, in terms of known nickel resources, laterites account for roughly 60% of the total, while sulfides account for 40% due to the complexity of processing nickel laterites compared to sulfides (Jessup and Mudd 2008).
Acidophilic bacteria and archaea play a role in metal leaching from sulfide minerals by increasing the oxidationreduction potential (ORP) through the oxidation of Fe 2þ and sulfur compounds, producing Fe 3þ and sulfuric acid, respectively (Abdollahi et al. 2021).In the bioleaching process, mixed cultures of iron and sulfur oxidizers are thought to be more efficient than pure cultures (Saneie et al. 2021).While mesophilic cultures can be inhibited or delayed when exposed to metal sulfide ore or concentrate, especially at high pulp densities (Haghshenas et al. 2009;Pourhossein and Mousavi 2018), On the other hand, it can be adapted to withstand some of these inhibiting conditions (Haghshenas et al. 2009).The bio-oxidation of sulfide minerals by accumulating Fe 3þ and SO 4 À2 can lead to the formation of jarosite, hindering the bioleaching process and causing precious metal deposition on the jarosite (Zhang et al. 2020).Secondary iron minerals, such as jarosite, are frequently formed in heap-leaching methods.Jarosite accumulates in the ore heap, lowering the ferric iron content of the leachate (Soe et al. 2021).Jarosite formation occurs when pH and ORP levels are raised, as ferric iron has an exceptionally low solubility at pH > 2.5 (Castro et al. 2017).The reaction for jarosite formation (with the general formula MFe 3 ,(SO 4 ) 2 (OH) 6 , where M is a monovalent cation) formation is given below (Nemati et al. 1998): Sulfides are typically concentrated and processed at high temperatures.However, conventional liberation and processing of nickel sulfide ores, particularly those containing pentlandite, ((Ni,Fe) 9 S 8 ) as exsolved lamella in pyrrhotite (Fe 1-x S) present challenges to liberate and process conventionally (Watling 2008).Many studies investigated the processing of nickel-sulfide ores (Cruz et al. 2010;Khandan et al. 2021;Sun et al. 2020).Pentlandite's sulfur species can undergo modification through the polysulfide mechanism, which is Nickel in pentlandite can be transformed from (Ni, Fe) 9 S 8 to NiS before dissolving in solution as Ni 2þ (Sun et al. 2020).The presence of accessory forsterite (Mg 2 SiO 4 ) in ophiolitic ore samples increases acid consumption, attributed to the protonation of magnesium silicate surfaces.Acid consumption increases in ore samples with low Cr 2 O 3 /MgO ratios (Khandan et al. 2021).Low-grade ores are treated under extremely mild conditions in order to minimize expenditures (Zare Tavakoli et al. 2017).According to research, heap bioleaching is a viable industrial option for metal recovery (Jalali et al. 2019).Laboratory-scale simulations of heap bioreactors (column bioleaching) enable in-depth examination and testing of leaching parameters and mineral-oxidizing microbial consortia (Yang et al. 2013).Because the conditions in the column are very similar to those in