The phosphate ore originated from Djebel Onk (Kef Essnoun region) was acquired courtesy of the National Company of Iron and Phosphate FERPHOS. Ore samples were collected from a site where no excavation or any other human activity was conducted. The material was collected into aseptic containers under sterile procedures, stored in air-tight conditions, and transported to the laboratory to isolate native microorganisms. The second part of the collected ore was used for physicochemical analysis and biosorption tests after proper mechanical treatment.
4.1. Characteristics of the raw ore (before mechanical preparation).
After transporting ore samples to the laboratory, a granulometric study was conducted first. The particle size analysis of the sample revealed that 3% of the particles were less than 80 μm (Fig. S1, Table S1). Seventeen granulometric classes were distinguished, which were characterized by their metal contents and mineralogical composition.
4.2. Measurements of metal concentration in phosphate ore.
In order to measure metal concentration, ore samples of each granulometric class were first mineralized in wet conditions, using a Multiwave 3000 Microwave Digestion of Perkin Elmer. After drying the material (temp 30°C, 0.5 h), 0.200 g samples were placed directly into Teflon vessels and covered with concentrated acids in the following volumes: 6 mL HNO3 (65%), 2 mL HCl (35%), and 4 mL HF (40%). The samples in the vessels were placed in a microwave oven and mineralized according to the program: the first step (raising the temperature to 220℃) – power: 1800 W, time: 25 min, temperature: 220℃; the second step (maintaining the temperature of 220℃) – power: 1800 W, time 20 min. After this step, samples were cooled, and 24 mL 4% H3BO3 was added to each dish to neutralize the remaining HF. Then, the vessels were placed again in the microwave oven according to the program: the first step (raising temperature to 120 ℃) – power 1200 W, time: 10 min; the second step (maintaining the temperature of 120 ℃) – power 1200 W, time 5 min. After mineralization, the samples were cooled and filled to 50 mL with deionized water. Metal contents in the samples were measured using iCE™3500 AAS atomic absorption spectrometer (ThermoFisher Scientific). Each sample was measured in two repetitions. The results of these analyses were collected in table 3. Following that, the mineralogical composition was assessed in each fraction, using the XRD method.
4.3. Mineralogical composition assessment: X-ray diffraction (XRD).
XRD analyses were performed on powdered samples using a PANalytical X’Pert Pro MPD (multipurpose diffractometer) powered by a Philips PW3040/60 X-ray generator and fitted with a 1D silicon strip detector (X’Celerator) with a 2.122° 2θ active length. The measurements were performed using Cu Kα-radiation with a wavelength of 0.1540598 nm, an acceleration voltage of 40 kV, a current of 40 mA, and with 0.02 °2θ step sizes between the angles of 5° and 70° 2θ and a 200 s measurement time per step. Powder diffraction analysis parameters are gathered in Table S2. The data obtained were processed using HighScore+ software (version 4.1), linked to the ICSD database (2015) and the PDF4+ ICDD database (2018).
For the standardless, quantitative phase analysis, the Rietveld method was used. Rietveld structure fit module is a part of the HighScore Plus program suite34. Quantitative phase analysis can be performed on multi-phase samples using the formalism described by Hill and Howard35. The Rietveld method is a full-pattern fit method. The measured profile and a profile calculated from crystal structure data are compared. By variation of many parameters, the difference between the two profiles is minimized. In order to obtain quantitative calculations, the semi-automatic Rietveld mode in HS+ was used. The refinement was carried out until good statistical parameters were obtained: R expected = 4,30; R profile = 6,50; Weighted R profile= 8,61; Goodness of Fit =4,01. The results of XRD analyses for each ore fraction are presented in table 4.
4.4. Mechanical preparation of ore for testing.
In the next stage, ore was prepared for biosorption tests. In order to achieve a homogenous fraction, mechanical procedures (crushing, grinding, and sifting) were conducted according to the algorithm presented in fig S2.
Homogenous fraction (80-160 μm) was subjected to the assessment of mineralogical composition using the XRD method. For this fraction the results were as follows: carbonate fluoroapatite (CFA; Ca5(PO4,CO3)3F) 87%, dolomite (CaMg(CO3)2) 10%, calcite (CaCO3) 1.5%, clinoptilolite (Ca2-3[Al3(Al,Si)2Si13O36]·12H2O) 1.0%, and quartz (SiO2) 0.5%. (Fig. 3). Before biosorption tests, ore was sterilized to eliminate any microorganisms interfering with the experiment. Samples of 1 g of processed ore were weighed out, autoclaved (121 °C, 2 h), and stored in sterile conditions for further use.
