Advanced wastewater treatment and membrane fouling control by electro-encapsulated self-forming dynamic membrane bioreactor

Membrane bioreactors (MBRs) achieve superior wastewater treatment but are hindered by membrane fouling and high cost. Self-forming dynamic membrane bioreactors (SFDMBRs) use inexpensive woven/non-woven supports on which a biofilm forms to act as the filtration layer, delivering high COD/DOC removal and good NH3-N removal but relatively poor total nitrogen (TN) and phosphorus (TP) removals. Prior work integrating electric fields into MBRs (e-MBRs) enhanced phosphate removal via anodic metal dissolution (Al/Fe) and reduced fouling by lowering soluble microbial products through electrocoagulation, electroosmosis, electrophoresis, and quorum quenching. The present study introduces an encapsulated self-forming dynamic membrane bioreactor with applied electric field (e-ESFDMBR), where the biofilm is protected between two polyester meshes and an Al anode–steel cathode pair induces electrochemical processes. The research goal is to compare e-ESFDMBR to an ESFDMBR (no electric field) for removal of nutrients (TN, TP), organic matter (COD, DOC), humic-like substances, and for membrane fouling control, and to elucidate differences in dynamic membrane morphology, fouling precursors, and microbial community structure that underpin performance differences.

• SFDMBRs reliably remove organic matter (COD/DOC reductions ~89–99%) and NH3-N (76–99%) but show limited TN (21–80%) and TP (6–27%) removals.
• Electro-assisted MBRs have achieved up to 96% PO4-P removal via anodic Al/Fe release and subsequent precipitation; co-precipitation aids removal of recalcitrant organics and pathogens.
• Electric fields in MBRs have reduced fouling by 30–67% through decreases in SMP/EPS and via electrocoagulation, electroosmosis, electrophoresis, and quorum quenching effects.
• Improved floc properties and particle size distributions under electric fields can stabilize dynamic membranes and mitigate fouling.

• Reactors: Two parallel 19 L PMMA cylindrical bioreactors operated aerobically: (i) ESFDMBR (no electric field) and (ii) e-ESFDMBR (with electric field). Each contained a central filtration module (PMMA frame) with Dacron mesh (30 µm pore size; effective area 0.021 m²), designed to form and encapsulate the self-forming dynamic membrane (SFDM) between two meshes.
• Operation: Continuous feed of synthetic municipal wastewater; effluent flux 30 LMH; HRT 25 h; cyclic operation 10 min (9 min filtration, 1 min backwash). Aeration via bottom diffusers. Permeate suction with Watson-Marlow Qdos 30 pumps; feed and backwash with 3235 peristaltic pumps.
• Electric field (e-ESFDMBR only): Concentric perforated electrodes: internal stainless-steel cathode near the membrane, external aluminium anode near bioreactor wall. DC power supply applied intermittent current density 0.5 mA/cm² (≈8.6 V), 5 min ON/20 min OFF.
• Synthetic wastewater: Prepared per literature, containing glucose (200 mg/L), sucrose (200 mg/L), proteins (68.3 mg/L), nitrogen sources (NH4)2SO4 (66.7 mg/L), NH4Cl (10.9 mg/L), phosphates (KH2PO4 4.44 mg/L; K2HPO4 9.0 mg/L), MgSO4·7H2O (21 mg/L), MnSO4·H2O (2.7 mg/L), NaHCO3 (30 mg/L), CaCl2·6H2O (19.74 mg/L), FeCl3·6H2O (0.14 mg/L). Inoculum: activated sludge from municipal WWTP (Salerno, Italy).
• Monitoring and analyses: Daily sampling of influent, bioreactor mixed liquor, and effluent. Continuous pH, temperature, DO, ORP. Turbidity (NTU), MLSS/MLVSS, COD, DOC, TN, NH3-N, NO3-N, PO4-P per standard methods. DOC via UV-persulfate oxidation/IR detection; NO3-N and PO4-P by ion chromatography. Humic-like substances by UV254 absorbance (A254). TMP continuously recorded (Omega PX409-015VI, Agilent 34972A data logger). Fouling rate calculated as ΔTMP/Δt (kPa/d). Fouling precursors TEP, EPS (carbohydrates/proteins), SMP (carbohydrates/proteins) measured three times weekly; concentrations normalized to MLVSS. Aluminium in mixed liquor by ICP-OES after mineralization.
• Particle size distribution (PSD): Mastersizer 3000 laser diffraction.
• Morphology: SEM (ZEISS LEO 1525). Samples fixed with 2.5% glutaraldehyde, ethanol-dehydrated, dried with supercritical CO2; both internal ESFDM and external cake layers imaged.
• Microbiology (culturable fraction): Serial dilutions plated on PCA; fungi on YEPD. Effluent filtered (0.45 µm) and cultured. Endospore-forming bacteria selected by heat (80 °C, 15 min) before plating. CFU normalized to g dry weight for layers/mixed liquor or mL for effluent. Isolates purified; 16S rDNA (V4–V5) and fungal ITS amplified and Sanger sequenced; taxonomic assignment via NCBI BLASTn. Diversity assessed by Shannon (H) and reciprocal Simpson (1/D).

