Development of Electroactive Biofiltration Dynamic Membrane for Enhanced Wastewater Treatment and Fouling Mitigation: Unraveling the Growth Equilibrium Mechanisms of Fouling Layer

Chengxin Niu; Wei Shi; Zhouyan Li; Zhiwei Qiu; Yun Guo; Zhiwei Wang

doi:10.1016/j.eng.2025.02.003

Engineering ›› 2025, Vol. 50 ›› Issue (7) :60 -71. DOI: 10.1016/j.eng.2025.02.003

Research

Article

Development of Electroactive Biofiltration Dynamic Membrane for Enhanced Wastewater Treatment and Fouling Mitigation: Unraveling the Growth Equilibrium Mechanisms of Fouling Layer

Author information +

History +

PDF (5421KB)

Abstract

We developed a strategy involving an electroactive biofiltration dynamic membrane (EBDM) for wastewater treatment and membrane fouling mitigation. This approach utilizes a cathode potential within an anaerobic dynamic membrane bioreactor to establish a growth equilibrium electroactive fouling layer. Over a 240 day operation period, the EBDM exhibited outstanding performance, characterized by an ultralow fouling rate (transmembrane pressure < 2.5 kPa), superior effluent quality (chemical oxygen demand (COD) removal > 93% and turbidity ∼2 nephelometric turbidity units (NTU)), and a 7.2% increase in methane (CH₄) productivity. Morphological analysis revealed that the EBDM acted as a biofilter consisting of a structured, interconnected, multilevel dynamic membrane system with orderly clogging. In the EBDM system, the balanced-growth fouling layers presented fewer biofoulants and looser secondary protein structures. Furthermore, the applied electric field modified the physicochemical properties of the biomass, leading to a decrease in fouling potential. Quartz crystal microbalance with dissipation monitoring analysis indicated that growth equilibrium promoted a looser fouling layer with a lower adsorption mass than did the denser, viscoelastic fouling layer observed in the control reactor. Metagenomic sequencing further demonstrated that continuous electrical stimulation encouraged the development of an electroactive fouling layer with enhanced microbial metabolic functionality on the EBDM. This approach selectively modifies metabolic pathways and increases the degradation of foulants. The EBDM strategy successfully established an ordered-clogging, step-filtered, and balanced-growth electroactive fouling layer, achieving a synergistic effect in reducing membrane fouling, enhancing effluent quality, and improving CH₄ productivity.

Graphical abstract

Keywords

Wastewater treatment / Electroactive biofiltration dynamic membrane / Growth equilibrium / Biomass properties / Microbial metabolism

Cite this article

Download citation ▾

Chengxin Niu, Wei Shi, Zhouyan Li, Zhiwei Qiu, Yun Guo, Zhiwei Wang. Development of Electroactive Biofiltration Dynamic Membrane for Enhanced Wastewater Treatment and Fouling Mitigation: Unraveling the Growth Equilibrium Mechanisms of Fouling Layer. Engineering, 2025, 50(7): 60-71 DOI:10.1016/j.eng.2025.02.003

登录浏览全文

4963

注册一个新账户忘记密码

1. Introduction

The development of urbanization, the depletion of resources, and the increasing scarcity of global freshwater resources are driving a critical shift in wastewater treatment plants (WWTPs) from traditional wastewater management to water reclamation, resource recovery, energy self-sufficiency, and carbon neutrality [1]. In contrast to conventional wastewater treatment methods, anaerobic membrane bioreactors (AnMBRs) have emerged as cutting-edge technologies that combine anaerobic digestion with ultrafiltration (1–100 nm) or microfiltration (0.1–10 μm) techniques [2], [3]. Offering independent control over hydraulic and sludge retention times, AnMBRs present significant technological benefits by delivering efficient solid–liquid separation and producing high-quality effluent through size exclusion mechanisms [4]. Nevertheless, severe membrane fouling, attributed to the adsorption of organic particles, pore-blocking/narrowing, and the deposition of inorganic particles, remains a major challenge hindering the broader application and development of AnMBRs [5], [6].

Dynamic membranes (DMs) strategically utilize a mixture of microorganisms, colloids, solutes, and cell debris to create a biocake layer, also known as a dynamic fouling layer, on support materials. This process enhances solid–liquid separation and results in the retention of biomass [7], [8]. Anaerobic dynamic membrane bioreactors (AnDMBRs) employ DMs as alternatives to traditional ultrafiltration or microfiltration membranes, providing benefits such as reduced membrane module costs, increased membrane flux, and superior protection against membrane fouling [9]. The efficacy of DMs hinges on the formation of a well-structured fouling layer; however, the build-up of foulants on the surface can disrupt long-term functionality by promoting the development of a thick and dense fouling layer [10], [11]. This excessive growth of the fouling layer can lead to increased fouling resistance and reduced membrane flux. Therefore, maintaining an optimal thickness or managing balanced growth of the dynamic fouling layer is crucial to ensure effective membrane filtration.

Over the past decade, the integration of electric fields into membrane bioreactors (MBRs) has been recognized as an innovative strategy to combat membrane fouling. This method leverages mechanisms such as electrochemical oxidation, electroosmosis, electrical repulsion, and electrophoresis to address fouling issues [12], [13], [14]. However, the adoption of electric fields within dynamic membrane filtration systems remains notably limited. Given the unique characteristics of DMs and the emerging evidence of electric fields in mitigating fouling, we propose a novel approach where electric fields enhance the growth and functionality of DMs. Specifically, we hypothesize that applying an electric field can not only support the balanced development of the dynamic fouling layer but also significantly alter microbial metabolic pathways and biomass properties [15], [16]. These changes are anticipated to enhance removal efficiency and optimize DM formation. By leveraging these synergistic effects, our research introduces the concept of an electroactive biofiltration dynamic membrane. This technique holds promise for advancing wastewater treatment technologies by reducing membrane fouling and enhancing overall system performance. Currently, the specific impacts on performance and the mechanisms driving EBDM formation through electric field application remain largely unexplored.

