Introduction
Microbes play a pivotal role in global geochemical cycling, biotransformations, biodegradation and the transport of contaminants and nutrients in natural environments (Zheng et al., Reference Zheng, Li, Long, Zhao, Jia, Li, Wang, Wang, Lu and Zhang2018; Borch et al., Reference Borch, Kretzschmar, Kappler, Cappellen, Ginder-Vogel and Voegelin2010). Their metabolic activities significantly influence surface mineralogy by driving the formation of biogenic minerals, commonly utilising transition metals to meet energy demands or to adapt to environmental stress (Cosmidis and Benzerara, Reference Cosmidis and Benzerara2022; Soldatova et al., Reference Soldatova, Tao, Romano, Stich, Casey, Britt, Tebo and Spiro2017; Rajkumar and Freitas, Reference Rajkumar and Freitas2008). Biogenic minerals commonly form in close association with bacteria, as either intracellular inclusions or extracellular precipitates on or near the cell surface. They typically exhibit distinct characteristics compared with abiogenic phases, including smaller particle size, lower crystallinity, greater surface area and higher chemical reactivity. As a result, biogenic minerals can strongly influence the biogeochemical cycling of contaminants and nutrients in natural environments (Kappler et al., Reference Kappler, Thompson and Mansor2023; Glodowska et al., Reference Glodowska, Schneider, Eiche, Kontny, Neumann, Straub, Kleindienst and Kappler2021; Abramov et al., Reference Abramov, Tejada, Grimm, Schädler, Bulaev, Tomaszewski, Byrne, Straub, Thorwarth, Amils and Kleindienst2020; Piepenbrock et al., Reference Piepenbrock, Schröder and Kappler2014; Muehe et al., Reference Muehe, Scheer, Daus and Kappler2013; Posth et al., Reference Posth, Canfield and Kappler2014). Biogenic minerals also play significant roles in contaminant remediation across engineered environments, including wastewater treatment systems and industrial or mining-impacted sites (Staicu et al., Reference Staicu, van Hullebusch and Ackerson2021).
Beyond its well-established role as a vital trace element for enzymatic and metabolic functions in humans, animals and plants (Chen et al., Reference Chen, Bornhorst and Aschner2018; Li and Yang, Reference Li and Yang2018; Lucchini et al., Reference Lucchini, Aschner, Kim and Šarić2015), manganese (Mn) plays a central role in microbial ecology and redox-driven biogeochemical processes. Many bacteria not only harness Mn for growth and protection against oxidative stress (Waters, Reference Waters2020; Culotta and Daly, Reference Culotta and Daly2013), but also actively oxidise Mn(II) to Mn(III/IV) oxides, generating reactive mineral phases (e.g. birnessite and pyrolusite) that influence elemental cycling and environmental contaminant dynamics. Enzymes with Mn cofactors are critical in catalysing nucleic acid and carbohydrate metabolism, underscoring the biochemical versatility and ecological importance of manganese (Martin et al., Reference Martin, Lisher, Winkler and Giedroc2017; Juttukonda and Skaar, Reference Juttukonda and Skaar2015). Certain bacteria oxidise Mn(II) both to mitigate potential toxicity and to harvest energy, precipitating Mn(III,IV) oxide minerals in the process (Yu and Leadbetter, Reference Yu and Leadbetter2020; Tebo et al., Reference Tebo, Bargar, Clement, Dick, Murray, Parker, Verity and Webb2004). This biologically mediated Mn oxidation is predominantly enzymatic. Despite the phylogenetic diversity of Mn(II)-oxidising microorganisms, many employ a conserved multicopper oxidase-like enzyme system to catalyse this redox transformation (Cai et al., Reference Cai, Yang, Qiu, Bi, Tian and Bi2023; Tebo et al., Reference Tebo, Johnson, McCarthy and Templeton2005). The oxidation of Mn(II) to Mn(IV) typically proceeds via a Mn(III) intermediate, which can exist transiently in solution or be stabilised through enzyme binding (Webb et al., Reference Webb, Dick, Bargar and Tebo2005; Tebo et al., Reference Tebo, Bargar, Clement, Dick, Murray, Parker, Verity and Webb2004). While Mn(II) oxidation by molecular oxygen is thermodynamically favorable (Diem and Stumm, Reference Diem and Stumm1984), it is kinetically slow under environmental conditions, making microbial catalysis the dominant pathway in most natural settings (Handley and Lloyd, Reference Handley and Lloyd2013; Tebo et al., Reference Tebo, Johnson, McCarthy and Templeton2005; Nealson et al., Reference Nealson, Tebo and Rosson1988). Microbial Mn oxidation is thus central to the formation of Mn oxides and the regulation of manganese redox cycling in aquatic and terrestrial environments. The efficiency and outcome of this process are strongly influenced by environmental parameters such as temperature, salinity and oxygen availability (Hansel, Reference Hansel2017).
In the environment, microbial Mn(II) oxidation is widespread across different microbial phylogenetic domains, including bacteria, archaea and eukarya (Hansel, Reference Hansel2017). Among known Mn oxidisers, Pseudomonas putida MnB1 (P. putida), a Gram-negative Gammaproteobacterium, has emerged as a model organism because of its ubiquity in soil and freshwater environments, and its robustness under diverse stresses, including heavy metal contamination (Tong et al., Reference Tong, Zhao, Sun, Li, Liu and Yuan2022; Galezowski et al., Reference Galezowski, Recham, Larcher, Miot, Skouri-Panet and Guyot2020; Zheng et al., Reference Zheng, Li, Long, Zhao, Jia, Li, Wang, Wang, Lu and Zhang2018; Toner et al., Reference Toner, Fakra, Villalobos, Warwick and Sposito2005; Villalobos et al., Reference Villalobos, Toner, Bargar and Sposito2003; Caspi et al., Reference Caspi, Tebo and Haygood1998). The biochemical mechanism of direct oxidation and acquisition of Mn(II) by P. putida is thought to involve a multicopper oxidase complex (Francis and Tebo, Reference Francis and Tebo2001; Tebo et al., Reference Tebo, Johnson, McCarthy and Templeton2005; Miller et al., Reference Miller, Dionísio, Sequeira-Braga, Hernández-Mariné, Afonso, Muralha, Herrera, Raabe, Fernandez-Cortes and Cuezva2012) and Mn transporters (Courville et al., Reference Courville, Chaloupka and Cellier2006; Nevo and Nelson, Reference Nevo and Nelson2006). Additionally, the involvement of extracellular superoxide radicals has been identified as an indirect Mn(II) oxidation pathway in P. putida (Miller et al., Reference Miller, Dionísio, Sequeira-Braga, Hernández-Mariné, Afonso, Muralha, Herrera, Raabe, Fernandez-Cortes and Cuezva2012; Learman et al., Reference Learman, Voelker, Vazquez-Rodriguez and Hansel2011). Despite extensive research, the exact mechanistic details of Mn(II) oxidation and extracellular mineral deposition remain incompletely understood (Webb et al., Reference Webb, Dick, Bargar and Tebo2005; Learman et al., Reference Learman, Voelker, Vazquez-Rodriguez and Hansel2011). Nonetheless, the efficiency of microbial Mn(II) oxidation occurring at rates several orders of magnitude faster than abiotic oxidation by oxygen (O2) (Tebo et al., Reference Tebo, Bargar, Clement, Dick, Murray, Parker, Verity and Webb2004) emphasises its dominant role in the formation of Mn oxide minerals in natural aquatic systems.