• Rapid SFDM formation and superior permeate quality: e-ESFDMBR effluent turbidity fell below 1 NTU on day 1 and remained near clean-water levels; ESFDMBR reached <5 NTU after ~7 days. Average turbidity: ESFDMBR 3.87 ± 1.34 NTU; e-ESFDMBR 0.21 ± 0.09 NTU (99.29 ± 0.31% reduction).
• Fouling mitigation: e-ESFDMBR maintained lower, more stable TMP (≈1.0–1.5 kPa; max ~1.5 kPa) versus ESFDMBR (rise to 2.30 kPa, with later fluctuations). Average fouling rates: 0.032 kPa/d (e-ESFDMBR) vs 0.105 kPa/d (ESFDMBR).
• Fouling precursors and sludge properties: Significant reductions in e-ESFDMBR of SMPc (65.4%), SMPp (78.1%), EPSc (76.1%), EPSp (72.3%), and TEP (25.4%) relative to ESFDMBR; sludge relative hydrophobicity lower (32.5% vs 63.4%).
• Morphology and PSD: e-ESFDMBR exhibited larger flocs (electrocoagulation) in the cake layer and a more porous cake morphology; ESFDM layers in both systems were compact. Electrokinetic effects (electrophoresis/electroosmosis) likely diverted charged foulants away from the membrane surface.
• Phosphorus removal: PO4 3−-P removal averaged 14.2 ± 9.3% in ESFDMBR vs >99.9% (complete) in e-ESFDMBR, attributed mainly to Al3+ release and formation of Al(OH)3/AlPO4 precipitates. Al concentration in mixed liquor increased from 20.64 mg/L to 1248 mg/L over runtime.
• Organic matter removal: COD removal 94.8 ± 3.4% (ESFDMBR) vs 99.4 ± 0.3% (e-ESFDMBR); DOC removal 93.08 ± 1.27% vs 97.47 ± 0.6%, respectively. Enhanced detention/adsorption on electrochemically generated solids aided biodegradation.
• Humic-like substances: A254 data showed 21.8% higher removal under electric field; effluent A254 remained low despite transient accumulation in the reactor during early operation.
• Nitrogen transformations: NH3-N removal improved from 46.9 ± 11% (ESFDMBR) to 96.0 ± 4% (e-ESFDMBR). NO3−-N concentrations in e-ESFDMBR remained low (<10 mg/L) with similar levels in mixed liquor and effluent. Overall TN removal reached 86.7 ± 8.2% in e-ESFDMBR. Turning on the electric field shifted conditions from aerobic (ORP ~268 mV; DO ~6.5 ppm) to anoxic (ORP ~2.7 mV; DO <1 ppm), favoring denitrification; intermittent operation induced alternating aerobic/anoxic regimes enhancing nitrification–denitrification.
• Microbiology (culturable): Total bacterial counts increased slightly (<1 log10) in e-ESFDMBR mixed liquor but decreased by ~2 log10 in the ESFDM layer versus ESFDMBR. Endospore-forming bacteria decreased under electric field. Proteobacteria dominated in both systems but were less abundant overall in e-ESFDMBR (69.84%) than ESFDMBR (85.25%). In e-ESFDMBR, γ-proteobacteria group (Pantoea/Klebsiella/Raoultella) increased in mixed liquor, while in the ESFDM layer this group decreased (39.13%→9.52%) and Enterobacter spp. increased (17.39%→42.86%). β-proteobacteria were promoted in the ESFDM layer under electric field. Presence of Bacillus spp. (AHL-lactonase producers) in ESFDM layer may have contributed to quorum quenching and fouling precursor reduction.

