Leveraging Biomineralization in Repurposed Stirred Reactors for Mn/Zn Removal from Mine Water: Insights from a Laboratory-Scale Study

Abstract

This study developed a semi-passive treatment system for manganese (Mn)- and zinc (Zn)-containing mine water by repurposing a neutralization tank into a biologically active stirred reactor. Laboratory-scale experiments demonstrated efficient removal of Mn²⁺ (>97%) and Zn²⁺ (>80%) with hydraulic retention times (HRTs) as short as 6 h—significantly faster than traditional passive systems. XRD and XANES analyses identified the predominant formation of birnessite, a layered Mn oxide, during Mn²⁺ oxidation, with Zn co-treatment promoting the precipitation of Zn-containing carbonates. Despite decreasing crystallinity of birnessite over time, microbial activity, dominated by Mn-oxidizing genera, such as Sphingomonas, Pseudonocardia, Sphingopyxis, Nitrospira, and Rhodobacter, persisted in the presence of Zn²⁺, ensuring system stability. Importantly, the low leachability of Mn and Zn from the resulting sludge in TCLP tests confirmed its environmental safety and potential for reuse. By leveraging existing infrastructure and microbial biomineralization, this system bridges the gap between passive and active treatments, significantly reducing treatment footprints and operational costs. These findings highlight the potential of repurposing mine water treatment tanks as a scalable, cost-effective solution for sustainable mine water remediation.

1. Introduction

Mine wastewater containing heavy metals is a persistent environmental problem, especially in regions where mining has ceased. After mine closure, the semi-permanent discharge of contaminated water necessitates continuous treatment to prevent environmental pollution. Among the heavy metals in mine wastewater, manganese (Mn) is particularly challenging to remove due to its stability as soluble Mn²⁺ over a wide pH and redox potential (Eh) range. Prolonged exposure to elevated Mn concentrations poses serious health risks, including manganism [1]. To mitigate these risks, effluent standards for Mn, such as Japan’s limit of 10 mg/L [2], have been established globally.

Conventional Mn removal methods for mine wastewater can be broadly classified into active and passive treatment approaches. Active treatment typically involves the continuous addition of chemical agents, such as pH adjusters, oxidants, and coagulants, to precipitate Mn²⁺ as Mn(III) or Mn(IV) oxides [3,4]. This method enables rapid and highly efficient Mn removal, as demonstrated in oxidative precipitation treatment systems using potassium permanganate (KMnO₄). Freitas et al. [5] evaluated the effectiveness of KMnO₄ in removing Mn²⁺ from acid mine drainage (AMD) under different pH conditions. Their study revealed that at pH 7.0, Mn²⁺ removal reached 99% within 20 min, effectively reducing Mn concentrations from an initial 90–105 mg/L to below 1.0 mg/L, forming MnO₂, Mn₃O₄, and MnOOH precipitates. Similarly, chlorine-based oxidative precipitation methods have also shown promising Mn removal efficiency. Li et al. [6] demonstrated that powdered activated carbon enhanced Mn²⁺ oxidation by chlorine, reducing Mn²⁺ concentrations from 200 μg/L to <10 μg/L within 20–30 min, significantly accelerating the oxidation process through adsorption-mediated electron transfer mechanisms. However, this approach is resource-intensive, requiring significant reagent and energy costs.

On the other hand, passive treatment is based on natural processes, including the biological oxidation of Mn²⁺ by Mn-oxidizing bacteria and autocatalytic reactions involving biogenic Mn oxides [7,8]. For instance, sand biofilters have been used for Mn removal in groundwater with Mn concentrations up to 3 mg/L, achieving high removal efficiency under hydraulic retention times (HRTs) of 12–24 h [9]. Araya-Obando et al. [10] demonstrated that sand biofilters effectively removed Mn from groundwater, achieving >90% removal over 107 days with an initial concentration of 0.57 mg/L. The study also showed that biological Mn removal remained effective even with a short HRT of approximately 12 min. By harnessing natural reactions and minimizing chemical inputs, passive treatment systems are gaining recognition as sustainable and cost-effective solutions for long-term wastewater management.

Microbial Mn oxidation is a critical process for Mn removal in passive systems. During this process, Mn²⁺ is enzymatically oxidized to Mn(IV) oxides via intermediate Mn(III) states [11], forming biogenic Mn oxides with unique mineralogical and functional properties. The general stoichiometry of this reaction is as follows [12]:

Mn²⁺ + 1/2O₂ + H₂O→Mn^III,IVO₂ + 2H⁺

(R1)

The microbiological oxidation of Mn²⁺ to Mn(IV) occurs as a two-step process. First, Mn²⁺ in the liquid phase temporarily forms mixed –valence Mn(III)-containing oxides (e.g., Mn₃O₄) or oxyhydroxides (e.g., β-MnOOH). Subsequently, it undergoes a protonation reaction to form Mn(IV) oxide (MnO₂) [13,14]. Biogenic Mn oxides such as birnessite ((Na, Ca, K)_0.5 Mn^III,IV₂O₄·1.5 H₂O), a layered Mn oxide, exhibit high adsorption capacities for various heavy metals, including Zn²⁺ [11,15]. This makes microbial Mn oxidation a promising tool not only for Mn removal but also for co-contaminant management.