The geochemistry of Mn is interlinked with iron (Fe), as these two metals commonly coexist in various forms, including as hydrated ions, in soluble complexes and in particles across diverse environments such as waters, soils and sediments (Kappler et al., Reference Kappler, Bryce, Mansor, Lueder, Byrne and Swanner2021; Liu et al., Reference Liu, Chen, Yang, Wei, Laipan, Zhu, He and Hochella2022; Bai et al., Reference Bai, Yang, Liang and Qu2016). Iron, the most abundant transition metal in the Earth’s crust, is ubiquitous in surface and groundwater environments. Fe(III) oxides form through the oxidative precipitation of Fe(II) under aerated conditions, primarily via chemical oxidation, although microbial processes can also facilitate the oxidation of Fe(II) to Fe(III) under anoxic conditions, particularly via nitrate-reducing and photoautotrophic bacterial metabolisms (Kappler et al., Reference Kappler, Bryce, Mansor, Lueder, Byrne and Swanner2021; Kappler and Straub, Reference Kappler and Straub2005). Moreover, microorganisms can also passively sorb Fe onto their cell walls, leading to Fe oxide nucleation on the cell surface. Microbial extracellular polymeric substances (EPS) can also act as templates for Fe oxide nucleation (Fortin and Langley, Reference Fortin and Langley2005). Fe and Mn oxides commonly coexist in environments and co-precipitate as Fe-Mn binary oxides (Zhao et al., Reference Zhao, Xie, Li, Chen and Zhong2022) and mixed oxides. Understanding the formation of biogenic Fe-Mn mixed oxides is critical because these minerals play essential roles in controlling the speciation, transport and fate of several redox-sensitive metals, thereby regulating their biogeochemical cycles (Zheng et al., Reference Zheng, Jiang, Li and Yang2024).
This research investigates a microbially driven pathway for Fe-Mn mixed oxides formation, emphasising how biogenic Mn oxides mediate Fe(II) oxidation under micro-oxic conditions, mimicking geochemical gradients at the oxic-anoxic interface. While numerous studies have examined the chemical synthesis of binary or Fe and Mn mixed oxides (Hu et al., Reference Hu, Zhang, Wang, Chen and Chen2022; Chen et al., Reference Chen, Liu, Li, Wu, Chen, Liu, Zhang, Xu, Li and Li2020; Zheng et al., Reference Zheng, Hou, Hartley, Ren, Wang, Tu and Tan2020, Reference Zheng, Tu, Hou, Ni, Wang, Ren, Wang, Cao, Xiong and Tan2021; Yang et al., Reference Yang, Zhou, Lou, Zhou, Liu, Li, Baig and Xu2018; Zhang et al., Reference Zhang, Qu, Liu, Liu and Wu2007, Reference Zhang, Liu, Liu, Qu and Liu2014; Liu et al., Reference Liu, Yang, Kang, Fan and Qu2012; Xu et al., Reference Xu, Wang, Liu, Zhao and Qu2011; Chang et al., Reference Chang, Qu, Liu, Liu and Zhao2009), investigations into the microbial processes underlying such mineral formation remain limited. Notably Bai et al. (Reference Bai, Yang, Liang and Qu2016) demonstrated the in situ formation of Fe-Mn binary oxides under oxic conditions using Pseudomonas sp. QJX-1 and Tong et al. (Reference Tong, Zhao, Sun, Li, Liu and Yuan2022) examined the effect of Fe(II) oxidation by O2 on Mn(II) oxidation by P. putida. However, the redox interaction between Fe(II) and biogenic Mn oxides under micro-oxic conditions has not been previously addressed. This research examines a two-step process in which Mn(II) is first bio-oxidised by P. putida under oxic conditions to form biogenic Mn oxides, after which Fe(II) oxidation under low-oxygen conditions is interpreted as being most consistent with reaction with these pre-formed Mn oxides. The study provides detailed mineralogical characterisation of the resulting Fe-Mn mixed oxides and evaluates their environmental implications in redox-dynamic aquatic systems.
Materials and methods
Synthesis of biogenic Fe-Mn mixed oxides
Biogenic Fe-Mn mixed oxides were produced in a two-step process: (1) synthesis of biogenic Mn oxides using P. putida; and (2) synthesis of biogenic Fe-Mn mixed oxides. Culturing and synthesis of biogenic Mn oxides were performed following a protocol slightly modified after Galezowski et al. (Reference Galezowski, Recham, Larcher, Miot, Skouri-Panet and Guyot2020). All solutions prepared and used in this experiment were sterilised either by autoclaving or filtering through 0.22 μm membrane filters. To culture the bacteria, a growth medium was prepared by dissolving nutrient broth (Sigma-Aldrich) in MilliQ water (8 g L–1, composed of 3 g L–1 beef extract and 5 g L–1 peptone), to which 50 μM MnSO4•H2O (manganese sulfate monohydrate) was added. Cells of P. putida strain MnB1 (ATCC 23483) were inoculated in the growth medium and cultured overnight under oxic conditions on a horizontal shaker at a speed of 120 rpm at 30°C and pH 6.8. On the next day, the cells were separated from the growth medium by centrifuging at 5000 rpm for 5 min. Cells were then washed with 10 mM 4-(2-hydroxyethyl) piperazine-1-ethane-sulfonic acid (HEPES) and 10 mM sodium chloride (NaCl) three times to rinse off any growth medium from the separated cells.
In the first step, a mineralisation medium composed of 10 mM HEPES, 2 mM ammonium sulfate (NH4)2SO4), 0.7 mM NaCl and 1 mM glucose was prepared, and the pH was adjusted to 7.4. Then, the rinsed P. putida cells were transferred at a cell density of 2.3 × 108 cells mL–1 to several 250 mL Schott bottles, each containing 150 mL of mineralisation medium. Next, 5 mL of a 5.58 mM Mn(II) (MnSO4•H2O) solution was added to each bottle to achieve a Mn(II) concentration of 180 μM, and the bottles, with lids loosely sealed, were placed on a horizontal shaker at a speed of 120 rpm at 30°C. Formation of blackish-grey precipitates was observed in the inoculated experimental media from the second day of the experiment, interpreted as the bio-oxidation of Mn(II) to poorly soluble Mn(III)/(IV) by P. putida. To assess spontaneous chemical formation of Mn oxide, an abiotic control was prepared by introducing Mn(II) at 180 μM in the same mineralisation medium, without inoculation with P. putida, and under the same experimental conditions mentioned above. After 12 days of Mn(II) bio-oxidation and mineralisation, each Schott bottle was sparged with nitrogen (N2) gas for 5 min to reduce dissolved oxygen prior to the second step. Because dissolved oxygen was not measured directly, the subsequent condition is referred to here operationally as low-oxygen or micro-oxic rather than as a quantitatively defined oxygen regime. To maintain these conditions, the reaction vessels were not hermetically sealed but were loosely capped to allow for a slow, continuous influx of atmospheric O2 by diffusion. This setup was intended to maintain a low but non-zero oxygen supply by diffusion, without establishing fully oxic conditions to support aerobic metabolism of P. putida without achieving full saturation. In each Schott bottle, Fe(II) (as FeSO4•7H2O) was added to the solution at a final concentration of 200 μM and the experiments were then continued for 6 days at 30°C, with agitation at a speed of 120 rpm. All solutions and cell handling were performed in a biosafety hood to prevent microbial contamination and maintain sterile conditions in the controls. After synthesis, minerals were separated from the mineralisation media by centrifuging at 8000 rpm for 10 min. To minimise the risk of secondary mineral formation from residual solutes, harvested precipitates were washed three times with oxic milli-Q water. This step was intended to remove the bulk of the remaining dissolved and weakly associated Fe and Mn. All experiments were conducted in independent quadruplicates (n = 4).
