Bacterial cellulose-graphene oxide composite membranes with enhanced fouling resistance for bio-effluents management

The study addresses the challenge of deploying bacterial cellulose (BC) as a free-standing membrane for pressure-driven water filtration, which is limited by low compressive strength, limited permeance, susceptibility to fouling, and poor selectivity. Native BC contains predominantly hydroxyl groups and lacks antibacterial properties, contributing to microbial and organic fouling. Under compression and dehydration, BC fibers stack and pores collapse, reducing flux. The authors hypothesize that integrating graphene oxide (GO) nanosheets, dispersed with PEG-400 as a stabilizing porogen, into BC during in-situ biosynthesis will reinforce the BC network, create a more favorable pore architecture (more pores in the X–Y plane and straighter channels along Z), enhance hydrophilicity and surface charge, and thereby improve mechanical resilience, permeability, and fouling resistance. The purpose is to develop an eco-friendly, in-situ fabrication of BC–GO composite membranes with optimized pore structure, improved wet compression strength, and enhanced antifouling/bacterial rejection suitable for bio-effluent management.

The paper outlines prior efforts and limitations in modifying BC for filtration, including the use of toxic crosslinkers (e.g., epichlorohydrin, glutaraldehyde) and complex chemical modifications (TEMPO oxidation, phosphorylation), as well as porogens (Ca-alginate, paraffin wax) that complicate removal and can degrade membrane fineness. PEG has been noted as a hydrophilic additive/porogen that can enhance flux and fouling resistance through chain mobility and hydration-layer formation. GO is highlighted for high surface area, mechanical strength, tunable surface chemistry, and facilitating ultrafast water transport. Previous techniques to incorporate inorganic particles into BC suffer from dispersion instability in fermentation media or produce small, unstable constructs under agitation. Studies in other polymer systems link broader pore size distributions and increased hydrophilicity to improved flux and antifouling. Reports of GO-modified PVDF/PES membranes show improved antifouling, but often require higher pressure and achieve lower flux than targeted here. The literature thus supports the rationale for in-situ BC-GO synthesis using PEG to achieve uniform dispersion, improved interfacial interactions, and enhanced performance without hazardous chemistries.

- Microorganisms and media: Komagataeibacter hansenii (ATCC 53582) used to biosynthesize BC; E. coli DH5α used as model bacterium for fouling/bacterial rejection tests. HS medium (peptone, yeast extract, disodium hydrogen phosphate, citric acid, glucose; pH 5.4) prepared and autoclaved. Cultures revived and maintained per ATCC protocols. - GO synthesis and dispersion: GO synthesized from graphite via modified Hummers method (graphite oxidized with KMnO4 in concentrated H2SO4, workup with H2O2, washing to pH ≥ 5). GO characterized by ATR-FTIR. Aqueous GO suspensions (0.5–50 mg/mL) prepared by sonication and stirring. Sterilized by ethanol rinse and drying, then re-dispersed in PEG-400 (0.5–50 mg/mL), sonicated and stirred to form stable dispersions. - In-situ BC–GO biosynthesis: Static cultivation at 26 °C in sterile 120 mm glass Petri dishes. HS medium (150 mL) inoculated with 4-day pre-grown K. hansenii (1:15 v/v). Two approaches tested: (i) adding 10 mL GO in DI water (0.5, 2, 30, 50 mg/mL) and (ii) adding 10 mL GO in PEG-400 at same concentrations. Incubations up to 16 days; samples taken every 2 days for yield/wet thickness. Control without GO. Composites labeled M 0.0 (pristine BC), M 0.5, M 2.0, M 30, M 50. - Purification and yield: Pellicles harvested, pressed, rinsed, treated with 0.1 M NaOH at 80 °C for 90 min, repeatedly rinsed to neutral pH. Dried at 40 °C to constant weight for yield: Yield (%) = mBC_dry / mCarbon source. Wet thickness monitored over time. - Morphology and pore structure: SEM (surface and cryo-fractured cross-sections, Au sputter coat). Pore size distribution by MWCO using PEG/PEO (10–300 kDa) at 2 bar, 25 °C in cross-flow. Stokes radii calculated; pore size distribution derived from log-normal fits. - Physicochemical characterization: ATR-FTIR (4000–500 cm⁻¹), Raman (532 nm; D and G bands), TGA (20–800 °C, N2, 10 °C/min), streaming potential/zeta potential (10 mM KCl, pH 2–9 with KOH titration), static water contact angle (sessile drop; time-resolved), compression testing (rotational rheometer, 25 °C, 25 mm plates with emery paper; compressive strain up to ~80%, 5–8 replicates). - Porosity: Gravimetric method on never-dried samples (wet/dry weights, thickness, area; porosity calculated from water content and geometry). - Filtration setup and protocols: Cross-flow cell (Sterlitech CF042, effective area 42 cm²). Never-dried membranes of 1.2 mm wet thickness used (thinner not self-standing). Pre-compaction 2 bar for 30 min. PWF measured over 0.1–3 bar; subsequent tests at 2 bar. Five 1 h PWF cycles recorded. - Antifouling tests with NOM: Synthetic NOM (mixtures of HA, SA, BSA; typically 200 ppm) filtered in five 1 h cycles with intermediate cleaning (DI water or 0.05 M NaOH in a separate set). TOC via NDIR detector used to compute rejection: R = (1 − Cp/Cf) × 100%. Flux recovery ratio FRR = (Jw_n / Jw_n−1) × 100%. - Bacterial filtration: E. coli broth (~1 × 10^6 CFU/mL) filtered in cross-flow mode at room temperature; sterilization and cleaning protocols applied. Five cycles with DI water cleaning between cycles. Enumeration by flow cytometry (dual staining with SYBR Green I and propidium iodide; Accuri C6+; computational gating using flowCore/OpenCyto) and plate counting calibration. Log reduction value LRV = log(Cf/Cp). Post-filtration SEM imaging assessed biofouling and cleanability.