a heap, leaching in columns, with or without recirculating the leaching liquid, simulates percolation leaching (Mousavi et al. 2006).Numerous parameters such as particle size, temperature, irrigation rate, aeration rate, and solution pH can affect the performance of column bioleaching (Mehta et al. 2010;Jalali et al. 2019;Zare Tavakoli et al. 2017).The formation of a jarosite layer on mineral surfaces, akin to shake flask bioleaching, hampers recovery by limiting microorganism contact (Hao et al. 2016).Ores and solid wastes are frequently pretreated prior to column bioleaching due to their acid/alkali consumption properties.Ores and solid wastes may contain acid-consuming minerals, such as sodium, potassium, calcium, and magnesium, which increase pH during bioleaching.By inhibiting the activity of acidophilic microbes, the significant increase in pH reduces bioleaching efficiency (Srichandan et al. 2020).Jarosite precipitation can be avoided during the bioleaching process by adding dilute sulfuric acid and maintaining an acidic condition in the leaching medium (Chen et al. 2015).Zhen et al. examined the bioleaching of a low-grade nickel-bearing sulfide ore containing relatively high olivine, chlorite, and antigorite (MgO 30-35%) in the main gangue minerals, using a mixed mesophile culture.After nearly two years of adaptation, the mixed bacteria's tolerance to Mg 2þ could significantly increase from 10 g/L to 25 g/L.A nickel recovery of 91% and a cobalt recovery of 81% were achieved in 312 days column leaching process, including 60 d acid preleaching stage and 252 d bioleaching stage (Zhen et al. 2009).Yang et al. investigated the copper and nickel bioleaching behavior of low-grade nickel, copper, and cobaltbearing sulfide ore.Although high recovery was achieved with the shake flask, 46% Ni, 39% Co, and 13% Cu were recovered after 139 d of column leaching, including 19 d of acid pre-leaching and 120 d of bioleaching.Potassium, sodium, and iron (III) ions in the leach solution precipitated jarosite within the ore bed, incorporating some nickel and copper ions into the precipitates, reducing overall metal recovery.Magnesium was preferentially leached from gangue minerals by acid, resulting in a magnesiumdepleted silicate skeleton, but discrepancies in residue analysis indicate that magnesium re-deposits into the silicate skeleton during bioleaching (Yang et al. 2011).
The bioleaching of two low-grade nickel-bearing samples with elevated magnesium concentrations was investigated in this study.Mineralogical studies were conducted to ascertain the mineral structure in order to determine the optimal method for nickel extraction.Both samples were adapted with a high pulp density to enhance bacteria' resistance to high magnesium levels in the environment.To determine the optimal method for nickel recovery, column leaching of samples with and without bacteria was investigated.Pretreatment methods such as acid washing were introduced to mitigate magnesium's adverse effects.Finally, residues were collected and identified.