In this study, we established an anaerobic conductive dynamic membrane bioreactor to delve into the growth dynamics and microbial metabolism in the EBDM system. The primary aims of our research include: ① Unraveling the mechanisms behind the formation of EBDM when subjected to an electric field; ② assessing the electroactive characteristics and biofiltration capabilities throughout extended operational periods; and ③ evaluating its efficacy in mitigating membrane fouling, enhancing effluent quality, and boosting methane (CH₄) productivity. This work demonstrates that the presence of electric fields enables the formation of EBDM, potentially paving the way of developing innovative approaches for wastewater treatment.

2. Materials and Methods

2.1. Experimental setup and operation conditions

Two lab-scale AnDMBRs, each with an effective volume of 1.5 L, were operated for synthetic wastewater treatment. One was electrochemical AnDMBR (E-AnDMBR), which was operated with an applied voltage, and the other was a control AnDMBR (C-AnDMBR), which functioned without voltage. A 2 mm thick layer of carbon fiber felt covered with a polyester mesh (∼20 μm pore size) was mounted onto the membrane frame to support the formation of a dynamic membrane. A titanium mesh assembled with carbon fiber felt was placed 1.5 cm away, facing the dynamic membrane module (13 cm × 9.5 cm) and acting as the anode. The anode and the cathode dynamic membrane were linked to a direct current (DC) power (UPT1305, UNIT, China) set at a voltage of 0.6 V. The membrane flux was 15 L·(m²·h)⁻¹, and the trans-membrane pressure (TMP) was monitored via a digital pressure gauge (YB80A, XSENR, China). To enhance mixing, a biogas sparging diffuser was installed at the bottom of the reactor, operating at a rate of 0.1 m³·(m²·h)⁻¹. Both the E-AnDMBR and C-AnDMBR were operated with a hydraulic retention time of 40 h at a temperature of (37 ± 0.5) °C and maintained by a circulating water bath (SC-6, Tianling Instruments Co., Ltd. China). The composition, physical, and chemical indicators, and trace element contents of the synthetic wastewater are provided in Tables S1–S3 in Appendix A.

2.2. Analytical methods

Chemical oxygen demand (COD) was determined via the standard methods outlined by American Public Health Association (APHA; 1998) [17]. The concentrations of proteins and polysaccharides were measured via the modified Lowry–Folin method and the Dubois method, respectively, as detailed by Xu et al. [18]. The extraction of extracellular polymeric substances (EPS) from biomass utilized heat extraction techniques (Text S1 in Appendix A). The conductivity of the fouling layer was assessed via the three-probe conductivity measurement method [19]. The composition of the biogas was examined using a gas chromatograph (6890 N, Agilent, USA) equipped with a thermal conductivity detector, whereas the permeate turbidity was determined via a turbidity meter (2100P, HACH, USA). Measurements of pH and oxidation–reduction potential (ORP) were conducted via pH/ORP meters. The characteristics of EPS were analyzed through three-dimensional (3D) excitation–emission matrix (EEM) fluorescence spectroscopy (Aqualog, Horiba, Japan). The particle size distribution (PSD) and viscosity of the biomass were evaluated via a laser granularity distribution analyzer (Mastersizer 3000, Malvern, UK) and a digital viscometer (NDJ-5S, Lichen, China), respectively. The zeta potential and capillary suction time (CST) of the biomass were measured via a Zetasizer Nano ZS (Malvern Instruments Ltd., UK) and a portable 304B instrument (Triton, UK), respectively.

The particles in the mixed liquor were negatively charged, experienced repulsive forces, and could be driven away from the cathode membrane. Therefore, such repulsive forces have effects similar to those of backwashing [5]. The equivalent backwash flux caused by the velocity of the charged particles induced by the electric field was calculated via the Smoluchowski kinetic equation (Eqs. (1), (2), (3)) [20]:

(1) $E=\frac{\Delta U}{d}$

(2)

v_{p} = \frac{∊ \cdot ξ}{μ} E

(3)

J_{b a c k w a s h} = \frac{Q_{p}}{A} \cdot v_{p}

where E is the electric field strength (V·m⁻¹), ΔU is the voltage applied (V), d is the distance between electrodes (m), v_p is the electrophoresis velocity of charged particles (m·s⁻¹), ε is the electrolyte constant, ζ is the zeta potential (V), J_backwash is the equivalent backwash flux (L·(m²·h)⁻¹), Q_p is the flow rate of water-conveying charged particles (m³·s⁻¹), μ is the viscosity of mixed liquor (Pa·s), and A is the effective membrane area (m²).

2.3. Time-resolved analysis of the biomass aggregation process

Multiple light scattering spectroscopy (MLiSSP; Turbiscan Tower, Formulaction, France) was employed to observe the real-time aggregation behavior of the biomass. This method enables automatic monitoring of both transmitted and backscattered light, providing insights into biomass aggregation dynamics [21]. The average transmission signal change rates (ΔTrans/Δt) were calculated via Eq. (4) to assess aggregation kinetics. The turbiscan stability index (TSI) of biomass was further calculated via Eq. (5).

(4)

\frac{Δ T r a n s}{Δ t} = \frac{\sum_{H_{i}}^{H_{n}} (\frac{{T r a n s}_{(t + Δ t, H)} - {T r a n s}_{(t, H)}}{Δ t})}{H_{n} - H_{i}}

(5)

TSI = \sqrt{\frac{\sum_{i = 1}^{n} {(x_{i} - \bar{x})}^{2}}{n - 1}}

where Δt represents the data collection interval, Trans₍_t ₊ _Δ_t,H₎ and Trans₍_t,H₎ represent transmission signals at height H, measured at time t + Δt and t, respectively. The height range is defined by H_i as the lower boundary and H_n as the upper boundary (15∼35 mm in this study), x_i represents the average backscattering values, iεN⁺,

\bar{x}

represents the average x_i, and n is the number of scans.