These microbes play pivotal roles in Mn geochemistry, contributing to Mn oxidation in both natural environments and engineered systems such as water treatment pipelines. The Mn oxides they produce are known to facilitate secondary reactions, such as disproportionation ((Reaction 2) (R2)) and symproportionation ((Reaction 3) (R3)). These reactions can occur under certain conditions, contributing to the geochemical cycling of Mn. The interplay between (Reactions 1 and 3 (R1) and (R3)) can influence the overall system dynamics. Depending on the balance of these processes, these secondary reactions can further contribute to Mn removal by promoting the transformation and stabilization of Mn species in solid phases [11,16].

2Mn^III(solid)→Mn^IV(solid) + Mn²⁺

(R2)

Mn²⁺ + Mn^IV(solid)→2Mn^III(solid)

(R3)

Mn-oxidizing microorganisms have been identified across diverse taxa, including the phylum Firmicutes (e.g., Bacillus sp. and Brevibacillus sp.); and classes within the phylum Proteobacteria, such as α-Proteobacteria (e.g., Pedomicrobium sp., Erythrobacter sp., Aurantimonas sp., Roseobacter sp., Rhizobium sp., Methylobacterium sp., Bosea sp., Sphingomonas sp., and Rhodobacter sp.), β-Proteobacteria (e.g., Leptothrix sp., Sphaerotilus sp., and Burkholderia sp.), and γ-Proteobacteria (e.g., Pseudomonas sp., Oceanospirillum sp., Halomonas sp., Citrobacter sp., Pantoea sp., and Aeromonas sp.). Additionally, members of the phylum Actinobacteria (Arthrobacter sp., Pseudonocardia sp., Lapillicoccus sp., Leifsonia sp., Terrabacter sp., and Mycobacterium sp.) have been implicated in Mn oxidation [17,18,19,20,21,22]. While well-documented studies strongly support the Mn-oxidizing activity of many of these genera, others have been included based on limited data, suggesting potential Mn oxidation capabilities. Such microbial Mn oxidation has been observed in both natural environments (e.g., soils and sediments) and engineered systems, where microbial biofilms play a key role in processes such as Mn oxidation and heavy-metal removal [23,24].

To address the challenges associated with the conventional Mn removal methods, researchers and policymakers have advocated for more sustainable approaches, including passive and semi-passive treatment systems that leverage existing infrastructure to reduce costs, while also ensuring effective performance. To date, most Mn bioprocesses have focused on treating groundwater and drinking water, aiming to reduce Mn concentrations below the WHO/EPA guidelines of 0.05 mg/L, typically using sand biofilters with HRTs of 12–24 h [9,25,26]. Some studies have investigated Mn removal from mine wastewater, where effluent discharge limits vary. For instance, Japan sets a regulatory limit of 10 mg/L for dissolved Mn in industrial effluents [3,27]. Recent studies have increasingly focused on advancing passive treatment systems for Mn²⁺ removal from wastewater. Research efforts have identified novel Mn-oxidizing microorganisms and explored innovative reactor designs to enhance removal efficiency [24]. Development efforts have been particularly geared towards applications in real-world settings, using mine wastewater samples to evaluate performance under practical conditions [28,29,30].

Previous research has investigated stirred reactor systems as a potential solution for Mn removal. One promising mechanical approach involves the application of stirring to enhance interactions between Mn²⁺, Mn-oxidizing microorganisms, and Mn oxides [31,32]. For example, Ginter and Grobicki (1997) investigated a stirred UASB reactor and achieved Mn removal efficiencies of up to 56% with HRTs of 24–48 h, primarily through adsorption and co-precipitation processes [31]. Similarly, Mao et al. (2023) evaluated a continuous stirred tank bioreactor (CSTB) for the treatment of Mn-rich acid mine drainage, reporting a maximum Mn removal efficiency of 82.4% at an HRT of 48 h and pH 5.5, with Mn oxidation facilitated by Acinetobacter and Azospirillum [32].

This study explores the potential scalability of the semi-passive treatment system by emphasizing the practical advantages of repurposing existing neutralization or agitation tanks commonly found at mining sites. To this end, exploratory experiments were conducted in small-scale laboratory tanks to assess system performance under controlled conditions, providing insights for potential large-scale implementation. These pre-installed infrastructures significantly reduce the initial capital investment required for establishing new treatment systems. While stirring power introduces an additional energy requirement compared to purely passive systems, it provides notable benefits, including enhanced microbial activity and accelerated Mn removal rates. By utilizing biologically active Mn-sludge within these tanks, this approach leverages the synergistic effects of microbial oxidation and controlled stirring conditions.

Our previous studies demonstrated that a passive bioreactor using bioactive Mn-sludge could achieve effective microbial Mn oxidation at an HRT of 8 h, using real mine water [33]. Here, we show that a semi-passive system, operated under controlled synthetic conditions, achieves >97% Mn removal and >80% Zn removal, while reducing the HRT by 25% (to 6 h). Additionally, we investigated the influence of Zn on Mn oxidation by assessing both its chemical effects and its impact on microbial community dynamics, providing further mechanistic insights into the potential repurposing of existing treatment infrastructure for cost-effective mine water remediation.

2. Materials and Methods

2.1. Bioactive Mn-Sludge

The Mn-sludge, a black precipitate, was collected from the N mine in Japan in September 2023 [33]. The sample was stored at 4 °C until use.

