Enhanced detoxification of Cr6+ by Shewanella oneidensis via adsorption on spherical and flower-like manganese ferrite nanostructures

Maximizing the safe removal of hexavalent chromium (Cr6+) from waste streams is an increasing demand due to the environmental, economic and health benefits. The integrated adsorption and bio-reduction method can be applied for the elimination of the highly toxic Cr6+ and its detoxification. This work describes a synthetic method for achieving the best chemical composition of spherical and flower-like manganese ferrite (MnxFe3−xO4) nanostructures (NS) for Cr6+ adsorption. We selected NS with the highest adsorption performance to study its efficiency in the extracellular reduction of Cr6+ into a trivalent state (Cr3+) by Shewanella oneidensis (S. oneidensis) MR-1. MnxFe3−xO4 NS were prepared by a polyol solvothermal synthesis process. They were characterised by powder X-ray diffraction (XRD), transmission electron microscopy (TEM), X-ray photoelectron spectrometry (XPS), dynamic light scattering (DLS) and Fourier transform-infrared (FTIR) spectroscopy. The elemental composition of MnxFe3−xO4 was evaluated by inductively coupled plasma atomic emission spectroscopy. Our results reveal that the oxidation state of the manganese precursor significantly affects the Cr6+ adsorption efficiency of MnxFe3−xO4 NS. The best adsorption capacity for Cr6+ is 16.8 ± 1.6 mg Cr6+/g by the spherical Mn0.22+Fe2.83+O4 nanoparticles at pH 7, which is 1.4 times higher than that of Mn0.8Fe2.2O4 nanoflowers. This was attributed to the relative excess of divalent manganese in Mn0.22+Fe2.83+O4 based on our XPS analysis. The lethal concentration of Cr6+ for S. oneidensis MR-1 was 60 mg L−1 (determined by flow cytometry). The addition of Mn0.22+Fe2.83+O4 nanoparticles to S. oneidensis MR-1 enhanced the bio-reduction of Cr6+ 2.66 times compared to the presence of the bacteria alone. This work provides a cost-effective method for the removal of Cr6+ with a minimum amount of sludge production.


Introduction
Chromium (Cr) is a common environmental pollutant coming from several industries such as wood preservation, 1 leather tanning, 2 steel production, 3,4 wool dyeing, 5 painting, 6 refractories, 4 lasers, 7 and electroplating, 8 among others. End-of-life products such as unwanted steel, wood, 1 leather and textiles, among other materials are extra sources of Cr leakage in the environment. The release of Cr in the environment was also attributed to mining activities, 9 and improper waste treatment associated with industrial processes. 10 Various Cr-bearing minerals, including chromite, are available in the soil, and bedrock also releases natural Cr into the environment. 10 It mainly occurs in two valence states, which are highly toxic carcinogenic 11 Cr 6+ and less toxic Cr 3+ . Various technologies have been developed to tackle the presence of Cr 6+ , including membranes, 12 coagulation, 13 photocatalysis, 14 electrochemical treatments, 13 adsorption 15,16 and biological treatments. 17,18 Integrating both adsorption and biological reduction of Cr 6+ together has been proposed as a promising solution. 19 Applying such combined methods can overcome the accessibility of certain technologies, 20 using less toxic chemicals and reducing the production of contaminated toxic waste. 19 The recovered chromium can be used in metallurgical industries and minimize the contaminated landll. 21,22 Microbial reduction of Cr 6+ has been regarded as a suitable Cr remediation approach because of being more eco-friendly than the conventional physico-chemical strategies, which are oen costly. Recently, many types of bacteria have been reported to detoxify Cr 6+ to less toxic Cr 3+ , including dissimilatory metal-reducing bacteria such as Shewanella oneidensis MR-1. 23,24 Under anaerobic conditions, S. oneidensis can use Cr 6+ as a terminal electron acceptor, 25 however cells exposed to Cr 6+ exhibited a loss in their enzymatic activity and cell lysis. 26 The bactericidal concentration of Cr 6+ was reported to be ∼42-65 mg L −1 for S. oneidensis MR-1. 26,27 A lethal effect of heavy metals on the microbes during respiration 26,[29][30][31] was considered as a potential limitation for the bio-remediation of Cr 6+ . 26 Compared with physical and chemical materials, the concentration of Cr 6+ that can be reduced by bacteria is much lower, and it is a great challenge to improve the efficiency of bioremediation. 28 Enhancing the bacterial tolerance to Cr 6+ is an effective way to improve the reduction of Cr 6+ . Zero-valent iron nanoparticles (ZVI NPs) can easily be oxidised to ferric oxides and hydroxides in water. The active surface of ZVI NPs can be decreased due to the attached layers of iron oxides and hydroxides. Shewanella, as iron-reducing bacteria, can reduce the adsorbed Fe 3+ to Fe 2+ , which reverses the oxidation of ZVI NPs, as shown in a review by Dong et al. 32 Hematite (a-Fe 2 O 3 ) particles enhanced the bioreduction of Cr 6+ bio-reduction by S. oneidensis MR-1, but they cause cytotoxicity to such kind of bacteria. 