2.4. Elucidation of the fouling layer morphology

Fourier transform infrared (FTIR) spectroscopy (Nicolet iS 10, Thermo Fisher Scientific, USA) was used to identify the major functional groups of the fouling layer [22]. Scanning electron microscopy (SEM; HitachiSu 8000, Japan) was used to observe the morphology of the EBDM. The sample preparation procedures were as follows: ① Washing three times with 0.1 mol·L⁻¹ phosphate-buffered saline (PBS; pH = 7.4); ② fixing with 2.5% glutaraldehyde for 4 h at 4 °C; ③ washing six times with PBS; ④ dehydrating with ethanol solutions at concentrations of 50%, 60%, 70%, 80%, 90%, 95%, and 100% (15 min each); and ⑤ air drying [23]. Confocal laser scanning microscopy (Leica, TCS SP8, Germany) was used to investigate the distributions of biofoulants deposited on the membrane surface [24]. ImageJ software was used to obtain the mean fluorescence intensity (MFI) of live/dead cells, proteins and polysaccharides via Eq. (6).

(6)

MFI = \frac{IntDen}{A_{s}}

where IntDen represents the cumulative fluorescence intensity within the scanning regime (a.u.), and A_s represents the area in the scanning regime (mm²).

2.5. Quartz crystal microbalance with dissipation monitoring

A quartz crystal microbalance with dissipation monitoring (QCM-D) was used to characterize the quantity and viscoelastic properties of the deposition layer by monitoring the changes in resonance frequency (Δf) and energy dissipation (ΔD) responses. QCM-D (E4, Q-Sense, Sweden) was utilized to explore the adsorption and desorption behaviors of the EPS layer on the sensor chips, aiming to elucidate the membrane fouling behavior. The QCM-D experiments were conducted under flow-through conditions via a peristaltic pump (IsmaTec, IDEX) at a flow rate of 150 μL·min⁻¹. The Δf detected by the QCM-D is correlated with the mass change (Δm), as determined by the Sauerbrey equation [25].

(7)

Δ m = - C_{f} \frac{Δ f}{r}

where Δm represents the mass adsorbed to the sensor, r represents the overtone number (r = 3, 5, 7, 9, and 11), Δf represents the shift in frequency, and C_f represents the mass sensitivity constant of the crystal.

2.6. Metagenomic sequencing analysis

The samples were collected from the mixed liquor, anode, and cathode dynamic membrane of the E-AnDMBR, and those from the mixed liquor and dynamic membrane of the C-AnDMBR were also sampled for metagenomic sequencing analysis, with each sample including three replicates. The functional genes and metabolic pathways were annotated via the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Further details regarding the metagenomic sequencing analysis can be found in Text S2 in Appendix A.

3. Results and discussion

3.1. Long-term wastewater treatment performance

Fig. 1 show the wastewater treatment efficacy of the two reactors over a 240-day period, highlighting the role of the DM in enhancing permeate quality, especially in terms of turbidity reduction. As shown in Fig. 1(a), both reactors experienced a rapid decrease in permeate turbidity. Permeate turbidity values less than 5 NTU are generally recognized as evidence of effective dynamic membrane formation [26]. Over the course of the study, both reactors consistently delivered high-quality effluent, with turbidity levels maintained at 2 NTU, indicating the efficient removal of the majority of the suspended and colloidal particles. Furthermore, as depicted in Fig. 1(b), both reactors achieved high COD removal efficiencies (> 93%). Specifically, the average effluent COD concentration was (41.4 ± 6.7) mg·L⁻¹ for the E-AnDMBR and (46.2 ± 6.8) mg·L⁻¹ for the C-AnDMBR. Despite the high COD removal efficiency achieved by both systems, the E-AnDMBR stands out for its superior performance in methane conversion and membrane fouling mitigation, which will be discussed in detail later.

The pH levels in the E-AnDMBR and C-AnDMBR systems were maintained at (6.6 ± 0.1) and (6.5 ± 0.1), respectively, as illustrated in Fig. 1(c). Furthermore, the ORP in both systems decreased from −100 to −250 mV, signifying an enhancement of the anaerobic conditions within the reactors [27]. As depicted in Fig. 1(d), the E-AnDMBR achieved a methane production rate of (127.5 ± 9.3) milliliter of methane produced per gram of removed chemical oxygen demand (mLCH₄·g⁻¹COD_re), whereas the C-AnDMBR rate was (118.9 ± 7.7) mLCH₄·g⁻¹COD_re, indicating a 7.2% increase in methane productivity.

Morphological analysis, presented in the inset of Fig. 1(d), revealed that the anode was covered with numerous rod/ball-like cells, forming a biofilm with a larger size and greater density on the electrode surface. Similarly, the cathode surface exhibited densely packed functional microbial clusters, suggesting the development of an electroactive fouling layer within the EBDM that benefits from enhanced metabolic activity induced by the cathode potential [28]. Thus, the observed increase in methane bioconversion and degradation of membrane foulants might be attributed to the facilitation of extracellular electron transfer by electroactive bacteria present on the electroactive fouling layer and anode. This process might also be associated with the amplification of cooperative interactions and synergistic relationships among diverse microbial communities [14].

3.2. Characteristics of the growth equilibrium in EBDM

3.2.1. Antifouling performance of the EBDM

Fig. 2(a) shows that the TMP changes over a 240 day operation period at a constant flux (15 L·(m²·h)⁻¹). Initially, within the first 125 days, the TMP of the C-AnDMBR increased from 1.56 to 2.61 kPa. A notable surge then occurred on day 180, jumping to 7.05 kPa, and further increased to 7.80 kPa by day 240. In contrast, with the application of the cathode potential to the conductive membrane, the EBDM system demonstrated significant fouling mitigation, leading to a dynamic growth equilibrium [12]. The system effectively maintained low levels of membrane fouling; the TMP gradually rose from an initial value of 1.39 to 2.49 kPa over the 240 day operation.