2.2. Mn/Zn Removal by Continuous Stirred Tank Reactor

2.2.1. Baseline Mn Oxidation Test (Mn-System)

The stirred tank reactor designed is shown in Figure 1. A cylindrical acrylic beaker (6.6 cm diameter × 9.1 cm height) was used as the Mn oxidation treatment tank and operated at room temperature. The reactor was filled with 100 g of Mn-sludge to form a 2 cm layer (pulp density, 50%; retention volume, 200 mL). The Mn-sludge and synthetic mine water were stirred continuously at 100 rpm, using a stirrer (EYELA ZZ-1200S; Tokyo Rikakikai Co., Ltd., Tokyo, Japan). To account for the small amount of Mn-sludge that overflowed during stirring, a secondary beaker of the same size was connected to the reactor to collect the escaping sludge. The collected sludge was returned to the Mn oxidation tank every three days to maintain the reaction conditions. The top of the reactor beaker was covered with parafilm (Bemis) throughout the experiment.

Synthetic mine water containing 70 mg/L Mn²⁺ (as MnSO₄), 430 mg/L Ca²⁺ (as CaCl₂), 240 mg/L Mg²⁺ (as MgSO₄), and 210 mg/L HCO₃⁻ (as NaHCO₃∙2H₂O) was prepared using distilled water as the base and continuously fed into the reactor at varying hydraulic retention times (HRTs): 10 h (Days 0–9), 8 h (Days 10–15), 6 h (Days 16–32), and 4 h (Days 33–44), before stabilizing at 5 h. The synthetic mine water used in this study was formulated to replicate the major elemental composition of water from the N-mine in Japan, which was the subject of prior studies [33]. While the composition was simplified by excluding certain trace elements, the concentrations of key components, particularly Mn and Zn, were retained. This approach ensured that the experimental conditions were representative, while allowing for controlled evaluation of the system’s performance. No organic matter was added to the feed water. Liquid samples were taken from the reactor inlet and outlet every three days, while solid samples were taken from the reactor occasionally.

2.2.2. Enhanced Mn Oxidation with Zn Co-Treatment (Mn/Zn-System)

The experimental design and procedure were identical to those described in Section 2.2.1, except for the addition of Zn²⁺ to the synthetic mine water. Zn²⁺ (as ZnSO₄) was introduced from Day 30 at an initial concentration of 2 mg/L. However, due to the high bicarbonate (HCO₃⁻) content in the feed water, Zn²⁺ precipitated in the feed water tank, making it difficult to maintain a stable inlet concentration. This precipitation phenomenon is more pronounced in the laboratory setup due to the controlled conditions and lack of continuous flow, which differs from field conditions. To mitigate this issue, we tested various initial Zn concentrations to determine an appropriate value that would result in a stable influent Zn²⁺ level under mine water conditions. It was found that setting the Zn²⁺ concentration to 10 mg/L allowed the influent Zn concentration to stabilize at approximately 2 mg/L [34]. Therefore, from Day 33, the Zn²⁺ concentration was increased to 10 mg/L. All other conditions, including HRTs and sampling protocols, remained as described in Section 2.2.1.

2.3. Liquid Analyses

Liquid samples were taken every three days from both tank reactors (Section 2.2.1 and Section 2.2.2) to monitor pH, Eh, and Mn/Zn concentrations, using inductively coupled plasma–optical emission spectrometry (ICP-OES; Optima8300, Perkin Elmer, Shelton, CT, USA). Calibration was performed using Zn and Mn standard solutions (0–20 mg/L) prepared from certified reference materials. Linear regression analysis confirmed calibration linearity (R² > 0.999). The detection limits were 0.2 μg/L for Zn and 0.1 μg/L for Mn, with measurement precision consistently below 2% relative standard deviation (RSD).

2.4. Solid Analyses

Solid Mn precipitate samples were periodically collected for various analyses, including X-ray diffraction (XRD; Ultima IV, Rigaku, Tokyo, Japan), X-ray absorption fine structure (XAFS; Section 2.4.1), adenosine triphosphate (ATP) concentration measurements (Section 2.4.2), and microbial community structure characterization (Section 2.4.3).

2.4.1. XANES Analysis

Mn precipitates were freeze-dried, finely ground in a mortar, and thoroughly mixed with boron nitride (BN) to achieve a Mn content of 2%. The prepared mixture (200 mg) was compressed into tablets at 15 MPa for over 10 min. These tablets were used for XANES analysis to determine the oxidation state and coordination environment of Mn in the solid samples.

XANES measurements were conducted at the Saga-LS, Kyushu University beamline (BL06) (SAGA-LS, Saga, Japan). Spectra were collected in transmission mode over the energy range of 6200–7800 eV, with a step size of 0.2 eV, near the Mn K-edge. Calibration was performed using Mn foil, and reference standards included MnSO₄ (Mn(II)), Mn₂O₃ (Mn(III)), and MnO₂ (Mn(IV)). Linear combination fitting (LCF) was conducted to determine the relative contributions of the Mn species using ATHENA software, with an R² value of less than 2% for all fits. This indicates high reliability of the oxidation state quantifications, and all data acquisition and analysis followed established protocols [35].