Sample preparation and characterization
Two low-grade nickel-bearing samples weighing 40 kg with d80 ¼ 15 cm were collected from ophiolite deposits in southern Iran.Initially, each sample was divided into two sections, with one section being retained for later use.The remaining portion was crushed using a jaw crusher followed by a cone crusher, resulting in particle sizes smaller than 10 mm.Sample size fractions were standardized to the (À9.5 þ 6.3, À6.3 þ 4.8, À4.8 þ 2.36, À2.36 mm) for the column bioleaching investigations, as illustrated in Figure S1.The samples' particle size distributions were quite similar, which meant that the samples' d 80 , d 50 , and d 25 fractions all had the same values and were equal to 7.26 mm, 3.87 mm, and 1.04 mm in diameter, respectively.
X-ray diffraction analysis of each sample was performed using a Bruker D8-Advance with a CuK beam.Analyzing the surface morphology of samples was done using scanning electron microscopy (SEM) with an EDAX and gold coating.The elemental composition and main oxide compounds of each sample were obtained by ICP-MS analysis and Lifusion.Microscopic studies were performed using the Leitz polarizing microscope of model SM-LUX-POL equipped with a digital imaging camera at the College of Mining Engineering, University of Tehran, Iran.

Microorganisms, culture media, and analytical methods
The R&D Center of the Sarcheshmeh Copper Complex (Sarcheshmeh, Kerman Province, Iran) provided mixed cultures of mesophilic bacteria (Acidithiobacillus ferrooxidans, Leptospirillum ferrooxidans, and Acidithiobacillus thiooxidans).Initial cultivation and adaptations were maintained in 250 ml shake flasks at 150 rpm and 34 C (mesophiles).As energy sources, Fe 2þ (9 g/l) as ferrous sulfate and elemental sulfur (10 g/l) were utilized for the cultures.The initial pH was adjusted to 1.8 using sulfuric acid.Adaptation experiments were carried out at 1, 2, 5, and 10% pulp densities.Adaptation at each step was achieved when the ORP and cell counts reached 700 mV and 10 8 cells/ml, respectively.A 20-liter reactor with a vertical impeller was used to cultivate the adapted bacteria in the same condition as shake flasks.A Neubauer counting chamber (0.1 1/400 mm 2 ) and a Zeiss optical microscope (1000 magnification) were used for measuring cell counts.An ORP (Pt vs. AgCl/AgCl) meter was used to detect the pH and oxidation-reduction potential (ORP, Pt vs. Ag/AgCl).Titration with one mM potassium dichromate was used to quantify the concentration of Fe 2þ .An atomic absorption spectrometer was used to determine the total Fe ion (Fe t ) concentration in the leachate.Scanning electron microscopy (SEM) with energy-dispersive X-ray spectroscopy (EDX) was used to evaluate the surface morphology and identify the materials' elemental composition of the samples.The concentration of Fe 3þ is obtained by subtracting Fe t from Fe 2þ according to the following formula: (2)

Column bioleaching test
To assess the effectiveness of bioleaching, each sample underwent three-column bioleaching tests and one controlcolumn leaching test, as outlined in Table 1.A column bioleaching test was conducted using a 10% bacteria-inoculated solution under the same conditions as the control test to determine the effect of bacteria.In test 3, a bacterial solution containing 10% by weight of the sample was sprayed on the sample before loading the materials to the column as a pretreatment method.By this method, bacterial contact with the sample would be more efficient, and irrigating non-inoculated acidic solution and jarosite precipitation would be substantially reduced.For test 4, a pretreatment step was performed using a pH ¼1 solvent to eliminate acid-consuming minerals before bioleaching.The leaching was continued till the pH of the solution output reached 1.8.afterward, the irrigation was performed using bacterial solution until the end of the process.A Model L100-1S (Longer Pump company) peristaltic pump irrigated the fresh acidic solution (which contained 10% inoculated bacteria in some tests) onto each column.Short isothermal laboratory columns of 50 cm (height) Â 7.5 cm (internal diameter), made of 5 mm thick plexiglass, were utilized in this study.A high irrigation rate of 25 L/m 2 h was applied to provide equal wetting of mineral particles due to the short column diameter to particle diameter ratio.At the top of the column, a 3 cm high layer of a sponge was placed on the samples to ensure uniform spraying of the solution.Air was fed into the column at 55 L/(m 2 min) through a plastic support plate with perforations (2 mm) at the bottom of the column, allowing the particle bed to be evenly distributed.The leachate was passed through the ore and collected in a plastic bucket.Samples were taken from the leachate at specified intervals.Column charges were rinsed with distilled water after the bioleaching processes were completed, and residues were dried and prepared for the final analysis.Kendall's Tau-b and spearman's rank correlation In Kendall's tau, the correlation coefficient between selected parameters is determined by finding their interdependencies or intercorrelations.The correlation between the two variables is assessed using s, a statistical measure ranging from À1 to 1.A value of s ¼ 1 indicates a perfect positive correlation, while s ¼ À1 signifies a complete negative association between the variables.A value of s ¼ 0 suggests no correlation between the variables (KENDALL 1938).
It is necessary for process optimization to determine the relationships between operational parameters in order to gain a deeper understanding of the underlying mechanisms.In this study, Kendall's Tau-b method available in IBM SPSS 26 was utilized to evaluate the correlation among the parameters.