The cathode potential initiated electrophoresis, repelling negatively charged foulants in the mixed liquor from the membrane surface, similar to conventional backwashing effects. As determined by Smoluchowski kinetic equations, the equivalent backwash flux in the E-AnDMBR was determined to be between (8.3 ± 0.1) and (16.6 ± 0.4) mL·(m²·h)⁻¹ (Fig. 2(b)). Although this equivalent backwash flux was relatively small compared with the operational flux, its continuous action played a crucial role in fouling mitigation [5]. Furthermore, electric field application potentially reduced the build-up of foulants such as proteins and polysaccharides, which will be discussed in subsequent analyses. Compared with the control reactor, the E-AnDMBR maintained a stable TMP profile, achieving growth equilibrium in the EBDM.

3.2.2. Microscopic characterization of the EBDM

To explore the effects of the cathode potential on membrane fouling, detailed microscopic analyses of the EBDM were conducted. Soluble microbial products (SMPs) and EPSs are known to significantly influence biomass characteristics and the development of the fouling layer [29]. Fig. 3(a) illustrate that the membrane surface of the C-AnDMBR exhibited a denser accumulation of foulants, with a pronounced presence of microbial clusters, SMPs, and EPSs, in comparison to those of the E-AnDMBR. The enhanced antifouling effect observed in the E-AnDMBR is attributed to the reduced deposition of negatively charged foulants on its membrane surface [30]. Additionally, the applied electric field likely facilitates the polarization of charged particles, aiding in diminishing the density of the fouling layer [31]. SEM images revealed that the polyester mesh surface was densely populated with sludge flocs, colloidal particles, and microbial aggregates, forming a dynamic cake layer that acted as the secondary layer within the EBDM for water purification (Fig. 3(b)). Within the carbon felt layer, substantial sludge accumulation was observed to fill the spaces between the carbon fibers, suggesting that an additional EBDM structural layer formed from substances not fully intercepted by the polyester mesh in the DM’s initial phase (Fig. 3(c)). Cross-sectional views of the carbon fibers validated the existence of this supplementary layer within the EBDM (Fig. 3(d)). The establishment of this extra layer might augment pollutant rejection and biodegradation capabilities [16].

3.2.3. Characterization of the biofouling on EBDM

To reveal the distribution of organic foulants within the EBDM, a combination of multiple staining techniques and confocal laser scanning microscopy (CLSM) was applied to examine the biomass and its metabolic byproducts in the fouling layers. Three-dimensional (3D) reconstruction images highlighted the prevalence of cells, proteins, and polysaccharides on the fouled membrane, with proteins and polysaccharides densely aggregating in biological clusters on the membrane surface (Figs. 4(a) and (b)). The mean fluorescence intensity (MFI) is directly related to the concentration of fluorescent compounds, serving as a gauge for the accumulation of biomass and metabolites [32]. Compared with those on the E-AnDMBR membrane, the biofoulants on the C-AnDMBR membrane resulted in more dead cells and higher concentrations of proteins and polysaccharides. The MFI values for proteins and polysaccharides on the E-AnDMBR membrane were (20.6 ± 4.3) and (20.2 ± 2.0) AU·mm⁻², respectively, lower than those on the C-AnDMBR membrane ((29.3 ± 1.8) AU·mm⁻² for proteins and (26.6 ± 0.9) arbitrary units per square millimeter (AU·mm⁻²) for polysaccharides). In the E-AnDMBR, the CLSM fluorescence intensity revealed 29.6% and 24.8% reductions in proteins and polysaccharides, respectively. Owing to their large size and gelation behavior, polysaccharides are key contributors to membrane fouling [33]. Additionally, polysaccharides can further interact with proteins within biopolymers, forming noncovalent networks with greater fouling potential, which exacerbates irreversible fouling [34].

Furthermore, the MFI for living cells in EBDM was 26.5% greater than that of the control reactor’s fouling layer. Moreover, the conductivity of the fouling layer in the E-AnDMBR ((5.65 ± 0.84) µs·cm⁻¹) was 1.87 times greater than that in the C-AnDMBR ((2.98 ± 0.96) µs·cm⁻¹; Fig. S2 in Appendix A), indicating the role of the cathode potential in the formation of an electroactive fouling layer. The underlying mechanisms include the electrophoretic repulsion of negatively charged foulants away from the membrane and the establishment of an electroactive fouling layer. These electroactive microorganisms could increase pollutant degradation and promote the balanced growth of EBDM.

Microbial aggregation and the formation of the fouling layer are significantly influenced by the secondary structures of proteins, notably through the amide I band (1700–1600 cm⁻¹) corresponding to C=O stretching vibrations in peptide groups [35]. Fig. 4(c) shows that the FTIR peaks in the 1600–1700 cm⁻¹ range, associated with amide I in proteins, have relatively high transmittance rates in the C-AnDMBR, suggesting a fouling layer with a potentially high protein content. Protein secondary structures include aggregated strands, β-sheets, α-helices, 3-turn helices, and antiparallel β-sheets/aggregated strands [36]. The tendency of microbial clusters to aggregate, attach, and clump may be facilitated by α-helix structures and reduced by β-sheet structures [35]. The presence of antiparallel β-sheets can decrease intercellular aggregation, as indicated by the fouling layer in the E-AnDMBR, which contains a greater proportion of antiparallel β-sheets (33.8% ± 8.0%) compared to C-AnDMBR ((27.6% ± 2.2%); Figs. 4(d) and (e)).

Moreover, the ratio of α-helix to (β-sheet + random coil) is a critical indicator of protein conformation, where a higher ratio suggests a more compact protein structure [37]. An increased α-helix/(β-sheet + random coil) ratio indicates more hydrogen bonding within the gel layer, which could lead to the formation of a denser fouling layer [38]. Table 1 shows that the growth-balanced fouling layer in the E-AnDMBR displayed a less dense protein structure, as evidenced by a lower α-helix/(β-sheet + random coil) ratio than that in the C-AnDMBR, confirming the antifouling capacity of the E-AnDMBR.