2.4.2. ATP Analysis

The ATP measurement was conducted to evaluate microbial activity within the Mn precipitates, serving as an indicator of biological contributions to Mn oxidation processes. ATP concentrations were measured using a Lumitester C-110 kit (Kikkoman Biochemifa, Tokyo, Japan). For each analysis, 10 mg of Mn precipitate samples was used. Calibration was performed using ATP standard solutions with concentrations ranging from 2 × 10⁻¹² to 2 × 10⁻⁸ M, ensuring accurate quantification.

2.4.3. Microbial Community Structure Analysis

Genomic DNA was extracted from the Mn precipitates and analyzed to characterize the microbial community associated with Mn oxidation processes. The extracted genomic DNA was subjected to 16S rDNA amplicon sequencing, targeting the V3–V4 region, using a MiSeq platform (Illumina, Tokyo, Japan). Sequence data were compared with reference databases using the Metagenome@KIN platform (TechnoSuruga Laboratory Co., Ltd., Shizuoka, Japan) for taxonomic classification.

2.4.4. Stability and Composition of the Resulting Mn Precipitates

The stability of the Mn precipitates recovered from both stirred reactors was evaluated using the toxicity characteristic leaching procedure (TCLP) test, following the method described by Tanaka and Okibe [36]. Freeze-dried Mn precipitate samples (0.05 g) were mixed with 1 mL of acetate buffer (pH 4.93) in screw-capped tubes and rotated at 30 rpm for 18 h at room temperature. After the leaching period, the Mn and Zn concentrations in the resulting leachate were analyzed using ICP-OES.

For compositional analysis, freeze-dried Mn precipitate samples (0.1 g) were treated with 10 mL of aqua regia (HCl:HNO₃ = 3:1) and heated to 210 °C at a rate of 7 °C /min, maintained at this temperature for 20 min, and then cooled to room temperature. Each sample was processed in triplicate, yielding consistent results and confirming the reliability of the method. No solid precipitation was observed after heating, indicating that Mn oxides in the samples were fully dissolved. The resulting solution was filtered through a 0.45 µm membrane to remove particulates. The Mn concentration in the filtrate was then measured using ICP-OES to determine the elemental composition of the precipitates.

3. Results and Discussion

3.1. Liquid Analysis and System Performance

The changes in Mn and Zn concentrations, Mn²⁺ removal efficiency, bicarbonate concentrations, pH, and Eh (vs. SHE) over time in the Mn-system and Mn/Zn-system are shown in Figure 2a–f. These parameters illustrate the system performance under varying HRTs and the effects of Zn²⁺ addition.

The experiment began with an HRT of 10 h (Days 0–9) to establish stable operating conditions and ensure sufficient Mn²⁺ removal (Figure 2a). During this initial phase, Mn²⁺ concentrations in the outlet decreased steadily, achieving removal rates above 90% (Figure 2d). This performance was consistent with previous findings in passive bioprocesses for Mn²⁺ removal at comparable concentrations [34]. Reducing the HRT to 8 h (Days 10–15) and further to 6 h (Days 16–32) demonstrated the system’s ability to maintain Mn²⁺ removal above 85%, highlighting its operational robustness during moderate HRT reductions.

From Day 30, Zn²⁺ was introduced into the Mn/Zn-system at a target concentration of 2 mg/L. Adjustments were made to stabilize Zn²⁺ input (Section 2.2.2), but precipitation during storage led to fluctuations, causing the actual inlet Zn²⁺ concentration to remain around the target range of 2–3 mg/L as ZnCO₃ [37]. This underlines the challenge of achieving consistent Zn²⁺ feed concentrations, a factor also reported in similar studies using carbonate-buffered systems [38].

When the HRT was reduced to 4 h (Days 33–44), Mn²⁺ removal efficiency declined significantly, with outlet Mn²⁺ concentrations exceeding the Japanese effluent standard of 10 mg/L (Figure 2a). Returning the HRT to 5 h (Days 45–54) slightly improved performance, although outlet Mn²⁺ concentrations remained above the effluent standard. Restoring the HRT to 6 h (Days 55–74) successfully re-established Mn²⁺ removal efficiency, achieving compliance with the effluent standard. However, a subsequent attempt to revert to an HRT of 5 h (Days 75–88) again resulted in Mn²⁺ concentrations exceeding the regulatory limit. These observations align with prior findings that shorter HRTs compromise the kinetics of Mn²⁺ oxidation and adsorption [33].

In contrast, Zn²⁺ was effectively removed under all tested HRT conditions (Figure 2b), due to a combination of adsorption onto Mn oxides and precipitation as ZnCO₃ or hydrozincite (Zn₅(CO₃)₂(OH)₆). The role of carbonate in facilitating Zn²⁺ removal was previously demonstrated in abiotic systems [37].

The pH remained stable within a neutral range of 7.0–8.0 throughout the experiment (Figure 2c), which is favorable for Mn²⁺ oxidation and precipitation. Deviations from this range, as reported in other Mn²⁺ treatment studies, are known to increase Mn²⁺ solubility and compromise removal efficiency [28]. Slight increases in pH observed during the addition of Zn²⁺ (from Day 30) can be attributed to CO₂ degassing triggered by ZnCO₃ precipitation. Specifically, as Zn²⁺ reacts with CO₃²⁻ to form ZnCO₃, the removal of CO₃²⁻ from solution disturbs the carbonate equilibrium, causing a reduction in carbonic acid (H₂CO₃).