33 The reduction of Cr 6+ by S. oneidensis was enhanced by goethite (a-FeOOH) and humic acid through the bio-reduction of Fe 3+ to Fe 2+ . The reactivity of magnetite (Fe 3 O 4 ) was increased by microbial Fe 3+ reduction to form Fe 2+ , which then can reduce Cr 6+ . 34,35 A biocompatible material such as manganese ferrite (MnFe 2 O 4 ) 36 was considered for enhancing microbial respiration of Cr 6+ . This ferrite was used to accelerate extracellular electron transfer in the microbial fuel cell, 37,38 and it showed the highest adsorption capacity among other ferrites for Cr 6+ . 39 The maximum adsorption capacity of MnFe 2 O 4 NPs for Cr 6+ was reported to be ranging from 31 to 35 mg g −1 . 36,39,40 The inuence of structural features of Mn x Fe 3−x O 4 NPs on Cr 6+ adsorption has not been thoroughly explored. The effect of the oxidation state of Mn precursors on the chemical structure, morphological and magnetic properties of Mn x Fe 3−x O 4 NPs prepared by scalable polyol solvothermal method has been studied in a few reports 41,42 but not in relation to their adsorption efficiency for heavy metals.
Herein, we report syntheses and characterization of the most suitable chemical structure of Mn x Fe 3−x O 4 NPs and nano-owers (NFs) for the best adsorption capacity of Cr 6+ . The impact of the oxidation states of Mn precursors and variation in Mn doping levels on the chemical structural and morphological characteristics of MnFe 2 O 4 NPs prepared by polyol solvothermal route has been investigated. The nature of Cr 6+ adsorption by doped and undoped ferrite NPs and, subsequently, the bio-detoxication of Cr 6+ by S. oneidensis have been studied.

Synthesis of nanomaterials
In polyol synthesis, metal precursors are reduced at a high temperature by alcohols (polyols), which can act as a capping agent, solvent and reductant. Then metal nuclei form, grow and controllably coalesce together to produce the desired particles. 43 In such a non-aqueous solvent, the metal oxide NPs were proposed to be formed via two steps. In the rst step, solvolysis of the precursor involved an interaction between tetraethylene glycol (TEG) and the selected metal acetylacetonate, causing the generation of metal carboxylate. 44 The second step is a condensation reaction in which carboxylate reacts with iron leading to the formation of an oxo-bridge between metal (metal-oxygenmetal clusters) and ultimately resulting in the formation of metal oxide nanocrystals. 44 3 ] # 3 have nearly spherical shape and are well dispersed on TEM grids, with sufficient interparticle distances as shown in Fig. 1A-C and S1. † D TEM (particle diameter determined by TEM) ranged from 5 to 12.5 nm with polydispersity indexes between 0.14 and 0.21, except for 0. 33 (Fig. 1D). In the case of using the divalent Mn precursor, a statistically insignicant change in D TEM was observed when increasing the ratio of precursors. This is in agreement with what was reported by Garcia-Soriano et al. 45 2.  3 ], a small shi of the peaks towards a lower 2q value (closer to the reference peak of MnFe 2 O 4 ) was observed in the XRD patterns ( Fig. 2 and S4 †). Using Mn(acac) 3 (Fig. S4 †) caused a relocation of XRD peak positions closer to the reference peak positions of MnFe 2 O 4 (2q = 40.8°). This can be attributed to a further inclusion of Mn 3+ into the spinel iron oxide lattice due to the similar ionic radii between Mn 3+ and Fe 3+ (0.64Å for both (ref. 49)), which are smaller than that of Mn 2+ (0.80Å). 49 With an increase in the Mn precursor concentration, the slight broadening of full-width half maximum (FWHM) of the most intense XRD peaks (311), was observed, which implies a reduction in crystal size. 50 Calculated size (in diameters) is summarized in Table S1 3 ] led to a slight broadening in the 311 peaks, which indicated a possible alteration of the crystal size, 50,53 as determined by measuring the FWHM and summarized in Table S1. † 42,53 The calculated crystal size obtained from XRD of samples (6.5-7 and 5-7.5 nm for NPs prepared from divalent and trivalent Mn precursors, respectively, see PDI of crystal size in Fig. S6 †) were within the range of the average particle size derived from TEM. Therefore, these NPs were considered to be single crystalline. However, the XRD analysis indicated the presence of MnCO 3 for NPs synthesised with precursor ratios in the range of 1 # [Mn(acac) 2 or 3 ]/  [Fe(acac) 3 ] # 3, and the crystal sizes determined by XRD were also within the size range observed by TEM.

