Elemental sulfur (S°) with a purity of 99.5% (Alfa Aesar) and particle size < 44 μm was initially used in bio-oxidation tests. Subsequently, a low-grade refractory sulfidic ore from a Canadian deposit was employed in further bio-oxidation experiments. The ore was acquired in a coarse form and later crushed and pulverized into fine particles. Manual rifflers were used to obtain representative samples. The particle size of 80% passing (P₈₀) of the sample prior to bio-oxidation was measured with laser particle size analyzer (Malvern Mastersizer 3000).

Thiobacillus Thioparus (DSMZ 505) and Starkeya Novella (DSMZ 506) were acquired in freeze-dried form from The German Collection of Microorganism and Cell Cultures (DSMZ). Both neutrophilic microorganisms are classified as biosafety level 1 (BSL1) sulfur-oxidizing bacteria. T. thioparus was grown in batch cultures of Thiobacillus medium DSMZ 36 (TM) which is composed of biological grade reagents including 0.1 g/L (NH₄)₂SO₄, 4 g/L K₂HPO₄, 4 g/L KH₂PO₄, 0.1 g/L MgSO₄·7H₂O, 0.1 g/L CaCl₂, 0.02 g/L FeCl₃·6H₂O, 0.02 g/L MnSO₄·H₂O, 10 g/L Na₂S₂O₃·5H₂O (Sigma-Aldrich and Fisher scientific) unless specified otherwise. Thiosulfate was used as primary energy sources for its metabolism and the phosphate salts in the medium conditioned the initial pH at 6.6 (Kelly et al., 2015). On the other hand, S. novella was also cultured in a modified Thiobacillus medium DSMZ 36 (TM1–6) supplemented with glucose (1.5 g/L, Fisher scientific) and yeast extract (0.3 and 5 g/L, Sigma-Aldrich) to favor its mixotrophic growth (Table 1), which considered the utilization of organic and inorganic substances (Leefeldt and Matin, 1980; Purnomo et al., 2019). Furthermore, S. novella was also grown in S. novella medium DSMZ 69 (SM) containing 10.6 g/L Na₂HPO₄·12H₂O, 1.5 g/L KH₂PO₄, 0.3 g/L NH₄Cl, 0.3 g/L yeast extract, 2 mL/L phenol red solution (0.1% w/v), 0.1 g/L MgSO₄·7H₂O, 5 g/L Na₂S₂O₃·5H₂O (Sigma-Aldrich and Fisher scientific), 5 mL/L trace element solution (MD-TMS, Cedarlane) as well as Trypticase soy broth medium DSMZ 535 (BM) comprised of 30 g/L trypticase soy broth (Fisher scientific).

Cultivation of these microbes was carried out in a rotatory shaking incubator (Biobase; BJPX-2102C) at 30°C and mixing speed of 150 RPM using 5% v/v inoculum and 500 mL Erlenmeyer flasks with 200 mL of working volume, unless stated otherwise. Before the experiments, all glassware was autoclaved at 121°C for 20 min and solutions were filtered with 0.2 μm Thermo Scientific Nalgene Rapid-Flow filter to maintain aseptic conditions. Optimal bacterial growth conditions for subsequent bio-oxidation tests were studied by monitoring the pH, oxidation-reduction potential (ORP), dissolved oxygen (DO), optical density at 600 nm (OD_600nm), bacterial density (cells/mL) and thiosulfate consumption (mg/L). Both microorganisms were subjected to an adaptation process to the sulfidic ore for 3 months following the method of Astudillo and Acevedo (2008) and were used in the experiments. Nevertheless, the adaptation was fruitless over the considered adaptation time compared to the original strains. All experiments were carried out in duplicates throughout this investigation.

In the first instance, elemental sulfur (S°) was used in bio-oxidation tests to evaluate the potential of both microorganisms to oxidize solid matrices. A concentration of 5,000 mg/L S° was initially employed as a sulfur material in bio-oxidation tests using a two-step methodology unless stated otherwise. A two-step bio-oxidation technique involves tests at which bacterial inoculum (5% w/v) of an existing culture is originally incubated in a fresh medium and allows to propagate using the medium components, e.g., thiosulfate, as an initial energy source to achieve an optimal growth. After the exponential phase of the bacterial growth (2–3 days) was accomplished, the sulfur material to be oxidized was introduced into the reactor, commencing the two-step reaction. Tests were conducted in a rotatory shaking incubator at 30°C and 150 RPM using 500 mL Erlenmeyer flasks with a working volume of 200 mL for a total experimental time of 10 days. pH, ORP, DO, and bacterial density were monitored throughout the tests. Control tests were aseptically carried out under identical testing conditions using the same bacterial growth medium with no microorganisms added.