Samples characterization
Table 2 shows both samples' elemental composition and key oxide compounds, Li-fusion (XRF analysis) results.It is evident that the first sample contains significantly lower amounts of nickel and magnesium compared to the second sample.Conversely, the first sample exhibits a considerably higher iron content than the second sample.Approximately half of the chemical composition of the second sample comprises magnesium oxide.Both samples also contain notable quantities of Cr, Al, and Si.The XRD patterns of both samples are shown in Figures S2 and S3, exhibit distinct peaks, highlighting their unique mineral compositions.Magnesium-bearing minerals have been observed in both samples; these minerals can consume acid, increasing pH, thus affecting the bioleaching process.Olivine and serpentine were the predominant minerals of sample 1, while Clinochlore and Magnesiochromite dominated in Sample 2. As the amount of nickel in samples was low, it was hard to detect them by XRD.Therefore, microscopic studies were conducted to characterize the nickel-bearing minerals.The microscopic images of both samples are shown in Figure 1.In Sample 1, nickel is primarily observed in heazlewoodite, with pentlandite present as well.Pentlandite is mostly found alongside nonmetallic minerals like chromite, and pentlandite minerals can be observed in certain chromite fractions.In Sample 2, nickel is predominantly present in pentlandite, with heazlewoodite occurring in negligible amounts.Pentlandite is interstitial, while heazlewoodite-millerite occurs as microfracture filling.

Adaptation
Figure 2 illustrates the bacterial counts during adaptation tests for both samples at various pulp densities of 0.5%, 1%, 2%, 3%, and 5%.If the log 10 (cell/ml) had reached around 8 on the 30th day of each test, the test would be halted, and the solution would be used for the next adaptation step or cultivation.It is obvious that the log phase of bacterial growth began later when the pulp density was increased.Additionally, the bacteria count in sample 2 reached the  desired level (10 8 cells/ml) later, which is evident in the higher percentage of solids.In the 10% adaptation test of sample 2, the bacteria count did not reach the appropriate level for 30 d, and thus the process continued until the 38th day.The main reason for this delay in sample 2 can be attributed to the high amount of magnesium in the sample, which increases the pH and delays the growth and multiplication of bacteria.

Column bioleaching of sample 1
Figure 3 illustrates the column bioleaching results for sample 1.As observed, the output pH was initially high in the control (test number 1) due to the dissolution of acid-consuming minerals.The pH decreased over time while the recovery increased.The level of ORP in the control test initially increased and then fluctuated around 400 mV.In the leach solution, the Fe 3þ /Fe 2þ ratio was the main redox factor.The mesophilic bacteria was primarily responsible for producing Fe 3þ via Fe 2þ oxidation and oxidizing sulfide entities.The Fe 3þ /Fe 2þ ratio is critical, and it is regulated by iron oxidation, and sulfur oxidation facilitates proton attack via acid production.Nickel dissolution requires simply proton attack, but because it may be contained inside a mineral's crystalline structure, its dissolution is significantly dependent on the oxidation rate.Compared to the control test, the pH decreased more rapidly in the bioleaching test (test number 2), indicating the effect of bacterial activity.Additionally, the ORP exceeded 600 mV after 14 d.Nickel dissolution has also increased in response to bacterial activity, indicating that bacterial activity in this column has increased nickel recovery.Additionally, the iron recovery rate, bacterial count, and Fe 3þ /Fe 2þ indicate bacterial oxidation activity.
It is clear that after approximately 20 d, the bacteria's activity reaches its peak, facilitating nickel dissolution.Due to the absence of irrigation of inoculated solution in test 3, iron oxidation was delayed, which is evident in ORP levels.Also, the cell count is lower in test 3 than in test 2.However, the dissolution of nickel was not significantly different.The slight difference in nickel recovery makes this method more desirable because it requires a less bacterial solution, hence less chemical consumption, and, consequently, the operational cost can be reduced.The conditions of test 4 are similar to those of test 2, except that this test begins with a pretreatment step.After seven days of pH 1 acid irrigation, the pH at the column output has decreased to less than 1.8, which is an acceptable level for bacterial activity.Afterward, the inoculated solution was irrigated.
The results indicate bacterial metabolism increased, which resulted in 95% nickel recovery in this test.This high recovery demonstrates the importance of pretreatment, as acidconsuming minerals can raise the pH and inhibit bacterial activity at the start of bacterial activity, but after a pretreatment step, the pH reaches the desired value, and it does not rise above the desired value due to bacterial activity.