Characterization of Mn x Fe 3−x O 4 NFs
The preparation of Mn x Fe 3−x O 4 NFs was implemented through a modied solvothermal method, and the morphology was precisely regulated by varying the ratios between [Mn(acac) 3 ]/ [Fe(acac) 3 ] precursors as well as the reaction temperature as shown in Fig. 3A Fig. 4 shows diffraction patterns of Mn x Fe 3−x O 4 with crystal size of NFs being in the range 6-8 nm. The small crystal size in comparison to D TEM of NFs (Fig. 3) implied the formation of primary nanocrystals, which do not grow signicantly. The primary nanocrystals aggregated into larger secondary particles and coarsening, as shown in Fig. 3A-C and as described by Gavilan. 56 The generation of MnCO 3 accelerated the hydrolysis of the precursors and promoted the formation of larger oxide clusters. 54 In our case, nano-clusters were prepared in a single step which included the synthesis of nanoparticles and their coalescence. Shiing in the peak of 311 from the reference     3 ] ratios were equal to 0.14 and 0.33. The faster thermal decomposition of Mn(acac) 3 than Mn(acac) 2 (ref. 52) resulted in more Mn-rich NPs that were prepared by the trivalent Mn precursor than those prepared by the divalent Mn precursor.
Overall, at 250°C, the change in the oxidation state and the ratios between the precursors did not show a variation in the morphology of NPs, but it signicantly affected the Mn doping level. While at a synthesis temperature of 200°C, the oxidation state and the ratios between the precursors affected both the Mn doping level and resulted in different shapes of Mn x Fe 3−x O 4 NPs and NFs.

Functionalisation of NPs and NFs
The advantage of applying a small molecule like citrate as a ligand is that a smaller hydrodynamic radius (D HD ) of NPs can be obtained compared to polymeric ligands, which are reected in hydrodynamic size (Fig. S8 †). Yet, the hydrodynamic shell size of citrate-coated NPs was large enough to maintain a physical barrier leading to good dispersibility. The obtained stable dispersions of nano-colloids were attributed to the negative charges induced by the citrate 59 as determined by z-potentials (Fig. S9 †).