Subsequently, the oxidative capacities of these neutrophilic microorganisms were tested with a refractory sulfidic ore following a similar experimental procedure used for S°, as specified in Table 2. Thus, once the exponential phase of the bacterial growth was successfully reached, a sulfidic ores sample at a solid to liquid ratio of 10–100 g/L was added into the reactor. Bio-oxidation tests with sulfidic ores were mainly performed in 500 mL Erlenmeyer flasks at the ideal conditions for bacterial growth mentioned in the previous section in a temperature-controlled incubator for 10 days of reaction time. Thymol (1% m/v) wash was used to ensure there was no other culture already present in the sulfidic ore before starting the tests (de Carvalho et al., 2019). 2 mL aliquots were withdrawn to monitor bacterial growth, pH, ORP and DO. Following tests in 500 mL reactors, bio-oxidation reactions with the refractory sulfidic ore were scaled up in a 1 L jacketed reactor (Chemglass) equipped with overhead mixing at 150 RPM as shown in Figure 1. The temperature was adjusted to 30°C, and the air was sparged at 0.1 L/min to maintain DO concentration throughout the test. A volume of 950 mL of fresh culture medium was added to the reactor, followed by 5% w/v inoculum of active bacterial culture, which was grown 2–3 days before starting the test with the ore. After 10 days of contact time, the reactor test was ended, and a liquid aliquot was taken for further analysis. The pulp was pressure-filtered, separating solid residue from the liquid phase and the solid residue was washed with deionized water and filtered one more time.

The minerology of the sulfidic ore was determined by X-ray diffraction analysis (XRD; Philips, Netherlands) and the morphology of solid surfaces was characterized with scanning electron microscopy (SEM; Bruker AXS Microanalysis GmbH, Germany). The concentration of elements in the sulfidic ores was determined with coupled plasma-optical emission spectrometry (ICP-OES; PerkinElmer ELAN 9000/NexION).

Cell growth was monitored with ultraviolet–visible spectroscopy (UV-VIS; Thermo Scientific GENESYS 10S) contrasted against aseptic culture medium at 600 nm and manual cell counting employing a Neubauer counting chamber and a Rebel microscope (ECHO) at 40–100 × magnification following the method of Abcam (2022) throughout the tests as shown in Supplementary Figure S1. Liquid samples were filtrated with 0.45 μm sterile syringe filter before further analysis. pH and oxidation-reduction potential (ORP) were monitored utilizing a saturated Ag/AgCl electrode on a multichannel pH/ORP probe (VWR; Symphony H30PCO). Dissolved oxygen (DO) was obtained using an optical DO sensor connected to a multimeter (METTLER TOLEDO, Fisher Scientific). Thiosulfate concentration was determined with iodometric titration using the method of Arima et al. (2004) and sulfate concentration was measured with ion chromatography (IC; Methrohm 930 Compact IC Flex).

After S° bio-oxidation, S° was washed with DI water and dried in a vacuum-sealed oven below 40°C (Liu et al., 2013). The oxidation rate in S° bio-oxidation was estimated by considering the concentration of initial elemental sulfur and the concentration of elemental sulfur in the final residue. Hence, sulfur oxidation was therefore calculated as shown in Equation 3:

where mS∘-S_initial represent the mass (g) of S° initially and mS∘-S__residue stands for mass (g) S°-S in the residue.

Moreover, sulfide oxidation rate in sulfidic ore was determined using carbon/sulfur analyzer (Eltra CS 200) (Ofori-Sarpong et al., 2011; Kaksonen et al., 2014a). After finishing the tests, the oxidized sample was separated and dried in a vacuum-sealed oven below 40°C. The treated sample was washed with toluene in a ratio of 1 g: 50 mL at 60°C for 60 min to remove any S° formed during bio-oxidation and dried under identical conditions (Marzoughi et al., 2022). Subsequently, the oxidized sample was washed with 3M hydrochloric acid (HCl) at 65°C for 90 min to remove sulfate (SO 42-) (Ofori-Sarpong et al., 2011). The washed residues were dried and further analyzed with carbon/sulfur analyzer to determine the sulfide oxidation rate. The oxidation rate was calculated based on the total sulfide concentration in the residue as displayed in Equation 4 (Kaksonen et al., 2014a):

wherein m_{sulfide − S_ore} stands for the mass (g) of sulfide-S in the original ore and m_{sulfide − S}__residue represent mass (g) sulfide-S in the residue after washing with toluene and HCl.