Chemical analyses
The concentrations of Fe and Mn in experimental solutions were determined by inductively coupled plasma mass spectrometry (ICP-MS) analysis using a PerkinElmer NexION 350D ICP mass spectrometer. A river water standard for trace metals (SLRS-5) from the National Research Council Canada was used as reference material for Fe and Mn concentrations, using rhodium as the internal standard. For the analysis, 200 μL aliquots were collected from the experimental solution immediately after the addition and mixing of Mn(II) in the solution, filtered to 0.22 μm and preserved in 2% nitric acid (HNO3) to determine the initial concentration of total dissolved Mn in the solution for bio-mineralisation. Sample aliquots were also collected from the experimental solutions before and immediately after the addition and mixing of Fe(II), as well as after 6 days of Fe(II) oxidation. The samples were filtered through 0.22 μm filters and dissolved in 2% HNO3 for ICP-MS analysis to determine the initial concentration of total dissolved Fe in the solution and concentrations of total dissolved Fe and Mn remaining in the solution after reaction. Concentrations of Mn and Fe in particulates were found by subtracting the concentrations of total dissolved Mn and Fe remaining in the mineralisation media from the concentrations of total dissolved Mn and Fe initially added. The results presented are the mean values of measurements performed on the experimental quadruplicates.
Mineralogical and morphological characterisation
XRD analysis
A portion of the biogenic solid was suspended in Milli-Q water, deposited onto a silicon low-background disk and air-dried overnight for X-ray diffraction (XRD) analysis. The XRD analysis was performed the next day using a Panalytical Empyrean Series 2 diffractometer with a cobalt (Co) anode Kα radiation source (Kα1 = 1.78901 Å and Kα2 = 1.79290 Å). The diffractometer was set up in a standard reflection geometry using a fixed slit at continuous scanning from 4° to 80° over a total estimated duration of 6.5 h in the 2θ position with a spinning holder (4 s per revolution) and a step size of 0.026°. Background noise was subtracted by manual fitting in HighScore software, which involves a set of base points, and subtracting that modeled line from experimental data.
SEM analysis
Scanning electron microscopy (SEM) imaging was used to examine the morphological features of the biogenic Fe-Mn mixed oxides using a high-resolution Zeiss Sigma 300 field emission gun scanning electron microscope at the Sir William Dunn School of Pathology, University of Oxford. Samples were prepared by filtering 2 mL of the homogeneously mixed mineralisation medium containing biogenic Fe-Mn mixed oxides through a 0.22 μm filter to retain the minerals. To minimise artifacts from the oxidation of residual solutes during preparation, harvested mineral precipitates were washed three times with deionised water to remove soluble ions. The samples were then air dried. While this method successfully preserves mineral distribution, we acknowledge that air drying leads to the dehydration and structural collapse of EPS, and morphology was interpreted with this limitation in mind. A small section of the filter containing biogenic minerals was affixed to a 13 mm aluminum stub covered with carbon tape. The stub was allowed to air dry overnight before being gold-coated for imaging. For SEM analysis, samples were coated with a 15 nm gold layer (Quorum Q150R ES, sputter current 20 mA, tooling factor 2.30) to ensure sufficient conductivity and minimise charging during imaging. While we recognise that this coating thickness may exceed the dimensions of the finest primary Fe-bearing nanoparticles, it provided the stability necessary to visualise the larger-scale mineral-cell associations. High-resolution imaging of individual mineral particles was achieved via scanning transmission electron microscopy (STEM)/transmission electron microscopy (TEM) as described in the following sections. Scanning electron microscopy images were acquired by the electron microscope operating in the secondary electron mode with an acceleration voltage of 2 kV and at a working distance of 8−9 mm.
STEM and EDX spectroscopy analyses
Scanning transmission electron microscopy (STEM) and energy dispersive X-ray spectroscopy (EDX) analyses of the biogenic Fe-Mn mixed oxides were performed at the David Cockayne Centre for Electron Microscopy of the Department of Materials at the University of Oxford. Scanning transmission electron microscopy images were acquired using a JEOL ARM200F equipped with a cold-field emission source and Cs-probe corrector, operating at 200 kV. The imaging conditions were set up to acquire high-angle annular dark field and bright field (BF) STEM images simultaneously. Scanning transmission electron microscopy EDX spectroscopy was performed in the same instrument, equipped with a 100 mm2 Centurion EDX detector. For the analyses, the separated particulate samples were diluted into Milli-Q water and a drop of the suspension (~3 μL) was placed on a 3 mm copper (Cu) TEM grid with formvar support. The grids were allowed to air dry overnight before analysis.
TEM-SAED analyses
Transmission electron microscopy selected area electron diffraction (TEM-SAED) was performed using a JEOL JEM-2100, 200 kV analytical electron microscope equipped with a Tungsten filament electron gun in combination with SAED. The SAED patterns were processed using ImageJ software (version 2.9.0/1.53t). Samples for TEM-SAED analysis were prepared in the same way as for STEM analysis (see previous section), but lacey carbon film-supported TEM grids were used instead of formvar supported grids.
Results and discussion
Synthesis of biogenic Fe-Mn mixed oxides
Bio-mineralisation of Mn oxides by P. putida was carried out over 12 days. A characteristic blackish-grey mineral deposit appeared at the bottom of the Schott bottles by the second day, indicating the bio-oxidation of Mn(II) to Mn(III) or Mn(IV). The initial concentration of Mn(II) introduced into the mineralisation medium was 187.5 μM. After 12 days, the concentration of remaining dissolved Mn was 60.8 (±0.6) μM, indicating that 67.6% of the added Mn had been bio-oxidised. No Mn oxides formation was observed in the abiotic control experiment.
Following 12 days of Mn oxide bio-mineralisation, Fe(II) was added to the medium, reaching a dissolved Fe concentration of 210.4 (±0.8) μM. The timing of Fe(II) addition was chosen carefully to follow the biosynthesis of biogenic Mn oxide particles, thereby avoiding potential Fe(II) toxicity to P. putida, which could have inhibited biogenic Mn oxide formation (Tong et al., Reference Tong, Zhao, Sun, Li, Liu and Yuan2022). To minimise chemical oxidation of Fe(II) by dissolved O2 and instead favor oxidation by biogenic Mn oxides, the medium was sparged with N2 gas for 5 min prior to Fe(II) addition. Biogenic Mn oxides are known to be among the strongest naturally occurring oxidants, exhibiting high redox reactivity (Fischel et al., Reference Fischel, Clarke and Sparks2024; Tebo et al., Reference Tebo, Bargar, Clement, Dick, Murray, Parker, Verity and Webb2004). Furthermore, chemical oxidation of Fe(II) by O2 proceeds slowly at low oxygen concentrations (Emerson, Reference Emerson and Lovley2000). Under these intended low-oxygen conditions, Fe(II) oxidation was interpreted as being most consistent with reaction with the pre-formed biogenic Mn oxides, although contributions from residual O2 and indirect cell-associated processes cannot be excluded fully. The Fe(II) oxidation and subsequent mineralisation process was then continued for an additional 6 days, extending the total experimental period to 18 days (Fig. 1). Oxidation of Fe(II) to Fe(III) by Mn oxides leads to the precipitation of poorly soluble Fe(III) mineral phases, as described by the following reaction:
where MnO2 represents Mn(III, IV) oxides such as pyrolusite or birnessite, and FeOOH denotes possible Fe(III) oxyhydroxides, such as ferrihydrite or goethite, although the exact mineral stoichiometry may vary depending on environmental conditions (Liu et al., Reference Liu, Chen, Yang, Wei, Laipan, Zhu, He and Hochella2022). Co-precipitation or physical mixing of these Fe and Mn mineral phases can result in the formation of Fe-Mn mixed oxides, defined here as co-occurring Fe- and Mn-bearing mineral phases without implying structural integration at the crystal lattice level.