The most negative value of z-potential was observed for Mn The FTIR measurements, as shown in Fig. S10, † conrmed that TEG ligand was exchanged by trisodium citrate, similarly observed by Chakraborty et al. 60 Carboxylates exhibited strong absorptions for infra-red spectrum due to their characteristic asymmetric (at 1620-1560 cm −1 ) and symmetric carbonyl stretching (1440 to 1310 cm −1 ). Bands in the region of 1280-1027 cm −1 represented the deformation of C-H. [61][62][63]   (x = 0.2, 0.4 and 0.6) induced a progressive, positive impact on the adsorption efficiency of Cr 6+ . Since Mn 2+ ions have larger ionic radii than Co 2+ (0.8Å vs. 0.7Å), the increase in x turned the overcoming of energy barriers for ion exchange interaction more difficult. 68 Given that the physical mechanism of Cr 6+ adsorption on the surface of oxide was reported to be a combination of electrostatic interactions between charged oxides and Cr 6+ and ion exchange in the aqueous solution, 40  The adsorption of Cr 6+ by the selected citrate-coated adsorbents that showed the best Q e at pH 7 at room temperature can be described by Langmuir isotherm model as a function of the initial Cr 6+ concentrations (Fig. 5C-E). Hence, the surface of nano-sorbents has homogeneous energy distribution via a monolayer sorption process. The calculated maximum adsorption capacity (Q max ) by Langmuir isotherm model tted the results of Q e .  Fig. 6C and D, a peak of Fe 2p 3/2 for Fe 3+ was spotted at 710 eV, and its satellite appeared at 718 eV. The asymmetric peaks are situated at 723.6 eV, attributed to Fe 3+ 2p 1/2 . For Fe 2p 1/2 , another satellite peak was observed at 729.5 eV. The peak of Fe 2p 1/2 was wider and weaker than Fe 2p 3/2 peak, and the FWHM of Fe 2p 1/2 peak is smaller than that of Fe 2p 3/2 because of spin-orbit (j-j) coupling. From the calculated FWHM (Table  S3 †), the FWHM of Fe 2p peaks at 712 eV was slightly smaller than its counterpart at 712 eV, which can serve as an indicator for the presence of both Fe 2+ and Fe 3+ in these two samples. 76,78 This interpretation matched the elemental analysis results by ICP (Fig. 5A) for the formation of non-stoichiometric Mn x -Fe 3−x O 4 . Furthermore, the absence of the satellite peak at 732 eV, as shown in Fig. 6D

The lethal dose of Cr 6+ for the tested Shewanella
Shewanella bacterial species are considered metal-reducing and resistant bacteria. 26 In our work, results revealed that the minimum inhibition concentration (MIC) of Cr 6+ for the tested wild-type Shewanella (S. oneidensis and S. loihica PV-4, see the molecular identication (Table S4 †)) was 60 mg L −1 and 70 mg L −1 respectively, being slightly higher than what was reported previously. 26 For S. oneidensis JG1486 and JG3355 (molecular identication at ESI †), MICs were 20 mg L −1 and 5 mg L −1 , respectively. The bactericidal effect of Cr 6+ was documented because of being taken up by Shewanella intracellularly and caused cell lysis. In fact, the toxic effect of Cr 3+ appeared to be associated with extracellular interactions, leading to stress-associated cell morphology and then to a lethal effect. 26,[81][82][83] Before reacting with Cr 6+ , the wild-type S. oneidensis MR-1 and S. loihica PV-4 cells were reported to be regular small rod-shaped with smooth surfaces. 26,[81][82][83] Meanwhile, the bacterial cells changed to be atrophic with a shrunken-surface shape and crack formation was also observed aer the reaction. 26,81,83 However, S. loihica PV-4 cells were reported to be elongated and exhibited a rough surface upon exposure to Cr 6+ , 82 which can explain why S. loihica PV-demonstrated higher resistance and reduction ability for Cr 6+ . As hazardous metal ions could damage microbial DNA when they entered the cells, extracellular reduction benetted Shewanella for their survival. 82 The low resistance of mutants was due to the inability of bioreduction of Cr 6+ for JG1486, 84

Bio-reduction of Cr 6+ by tested Shewanella
For safe removal of Cr 6+ , such hexavalent cations should be detoxied by reduction to Cr 3+ . The capability of S. oneidensis 83 to respire Cr 6+ was affected by the initial concentration of the heavy metal. 83,88,89 These results were credited for the chromate dose-dependent toxicity, which causes growth and viability inhibition. 83,88 This occurred in the presence of Cr 6+ alone 83,88,89 or in the presence of goethite and humic acid 34 or ferric oxyhydroxide mediators. 83,88 Hence, our experiments were designed at a high concentration of Cr 6+ , i.e. sub-lethal dose.
Our results (Fig. 7B) revealed a signicant drop in the concentration of Cr 6+ in media supplemented by both wild-type of Shewanella which included the strain of interest (S. oneidensis MR-1) and positive control (S. loihica PV-4); this was attributed to the respiration of Cr 6+ into Cr 3+ form 81,90,91 or bio-sorption 92,93 by bacterial cells. The drop in the concentration of Cr 6+ in the medium exposed to S. oneidensis JG1486 and JG3355 was signicantly lower than those supplemented by both wild type bacteria.