The composition of the sample is presented in Table 3. The low-grade refractory sulfidic ore sample contained 5.7% w/w total sulfur and 1.7 ppm Au with a minor content of heavy metals. Moreover, Figure 2A displays the XRD pattern of the refractory sulfidic ores used in bio-oxidation. The main phases in the ore were quartz, muscovite, greigite, spangolite, and pyrite. Further, Figure 2B shows the particle size of 80% passing (P80) which was determined to be 55 μm.

The bacterial growth of both neutrophilic bacteria was evaluated in response to changes in the concentration of certain ingredients in the culture media. Figures 3, 4 show the changes in bacterial growth as well as the fluctuations in pH, ORP (mv), and measured thiosulfate concentration (mg/L) over time for T. thioparus and S. novella, respectively.

In the case of T. thioparus (Figure 3), a significant impact was produced by the change in the initial concentration of sodium thiosulfate in the culture medium, influencing the main parameters measured during its growth. From Figures 3A, B, the exponential phase of the bacterial growth was achieved after 3 days in all the experimental conditions, abruptly increasing optical density and bacterial density, correspondently. Optical density reflects the bacterial growth based on the turbidity of the bacterial solution (Widdel, 2010), which is normally determined at a wavelength of 600 nm in the case of the selected bacteria (Toth et al., 2015; Peighami et al., 2022). Thus, an increase in OD₆₀₀ indicates that the bacterial population increases over time which is reflected in more turbid solutions compared to the original solution at day 0 in Figure 3. An increase in sodium thiosulfate concentration from 2 to 10 g/L led to a substantial rise of 2–3 folds in OD₆₀₀ and bacterial density as thiosulfate is utilized as the principal energy source for the bacteria, favoring its metabolism at higher concentration in the range 2–10 g/L.

The more significant growth using 10 g/L sodium thiosulfate could be attributed to its metabolic characteristic since T. thioparus is an obligately chemolithotrophic bacteria that obtains its energy mainly from the oxidation of reduced sulfur compounds such as thiosulfate presented in the culture medium (Kelly et al., 2015). Reduced sulfur species are used as electron donors and oxygen is utilized as an electron acceptor to generate an electron flow in the cell membranes required to satisfy the bacterial metabolism through phosphorylation reactions (Kelly et al., 2015; Lee et al., 2021). Therefore, an increase in the concentration of sodium thiosulfate in the range 2–10 g/L seems to have helped to enhance the metabolism of T. thioparus, which resulted in an augment in OD₆₀₀ > 0.6 with a bacterial density of 1.2 × 10⁶ cells/mL.

A concentration higher than 10 g/L sodium thiosulfate was also employed to enhance the bacterial growth of T. thioparus. Nevertheless, a concentration of 20 g/L sodium thiosulfate (Supplementary Figure S2) had a negative impact, which seems to be related to a possible salinity stress generating a opposite effect on bacterial growth. Higher salinity in the medium affects the water balance of bacterial membranes due to its osmotic properties, which depends on the class of bacteria and environmental conditions (Boziaris et al., 2007; Kumar et al., 2021). Keller (1969) noted that the growth of T. thioparus ATCC 8158 was entirely inhibited at 0.3 M NaCl due to low tolerance to high saline environments.

Furthermore, Figures 3C, D show that a higher concentration of sodium thiosulfate (2–10 g/L) caused a more abrupt decrease in the solution pH and an increase in ORP during the evaluated period. A drop in pH was observed from initial pH 6.6 to pH 4.0–5.0 when the exponential phase of the growth was reached. ORP increased during the exponential phase and then remained in the range 130–230 mV depending on the initial sodium thiosulfate provided.

The decrease in pH and increase in ORP could be caused by the metabolism of thiosulfate during the bacterial growth and the release of protons into the medium by T. thioparus (Kelly, 2006; Tang et al., 2009). Thiosulfate was undetectable at the end of the bacterial growth in all the culture media as shown in Figure 3E. Control tests kept a near-neutral pH and ORP < 50 mV with insignificant changes in thiosulfate concentration compared with the test (data not shown), demonstrating that the T. thioparus was responsible for the changes observed during the evaluated interval. A concentration of 10 g/L sodium thiosulfate was preferred during the evaluation of its oxidative capacities of solid matrices due to a more robust growth was obtained under these conditions.