Percentages of Mn and Fe mineralisation over 18 days calculated from four independent replicates. Error bars have been calculated as the standard deviation of four replicates.
Figure 1. Long description
The figure consists of two panels.
Left panel: A photo shows five glass bottles with blue caps containing a cloudy liquid and dark sediment at the bottom. A label below identifies these as Synthesized biogenic Fe-Mn mixed oxides.
Right panel: A line graph with two Y-axes. The X-axis is Mineralization Time (Days) from 0 to 20. The primary Y-axis (left) is Dissolved M n and F e Concentration in Solution (mu M) from 0 to 250. The secondary Y-axis (right) is percentage of Mineralization from 0 to 100.
Data trends:
* Dissolved M n Concentration (blue line): Starts at approximately 190 mu M on day 1 and decreases linearly to 0 mu M by day 18.
* M n percentage Mineralization (green line): Starts at 0 percent on day 1 and increases steadily to 100 percent by day 18.
* F e (I I) addition: Indicated by a horizontal arrow pointing to day 14.
* Dissolved F e Concentration (red line): Appears at day 14 at approximately 210 mu M and drops sharply to about 40 mu M by day 18.
* F e percentage Mineralization (orange line): Starts at 0 percent on day 14 and increases sharply to approximately 82 percent by day 18.
After 18 days of mineralisation, more than 99% of Mn(II) (186.2 ± 1.8 μM) had been removed from solution and associated with the solid phase. Although Mn(II) can also be subject to surface sorption, the observed sustained decline in dissolved Mn over this timeframe is most consistent with continued biologically mediated Mn transformation within the system. At circumneutral pH and under the intended low-oxygen conditions following N2 sparging, purely abiotic oxidation of dissolved Mn(II) by O2 would be expected to be strongly limited. However, because a dedicated post-sparging control without Fe(II) addition was not maintained, contributions from abiotic sorption or other coupled removal pathways cannot be excluded fully. Simultaneously, more than 81% (171.6 ± 2.3 μM) of the initial Fe was precipitated, making the Fe:Mn ratio in the synthesised biogenic Fe-Mn mixed oxides ~0.9:1. Despite this extent of Fe oxidation, dissolved Mn concentrations did not show a corresponding increase, suggesting that any Mn released during Fe-Mn redox coupling was re-oxidised and/or retained within the solid phase. However, our data do not permit a precise determination of manganese behavior during the second stage of the experiments.
The interpretation of a low-oxygen, but not fully anoxic, condition is inferred from the sustained activity of P. putida together with incomplete Fe oxidation. The sustained Mn transformation is consistent with the continued availability of low levels of O2. The incomplete oxidation of Fe(II) indicates that the system was oxidant-limited under micro-oxic conditions; the combined availability of biogenic Mn(III,IV) oxides and the trace supply of O2 was insufficient to oxidise the Fe pool to completion.
Throughout the mineralisation process, the solution pH remained relatively stable. During the initial 12 days of Mn(II) bio-oxidation, the pH decreased from 7.4 to 7.3 (±0.2). This modest acidification may be attributed to the production of organic acids (e.g. uronic acids) by bacterial cells and EPS (Behbahani et al., Reference Behbahani, Kiridena, Wijayaratna, Taylor, Anker and Tzeng2022), and protons released during Mn(II) biooxidation, following the reaction:
Some variability in pH among different bio-mineralisation replicates was also observed, probably due to differences in acid production rates and O2 consumption by bacterial inoculates in each replicate (Behbahani et al., Reference Behbahani, Kiridena, Wijayaratna, Taylor, Anker and Tzeng2022). On the addition of Fe(II), the pH declined further to 7.15 (±0.2). This drop is consistent with proton release during Fe(II) oxidation and Fe(III) oxyhydroxide precipitation (Equation 1), as well as biochemical activity linked to bacterial cell turnover and EPS degradation. By the end of the 18 days of mineralisation period, the pH had stabilised at 7.2 (±0.2). In contrast, the abiotic controls maintained a consistent pH range of 7.39–7.41 throughout the experiment.
Mineralogical and morphological characterisation
X-ray diffraction analysis
The XRD patterns for the biogenic Mn oxides and Fe-Mn oxides samples were measured and analysed. The XRD data of the biogenic Mn oxides closely resemble amorphous birnessite structure (details in Supplementary Fig. S1).
The XRD pattern for the biogenic particles collected from the experimental medium after 18 days is shown in Fig. 2. Even though the diffraction patterns were not well defined, suggesting the biogenic precipitates were either poorly or nano crystalline (Post, Reference Post1999; Billinge and Levin, Reference Billinge and Levin2007), the measured data showed a plausible match with several entries in the International Centre for Diffraction Data (ICDD). The background corrected diffractogram (Fig. 2b) shows a broad peak at the ~14.2° 2θ position (d spacing 7.26 Å) and weak peaks at the 28.7°2θ (d spacing 3.59 Å), 43°2θ (d spacing 2.46 Å) and 78.6°2θ (1.42 Å) positions matching d spacing of ICDD entries (00-013-0105, Fig. 2c; 01-073-7862, Fig. 2d), suggesting the presence of poorly crystalline birnessite, similar to previously published data (Cieschi et al., Reference Cieschi, de Francisco, Herrero, Sánchez-Marcos, Cuevas, Esteban, Lucena and Yunta2021; Bondar and Alekseev, Reference Bondar and Alekseev2020; Gao et al., Reference Gao, Shen, Jia, Qiu, Liu, Zhang, Feng and Cai2015). The patterns closely resemble the measurements of poorly ordered birnessite produced using P. putida (Villalobos et al., Reference Villalobos, Toner, Bargar and Sposito2003; Galezowski et al., Reference Galezowski, Recham, Larcher, Miot, Skouri-Panet and Guyot2020). The XRD diffractogram also shows weak, broad scatterings consistent with the presence of a poorly crystalline Fe oxyhydroxide phase. These features are compatible with ferrihydrite-like material (Fig. 2e) in line with the evidence for poorly crystalline ferrihydrite found by other researchers (Valášková et al., Reference Valášková, Leštinský, Matějová, Klemencová, Ritz, Schimp, Motylenko, Rafaja and Bělík2022; Schwertmann et al., Reference Schwertmann, Friedl and Stanjek1999), but are not diagnostic, and definitive phase identification would require Fe-specific spectroscopy. While XRD and EDX provide a primary characterisation of the mineral association, advanced spectroscopic techniques such as Fe K-edge Extended X-ray absorption fine structure, along with X-ray absorption near-edge structure or Mössbauer spectroscopy are required to unambiguously resolve the Fe coordination environment and phase mineralogy. We have framed our current results as a baseline characterisation that invites further spectroscopic investigation.