The tested Shewanella oxidized lactate (electron source), and the liberated electrons are transferred via the respiratory chain to be directed to an externally available terminal electron acceptor (Cr 6+ ). The redox potential of Cr 6+ (1.33 V vs. standard hydrogen electrode; SHE) has been reported to be higher than the redox potential of oxygen (1.23 V) and the electron source (−0.19 V). 94 So, Cr 6+ was considered a favourable electron acceptor for bacteria in the process of respiration as bacteria gain more energy. 94 Cr 6+ can be reduced extracellularly 82 and also transported into the cell interior and then reduced in the cytoplasm. 82 The ability of Shewanella to transfer electrons to metal ions was known to take place via one of four porin-cytochrome conduits; the MtrCAB complex, 95 the MtrFED complex, 96 the DmsEFA dimethyl sulfoxide reductase system 97 and the SO4359-SO4360 system. 96 The superiority of S. loihica PV-4 in the respiration of Cr 6+ in our experiment was thanks to their higher content of ctype cytochrome genes in the metal reductase-containing locus than S. oneidensis MR-1. 98 S. oneidensis JG1486 (DmtrCAB/ DmtrFED/DomcA/DdmsE/DSO4360/DcctA/DrecA) lacks the responsible genes for extracellular metal reduction. 84 The recombination between the expression of outer membrane cytochromes (controlled by lac promoters) and periplasmic electron carriers was stopped by the deletion of RecA gene. 84 Such mutants showed the lowest removal of Cr 6+ as a result of the inability of bioreduction. 84 S. oneidensis JG3355 lacked both ClpX and ClpP genes. 86 The role of ClpXP has been revealed for regulating Fe 2+ stress in anaerobic bacteria 86 and stress regulation (ClpP) in response to 24 h Cr 6+ exposure. 85 Therefore, the inability of S. oneidensis JG3355 to respire metal could be the result of losing the bacterial viability as was reported before [29][30][31] and presented in Fig. 7A.

Enhancement of respiration of Cr 6+ by Shewanella in the presence of selected materials
The respiration of Cr 6+ (at sub-lethal concentration) was improved in the tested groups supplied by citrate alone (only 0.83 folds), as illustrated in Fig. 7B. Bencheikh-Latmani et al. 87 explained a similar observation as a result of the possible complexation between the product of bio-reduction (Cr 3+ ) and citrate, which consequently limits the availability of the toxic metal to bacterial cells. 87 In the presence of a sublethal concentration of Cr 6+ , the alive cell extent was the highest in the group of citrate coated Mn 0. microbial survival, which was positively related to enhanced Cr 6+ bio-reduction by 2.5-3.6 folds. The increase in the percentage of bacterial viability may be attributed to the adsorption of Cr 6+ by NPs, which led to the decrease of stress on the strains themselves. In addition, the possible continuous adsorption-desorption rate of Cr 6+ was based on Langmuir adsorption isotherm of the equilibrium between the adsorbate and adsorbent system (Fig. 5C-E).
The presence of manganese in the chemical structure of NPs improved the antioxidant activity and, in turn, the viability of cells and the ability to respire metal. 99 Mn 2+ ions can act as antioxidants which helps enzymatic systems to act against oxidative stress. For Fe-rich and Mn-poor cells such as S. oneidensis MR-1, death at low doses of ionizing radiation might not be caused by DNA damage inicted during irradiation but instead by the release of Fe 2+ and the subsequently formed toxic by-product of energy-metabolism aer irradiation. 100 The electron transfer from cells to the acceptor 101 occurred via redox cycling of the electron-donating and accepting functional groups via direct electron transfer through NPs. The affinity of MnFe 2 O 4 NPs to bind proteins on the bacterial outer membrane can improve the contact area between a single bacterium cell and an external electron acceptor. 37 There are some explanations for NP-enhanced bio-reduction of Cr 6+ to Cr 3+ by S. oneidensis MR-1; however, the exact mechanism is not fully unravelled. 102 NPs can act as a bridge between the bacterial cell and Cr 6+ to promote electron transfer. 102  Colonial growth from all cultures was stored aer being preserved on cryobeads at −20°C.  3 ] equal to 0, 0.14, 0.33, 0.6, 1, 1.66 and 3 were mixed in 20 mL of TEG as a solvent. The resulting mixture was processed by vortexing for 10 min, then sonicated for 30 min to be homogenized, followed by its transfer into a 45 mL Teon-lined stainless-steel autoclave. The autoclave was placed in an oven (Memmert, model UFP400) at room temperature, and the reaction temperature was raised for 30 min to 250°C, which was maintained for 6 h. For the ratio [Mn(acac) 2 or 3 ]/[Fe(acac) 2 ] equal to 0 and 0.33, the temperature was raised up to 200°C only. The resulting black dispersion was separated by a magnet and washed with 1 : 10 v/v of acetone, followed by ethanol and water three times for each solvent. Then, the nanomaterials were ready for characterization and functionalization.