SEM and EDAX analysis of sample 1 solid residues
Because one of the explanations for the limited recovery in certain tests might be the presence of jarosite, SEM and EDAX tests were taken from sample 1. Figure 4 shows SEM micrographs of Sample 1 bioleaching tests residue.After 60 d of leaching or bioleaching, porous and degraded surfaces are clearly visible in mineral surfaces in both samples.As sample 1 had high iron content, detecting jarosite on the mineral surfaces is difficult in these tests.Also, SEM cannot show the mineral composition due to only measuring a part of the mineral surface.However, the amount of magnesium in tests 2 and 3 is much higher than in test 4, indicating the effectiveness of the pretreatment method.Meanwhile, the amount of nickel in the test is as high as expected, showing that this method cannot completely recover Ni content.

Column bioleaching of sample 2
Figure 5 illustrates the column bioleaching results for sample 2. As previously stated, the magnesium concentration in this sample was significantly higher than in sample 1, causing the pH to decrease at a slower rate in the control test.Additionally, this sample contained more nickel than the previous sample.Due to the increased concentration of acid-consuming substances and higher nickel concentration in the control test, the amount of nickel recovered at the end of 60 d is less than in the previous sample.The ORP level fluctuates around 400 mV as before in the control test after an initial increase.Compared to the control test, the pH decreased more rapidly in the bioleaching test.Additionally, a higher recovery rate was observed, indicating a beneficial effect on bacterial activity.However, in comparison to the previous sample, the ultimate recovery has decreased in this test, which could be due to the high magnesium content of this sample, which is detrimental to bacterial activity.Additionally, jarosite can be formed because the ORP has exceeded 600 mV after only 11 d, but the pH remains greater than 3 simultaneously.Jarosite sedimentation can close the column pathways and co-precipitate nickel, both of which reduce recovery.In contrast to the previous sample, the nickel recovery in test 3 was higher than that in test 2.This increased recovery may be attributed to the delayed complete oxidation (Fe 3þ /Fe 2þ reaching 100%), resulting in less jarosite sedimentation.As can be seen, the ORP in test 3 exceeded 600 mV after approximately 42 d, when the pH was around 2.2 the final recovery of test 3 was slightly lower than sample 1.The pretreatment period in test number 8 required 14 d due to the higher level of acid-consuming minerals in sample 2. The pH and ORP fluctuated around 1.9 and 700 mV, respectively.The Ni recovery in this test reached more than 95%.Although the recovery in tests 6 and 7 was smaller than in tests 2 and 3, the recovery in test 8 has stayed relatively consistent and has achieved an acceptable level, indicating the success of the pretreatment method.

SEM and EDAX analysis of sample 2 solid residues
Figure 6 showcases SEM micrographs of the residue from the bioleaching tests conducted on sample 2. Similar to sample 1, mineral surfaces in both samples show porous and eroded surfaces after 60 d of leaching or bioleaching.The presence of a higher magnesium content in sample 2, as expected, is evident.Tests 2 and 3 also show elevated levels of iron and sulfur, which suggests the deposition of jarosite.Additionally, the nickel content in test 5 is consistent with expectations.
A comparison of the SEM images between the two samples indicates that the high magnesium concentration in the second sample can contribute to the precipitation of jarosite.Moreover, the successful dissolution of magnesium as an acid-consuming mineral can be attributed to the effectiveness of the pretreatment method.It is worth noting that the deposition of jarosite in tests 4 and 8 aligns with expectations.