(a) Raw XRD pattern of the microbial precipitates collected at 18 days. (b) Background-corrected XRD spectrum with corresponding d spacings. (c–e) International Centre for Diffraction Data reference line plots for birnessite and ferrihydrite, showing major reflection positions.
Figure 2. Long description
A multi-panel line graph with a shared horizontal axis representing 2 theta in degrees using a Cobalt [C o] source, ranging from 0 to 80.
Panel A (Top): Raw X R D data shown as a red line. It starts with a high intensity near 150,000 counts at low angles, followed by a broad hump peaking around 15 degrees and a gradual downward slope toward 25,000 counts at 80 degrees.
Panel B: Background corrected intensity in counts. A prominent broad peak is centered at 14 degrees labelled 7.26 Angstroms. Smaller peaks follow at 29 degrees (3.59 Angstroms), 42 degrees (2.46 Angstroms), 64 degrees (1.63 Angstroms), and 77 degrees (1.42 Angstroms).
Panel C: Reference line plot for Birnessite (N a-C a) with I C D D code 00-013-0105. Major vertical lines appear at 14, 29, 43, and 79 degrees.
Panel D: Reference line plot for Birnessite (K) with I C D D code 01-073-7862. It shows a dense cluster of lines between 40 and 45 degrees, with additional lines at 14, 29, 48, 52, 76, and 79 degrees.
Panel E (Bottom): Reference line plot for Ferrihydrite with I C D D code 00-058-0898. It displays numerous lines across the spectrum, with the highest intensity peaks located between 30 and 45 degrees, and significant reflections at 62, 66, and 74 degrees.
SEM analysis
Scanning electron microscopy imaging was performed to constrain the morphology and texture of the biogenic precipitates collected on day 18 of the experiment (Fig. 3a–d). The images in Fig. 3 indicate that these minerals were complexly interwoven with bacterial cells and EPS.
(a, b) Scanning electron microscopy images of the biogenic precipitates showing co-localisation of bacterial cells, extracellular polymeric substances (EPS) and mineral particles where the EPS morphology shown is a dehydrated/collapsed state. (c, d) Magnified images of the biogenic Fe-Mn mixed oxide precipitates.
Figure 3. Long description
A four-panel SEM micrograph series labeled a through d.
Panel a, top-left. A wide view with a 1000 n m scale bar. On the left, a red circle highlights a cluster of rod-shaped structures labelled Bacteria cells and E P S. To the right, another red circle highlights a porous, flaky texture labeled Biogenic F e and M n oxide precipitates.
Panel b, top-right. A higher- magnification view with a 1000 n m scale bar showing a complex, web-like network of dehydrated E P S with thin, interconnected membranes and irregular voids.
Panel c, bottom-left. A magnified view of the mineral precipitates with a 2000 n m scale bar. The structure appears as a dense, sponge-like matrix of irregular, light-coloured mineral flakes.
Panel d, bottom-right. A similar high-magnification view with a 2000 n m scale bar, showing the intricate, crinkled surface of the F e- M n mixed oxide precipitates, characterized by high surface area and jagged edges.
Extracellular polymeric substances are produced during the growth of bacteria and include polysaccharides, humic acids, proteins, nucleic acids and lipids (Wingender et al., Reference Wingender, Neu and Flemming1999; Ras et al., Reference Ras, Girbal-Neuhauser, Etienne and Lefebvre2008; Guibaud et al., Reference Guibaud, van Hullebusch, Bordas, d’Abzac and Joussein2009; Tebo et al., Reference Tebo, Bargar, Clement, Dick, Murray, Parker, Verity and Webb2004). The SEM images show that the minerals, mixed with the EPS components, are located extracellularly or on the surface of the cells. The SEM image of the biogenic Mn oxides (see Supplementary Fig. S2) also shows close aggregation of mineral particles and bacterial cells, consistent with previous findings (Liang et al., Reference Liang, Yang, Wu, Jiang and Xu2017; Cömert and Tepe, Reference Cömert and Tepe2020). The association between the Fe-Mn mixed oxides and the organic matrix is evident in the electron microscopy images; however, because of the dehydrating nature of air drying, the EPS appears as a collapsed film rather than the hydrated gel-like structure present in the aqueous environment. While the air-drying process presents a theoretical risk of secondary abiotic oxidation of residual species, the precipitates were thoroughly washed to remove soluble Fe(II) and Mn(II) prior to imaging. Given that the majority of iron (~81%) had already precipitated during the 18-day incubation, the observed morphologies are considered representative of the primary biogenic minerals rather than artifacts of sample preparation.
Elemental composition and distribution
Scanning transmission electron microscopy EDX spectroscopy analysis of the biogenic mineral precipitates collected after 18 days revealed a heterogeneous distribution of Fe and Mn within the particles. Elemental mapping across the investigated mineral area of interest indicated that Mn oxides and Fe oxides exist as distinct separate mineral phases (Fig. 4b–d), heterogeneously distributed relative to one another rather than forming a homogeneous mixture (Fig. 4). Several studies (Zheng et al., Reference Zheng, Hou, Hartley, Ren, Wang, Tu and Tan2020, Reference Zheng, Tu, Hou, Ni, Wang, Ren, Wang, Cao, Xiong and Tan2021; Li et al., Reference Li, Gyoten, Sonoda, Feng and Xue2017; Zhang et al., Reference Zhang, Liu, Liu, Qu and Liu2014) have demonstrated that the synthesis of abiogenic Fe-Mn binary oxides or mixed oxides typically results in similar elemental distributions.
(a) High-angle annular dark field scanning transmission electron microscopy (STEM) image of the biogenic Fe-Mn mixed oxides, (b) STEM energy dispersive X-ray (EDX) spectroscopy Mn map, (c) STEM-EDX Fe map, (d) STEM-EDX oxygen (O) map, (e) STEM-EDX phosphorus (P) map and (f) STEM-EDX P and carbon (C) overlaid map in biogenic Fe-Mn mixed oxides.
Figure 4. Long description
A multi-panel microscopy series arranged in a two-by-three grid. All panels share a 1 micrometer scale bar located at the bottom right of panel f.
* Panel a: A greyscale H A A D F- S T E M electron image showing a complex, porous, and fibrous morphology of biogenic oxides with varying brightness indicating density.
* Panel b: M n map in yellow overlaid on the electron image. The manganese signal is concentrated in the upper-left and central regions, following the fibrous structure.
* Panel c: F e map in red. The iron signal shows a high density across the top and right-hand side, appearing more continuous than the manganese.
* Panel d: O map in dark blue. Oxygen is distributed heavily across the left and central portions of the sample area.
* Panel e: P map in cyan. Phosphorus is concentrated primarily in the upper-left quadrant, showing a more localized distribution compared to iron or oxygen.
* Panel f: Overlaid map of C in green and P in cyan. The carbon signal is widespread across the fibrous structure, while the phosphorus remains localized in the upper-left, showing the spatial relationship between organic carbon and mineralized phosphorus.
The heterogeneity of Fe and Mn distributions was further investigated by analysing the elemental compositions of specific areas within the sample (Fig. 5). For instance, in Fig. 5, the EDX maps and spectra indicate that the distribution and concentration of Fe and Mn were different in areas 1 and 2, with spectrum 1 showing mainly Fe oxide composition and spectrum 2 showing both Fe and Mn oxide compositions.