The Eh values remained between 300 and 400 mV across all conditions (Figure 2f), consistent with redox potentials favorable for Mn²⁺ oxidation. This range also overlaps with the reported thresholds for Mn-oxidizing bacterial activity, suggesting that microbial contributions can supplement abiotic Mn²⁺ oxidation in both systems [11]. Notably, the decline in Mn²⁺ removal efficiency at shorter HRTs was not accompanied by significant changes in Eh, further supporting the hypothesis that insufficient reaction time, rather than unfavorable redox conditions, was the limiting factor.

Bicarbonate concentrations in the inlet remained stable at ~100–120 mg/L, with consistent consumption in the outlet (Figure 2e). This pattern indicates that bicarbonate played a dual role: it participated in reactions such as MnCO₃ formation and served as a pH buffer to stabilize the system. The role of bicarbonate in forming transient MnCO₃ precursors, which are subsequently oxidized to Mn oxides, has been demonstrated in both field and laboratory studies [39]. The current results align with this mechanism.

The observed performance improvements with stepwise HRT adjustments and carbonate buffering in this study mirror findings from our prior work [34], where passive bioprocesses for Mn²⁺/Zn²⁺ removal achieved high efficiency under controlled HRT conditions. Additionally, the robust Mn²⁺ removal observed in this study at HRT 6 h demonstrates the system’s potential for treating high-strength mine waters.

3.2. Mineralogical and Biochemical Dynamics of Mn Oxidation in Two Stirred Tank Reactors

XRD analysis identified birnessite as the predominant Mn oxide in the Mn-sludge during the initial phase of the experiment. The evolution of the XRD peaks in the Mn/Zn-system is shown in Figure 3. Peaks corresponding to birnessite (B) were prominent early in the experiment but weakened progressively after Day 58. Notably, Zn-containing minerals (Z) and calcium carbonate (C) were also detected at Day 0. This observation can be attributed to the presence of trace amounts of Zn²⁺ in the water from which the Mn-sludge was collected, leading to the initial adsorption or precipitation of Zn-containing minerals on the sludge surface.

The decline in birnessite peaks can be attributed to the progressive accumulation of amorphous carbonate precipitates, primarily MnCO₃ and ZnCO₃, on the Mn-sludge surface. García-Sánchez and Álvarez-Ayuso [38] demonstrated that carbonate systems effectively facilitate Zn precipitation as ZnCO₃, which supports the formation of Zn-containing phases observed here.

The presence of Zn²⁺ did not significantly affect birnessite crystallinity, as Zn precipitated as ZnCO₃ or hydrozincite under carbonate-buffered conditions [37]. This is consistent with the sustained HCO₃⁻ consumption observed in liquid analyses (Figure 2e), supporting the hypothesis that carbonate-buffered systems facilitate the co-precipitation of Mn and Zn carbonates over time. This finding aligns with Giannetta et al. (2022), who reported hydrozincite as a common phase in mining environments [40]. While abiotic precipitation remains a dominant pathway, the microbial community identified in this study may have contributed to Zn removal through biologically mediated carbonate production and pH regulation. The results imply that Zn removal in the Mn/Zn-system involved both precipitation as hydrozincite and adsorption onto biogenic Mn oxides. These two mechanisms worked synergistically to enhance Zn removal efficiency and stability, as supported by consistent Zn retention and mineralogical analyses.

Despite the decline in birnessite crystallinity, XANES analysis demonstrated that Mn oxidation continued throughout the experiment (Figure 4). The initial Mn-AOS in the sludge was 3.6, with 80% of Mn present as Mn(IV). By Day 20, the Mn-AOS dropped to 2.6 in the Mn-system and to 3.1 in the Mn/Zn-system, due to the reduced activity of Mn-oxidizing bacteria during the early phase of the experiment. This aligns with findings by Tebo et al. [11], who noted that microbial Mn oxidation is highly dependent on initial bacterial adaptation to the reactor environment. The Mn spectrums of the Mn-system and Mn/Zn-system can be found in Supplementary Materials Figure S1.

Over time, the Mn-AOS gradually increased, reaching 3.7 by Day 85, confirming sustained Mn oxidation. ATP concentration measurements, used as an indicator of microbial activity, showed significant decreases on Day 40 when the HRT was reduced from 6 to 4 h (Figure 4). This reduction in microbial activity is attributed to increased physiological stress induced by the shorter HRT, as supported by prior studies demonstrating that insufficient reaction times can suppress microbial Mn oxidation [32]. However, when the HRT was restored to 5 and then 6 h, ATP concentrations recovered, indicating that microbial activity and Mn oxidation levels are closely tied to HRT adjustments. This is consistent with Hallberg and Johnson’s [27] findings, which emphasized the critical role of HRT optimization in maintaining active microbial communities in bioreactors.

The similar trends in XRD and XANES results between the Mn-system and Mn/Zn-system suggest that the presence of Zn²⁺ did not inhibit the mineralogical or biochemical processes underlying Mn oxidation. This observation contrasts with earlier reports where Zn²⁺ acted as an inhibitor to microbial Mn oxidation under low-pH conditions [33]. In the current study, the neutral pH (~7.5) maintained by carbonate buffering can mitigate Zn²⁺ toxicity, facilitating microbial activity and Mn oxidation. Moreover, Zn²⁺ removal occurred primarily through co-precipitation with Mn oxides or as hydrozincite, as indicated by XRD and supported by Tajima et al. [37]. The recovery of microbial activity at optimized HRTs highlights the resilience of Mn-oxidizing bacteria in adapting to operational changes, as also observed by Hallberg and Johnson [27]. Furthermore, the results suggest that carbonate buffering not only stabilizes pH but also plays a crucial role in enhancing the kinetics of Mn²⁺ oxidation and the stability of resulting Mn oxides, supporting the findings of Namgung and Lee [39].