Correlation coefficient of bioleaching parameters
In this research, the correlation of the final results of three main parameters (nickel recovery, pH, and ORP) in all leaching and bioleaching tests was examined and the result is shown in Table S1 and S2.The objective was to compare the effective of each parameters on each other.Kendall's tau-b is a non-parametric correlation measure that does not assume linearity and is suitable for ordinal or ranked data.The results show that nickel recovery and ORP have a moderate positive relationship (correlation coefficient % 0.571, p < 0.05).Also, nickel recovery and pH have a moderate negative relationship (correlation coefficient % À0.643, p < 0.05).However, ORP and pH show a weak negative relationship (correlation coefficient % À0.214), and it is not statistically significant (p > 0.05).Hence, as was expected, the more pH decreases and ORP increases, the recovery of nickel would be enhanced.It shows why the bacterial activity (increase of ORP) and pretreatment methods (decrease of pH) had a positive impact on nickel extraction.Also, the presence of magnesium would have a negative effect on nickel recovery.

Bio-heap operation feasibility
Jarosite is a sulfate mineral found naturally or produced as a waste product in industrial operations.Fe 3þ precipitates mainly as jarosites in sulfate-rich conditions (Castro et al. 2013).a layer of jarosite on particle surfaces that is encased in sulfide grains and thick enough can prevent reactants and reaction products from transferring between the primary solution and mineral surfaces (Kaksonen et al. 2014).High ferric concentrations can facilitate nickel extraction; excessive jarosite production might deplete the solution of ferric ions and hinder nickel leaching (Yang et al. 2011).Moreover, the research conducted confirmed that the formation of jarosite leads to reduced recovery.For these reasons, the main factor for operating a bio-heap would be to prevent the formation of jarosite.The pretreatment method showed the best ability for preventing the formation of jarosite and Ni recovery.As acid leaching was initially started in this method, the pH decreased.After the dissolution of a considerable amount of sulfide-consuming minerals, bioleaching started.For this reason, when Fe 3þ /Fe 2þ reached its limit, the pH was lower than 2. This condition is not suitable for jarosite formation compared to other test conditions.The importance of this method was more evident when a high amount of acid-consuming minerals were bioleached.The only problem with this method may be its high acid sulfuric consumption.However, when compared to other acids, acid sulfuric has historically been one of the most chosen acids due to its low price, low corrosivity, toxicity, and excellent efficiency (Wolfaardt et al. 2021).Finally, the pretreatment method demonstrates its feasibility for bioheap operations.Enhancements can be made by utilizing thermophilic cultures instead of mesophiles.Thermophilic microorganisms have a greater capacity to remove the passive sulfur layer on mineral surfaces, and higher bioleaching temperatures can accelerate the oxidation rate (Khodadadmahmoudi et al. 2022).If the requisite temperature conditions for thermophiles can be created and sustained within the heap, thermophilic bio-heap leaching may be a viable low-cost, low-maintenance option (Pradhan et al. 2008).

Conclusions
It was discovered that sample 1 contained a lower concentration of nickel and magnesium than sample 2. Bacteria were able to adapt to both samples up to a pulp density of 10%.It took 30 d for sample 1 to reach a bacterial count of 10 8 counts/ml, whereas sample 2 required 36 d, likely due to its higher magnesium content.The column tests involved leaching, bioleaching, bioleaching by sprayed bacteria, and bioleaching by pretreatment of acid washing.It was concluded that the formation of jarosite is the most critical inhibitor of bioleaching.The pretreatment method using acid washing showed the most promising outcomes for both samples, achieving over 95% nickel extraction.Because of the significant amount of magnesium present in sample 2, the pH of the bioleaching solution rose, resulting in the precipitation of jarosite in the column and a reduction in efficiency.SEM tests were conducted on the residues from each bioleaching process, confirming that the presence of abundant magnesium in sample 2 contributed to jarosite formation.In conclusion, the pretreatment method proves to be a viable approach for Bio-heap Operations.

Figure 1 .
Figure 1.Polarizing microscopic images of samples 1 and 2 with on a 100 mm scale.

Table 1 .
Conditions of column leaching and bioleaching tests of two samples.

Table 2 .
XRF Analyses of the representative samples.