STEM-EDX spectra of specific regions of biogenic Fe-Mn mixed oxides shown on the overlaid EDX maps of Fe and Mn (spectrum 1 and spectrum 2).
Figure 5. Long description
The top section contains two side-by-side scanning electron microscopy maps with a 500 nanometer scale bar. The left map shows iron in red and manganese in yellow overlaid on a grey electron background. The right map shows only manganese in yellow. Two rectangular regions are labelled Spectrum 1 in the upper left and Spectrum 2 in the center right.
The bottom section is an E D X spectrum graph with Energy in k e V on the x-axis from 0 to 12 and Intensity in arbitrary units on the y-axis. Two data lines are shown: a blue line for Spectrum 1 and a red line for Spectrum 2.
From left to right, the peaks are identified as follows:
* Between 0 and 1 k e V: High intensity peaks for C, N, O, Fe, and Cu.
* At 1.5 k e V: A small Al peak.
* At 1.8 k e V: A small Si peak.
* At 2.0 k e V: A distinct P peak, significantly higher in Spectrum 1.
* At 2.3 k e V: A small S peak.
* At 5.9 k e V: A Mn peak, present in both spectra but relatively higher in Spectrum 2.
* At 6.4 k e V: A prominent Fe peak, significantly higher in Spectrum 1.
* At 7.1 k e V: A secondary Fe peak.
* At 8.0 k e V: A Cu peak, higher in Spectrum 1.
* At 8.9 k e V: A secondary Cu peak.
The presence of phosphorus (P) associated with the biogenic minerals was observed in the EDX maps and spectra (Figs 4e, 5). The detected carbon (C) distribution in the EDX maps correlated closely with the P distribution. While the formvar coating of the TEM grid may contribute some of the C signal, these distributions suggest that P and C probably originated from the bacterial cells and EPS. The EDX spectra (Fig. 5) show higher P and C signals in area 1 compared to area 2, supporting the claim that the presence of P is mainly associated with organic carbon (bacteria cells and EPS) and the EDX mapping (Fig. 4f) also supports the co-distribution of these elements. Some of the P signals may be explained by the formation of Mn phosphates, although such phases were not identified in our XRD experiments.
It has been previously reported that inorganic phosphate may originate from the hydrolysis of cells, increasing concentration in the extracellular environment, which could promote its reaction with Mn adsorbed on the bacterial cell surfaces and the formation of manganese phosphate precipitates close to the cells (Parikh and Chorover, Reference Parikh and Chorover2005; Galezowski et al., Reference Galezowski, Recham, Larcher, Miot, Skouri-Panet and Guyot2020). A similar finding was demonstrated by Galezowski et al. (Reference Galezowski, Recham, Larcher, Miot, Skouri-Panet and Guyot2020) in biogenic Mn oxides produced by P. putida.
Energy dispersive X-ray analysis also identified other elements such as nitrogen (N), sulfur (S), aluminium (Al), silicon (Si) and negligible amounts of calcium (Ca) and potassium (K), consistent with findings reported by Morales et al. (Reference Morales, Formanski, Shaner and Stone2024) for biogenic Mn oxides produced by P. putida. Miller et al. (Reference Miller, Dionísio, Sequeira-Braga, Hernández-Mariné, Afonso, Muralha, Herrera, Raabe, Fernandez-Cortes and Cuezva2012) also reported the presence of Al, Si, Ca and P in biogenic Mn oxides. These elements probably originate from borosilicate bottles and chemicals used for preparation of the experimental media. The high Cu signal originates primarily from the TEM grid.
STEM microscopy and TEM-SAED analyses
The STEM images revealed key insights into the relative sizes, crystalline properties and morphologies of the biogenic Fe-Mn mixed oxides. Although precise particle-size measurements were complicated by aggregation effects, the individual particles observed appeared to be predominantly <100 nm in diameter (Fig. 6a–f), consistent with classification as nanoparticles.
Bright-field scanning transmission electron microscopy images of the biogenic Fe-Mn mixed oxide (a, b) nanoparticles of various morphologies mixed with bacteria cells, (c) ultrathin layers, (d, e) chain-like particle aggregates and (f) a large granular particle cluster.
Figure 6. Long description
A six-panel B F- S T E M micrograph series.
* Panel A: Top-left. A dark, porous network of biogenic Fe- Mn mixed oxide. Three red ovals with arrows point to lighter, oval-shaped bacterial cells embedded within the mineral matrix. Scale bar is 1000 n m.
* Panel B: Top-center. A higher magnification view of the mineral matrix showing thin, overlapping sheet-like morphologies. Scale bar is 200 n m.
* Panel C: Top-right. Ultrathin layers of the oxide appearing as translucent, folded sheets with dark edges where the layers are perpendicular to the beam. Scale bar is 100 n m.
* Panel D: Bottom-left. A wide view of chain-like particle aggregates and loose mineral clusters. Scale bar is 1000 n m.
* Panel E: Bottom-center. A detailed view of the chain-like aggregates, showing elongated, interconnected mineral particles. Scale bar is 200 n m.
* Panel F: Bottom-right. High-magnification view of a large granular particle cluster with a dense, textured surface. Scale bar is 20 n m.
Bright-field scanning transmission electron microscopy (BF-STEM) was used to evaluate the morphology and crystalline characteristics of the biogenic Fe-Mn mixed oxides. The BF-STEM image in Fig. 6a shows rod-shaped P. putida cells in close association with Fe and Mn oxide particles. Morphological variation of the mixed oxides across the field of view is evident (Fig. 6b). The particles frequently appear as ultrathin, layered clusters with individual sheets folded at various angles or stacked atop one another (Fig. 6c), a morphology previously described in biogenic Mn oxides (Villalobos et al., Reference Villalobos, Toner, Bargar and Sposito2003; Miller et al., Reference Miller, Dionísio, Sequeira-Braga, Hernández-Mariné, Afonso, Muralha, Herrera, Raabe, Fernandez-Cortes and Cuezva2012; Galezowski et al., Reference Galezowski, Recham, Larcher, Miot, Skouri-Panet and Guyot2020). Other particles exhibit chain-like morphologies (Fig. 6d,e). Nanoparticles also form large aggregates and granular clusters (Fig. 6f). While the precise mechanisms underlying the formation of these morphologies remain unclear, electrostatic interaction and binding between mineral surfaces and charged functional groups at the bacterial cell walls or EPS probably promote particle aggregation and structural organisation (Fortin and Langley, Reference Fortin and Langley2005).
High-resolution STEM imaging did not consistently reveal lattice fringes in the biogenic Fe-Mn mixed oxides. This may be attributed to several factors: (1) the highly disordered or amorphous structure of the biogenic minerals (Smith and Horgan, Reference Smith and Horgan2021); (2) the presence of organic matter or diffuse electron scattering at particle surfaces, which reduces image contrast and obscure lattice fringes; or (3) beam-induced radiation damage resulting in partial reduction of Mn(III)/Mn(IV) to Mn(II) species (Galezowski et al., Reference Galezowski, Recham, Larcher, Miot, Skouri-Panet and Guyot2020). Organic matter commonly associated with biogenic Mn oxides is known to generate free-radical species when exposed to radiation beams, leading to the photo-reduction of Mn(IV)/Mn(III) (Toner et al., Reference Toner, Fakra, Villalobos, Warwick and Sposito2005). These effects, in conjunction with the high electron acceleration voltages used during imaging, probably contributed to the observed difficulty in resolving lattice fringes. Furthermore, biogenic minerals have previously been shown to exhibit lattice distortions relative to their synthetic or naturally occurring abiogenic counterparts, which may further complicate structural visualisation (Deng et al., Reference Deng, Jia and Li2022).