3.3. Microbial Community Composition and Dynamics in Mn- and Mn/Zn-Systems

The taxonomic composition of the microbial communities at the order level is presented in Figure 5, with the detailed numerical data provided in Supplementary Table S1. The relative abundance of potential Mn-oxidizing microorganisms at the genus level is summarized in

Table 1. Change over time in the relative percentage of potential Mn-oxidizing microorganisms in Mn-sludge. The microbial community composition at the genus level in raw Mn-sludge [33] and during Day 12 and Day 43 of the Mn-system and Mn/Zn-system. The listed genera were previously associated with Mn oxidation in various studies [17,18,19,20,21,22,41,42,43], but direct experimental evidence for Mn oxidation may be lacking for some of them.

Genus	Relative Abundance of Potential Mn-Oxidizing Microorganisms (%)
	Day 0	Mn-System		Mn/Zn-System
		Day 12	Day 43	Day 12	Day 43
Hyphomicrobium	2.08	8.16	4.74	5.28	3.66
Nitrospira	0.88	1.56	1.92	0.61	0.95
Sphingomonas	-	0.24	0.31	0.29	1.97
Sphingopyxis	0.67	1.31	0.727	1.27	1.00
Rhodococcus	0.39	0.07	0.35	0.05	0.24
Pseudonocardia	0.36	1.31	0.58	1.19	1.14
Pedomicrobium	0.34	0.04	0.21	0.07	0.17
Pseudomonas	0.25	0.22	0.78	0.19	0.22
Variovorax	0.18	0.11	0.09	0.1	0.07
Mycobacterium	0.17	0.08	0.00	0.05	0.13
Streptomyces	0.12	0.03	0.19	0.04	0.10
Rhodobacter	0.11	1.83	0.77	1.14	0.47
Caulobacter	0.11	0.30	0.24	0.11	0.14
Terrimonas	0.10	0.04	0.15	0.03	0.08
Afipia	0.08	0.13	0.227	0.17	0.19
Paracoccus	0.06	0.08	0.03	0.03	0.04
Cupriavidus	0.05	0.09	0.05	0.05	0.05
Shinella	0.04	0.03	0.055	0.02	0.08
Microbacterium	0.02	0.004	-	-	0.01
Corynebacterium	0.02	0.004	-	-	0.004
Propionibacterium	0.02	0.004	-	-	0.004
Achromobacter	0.02	0.01	0.01	-	0.04
Chromobacterium	0.02	-	-	-	-
Brevibacillus	-	-	0.004	-	-
Methylobacterium	-	-	-	0.07	0.01
Bosea	-	-	0.05	0.04	0.02
Bacillus	-	0.03	0.06	0.06	0.04
Erythrobacter	-	0.13	0.04	0.05	0.03
Aurantimonas	-	0.12	0.08	0.05	0.02
Rhizobium	-	0.11	0.26	0.13	0.16
Aeromonas	-	0.16	0.31	0.08	0.02
Arthrobacter	-	0.02	-	0.02	0.01
Lapillicoccus	-	-	-	0.004	-
Leifsonia	-	-	0.03	0.004	-
Terrabacter	-	-	0.07	-	0.05
Cellulomonas	-	-	0.01	0.004	0.004
Nocardia	-	0.02	0.09	-	0.03
Geodermatoiphilus	-	-	-	-	0.004
Cytophaga	-	0.004	-	-	-
Flavobacterium	-	0.30	0.35	0.19	0.18
Sphingobacterium	-	0.004	0.01	-	-
Lysinibacillus	-	0.01	0.03	0.04	0.04
Staphylococcus	-	-	-	0.004	-
Agrobacterium	-	0.01	-	0.01	-
Sulfitobacter	-	0.06	0.03	0.02	0.09
Duganella	-	0.02	-	-	-
Total	6.1	16.6	12.8	11.5	11.4

, while the full list of genera is available in Supplementary Table S2. The microbial community in both Mn- and Mn/Zn-systems showed notable compositional changes at both order and genus levels. The interplay of environmental factors, such as HRT and the presence of Zn²⁺, influenced the abundance and dynamics of potential Mn-oxidizing microorganisms.

The proportion of potential Mn-oxidizing microorganisms increased from 6.1% (Day 0) to 17% (Day 12) in the Mn-system and then stabilized at 13% by Day 43 (Table 1). In the Mn/Zn-system, this increase was slightly more moderate, reaching 12% by Day 12 and stabilizing at 11% by Day 43. Despite Zn presence, Mn-oxidizing bacteria increased in both systems, indicating a strong microbial response to Mn exposure. Distinct microbial community characteristics were observed between the Mn- and Mn/Zn-systems, as follows:

3.3.1. Microbial Community Composition in the Mn-System

In the raw Mn-sludge (Day 0), Hyphomicrobiales (5.1%) and Burkholderiales (2.9%) were the dominant orders (Figure 5 and Supplementary Table S1). Hyphomicrobiales include genera such as Hyphomicrobium and Pedomicrobium (Table 1). P. manganicum [44] and P. australicum [45] are well-documented Mn-oxidizing bacteria, actively contributing to Mn removal in biogeochemical systems. In contrast, while Hyphomicrobium is frequently detected in Mn-rich environments, its direct Mn-oxidizing ability remains unconfirmed [24]. Burkholderiales include genera such as Variovorax and Cupriavidus (Table 1). Variovorax paradoxus was identified among bacterial strains capable of Mn oxidation in Antarctic soils [41]. Bai et al. [42] isolated and identified Cupriavidus sp. HY129 as an efficient denitrifying and Mn-oxidizing bacterium.