4.5. Isolation and identification of microorganisms from raw ore.
In order to isolate native bacteria from the phosphate ore from Djebel Onk mine (Algeria) 10 g of ore was suspended in 90 mL of sterile NaCl, shaken for 1 hour at 120 rpm, and serially diluted. 0.1 mL of each ore dilution (from 10-1 to 10-6) was spread on the solid culture media LB and R2A (BTL, Łódź, Polska). The cultures were incubated for 96 h at 28 °C. Single bacterial colonies were passaged on fresh media to obtain pure cultures. Among the isolated bacteria, four morphologically different strains were selected for identification and biosorption testing. Molecular identification of bacterial strains was based on the sequence of 16S rDNA gene fragment. PCR with primers 8F and 1492R36 was performed as described by Pacwa-Płociniczak et al.37. 1484 bp PCR products were cloned with the use of pGEM®-T Easy Vector System (Promega, Madison, Wisconsin, USA) and sequenced at Genomed S.A. (Warsaw, Poland). The edition of sequences was conducted manually using Chromas Lite 2.01 (Technelysium Pty Ltd, Brisbane, Queensland, Australia), and chimera detection was performed using Decipher 2.19.238. Obtained sequences were aligned to the reference sequences of 16S rRNA gene available in the GenBank database (National Centre for Biotechnological Information) using BlastN. The 16S rDNA sequences of strain HK1 showed 99.80% of identity to the sequence of Lysinibacillus sphaericus (CP026120.1) and 99.74% of identity to the sequences of L. sphaericus (KF228905.1) and Lysinibacillus fusiformis (CP010820.1, EU545408.1). The sequence of strain HK2 showed 99.12% of identity to the sequences of Pseudarthrobacter oxydans (KR085945.1, KR085776.1) and Pseudarthrobacter scleromae (KR085778.1), and 99.06% of identity to the sequence of Pseudarthrobacter psychrotolerans (CP047898.1). The sequence of strain HK3 showed 99.93% identity to the sequences of Bacillus mycoides (MT827167.1, CP031071.1). The sequence of strain HK4 was 99.93% identical to the sequences of Bacillus subtilis (KX281166.1, KU551251.1), Bacillus amyloliquefaciens (KU551122.1) and Bacillus velezensis (CP053764.1). The sequences were deposited in GenBank with accession numbers: MZ046078 (USK1), MZ046079 (HK1), MZ046080 (HK2), MZ046081 (HK3), MZ046082 (HK4).
4.6. Biosorption testing.
Native microorganism strains chosen for tests and Bacillus subtilis USK1 from the collection of the University of Silesia in Katowice (as a reference strain) were cultured in LB medium for 24 hours. After incubation, bacterial cultures were washed and suspended in 0.98% NaCl to get the concentration of 3 g of biomass L-1 for each strain. Suspensions were portioned into 50 mL aliquots for biosorption analysis. Sterilized 1 g ore samples were incubated (28°C with constant shaking) with microorganisms (50 ml suspension) for 20 minutes at different pH values (4, 7, and 10). After incubation, samples were set aside for 5 min to facilitate sedimentation of ore particles. Next, the solution containing microorganisms was gently poured into sterile 50 ml centrifuge tubes, and metal concentrations in biomass were assessed according to a procedure described elsewhere15. In short, the suspensions of microorganisms were centrifuged (Ultracentrifuge Beckman Optima LE-80K) at 8000 rpm (4 °C; 15 min). Biomass was frozen at -70°C and lyophilized (Freeze dryer Alpha 1–4; Christ, Germany) at −35°C and 0.2 mBar for 24 h. Portions of ~0.02 g of dried microorganisms were mineralized with 0.5 mL of ~65% HNO3 at 110°C for 48 h. After mineralization, the samples were diluted with deionized water to a total volume of 5 mL. Cd and Mg contents were measured by AAS methods with an iCE™ 3500 AAS atomic absorption spectrometer (Thermo Fisher Scientific). The quality of the analytical procedure was confirmed using standard solutions from Merck at the initial concentration of 1 g of metal L−1 of water. Metal content was expressed as μg g−1 of biomass. The metal analyses were done in two technical replications. All tests in each experimental group were done in triplicates. Results were expressed as mean ± SD. The analysis of variance, followed by the Fishers Least Significant Difference test (LSD, ANOVA; p < 0.05) was performed to assess the significance of differences in metal biosorption among the studied bacterial strains and different pH. Statistical analysis was conducted using Statistica 13.1 software.