Synthetic methodology
3 In order to exchange the initial ligand TEG, 1 mL of the dispersions of the prepared nanomaterials and 10 mL of 1 M aqueous tri-sodium citrate solution was mixed for 48 h at room temperature under stirring. Immobilization of citrate on the surface of the commercially available Mn 3 O 4 (used as Mn-rich and iron-free ferrite control) was carried out by dispersing 0.1 g of the metal oxide in 10 mL of 1 M aqueous tri-sodium citrate solution under similar mentioned conditions. The functionalized nanostructures were puried by magnetic separation, followed by washing with acetone three times, aer that ethanol washing was performed three times, and nally the particles were dispersed in de-ionized water.
3.2.4 Characterization of Mn x Fe 3−x O 4 NS. For the prepared nanomaterials, the shape and diameter of the core were determined by a JEOL JEM 1200-EX microscope operating at an acceleration voltage of 120 kV. The crystal phase and the average crystallite size were analyzed by X-ray diffractometer (XRD; PANalytical XPERT PRO MPD) coupled with Co Ka radiation source (l = 1.789Å) and an X'Celerator detector operated at 40 kV and 40 mA. The crystalline phases were identied using the International Centre for Diffraction Data Powder Diffraction File (ICDD PDF) database. The crystal domain size (D XRD ) was calculated using Scherrer's equation at the most intense X-ray peaks (311). The chemical composition of Mn x Fe 3−x O 4 was determined by an Optima 3100 XL PerkinElmer Inductively Coupled Plasma Atomic Emission (ICP-AES) spectrometer. The oxidation states of Mn and Fe in selected Mn x Fe 3−x O 4 NPs that produce the lowest and the highest Q e were analysed by X-ray photoelectron spectrometry (XPS), a Kratos Analytical AXIS Ultra DLD system with aluminium monochromatic X-ray source (Al Ka = 1486.6 eV), under ultra-high vacuum conditions (10 −9 Torr). The experimental curves were best tted by a combination of Gaussian (70%) and Lorentzian (30%) distributions. Over the range 150-2000 cm −1 , Raman spectra were collected for powder samples of the selected Mn x Fe 3−x O 4 materials prepared using [Mn(acac) 2 ]/[Fe(acac) 3 ] equal to 0 and 0.33 at 250°C for 6 h. A Renishaw InVia micro-Raman spectrometer was used and experiments were conducted at room temperature and excited by green Ar-laser for excitation (l = 514.5 nm) of photon energy 2.4 eV and diffraction grating 2400 grating.
3.2.5 Surface characterization of Mn x Fe 3−x O 4 NS. In order to study the surface coordination of the capping agents, a Per-kinElmer FTIR; Spectrum 100 instrument with a Ge/Ge universal attenuated total reectance (ATR) was used. The samples were prepared by air drying at room temperature overnight to yield a ne powder and then directly placed on an ATR crystal. The measurement window for the recorded spectra was in the range 4000-600 cm −1 , with a 2 cm −1 resolution, using 40 scan accumulation. The hydrodynamic diameter (D HD ) of citrate-functionalized NPs was evaluated by DLS measurements performed with a Nanosizer ZS instrument (He-Ne 633 nm laser) from Malvern Instruments Ltd, Worcestershire, UK). The z-potentials of the functionalized NPs were determined using a disposable capillary cell (DTS1070) at 25°C by DLS. For iron content quantication of the functionalized nanomaterials dispersed in water, a colourimetric phenanthroline method was applied for the acid-digested tested agent. 105 The concentration of Mn in the Mn 3 O 4 dispersion was estimated by inductively coupled plasma atomic emission spectroscopy (ICP-AES) spectrometer.