Despite limitations in lattice resolution, the TEM-SAED analysis provided critical insights into the crystalline properties of these biogenic minerals, revealing spatial heterogeneity across the sampled regions (Fig. 7). An amorphous halo is observed in the SAED diffraction patterns (Fig. 7b,c,h,i) of some areas (Fig. 7a,g) alongside several weak diffraction speckles (Fig. 7b), which suggest the presence of nano-sized polycrystals sharing a common orientation within the measured area. In contrast, a separate region (Fig. 7d) displays a SAED pattern with a faint diffraction ring (Fig. 7e,f), suggesting the presence of randomly oriented polycrystalline domains. The diffraction speckles and d spacing values are broadly consistent with the presence of nano-sized birnessite-like material together with a poorly crystalline Fe-bearing oxyhydroxide phase (Lee and Xu, Reference Lee and Xu2019; Arias et al., Reference Arias, Becerra and Giraldo2019; Powell et al., Reference Powell, Bruggraber, Faria, Poots, Hondow, Pennycook, Latunde-Dada, Simpson, Brown and Pereira2014), but do not permit definitive assignment of a specific Fe mineral. Differences between the XRD and TEM-SAED observations probably reflect the distinct sampling scales and structural sensitivity of the two techniques. X-ray diffraction captures the bulk-average long-range crystallographic order of the sample, whereas TEM-SAED is sensitive to local nanocrystalline domains, which can yield diffraction features even in materials that appear only weakly crystalline at the bulk scale. The amorphous halos observed in all the SAED diffraction patterns probably originate from the presence of organic matter, specifically bacterial cells and EPS, within the biogenic samples (Shi et al., Reference Shi, Wang, Zhang and Sun2024; Galezowski et al., Reference Galezowski, Recham, Larcher, Miot, Skouri-Panet and Guyot2020).
Transmission electron microscopy images of three representative regions of the biogenic Fe-Mn mixed oxides sample (a, d, g) with corresponding selected area electron diffraction (SAED) patterns (b, e, h) and measured d spacings (c, f, i). The observed diffraction features are broadly consistent with nano-sized birnessite-like Mn oxide domains together with a poorly crystalline Fe-bearing oxyhydroxide phase; however, these SAED data do not allow definitive assignment of a specific Fe mineral.
Figure 7. Long description
A grid of nine panels organized into three rows.
Top Row:
Panel A is a T E M micrograph showing dark, irregular mineral aggregates on a lacey carbon support with a 500 nanometer scale bar.
Panel B shows the corresponding S A E D pattern with a bright central spot and diffuse diffraction spots.
Panel C annotates the pattern with two yellow concentric rings. The inner ring corresponds to a d-spacing of 3.4 Angstroms and the outer ring to 1.9 Angstroms.
Middle Row:
Panel D is a T E M micrograph of a second region showing thinner, more dispersed mineral clusters.
Panel E shows a diffuse S A E D pattern with a faint halo.
Panel F annotates this halo with a single yellow ring corresponding to a d-spacing of 2.7 Angstroms.
Bottom Row:
Panel G is a T E M micrograph of a third region showing small, dense mineral particles.
Panel H shows the S A E D pattern with distinct diffraction spots.
Panel I annotates the pattern with two yellow concentric rings. The inner ring corresponds to a d-spacing of 4.4 Angstroms and the outer ring to 1.9 Angstroms.
All S A E D patterns (B, C, E, F, H, I) include a scale bar of 2 inverse nanometers in the bottom right corner.
Environmental implications: a proposed model for mixed oxides formation
The experimental results presented here, particularly the observed decoupling of Mn and Fe mineral phases under restricted O2, provide a basis for refining models of mineral formation in redox-stratified environments. Previous studies have documented the formation of biogenic Fe-Mn binary oxides or mixed oxides via pathways distinct from that presented in this study. For instance, Bai et al. (Reference Bai, Yang, Liang and Qu2016) investigated aerobic Fe(II) oxidation coupled with Mn(II) bio-oxidation by Pseudomonas sp. QJX-1, producing lepidocrocite and birnessite. Similarly, Yang et al. (Reference Yang, Yu, Zeng, Liang, Rong, Li and Qu2023) also reported biogenic Fe-Mn binary oxides formation under fully oxic conditions. In contrast, the current study examines a low-oxygen phase in which Fe(II) oxidation is interpreted as being consistent with reaction with pre-formed biogenic Mn oxides, providing a conceptual basis for considering one possible pathway of Fe-Mn mixed oxides formation at redox interfaces.
Biogenic Fe and Mn oxides are commonly found across various environmental systems (Glodowska et al., Reference Glodowska, Schneider, Eiche, Kontny, Neumann, Straub, Kleindienst and Kappler2021; Estes et al., Reference Estes, Andeer, Nordlund, Wankel and Hansel2017; Tebo et al., Reference Tebo, Bargar, Clement, Dick, Murray, Parker, Verity and Webb2004; Villalobos et al., Reference Villalobos, Toner, Bargar and Sposito2003; Sowers et al., Reference Sowers, Harrington, Polizzotto and Duckworth2017), typically forming mixed deposits in response to dynamic redox conditions. These Fe- and Mn-rich zones are common at oxic–anoxic interfaces in both terrestrial and aquatic settings. For instance, Bábek et al. (Reference Bábek, Sracek, Všianský and Holá2023) reported micrometer-scale Fe-Mn laminae in a groundwater aquifer, probably of microbial origin. Similar Fe- and Mn-enriched strata have also been found in shallow and deep marine sediments, as well as freshwater systems, under both oxic and transitional redox conditions (Tazaki, Reference Tazaki2000; Palermo and Dittrich, Reference Palermo and Dittrich2016).
Mn(II) bio-oxidation followed by abiotic Fe(II) oxidation, as described below, may represent a plausible pathway for forming biogenic Fe-Mn mixed oxide precipitates at the oxic-anoxic interface in natural aquatic environments. In redox-stratified aquatic environments such as in stratified lakes, Fe and Mn cycling is primarily governed by the availability of dissolved oxygen, which in turn shapes the redox gradients that drive mineral transformations (Rincón-Tomás et al., Reference Rincón-Tomás, Khonsari, Mühlen, Wickbold, Schäfer, Hause-Reitner, Hoppert and Reitner2016). Multiple formation pathways for Fe and Mn oxides exist, depending on local redox conditions (Nelson et al., Reference Nelson, Lion, Shuler and Ghiorse2002). A conceptual model illustrating these processes is shown in Fig. 8. In the oxygenated surface layer, Fe(II) undergoes abiotic oxidation by O2 to form Fe(III) oxides (step 1). Under micro-oxic conditions, however, biotic Fe(II) oxidation by microaerophilic bacteria may outpace abiotic processes, yielding biogenic Fe(III) oxyhydroxides (step 2) (Emerson et al., Reference Emerson, Fleming and McBeth2010). In contrast, Mn(II) is poorly oxidised abiotically as a result of kinetic constraints (step 3), but is efficiently oxidised by Mn-oxidising bacteria and fungi to form biogenic Mn(III/IV) oxides (step 4) (Madison et al., Reference Madison, Tebo, Mucci, Sundby and Luther2013; Tebo et al., Reference Tebo, Bargar, Clement, Dick, Murray, Parker, Verity and Webb2004), which we have shown in our study. These microbially mediated processes may set the stage for Fe(II) oxidation by Mn oxides at deeper, low-oxygen levels, thereby providing one plausible route to Fe-Mn mixed oxide formation.