3.3. Impact of biomass characteristics on the fouling behavior of the EBDM

Exploring the physicochemical characteristics of biomass offers a deeper understanding of strategies for mitigating membrane fouling and achieving growth equilibrium in EBDM. Fig. 5(a) presents comparative data on the zeta potential, mean particle size, viscosity, and CST values for biomass in both the C-AnDMBR and the E-AnDMBR. The zeta potential of sludge in the E-AnDMBR ((−25.8 ± 2.9) mV) was less negative than that in the C-AnDMBR ((−29.4 ± 3.1) mV), indicating a decrease in biomass repulsion and an increase in flocculation capabilities [19]. PSD analysis revealed that electric field application successfully increased the sludge floc size, leading to a greater mean particle size ((68.0 ± 6.9) µm in the E-AnDMBR vs (62.2 ± 6.0) µm in the C-AnDMBR). Additionally, the viscosity of the biomass decreased in the presence of an electrical field, with values of (211.3 ± 21.7) mPa·s in the E-AnDMBR and (238.2 ± 28.2) mPa·s in the C-AnDMBR. Consequently, the CST for biomass in the E-AnDMBR was found to be 16.4% lower than that in the C-AnDMBR. The change in biomass properties diminished the fouling potential and contributed to the development of a more porous fouling layer [31], [39], [40].

The dynamic aggregation behavior of the biomass was further explored through MLiSSP analysis. Fig. 5(b) presents the transmission spectra, the rate of change in transmission (ΔTrans/Δt), and the Turbiscan stability index (TSI) for the biomass. In the E-AnDMBR, the transmission curves of the biomass displayed marked fluctuations, suggesting that the electric field enhances the flocculation dynamics of the biomass [41]. Notably, within 15 min, the ΔTrans/Δt value for the biomass in the E-AnDMBR peaked at a value greater than that observed in the C-AnDMBR [42]. The TSI further revealed that the biomass in the C-AnDMBR exhibited greater stability, which may lead to more significant membrane fouling.

EPS produced by microorganisms are a significant factor contributing to fouling [43]. The organic compounds within EPS can be categorized into five spectral regions, labeled I–V. These regions are associated with tyrosine-like, tryptophan-like, fulvic acid-like, SMP-like, and humic-like substances, respectively. Fig. 5(c) shows that the fluorescence intensities of various EPS layers were notably high in Regions II and IV, indicating a greater presence of tryptophan-like and SMP-like substances. The EPS fluorescence intensity in the C-AnDMBR was greater than that in the E-AnDMBR. Specifically, the concentrations of soluble EPS (S-EPS), loosely bound EPS (LB-EPS), and tightly bound EPS (TB-EPS) in the C-AnDMBR were measured to be (5.7 ± 2.1), (3.4 ± 1.6), and (6.5 ± 3.1) milligram per gram of volatile suspended solids (mg·g⁻¹-VSS), respectively. In contrast, in the E-AnDMBR, the concentrations were lower, at (4.2 ± 2.3), (2.8 ± 1.5), and (5.1 ± 2.0) mg·g⁻¹-VSS. Thus, the application of an electric field has been shown to reduce the EPS concentration, thereby mitigating EPS-induced membrane fouling [44].

Furthermore, in the C-AnDMBR, the polysaccharide-to-protein ratios (EPSc/EPSp) for the three EPS types were observed to be 1.35, 1.26, and 0.85, which are significantly greater than those in the E-AnDMBR (1.17, 1.05, and 0.82, respectively). Elevated EPSc/EPSp values tend to create a more adhesive fouling layer on the membrane surface, resulting in increased filtration resistance [15], underscoring the effectiveness of electric fields in managing membrane fouling dynamics.

3.4. Adherence and rigid properties of the adhered biomolecules in the fouling layer

To elucidate the mechanisms underlying the formation of a growth-balanced fouling layer by adhered biomolecules, QCM-D was applied to analyze the adhesion, accumulation and fluidity behaviors of these adhered biomolecules quantitatively. Figs. 6(a) and (d) display the results of the resonance frequency (Δf) and energy dissipation (ΔD) measurements for S-EPS, LB-EPS and TB-EPS. Higher |Δf| values represent stronger adsorption capacity and adhesion of the biomass [45]. The |Δf| values of various EPS layers in the C-AnDMBR were greater than those in the E-AnDMBR, indicating that the biomass in the C-AnDMBR tended to form a fouling layer. Additionally, in the C-AnDMBR, the |Δf| values for EPS followed the order TB-EPS > S-EPS > LB-EPS, implying that TB-EPS and S-EPS might result in a more severe fouling layer [46]. Specifically, according to the Sauerbrey equation, the deposition quantities of S-EPS, LB-EPS, and TB-EPS in the C-AnDMBR were greater ((166.0 ± 0.6), (134.3 ± 0.1), and (201.7 ± 1.8) ng·cm⁻², respectively), significantly exceeding those in the E-AnDMBR ((49.9 ± 0.9), (122.8 ± 0.2), and (139.3 ± 0.7) ng·cm⁻², respectively). Fig. 6(c) further illustrate the |ΔD/Δf| values of different EPS layers, with |ΔD/Δf| representing the viscoelastic properties of the adsorption layer. The |ΔD/Δf| values of EPS in the C-AnDMBR were lower than those in the E-AnDMBR, particularly for S-EPS, indicating that EPS components in the C-AnDMBR were more likely to form a denser and more rigid fouling layer [13]. Moreover, the higher values of |ΔD/Δf| in the E-AnDMBR suggest that the application of an electric field can assist the EBDM in developing a relatively loose electroactive fouling layer structure. The results obtained from QCM-D suggested that the application of an electric field mitigated the adsorption of adhered biomolecules, leading to a reduction in the fouling layer thickness, alleviation of membrane fouling, and maintenance of growth equilibrium in EBDM.