- In-situ incorporation: GO in DI water led to poor incorporation or weak/inconsistent membranes; GO dispersed in PEG-400 enabled uniform, stable BC–GO composites at low GO loadings (0.5–2 mg/mL). High GO (30–50 mg/mL) inhibited BC fiber secretion and reduced yields. - Optimal loading and growth: Desired 1.2 mm wet thickness achieved after ~3 days. Highest yields at 0.5–2 mg/mL; yields decreased with 30–50 mg/mL. Optimal membrane properties observed at 2 mg/mL GO. - Morphology and pore structure: Uniform GO dispersion within BC interlayers at 0.5–2 mg/mL with perforated network and uniform pores through the membrane body; agglomeration at higher GO. Mean pore diameter reduced from 64 nm (pristine) to 53 nm (0.5 mg/mL) and 42 nm (2 mg/mL), while pore size distribution broadened (~20–90 nm). Porosity increased by ~32% (M 0.5) and ~35% (M 2.0) over pristine. - Physicochemical properties: FTIR indicated decreased BC –OH band intensity and increased carboxylate band (1650 cm⁻¹) up to 2 mg/mL, consistent with hydrogen bonding interactions; diminished at higher GO due to agglomeration. Raman showed D and G bands with higher ID/IG for 0.5 (1.02) and 2 mg/mL (1.44), indicating increased disorder/intercalation; lower ID/IG at 30–50 mg/mL. TGA showed enhanced thermal stability; residual ash increased from 0.5% (BC) to 23%, 28%, 59%, 69% for 0.5, 2, 30, 50 mg/mL, respectively. - Surface charge and wettability: Zeta potential at pH 9 changed from −2.3 mV (pristine) to −23.1 mV (0.5 mg/mL) and −27.2 mV (2 mg/mL). Contact angle dropped from 55.4° (initial) to 47.2° (90 s) for pristine; to 24°→13.2° (0.5 mg/mL) and 21.3°→9.2° (2 mg/mL), indicating superhydrophilicity. - Mechanical strength: Compression modulus at ~80% strain increased from 73 kPa (pristine) to 315 kPa (0.5 mg/mL) and 434 kPa (2 mg/mL)—approximately sixfold improvement at 2 mg/mL. - Permeation: PWF increased from 170.45 LMH (pristine) to 295.3 LMH (0.5 mg/mL) and 394.56 LMH (2 mg/mL) at 2 bar; flux stable over five cycles. Earlier summary noted nearly twofold higher water flux (~380 LMH) and MWCO ~100–200 kDa. - Antifouling with NOM: Average FRR over five cycles: 77% (pristine), 89% (0.5 mg/mL), 95% (2 mg/mL). BC–GO maintained stable flux across cycles; pristine BC showed progressive decline. - NOM rejection: >95% rejection for composite membranes (as indicated in abstract and results) alongside high flux recovery. - Bacterial filtration: BC–GO membranes achieved LRV ≈ 5 (≈99.99% rejection) vs pristine BC LRV ≈ 4. Flux recovery with E. coli broth: 78.2% (pristine), 89.72% (0.5 mg/mL), 95.6% (2 mg/mL). SEM showed abundant E. coli clusters on pristine vs sparse individual cells on BC–GO; easy DI-water cleaning restored flux. - Mechanisms: Enhanced hydrophilicity and more negative surface charge formed a robust hydration layer and electrostatic repulsion, reducing NOM/biofouling. GO reinforcement mitigated pore collapse under compression, improving pore interconnectivity and sustained permeance.