Putative pathways of biogenic Fe-Mn mixed oxides formation at the oxic–anoxic interface of a redox-stratified water column, as informed by the experimental results herein.
Figure 8. Long description
A vertical cross-section diagram divided into three horizontal zones: a light blue Oxic layer at the top, a grey Oxidation zone in the middle, and a tan Anoxic layer at the bottom.
In the top Oxic layer:
- Pathway 1: Iron F e (II) undergoes aerobic oxidation with O sub 2 to form F e III) oxides.
- Pathway 2: Microaerobic F e (II) oxidizers facilitate the formation of F e (III) oxides.
- Pathway 3: Manganese M n (II) undergoes very slow aerobic oxidation with O sub 2 to form M n III, IV) oxides.
- Pathway 4: M n oxidizers facilitate faster microbial oxidation of M n (II) to M n (III, IV) oxides.
- Physical mixing arrows connect the iron and manganese pathways in the center.
In the middle Oxidation zone:
- F e (III) oxides and M n (III, IV) oxides sink from the upper layer.
- Pathway 9: Microbial oxidation converts M n (II) to M n (III, IV) oxides.
- Pathway 10: Anaerobic F e (II) oxidizers and nitrate reducers facilitate the reaction between M n (III, IV) oxides and F e (II) to produce F e (III) oxides.
In the bottom Anoxic layer:
- Pathway 5 and 6: Reduction processes convert F e (III) oxides back to F e (II) and M n (III, IV) oxides back to M n (II).
- Pathway 7 and 8: F e (II) and M n (II) move upward via diffusion.
- Vertical arrows labeled Sinking and Physical mixing lead to the bottom-most deposits labeled F e - M n mixed oxides, F e (III) oxides on the left, and F e - M n mixed oxides, M n (III, IV) oxides on the right.
Microbial Mn(II) oxidation at oxic–anoxic interfaces has been evidenced by many researchers (Oldham et al., Reference Oldham, Jones, Tebo and Luther2017; Jones et al., Reference Jones, Nico, Ying, Regier, Thieme and Keiluweit2018; Clement et al., Reference Clement, Luther and Tebo2009) in natural environments. Biogenic Fe-Mn mixed oxides can form through co-precipitation and/or mixing of biogenic Fe and Mn oxide mineral phases. Their formation depends on the relative concentrations of Fe(II) and Mn(II), as well as the activity of Fe- and Mn-oxidising microorganisms within redox transition (Nelson et al., Reference Nelson, Lion, Shuler and Ghiorse2002). For example, Fe oxide surfaces can serve as substrates for Mn-oxidising bacteria, followed by precipitation of biogenic Mn oxides. Conversely, Mn oxide surfaces can become coated with Fe oxides, resulting in Fe-Mn oxide composites (Nelson et al., Reference Nelson, Lion, Shuler and Ghiorse2002). These oxides may then sink into the micro-oxic zone, where they are subjected to microbial or abiotic reduction to soluble Fe(II) and Mn(II) (steps 5 and 6). The reduced species can diffuse upwards from anoxic sediments or deeper water layers into the oxic or micro-oxic zones (steps 7 and 8). Mn(II), once reintroduced into these zones, can again undergo microbial oxidation to Mn(III/IV) oxides (step 9), which in turn can oxidise Fe(II) abiotically, yielding Fe(III) (oxyhydr)oxides (step 10). This Mn-mediated oxidation of Fe(II) competes with other Fe(II) oxidation pathways, including direct oxidation by O₂, nitrite-driven chemodenitrification and microbial oxidation by microaerophilic or anaerobic Fe(II)-oxidising bacteria, such as nitrate-reducers and phototrophs (Kappler and Straub, Reference Kappler and Straub2005; Kappler and Newman, Reference Kappler and Newman2004; Straub et al., Reference Straub, Schönhuber, Buchholz-Cleven and Schink2004; Hafenbradl et al., Reference Hafenbradl, Keller, Dilmeier, Rache, Roßnagel, Burggraf, Huber and Stetter1996; Myers and Nealson, Reference Myers and Nealson1988).
Depending on environmental conditions, particularly the steepness and stability of the redox gradient, two distinct scenarios may govern Fe-Mn oxides formation in natural systems. In one pathway (1) Fe(II) oxidation and Mn(II) bio-oxidation occur independently, resulting in the formation of discrete Fe(III) and Mn(III,IV) mineral phases that subsequently undergo physical mixing during sedimentation. In the alternative pathway (2), biogenic Mn oxides form first and subsequently oxidise Fe(II), leading to the in situ co-precipitation of finely intermixed Fe-Mn mixed oxides. This second mechanism aligns with the findings of this study, where Fe and Mn redox transformations and mixed oxides formation were observed primarily within the oxic to micro-oxic zones of a stratified redox environment.
While this conceptual pathway is informed by our laboratory observations of P. putida-driven Mn and Fe redox transformations, further in situ studies are required to validate the prevalence of these specific mixed-oxide mineral formation proecesses in diverse aquatic systems.
Conclusions
The results support a novel pathway for Fe-Mn mixed oxides formation, driven by microbial Mn(II) oxidation followed by abiotic Fe(II) oxidation mediated by biogenic Mn oxides under low-oxygen conditions. Through a comprehensive suite of analytical techniques (XRD, SEM, STEM, STEM-EDX and TEM-SAED), we characterised the resulting minerals as poorly crystalline birnessite and Fe-oxyhydroxide nanocrystals, intimately associated with bacterial cells and EPS. High-resolution imaging revealed morphological diversity, granular, chain-like and layered nanoparticle assemblies, while EDX mapping confirmed heterogeneous Fe-Mn distributions and SAED patterns highlighted structural disorder and organic content.
These structural and compositional features suggest that the synthesised Fe-Mn mixed oxides are reactive redox-active materials the environmental behaviour of which warrants further investigation. More broadly, the present work advances understanding of microbial Fe-Mn biomineral formation under redox gradients and provides a basis for future studies evaluating the environmental behaviour and potential utility of such materials.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1180/gbi.2026.10010.
Acknowledgements
The authors wish to thank Laura Galezowski and Feriel Panet (Institute of Mineralogy, Physics of Materials, and Cosmochemistry at Sorbonne University) for providing the Pseudomonas putida MnB1 strain. We gratefully acknowledge the assistance of Errin Johnson (microscopy facilities at the Sir William Dunn School of Pathology, University of Oxford) for SEM analyses. We also thank Neil Young and Ruomu Zhang (David Cockayne Centre for Electron Microscopy at the Department of Materials, University of Oxford) for their support with TEM imaging. Finally, we are indebted to Luke Williams (Department of Earth Sciences, University of Oxford) for his laboratory assistance. This research was supported by the Commonwealth Scholarship Commission in the United Kingdom through the award of a Commonwealth PhD Scholarship 2021.
Competing interests
The authors declare no competing interests.
Open access