3.5. Changes in the metabolic pathways in EBDM

The metabolic characteristics of the fouling layer are crucial determinants of membrane fouling behavior [19]. Figs. 7(a) and (b) display the bacterial communities in the mixed liquor, anode, and electroactive fouling layers of the E-AnDMBR, along with those in the mixed liquor and fouling layers of the C-AnDMBR, categorized at both the phylum and genus levels. Compared with those in the C-AnDMBR, the relative abundances of Proteobacteria (19.3%), Chloroflexi (18.2%), and Actinobacteria (6.5%) in the E-AnDMBR notably increased by 12.7%, 21.8%, and 23.0%, respectively. At the genus level, the enrichment of Mesotoga in the EBDM was 8.1 times greater than that in the fouling layer in the C-AnDMBR. Mesotoga is associated with hydrolysis and acidification processes, indicating its potential to degrade organic foulants within the electroactive fouling layer. Notably, the presence of an electric field specifically increased the relative abundance of Geobacter on the anode to 8.4%, which was substantially greater than that in the other samples (ranging from 0.02% to 0.32%). Geobacter species, known for their ability to reduce the activation energy barrier and facilitate direct electron transfer to methanogens such as Methanosaeta and Methanosarcina, increase methane production through the use of electrically conductive pili (e-pili) or c-type cytochromes for extracellular respiration [47]. Moreover, in the electroactive fouling layer of the E-AnDMBR, there was a significant presence of H₂-producing bacteria, particularly unclassified_p_Anaerolineae populations (11.9%), suggesting the formation of H₂-mediated syntrophic communities. In contrast, the cake layer of the control reactor lacked an increased abundance of functional microbes, indicating that continuous electrical stimulation modifies the ecological niche of the dynamic fouling layer [16].

For the archaeal communities, Euryarchaeota and Archaea constituted the primary archaeal phyla in both systems (Figs. 7 (c) and (d)). The predominant archaeal genera within the mixed liquors of both reactors were Methanothrix (formerly known as Methanosaeta) and Methanoregulaceae. In the mixed liquor of the E-AnDMBR, Methanothrix represented a larger fraction, comprising 41.1%, compared with 34.0% in the C-AnDMBR. Methanothrix, as acetoclastic methanogens, synergizes with Geobacter via direct interspecies electron transfer (DIET), where Geobacter supplies electrons through e-pili to Methanothrix to convert CO₂ into methane [48]. This enrichment of Methanothrix suggests that electric field application enhances the acetoclastic methanogenesis pathway in the mixed liquor. Conversely, hydrogenotrophic Methanobacterium emerged as the most metabolically active methanogen in the electroactive fouling layer, accounting for a significantly larger proportion of 50.5%, compared with 17.5% at the anode and merely 1.1%–3.3% in the other samples. This distribution indicates a distinct shift toward hydrogenotrophic methanogenesis, particularly within the EBDM.

To further explore the metabolic characteristics of the EBDM, we investigated the primary methane metabolism pathways (i.e., M00567, M00357, M00356, and M00563) and their associated enzyme-encoding genes. As depicted in Fig. S1, EBDM presented the highest relative abundance of these metabolic pathways, underscoring the effectiveness of EBDM in the E-AnDMBR as a highly electroactive fouling layer driven by the cathode potential. Notably, the proportions of M00357 (acetate decarboxylation) and M00567 (CO₂ reduction) in the EBDM were significantly greater than those in the fouling layer in the C-AnDMBR (12.6% vs 9.6% and 11.5% vs 6.7%, respectively), indicating an increase in both the acetoclastic and hydrogenotrophic pathways.

Fig. 7(e) highlights the dominance of the acetoclastic methanogenesis pathway in mixed liquors, the EBDM, and the fouling layer in the control test, with related genes present in proportions ranging from 0.19% to 0.21%. Metagenomic sequencing further revealed a marked increase in CO₂ reduction and the methylotrophic pathway within the EBDM, resulting in higher gene abundances than those in the control reactor fouling layer (0.11% vs 0.06% and 0.13% vs 0.07%, respectively). Additionally, crucial enzymes of M00567 involved in the hydrogenotrophic pathway within the EBDM, including formylmethanofuran dehydrogenase [EC: 1.2.7.12], tetrahydromethanopterin methyltransferase [EC: 2.1.1.86], formylmethanofuran-tetrahydromethanopterin N-formyltransferase [EC: 2.3.1.101], methenyltetrahydromethanopterin cyclohydrolase [EC: 3.5.4.27], and coenzyme F420 hydrogenase [EC: 1.2.98.1], presented greater relative abundances than did the fouling layer of the control group (0.23% vs 0.11%, p < 0.01). These findings suggest that electric field application alters the metabolic pathways associated with EBDM. This, in turn, stimulates the electrochemical activity of the EBDM and enhances foulant degradation. Consequently, the E-AnDMBR is more effective at controlling fouling layer properties and achieving the growth equilibrium of the EBDM.

Based on the abovementioned results, an efficiently structured, growth-balanced EBDM is schematically shown in Fig. 8(a). EBDM resulted in a greater quantity of viable cells and fewer biological clusters of proteins and polysaccharides, which was quite different from the dense fouling layer in the control reactor depicted in Fig. 8(b). Furthermore, the electric field promoted modifications in the biomass characteristics, such as the zeta potential, PSD, flocculation ability, and EPS distribution, which all play a role in reducing the degree of membrane fouling. Additionally, the cathode potential notably influenced the metabolic behavior of the biofilm, particularly by increasing hydrogenotrophic metabolism. In conclusion, by integrating a cathode potential within an AnDMBR, we created an electroactive dynamic membrane that significantly enhanced wastewater treatment efficiency. This study emphasizes the potential of combining an electric field with AnDMBR technology to create an EBDM system capable of superior treatment efficiency and fouling control. The findings also contribute to a better understanding of the interactions between electric fields and electroactive biofilms, presenting a promising engineering solution for enhancing membrane performance and process sustainability in wastewater treatment systems.

4. Conclusions

We created a novel EBDM system that significantly enhances wastewater treatment and reduces membrane clogging. By integrating a cathode potential within a bioreactor, this system forms a self-regulating, electroactive layer. Over 240 days, the EBDM showed remarkable results with minimal clogging, resulting in high-quality water with over 93% COD removal and turbidity of approximately 2 NTU. Additionally, it increased methane production by 7.2%, highlighting its dual benefit in both wastewater treatment and energy recovery. EBDM operates as an advanced biofilter with a well-structured, interconnected, multilevel dynamic membrane. The electric field also adjusted the properties of the biomass, leading to a decreased clogging potential and a less dense fouling layer. Overall, EBDM effectively controlled clogging, enhanced water quality, and increased methane production through a structured and electroactive layer, representing a significant advancement in wastewater treatment with practical implications for future engineering applications.