Figure 4 displays the bacterial growth of S. novella cultured in modified TM (Table 1), BM, and SM. S. novella is a facultative chemolithotrophic sulfur-oxidizing bacterium that can thrive using both organic and inorganic reagents to obtain energy required for its metabolism (Perez and Matin, 1980; Kelly and Wood, 2000). Media TM1–6 and SM favored a mixotrophic growth using organic and inorganic compounds, while medium BM supported a heterotrophic growth based on the utilization of only organic substrates.

The exponential phase of the bacterial growth was reached after 1–2 days of incubation in all the culture media as shown in Figure 4A, which resulted in an abrupt increase in OD in all media and thus more turbid solutions compared to the respective original culture medium at day 0 owing to a more significant number of microorganisms. In media TM1–6, the bacterial growth was increased almost 3 folds as the concentration of yeast extract was augmented from 0.3 to 5 g/L. Organic reagents such yeast extract, are used by S. novella as a source of carbon required for its growth (Kelly et al., 2000). In particular, yeast extract is a nourishing nutrient with a high diversity of proteins, amino acids, and minerals (Pasupuleti and Demain, 2010). Thus, adding a higher concentration of yeast extract in the culture medium could stimulate bacterial growth notably in terms of OD₆₀₀ and bacterial density as shown in Figures 4A, B, respectively.

However, even though the bacterial growth of S. novella resulted in higher OD in culture media with 5 g/L yeast extract (TM4–6), the utilization of thiosulfate was slower compared to cultures in TM1–3 as illustrated in Figure 4E. The ability of S. novella to oxidize reduced sulfur species could be affected by an excess of yeast extract causing an inhibition of substrate utilization, as the microorganism might prefer other substrates for its metabolism such as organic reagents instead of thiosulfate (Leefeldt and Matin, 1980). When organic reagents were not provided in TM, the growth of S. novella was completely inhibited (data not shown) due to the requirement of organic reagents (Leefeldt and Matin, 1980; Kelly and Wood, 2000). Contrasting T. thioparus, the variation in sodium thiosulfate concentration in the range of 2–10 g/L in the modified TM1–6 was not an influential parameter for S. novella, producing a slight change in OD₆₀₀ and bacterial density (Figures 4A, B).

Medium SM was also evaluated under similar conditions; nonetheless, the bacterial growth in OD₆₀₀ and bacterial density was 4–6 times lower than TM1–6, which is reflected in the bacterial density (cells/mL). The difference between TM1–6 and SM seems to have been attributed to the addition of another exogenous organic matter in TM1–6, e.g., glucose, which has been noted to support bacterial metabolism and cell growth of S. novella (Rittenberg, 1969; Leefeldt and Matin, 1980; Kelly and Wood, 2000). The bacterial growth was also investigated in a pure organic medium (BM) containing trypticase soy broth, which is a substrate mainly composed of tryptone, a rich source of amino acid (Pasupuleti and Demain, 2010). This culture medium produced a noticeably more significant growth than TM1–6 and SM, which resulted in an OD₆₀₀ > 2. However, the behavior of the bacteria under heterotrophic conditions could be different to mixotrophic growth due to dissimilar enzyme production as demonstrated by Aleem (1965) and Leefeldt and Matin (1980).

As shown in Figures 4C, D, pH and ORP generally increased during exponential phase and then stayed in the range 7.0–8.5 and 20–60 mV in most of the cases. Culture media containing yeast extract and tryptone showed an increase in pH due to the buffer capacity of these reagents, which seems to be associated with the decomposition of organic components and formation of alkaline products generated during bacterial growth (Losen et al., 2004). The increase in ORP in most of the culture media could be attributed to the metabolism of thiosulfate, as explained in the previous section. Additionally, it is worth mentioning that S. novella consumed up to 1,000 mg/L thiosulfate in all the experimental conditions, which was ca. 4 times lower than T. thioparus, causing a minor change in pH and ORP. Therefore, T. thioparus might have a higher potential to oxidize solid matrices than S. novella. Besides, the absence of thiosulfate in BM composition (not shown in Figure 4E) could have influenced the decreasing trend in ORP, reaching ca −80 mV at the end of the incubation.

Although all the culture media were tested for sulfide oxidation, TM3 was preferred in future experimentation with solid matrices since S. novella metabolized most of the initial thiosulfate presented in the solution after the exponential compared to the other cases, which could lead to increase the possibilities of oxidizing the sulfidic ore.

A preliminary assessment was performed with elemental sulfur (S°) to explore the oxidative capacities of T. thioparus and S. novella on solid matrices. Although there is no extensive research on S° oxidation with these specific microorganisms, neutrophilic bacteria such as T. thioparus and S. novella have been commonly encountered in soils, showing a more notorious activity in soils containing S° due to their bacterial metabolism (Deluca et al., 1989; Chapman, 1990; Germida and Janzen, 1993; Pourbabaee et al., 2020).