By Day 12, the relative abundance of Hyphomicrobiales increased significantly to 12.4% (Figure 5), primarily due to the enrichment of Hyphomicrobium (from 2.1% to 8.2%; Table 1). Similarly, Burkholderiales increased to 7.8% (Figure 5), though this rise was not strongly driven by known Mn-oxidizing genera (Table 1). Among other potential Mn-oxidizing genera, Pseudonocardia, Nitrospira, Sphingopyxis, and Rhodobacter also increased noticeably (Table 1). Nitrospira has been implicated in Mn oxidation in aquatic and sedimentary environments, contributing to the biogeochemical cycling of Mn. Wang et al. [46] demonstrated that comammox Nitrospira was significantly correlated with Mn-oxidizing microbes in the Phragmites rhizosphere, suggesting its potential involvement in Mn oxidation. Liang et al. (2017) [47] found that a co-culture of Arthrobacter sp. and Sphingopyxis sp. exhibited Mn-oxidizing activity, despite neither being able to do so individually. Kalman et al. (2003) [48] demonstrated that modifying the reaction center of Rhodobacter sphaeroides enables the photochemical oxidation of Mn²⁺, although its role in natural Mn oxidation remains unclear.

By Day 43, the microbial community underwent further compositional changes, influenced by the shortened HRT (4–5 h; Figure 2). While Hyphomicrobium declined to 4.7%, Nitrospira exhibited greater persistence (0.9% on Day 0, 1.6% on Day 12, and 1.9% on Day 43) than other potential Mn-oxidizing genera (Table 1). This suggests that taxa capable of maintaining Mn oxidation under fluctuating conditions were favored.

3.3.2. Microbial Community Composition in the Mn/Zn-System

The microbial community composition in the Mn/Zn-system demonstrated significant temporal changes, reflecting the combined influence of Mn and Zn, as well as the shortened HRT by Day 43. As in the Mn-system, Hyphomicrobiales increased from 5.1% to 10% by Day 12, followed by a moderate decline to 7.3% by Day 43 (Figure 5 and Supplementary Table S1). This trend was primarily driven by the population shift of Hyphomicrobium (5.3% on Day 12 and 3.7% on Day 43) (Table 1). Similarly, Burkholderiales increased from 2.9% to 5.9% on Day 12 and remained at 5.7% on Day 43 (Figure 5 and Supplementary Table S1). However, as in the Mn-system, this rise was not strongly driven by known Mn-oxidizing genera, such as Variovorax and Cupriavidus (Table 1).

Among other potential Mn-oxidizing genera, Sphingomonas, Sphingopyxis, Pseudonocardia, and Rhodobacter exhibited notable development in the Mn/Zn-system (Table 1). The key difference between the Mn- and Mn/Zn-systems was the persistence of Sphingomonas, Sphingopyxis and Pseudonocardia on Day 43 (Table 1).

Sphingomonas and Sphingopyxis, both belonging to the order Sphingomonadales, exhibited higher persistence in the Mn/Zn-system, likely due to their Zn resistance mechanisms. The relative abundance of Sphingomonadales increased from 1.4% to 4.8% by Day 12 and further to 7.2% by Day 43. Sphingomonas has been reported to carry a substantial number of zinc resistance genes, particularly those involved in key transport and efflux systems for Zn. Additionally, Lombardino et al. (2022) [49] showed that Sphingomonas spp. possess multiple genes responsible for oxidative stress response and heavy-metal resistance, which may provide an advantage for survival in Mn/Zn-rich environments.

For Pseudonocardia, which belongs to the order Actinomycetales, direct Zn resistance mechanisms remain unclear. However, this genus is known for its production of various secondary metabolites, which facilitate symbiotic relationships with other life forms [50]. These metabolites may contribute to protective mechanisms that enhance Zn tolerance. In contrast, in the Mn-system, where Zn was absent, other Mn-oxidizing bacteria may have had a competitive advantage, leading to the decline of Sphingomonas and Pseudonocardia over time.

Although Nitrospira was less dominant than in the Mn-system, it remained detectable at 0.95% on Day 43, indicating its continued presence in the Mn/Zn system (Table 1).

3.4. Composition and Stability of the Resulting Mn Minerals

The composition and stability of Mn-sludge collected from the Mn- and Mn/Zn-systems on Day 88 were evaluated using acid dissolution tests and the toxicity characteristic leaching procedure (TCLP). The results are summarized in

Table 2. TCLP test results and composition of Mn precipitates collected from the Mn-system and Mn/Zn-system on Day 88. MnO₂(138-09675; Wako) was used as a control. BDL: below the detection limit (<0.01 mg/L).