CRediT authorship contribution statement

Chengxin Niu: Methodology, Investigation. Wei Shi: Methodology, Investigation. Zhouyan Li: Methodology, Investigation. Zhiwei Qiu: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization. Yun Guo: Methodology, Investigation. Zhiwei Wang: Conceptualization, Funding acquisition, Resources, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Financial support by Natural Science Foundation of China (52430001) is acknowledged.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.eng.2025.02.003.

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Vinardell S, Astals S, Peces M, Cardete MA, Fernández I, Mata-Alvarez J, et al. Advances in anaerobic membrane bioreactor technology for municipal wastewater treatment: a 2020 updated review. Renew Sustain Energy Rev 2020; 130:109936.

[2]	Cheng H, Li Y, Guo G, Zhang T, Qin Y, Hao T, et al. Advanced methanogenic performance and fouling mechanism investigation of a high-solid anaerobic membrane bioreactor (AnMBR) for the co-digestion of food waste and sewage sludge. Water Res 2020; 187:116436.

[3]	Wang C, Ding M, Ng TCA, Ng HY. Mechanistic insights into simultaneous reversible and irreversible membrane fouling control in a vibrating anaerobic membrane bioreactor for sustainable municipal wastewater treatment. Chem Eng J 2023; 466:143226.

[4]	Wang L, Wu Y, Fu Y, Deng L, Wang Y, Ren Y, et al. Low electric field assisted surface conductive membrane in AnMBR: strengthening effect and fouling behavior. Chem Eng J 2022; 431:133185.

[5]	Zhang S, Yang K, Liu W, Xu Y, Hei S, Zhang J, et al. Understanding the mechanism of membrane fouling suppression in electro-anaerobic membrane bioreactor. Chem Eng J 2021; 418:129384.

[6]	Guo Y, Liu C, Liu H, Zhang J, Li H, Zhang C. Contemporary antibiofouling modifications of reverse osmosis membranes: state-of-the-art insights on mechanisms and strategies. Chem Eng J 2022; 429:132400.

[7]	Millanar-Marfa JMJ, Corpuz MVA, Borea L, Cabreros C, De MDG Luna, Ballesteros F, et al. Advanced wastewater treatment and membrane fouling control by electro-encapsulated self-forming dynamic membrane bioreactor. npj Clean Water 2022; 5(1):38.

[8]	Siddiqui MA, Biswal BK, Saleem M, Guan D, Iqbal A, Wu D, et al. Anaerobic self-forming dynamic membrane bioreactors (AnSFDMBRs) for wastewater treatment—recent advances, process optimization and perspectives. Bioresour Technol 2021; 332:125101.

[9]	Ersahin ME, Ozgun H, Tao Y, van Lier JB. Applicability of dynamic membrane technology in anaerobic membrane bioreactors. Water Res 2014; 48:420-429.

[10]	Saleem M, Alibardi L, Cossu R, Lavagnolo MC, Spagni A. Analysis of fouling development under dynamic membrane filtration operation. Chem Eng J 2017; 312:136-143.

[11]	Sun M, Wang X, Winter LR, Zhao Y, Ma W, Hedtke T, et al. Electrified membranes for water treatment applications. ACS EST Eng 2021; 1(4):725-752.

[12]	Chen M, Li Y, Sun X, Dai R, Zheng J, Wang X, et al. Recent advances in electrochemical processes integrated with anaerobic membrane bioreactor in wastewater treatment. Chem Eng J 2023; 468:1438222.

[13]	Deng H, Ren H, Fan J, Zhao K, Hu C, Qu J. Membrane fouling mitigation by coagulation and electrostatic repulsion using an electro-AnMBR in kitchen wastewater treatment. Water Res 2022; 222:118883.

[14]	Zhang Y, Gu K, Zhao K, Deng H, Hu C. Enhancement of struvite generation and anti-fouling in an electro-AnMBR with Mg anode-MF membrane module. Water Res 2023; 230:119561.

[15]	Niu C, Li B, Wang Z. Using artificial intelligence-based algorithms to identify critical fouling factors and predict fouling behavior in anaerobic membrane bioreactors. J Membr Sci 2023; 687:122076.

[16]	Cheng P, Wang C, Zhao X, Liu J, Xu X, Zhu L. Electroactive biocake layer-driven advanced removal of dissolved organic matter at membrane interface of anaerobic electrochemical membrane bioreactor. Water Res 2023; 245:120618.

[17]	American Public Health Association (APH A). Standard methods for the examination of water and sewage.9th ed. Washington, D C: APHA Press; 1998.

[18]	Xu B, Ng TCA, Huang S, He M, Varjani S, Ng HY. Quorum quenching affects biofilm development in an anaerobic membrane bioreactor (AnMBR): from macro to micro perspective. Bioresour Technol 2022; 344:126183.

[19]	Wang L, Yu Q, Sun C, Zhu Y, Wang Z, Zhang Y. Intermittent voltage induced sludge polarization to enhance anaerobic digestion. Water Res 2022; 224:119071.

[20]	Zhang J, Satti A, Chen X, Xiao K, Sun J, Yan X, et al. Low-voltage electric field applied into MBR for fouling suppression: performance and mechanisms. Chem Eng J 2015; 273:223-230.

[21]	Zhang J, Wang Z, Wang Q, Ma J, Cao J, Hu W, et al. Relationship between polymers compatibility and casting solution stability in fabricating PVDF/PVA membranes. J Membr Sci 2017; 537:263-271.

[22]	Mahat SB, Omar R, Che H Man, Md AI Idris, Mustapa SM Kamal, Abdullah LC, et al. Dynamic anaerobic membrane bioreactor (DAnMBR) with phase separation for food processing wastewater treatment at mesophilic temperature: characterization of cake layer. J Environ Chem Eng 2021; 9(4):105718.