The e-ESFDMBR architecture—an encapsulated dynamic membrane combined with intermittent low-current electrochemical stimulation—addressed the dual objectives of enhanced nutrient/organic removal and fouling control. Electrocoagulation by anodic Al release generated larger, settleable flocs and Al(OH)3/AlPO4 solids, which adsorbed organics and precipitated phosphate, yielding near-complete PO4-P removal and higher COD/DOC removals. Electrokinetic phenomena (electrophoresis/electroosmosis) redistributed charged foulants, limiting their deposition on the membrane and contributing to low, stable TMP and reduced fouling rates. The electric field shifted redox conditions to anoxic during ON periods, enabling effective denitrification; intermittent operation fostered alternating aerobic/anoxic conditions that, together with improved biomass retention within the ESFDM, achieved high NH3-N and TN removals. Microbial community shifts (e.g., enrichment of Enterobacter in the ESFDM layer, reductions in endospore-forming Bacillus in mixed liquor, and potential AHL-lactonase activity within the layer) are consistent with observed decreases in SMP/EPS/TEP and the improved fouling resistance. Overall, the findings demonstrate that integrating controlled electrochemical processes into an encapsulated SFDMBR substantially improves effluent quality and operational stability relative to a non-electrified ESFDMBR.

A novel electro-encapsulated self-forming dynamic membrane bioreactor (e-ESFDMBR) was developed and benchmarked against an ESFDMBR. Intermittent application of a low current density (0.5 mA/cm²) enhanced PO4-P (>99.9%), NH3-N (to 96.0 ± 4%), TN (to 86.7 ± 8.2%), and humic-like substance removals (by 21.8%), while lowering fouling indicators: turbidity (~0.21 NTU), TMP (~1–1.5 kPa), and fouling rate (0.032 kPa/d vs 0.105 kPa/d). The improvements stem from electrocoagulation/precipitation, electrokinetic foulant control, and beneficial shifts in microbial communities and redox conditions. The e-ESFDMBR shows strong potential as an efficient, stable alternative to conventional MBRs for advanced wastewater treatment. Future research should decouple and quantify the individual contributions of electrocoagulation, electroosmosis, and electrophoresis; apply next-generation sequencing for comprehensive microbiome analysis; determine actual specific energy consumption with process-wise breakdown; assess electrode longevity and aluminium residuals in effluents; and optimize intermittent operation to minimize energy use.

• Energy assessment was not directly measured in this study; a calculated specific energy for electrochemical processes (~4.42 kWh/m³ under the tested intermittent regime) was provided without full process energy breakdown.
• Mechanistic contributions (electrocoagulation vs electroosmosis vs electrophoresis) were not isolated; their combined effect was evaluated.
• Microbial analyses targeted the cultivable fraction via Sanger sequencing of isolates; comprehensive community profiling (e.g., NGS) was not performed.
• Study used synthetic municipal wastewater and lab-scale reactors; performance and energy metrics may differ under real wastewater and larger-scale conditions.
• Electrode lifespan and potential aluminium residuals in effluent were not assessed; pH effects on electrocoagulation costs warrant evaluation.