3.2.6 Measurement of the Cr 6+ adsorption capacity of nanostructures. Equal volumes of aqueous dispersed citratecoated nanostructures (adsorbents) and Cr 6+ aqueous solution were mixed and incubated for 6 h at pH 7 at room temperature. Both citrate-capped Fe 3 O 4 NPs and Mn 3 O 4 NPs were used as the control group, and 0.01 M of trisodium citrate served as a background. The amounts of adsorbed Cr 6+ per unit mass of adsorbent (Q e ; mg g −1 ) were calculated using eqn (1): C i and C ad were the initial concentration of Cr 6+ , which was equal to 30 mg L −1 , and the concentration of Cr 6+ in the solution at the equilibrium, respectively. The total volume of the reactants mixture (V) was 2 mL, and the mass of adsorbents (m) was represented in g with respect to Mn mass fractions. The concentration of Cr 6+ was quantied by measuring the optical density of the colour generated by the Cr 3+ DPC (diphenylcarbazide) complex method at l 545 nm. 106 O 4 , the mass of the adsorbent was calculated in respect to both Fe and Mn fractions which were 0.68 and 0.05, respectively. Adsorption isotherms were tted by both Langmuir and Freundlich models. 108 3.2.7 Bacteria identication. The genome DNA of bacteria was extracted by boiling a single colony in ultra-pure water for 10 min at 95°C. To deliver the highest DNA yield from the tested colony, FastDNA Spin Kit for Soil was applied to the boiled broth following the manufacturer's instructions. Using a Qubit 3.0 uorometer (Life Technologies, UK), the quantity of DNA was evaluated using Qubit dsDNA broad range (2 to 1000 ng) assay kit from Invitrogen (UK). 16S rRNA genes of all strains were amplied by polymerase chain reaction (PCR) using pair primers; 1369F and 1492R primers and Luna Universal qPCR Master Mix. 109 Using thermocycler (Cole-Parmer), cycling conditions included initial denaturation at 94°C for 5 min, followed by 30 cycles of denaturation at 94°C for 45 s, heating at 52°C for 45 s and extension at 72°C for 90 s. The nal extension was tested at 72°C for 90 s. Both the purication of amplied PCR products and Sanger sequencing were implemented using the commercial service of Source Bioscience, Cambridge, UK. The sequenced data were assigned for matching identity for species with the highest tting 96-100% by nucleotide BLAST (Basic Local Alignment Search Tool) from the National Centre for Biotechnology Information (NCBI) database (https:// blast.ncbi.nlm.nih.gov).
3.2.8 Minimum inhibition concentration of Cr 6+ . To assess the impact of Cr 6+ on the viability of the tested Shewanella, a Guava easyCyte® ow cytometer (Merck, UK) was used. 10 mL of homogeneous bacterial cell suspensions with OD measured at the wavelength of 600 nm equal to 0.1 was added to 80 mL of M9 minimal salts (2×) medium. 110 This medium was supplemented by 20 mM sodium lactate as a sole electron source, 5 mL L −1 each of vitamins and minerals and pH was adjusted to 7.2 by 10 mM HEPES buffer. 111 The viability of cells was counted in response to serial dilutions of Cr 6+ (1 to 100 mg L −1 ) as a terminal electron source alone. Sodium fumarate (20 mM) was used as an alternative terminal electron acceptor to Cr 6+ . In all cases, media were purged with nitrogen gas for 5 min aer bacterial inoculation. The proportion of live cells was quantied in relation to the total number of cells via the Live Dead Bac-Light Bacterial viability assay. Populations of living and/or dead bacteria were gated according to uorescence minus one (FMO) controls using single stains of SYTO9 and propidium iodide (PI). 112 All data are expressed as means ± standard deviation. The MIC of any agent was dened as its lowest concentration that inhibits the growth of bacteria aer overnight incubation. 3