Bio-oxidation tests with S° were used as a baseline for subsequent bio-oxidation tests with sulfidic ores. Two different approaches were implemented to confirm the ability of both microorganisms to oxidize S°. On the one hand, the bacteria were first grown with thiosulfate as an initial energy source until the exponential phase was achieved, and subsequently, 5,000 mg/L S° was added into the reactor. On the other hand, the bacteria were directly grown with S° (5,000 mg/L) as sole energy source material with no addition of thiosulfate in the medium. Figure 5 shows a comparison between both strategies along with the control tests for T. thioparus and S. novella.

Thiobacillus thioparus showed more significant potential than S. novella in oxidizing S° using both approaches. When S° was used as the only substrate, 2,830 mg/L sulfate was measured in the presence of T. thioparus, which was almost 2 folds more sulfate compared to S. novella under similar conditions. A total S° oxidation rate of ca. 18% was oxidized by T. thioparus after 10 days of reaction time, while 10% of total S° were oxidized by S. novella (Figure 5B). During the tests with S° as the only energy source, T. thioparus increased solution acidity to pH 4, while S. novella maintained a near-neutral pH (data not shown). Chapman (1990) found that the addition of S° increased T. thioparus bacterial density, reaching 10⁵ cells/g soil and more acid pH solution, and Lindemann et al. (1991) showed similar results with the same microorganism. In another study, Boretska and Bellenberg (2013) pointed out that S° stimulated T. thioparus to produce different extracellular polymeric substances (EPS), improving bacterial attachment in solid substrates. On the other hand, Pourbabaee et al. (2020) reported a closely related strain to S. novella that produced ca. 12–15% S oxidation with 0.3-unit pH variation over 2 weeks, which agreed with the results shown in Figure 5B.

By using thiosulfate as an initial substrate for bacterial growth before the addition of S°, T. thioparus formed ca. 6,000 ppm (Figure 5A). However, the oxidation rate of S° was 14%, which was slightly lower than the previous approach. The decrease in oxidation could be produced by the presence of thiosulfate, which also served as a source of energy for the bacterium (Kelly et al., 2015). S. novella cultured with 2 g/L sodium thiosulfate in TM3 showed that ca. 7% of the initial S° was oxidized in the medium. Blank tests showed a negligible formation of sulfate in the range 150–250 mg/L, which could be caused by the presence of traces of sulfate salts in the culture medium and minor chemical oxidation. The oxidation of S° produced by abiotic factors has been noticed to be limited in soils compared to biological reactions (Germida and Janzen, 1993), which agreed with the results displayed in Figure 5.

From Figure 5, it is clear that both microorganisms are able to oxidize solid matrices, achieving a maximum of 18% oxidation with T. thioparus over 10 days. Although there was no complete oxidation of S°, the sulfur content of the refractory gold ore used in this study is 5.7%, which is lower than the amount of S° utilized in this section. Thus, this opens an option to achieve higher oxidation when working with the refractory sulfidic ore compared to S°.

The surface morphology of S° before and after bio-oxidation was evaluated with SEM images (Figure 6). The surface of the S° before bio-oxidation was characterized by a smooth texture, and the S° surface obtained after sterile control showed subtle surface alterations compared to the as-received S° sample (Figures 6A, B). Liu et al. (2013) also confirmed that the S° surface in control tests remained almost unchanged and that the roughness of the S° surface was produced by bacterial action.

Figures 6C–F show that the S° surface after contact with T. thioparus and S. novella became rougher compared to as received S° (Figure 6A). It was observed that T. thioparus led to more severe corrosion than S. novella, with more alteration and pits. The changes on S° surface were similar to the alteration produced by common acidophilic microorganisms (Briand et al., 1999). The bacterial action was generally irregular on the S° surface, causing non-uniform distortions (Espejo and Romero, 1987), which was less noticeable in the bio-oxidation test with S. novella due to the lower oxidation rate.

The more remarkable bacterial alteration produced by T. thioparus seems to be related to the higher sulfate production compared to S. novella. Briand et al. (1999) also noted more corroded surfaces led to a higher sulfate generation due to S° utilization in average 10–14 mmol/day. However, the sulfur consumption rate with neutrophilic bacteria was observed to be lower, which was reflected in the less corroded surfaces compared to acidophilic bacteria such a T. thiooxidans and A. ferroxidans.