	TCLP Test				Composition
	Mn Leached		Zn Leached		Mn	Zn
Mn-system Mn/Zn-system Control: MnO2	mg/L	%	mg/L	%	mg/g	mg/g
	1.72	14.7	BDL	<0.01	234.5	BDL
	1.34	11.1	BDL	<0.01	241.4	0.94
	BDL	<0.01	-	-	537.0	-

.

The Mn content in the sludge was 234.5 mg/g in the Mn-system and 241.4 mg/g in the Mn/Zn-system, accounting for approximately half of the theoretical Mn content of birnessite (509 mg/g) or MnO₂ (537 mg/g). This discrepancy can be attributed to multiple factors, including the presence of biomass, structural water in birnessite, and amorphous phases within the sludge matrix. Biomass, consisting of microbial cells and EPS, diluted the Mn content by becoming integrated into the sludge during the biological oxidation process [11]. Additionally, mixed Mn phases, which are commonly observed in biogenic Mn oxides, may further reduce the measured Mn concentration compared to the theoretical Mn content of MnO₂ and birnessite [37].

XRD analysis (Figure 3) confirmed the presence of impurities such as silica (SiO₂), which appears to be physically associated with Mn oxides. The formation of mixed phases and amorphous components could account for both the lower Mn content and the slight increase in Mn leachability under acidic conditions. Such structural imperfections can increase the solubility of Mn oxides, as previously reported by Jensen et al. [51].

The introduction of Zn²⁺ into the Mn/Zn-system did not significantly alter the sludge composition, as indicated by the comparable Mn content between the two systems. Zn²⁺ was successfully captured within the sludge, with a Zn content of 15.2 mg/g, as determined by ICP-OES. This suggests that Zn²⁺ primarily interacts with Mn oxides through adsorption or co-precipitation rather than replacing Mn in the oxide lattice, consistent with findings by Namgung and Lee [39]. Furthermore, the presence of Zn in the Mn/Zn-system did not appear to inhibit Mn oxidation or compromise sludge formation.

The TCLP test results revealed Mn leaching concentrations of 1.7 mg/L in the Mn-system and 1.3 mg/L in the Mn/Zn-system, which are well below the Japanese effluent standard for Mn²⁺ (10 mg/L). However, compared to the pure MnO₂ control, where Mn leaching was below the detection limit, the higher leachability in the sludge reflects the influence of biomass, structural defects, and amorphous Mn phases, as evidenced by XRD analysis and previous studies [11]. Additionally, the slightly acidic conditions in the TCLP test could have caused the dissolution of Mn. Importantly, Zn leaching concentrations were below detection limits (<0.01 mg/L) in the Mn-system and remained low (0.94 mg/g) in the Mn/Zn-system, well within the effluent standard for Zn²⁺ (2 mg/L). The strong association of Zn with Mn oxides through inner-sphere complexation contributed to the environmental stability of the sludge. In addition, Zn removal in the Mn/Zn-system likely involved a combination of precipitation as hydrozincite and adsorption onto biogenic Mn oxides. These two mechanisms may have worked synergistically, enhancing Zn removal efficiency and stability.

Overall, the semi-passive system developed in this study demonstrates a significant advancement in the field of Mn and Zn removal from mine waters. Compared to previously reported microbial Mn oxidation systems, which often require longer HRTs (e.g., 24–48 h; [27,32]), the system described here achieved >97% Mn removal and >80% Zn removal at an HRT as short as 6 h. This substantial reduction in HRT not only enhances the efficiency of the treatment process but also reduces the footprint and operational costs, making it a more practical option for real-world applications. Moreover, while the Zn concentration used in this study (~10 mg/L) is moderate compared to the levels observed in certain mine waters, it reflects a realistic condition for regulatory compliance and controlled evaluation.

While the semi-passive system demonstrated high Mn and Zn removal efficiencies in lab-scale experiments, its applicability to field conditions remains to be assessed. Factors such as variable flow rates, seasonal fluctuations, and microbial stability may influence performance in large-scale systems [52]. Further investigations should focus on evaluating long-term stability, optimizing HRTs, improving reactor design, and refining operational strategies to enhance scalability. These efforts will provide insights into the system’s potential for practical mine water treatment.

4. Conclusions

This study demonstrates the successful development of a lab-scale semi-passive treatment system inspired by the design and functionality of existing neutralization tanks, which could potentially be repurposed for the removal of Mn and Zn from mine waters. By leveraging biomineralization processes, the microbial oxidation of Mn, driven by genera such as Nitrospira and Pseudonocardia, facilitated the formation of stable Mn oxides, predominantly birnessite. These biogenic Mn oxides not only enabled efficient Mn removal (>97%) but also exhibited resilience to Zn co-treatment. The system achieved high removal efficiencies within a short HRT (6 h), offering a significant improvement over traditional passive systems. While this study utilized a lab-scale reactor, the results highlight the feasibility of adapting neutralization tank designs to semi-passive systems, where microbial activity and engineered processes could bridge the gap between passive and active treatments, enhancing both operational efficiency and environmental sustainability. XRD and XANES analyses confirmed the formation of stable Mn oxides, while microbial community dynamics revealed the adaptability of Mn-oxidizing microorganisms. Additionally, the low leachability of Mn and Zn from the resulting sludge validates the environmental stability of these biominerals, supporting their safe disposal or potential reuse.