Integrating GO nanosheets into BC via an in-situ PEG-assisted route effectively addresses the core limitations of pristine BC membranes in pressure-driven filtration. Uniformly intercalated GO at low loadings (0.5–2 mg/mL) strengthened the BC network, reduced mean pore size while broadening the pore size distribution, and increased porosity, yielding straighter and more interconnected transport pathways. The membranes became markedly more hydrophilic and negatively charged, promoting the formation of a hydration layer and electrostatic repulsion that suppressed adsorption of NOM and bacterial adhesion. Mechanically, GO reinforcement increased wet compression strength and prevented pore collapse during compaction, enabling sustained high flux at operational pressure (2 bar) and stable performance over repeated cycles. Consequently, BC–GO membranes exhibited substantially higher PWF, >95% NOM rejection with high FRR, and strong bacterial removal (LRV ≈ 5) with superior anti-biofouling behavior and facile DI-water cleanability. High GO loadings (≥30 mg/mL) impaired biosynthesis and membrane uniformity, underscoring a critical balance between filler content and biological production. Overall, the findings validate the hypothesis that PEG-stabilized, in-situ GO incorporation enhances filtration performance by synergistically tuning structure, surface chemistry, and mechanics.

The study presents an eco-friendly, in-situ biosynthesis route for BC–GO composite membranes using PEG-400 to disperse GO and promote uniform intercalation within the BC network. At optimal GO loading (2 mg/mL), the membranes achieve significantly improved wet compression strength (~6×), increased porosity, reduced mean pore size with broader distributions, enhanced hydrophilicity and negative surface charge, leading to high pure water flux (~395 LMH at 2 bar), stable repeated-cycle performance, >95% rejection of synthetic NOM, high flux recovery (≈95%), and excellent bacterial removal (LRV ≈ 5) with low biofouling propensity and simple DI-water cleanability. These outcomes highlight the potential of BC–GO membranes for high-throughput, energy-efficient treatment of bio-effluents. Future work should optimize biosynthesis conditions and microbial growth kinetics to further enhance yield and scalability, and expand testing to a broader range of real wastewater matrices and long-term durability assessments.

- High GO concentrations (30–50 mg/mL) inhibited BC fiber formation, reduced yields, and led to weak/inconsistent membranes, limiting the useful GO loading range. - Filtration performance and fouling tests were conducted on never-dried membranes at a fixed wet thickness of ~1.2 mm and primarily at 2 bar; thinner membranes were not self-standing, and performance outside these conditions was not evaluated. - Antifouling/rejection studies used synthetic NOM mixtures and E. coli broth; broader validation with diverse real wastewater matrices and long-term continuous operation was not reported. - Only select GO loadings (0.5 and 2 mg/mL) were evaluated for detailed filtration performance; results may not generalize to other formulations without further study.