[23]	Niu C, Pan Y, Lu X, Wang S, Zhang Z, Zheng C, et al. Mesophilic anaerobic digestion of thermally hydrolyzed sludge in anaerobic membrane bioreactor: long-term performance, microbial community dynamics and membrane fouling mitigation. J Membr Sci 2020; 612:118264.

[24]	Yu Q, Zhang Y. Fouling-resistant biofilter of an anaerobic electrochemical membrane reactor. Nat Commun 2019; 10(1):4860.

[25]	Sauerbrey G. The use of quartz oscillators for weighing thin layers and for microweighing. Z Phys 1959; 155:206-222.

[26]	Mohan SM, Nagalakshmi S. A review on aerobic self-forming dynamic membrane bioreactor: formation, performance, fouling and cleaning. J Water Process Eng 2020; 37:101541.

[27]	Wang R, Lv N, Li C, Cai G, Pan X, Li Y, et al. Novel strategy for enhancing acetic and formic acids generation in acidogenesis of anaerobic digestion via targeted adjusting environmental niches. Water Res 2021; 193:116896.

[28]	Jiang Z, Yu Q, Sun C, Wang Z, Jin Z, Zhu Y, et al. Additional electric field alleviates acidity suppression in anaerobic digestion of kitchen wastes via enriching electro-active methanogens in cathodic biofilms. Water Res 2022; 212:118118.

[29]	Flemming HC. Biofouling and me: my Stockholm syndrome with biofilms. Water Res 2020; 173:115576.

[30]	Choi G, Choi H, Park J, Lee C. Effects of applying external voltage on development of anaerobic dynamic membrane under high suspended solids conditions. Chem Eng J 2022; 438:135528.

[31]	Sun J, Hu C, Tong T, Zhao K, Qu J, Liu H, et al. Performance and mechanisms of ultrafiltration membrane fouling mitigation by coupling coagulation and applied electric field in a novel electrocoagulation membrane reactor. Environ Sci Technol 2017; 51(15):8544-8551.

[32]	Shao Y, Zhou Z, Jiang J, Jiang LM, Huang J, Zuo Y, et al. Membrane fouling in anoxic/oxic membrane reactors coupled with carrier-enhanced anaerobic side-stream reactor: effects of anaerobic hydraulic retention time and mechanism insights. J Membr Sci 2021; 637:119657.

[33]	Seviour T, Pijuan M, Nicholson T, Keller J, Yuan Z. Gel-forming exopolysaccharides explain basic differences between structures of aerobic sludge granules and floccular sludges. Water Res 2009; 43(18):4469-4478.

[34]	Neemann F. Non-covalent protein—polysaccharide interactions and their influence on membrane fouling. J Membr Sci 2013; 446:310-317.

[35]	Yin C, Meng F, Chen GH. Spectroscopic characterization of extracellular polymeric substances from a mixed culture dominated by ammonia-oxidizing bacteria. Water Res 2015; 68:740-749.

[36]	Li Y, Zhu Y, Wang D, Yang G, Pan L, Wang Q, et al. Fe(II) catalyzing sodium percarbonate facilitates the dewaterability of waste activated sludge: performance, mechanism, and implication. Water Res 2020; 174:115626.

[37]	Hou X, Liu S, Zhang Z. Role of extracellular polymeric substance in determining the high aggregation ability of anammox sludge. Water Res 2015; 75:51-62.

[38]	Lei Z, Wang J, Leng L, Yang S, Dzakpasu M, Li Q, et al. New insight into the membrane fouling of anaerobic membrane bioreactors treating sewage: physicochemical and biological characterization of cake and gel layers. J Membr Sci 2021; 632:119383.

[39]	Bani-Melhem K, Elektorowicz M. Development of a novel submerged membrane electro-bioreactor (SMEBR): performance for fouling reduction. Environ Sci Technol 2010; 44(9):3298-3304.

[40]	Yu B, Sun J, Zhao K, Tian J, Hu C. Low-maintenance anti-fouling and phosphorus removal of an electro-MBR with Fe anode–cathodic membrane. J Membr Sci 2023; 672:121417.

[41]	Wu W, Ma J, Xu J, Wang Z. Mechanistic insights into chemical conditioning by polyacrylamide with different charge densities and its impacts on sludge dewaterability. Chem Eng J 2021; 410:128425.

[42]	Hu D, Liu L, Liu W, Yu L, Dong J, Han F, et al. Improvement of sludge characteristics and mitigation of membrane fouling in the treatment of pesticide wastewater by electrochemical anaerobic membrane bioreactor. Water Res 2022; 213:118153.

[43]	Zhou Z, Meng F, He X, Chae SR, An Y, Jia X. Metaproteomic analysis of biocake proteins to understand membrane fouling in a submerged membrane bioreactor. Environ Sci Technol 2015; 49(2):1068-1077.

[44]	Zhao K, Su F, Gu K, Qi J, Liu R, Hu C. Antifouling potential and microbial characterization of an electrochemical anaerobic membrane bioreactor utilizing membrane cathode and iron anode. Bioresour Technol 2021; 334:125230.

[45]	Gutman J, Walker SL, Freger V, Herzberg M. Bacterial attachment and viscoelasticity: physicochemical and motility effects analyzed using quartz crystal microbalance with dissipation (QCM-D). Environ Sci Technol 2013; 47(1):398-404.

[46]	Xiao K, Wang K, Yu S, Yuan Y, Qin Y, An Y, et al. Membrane fouling behavior in membrane bioreactors for nitrogen-deficient wastewater pretreated by ammonium ion exchange. J Membr Sci 2023; 665:121087.

[47]	Lovley DR, Holmes DE. Electromicrobiology: the ecophysiology of phylogenetically diverse electroactive microorganisms. Nat Rev Microbiol 2022; 20(1):5-19.

[48]	Yan Y, Zhang J, Tian L, Yan X, Du L, Leininger A, et al. DIET-like mutualism of Geobacter and Methanogens at specific electrode potential boosts production of both methane and hydrogen from propionate. Water Res 2023; 235:119911.