Excessive membrane biofilm growth on membrane fibers depends on various factors, with membrane properties playing a pivotal role in influencing microbial affinity for the membrane. To investigate the antibacterial impact of nano-sized zero-valent iron (nZVI) on membrane biofilm structure, pristine (polyvinylidene fluoride (PVDF)) only: HF-0 (PVDF:20/nZVI:0 w/w) and four gas transfer membranes (PVDF:nZVI at different concentrations: HF-1 (PVDF:20/nZVI:0.25 w/w), HF-2 (PVDF:20/nZVI:0.50 w/w), HF-3 (PVDF:20/nZVI:0.75 w/w), HF-4 (PVDF:20/nZVI:1.0 w/w)) were produced. These membranes were assessed for surface morphology, porosity, gas permeability, and biofilm thickness, which ultimately affect biochemical reaction rates in membrane biofilm reactors (MBfRs). Various MBfRs utilizing these gas transfer membranes were operated at different hydraulic retention times (HRTs) and oxygen pressures to assess chemical oxygen demand (COD) removal efficiency and nitrification performance. Incorporating nZVI into the PVDF polymer solution increased surface hydrophilicity and porosity but reduced Young's Modulus and oxygen diffusion coefficients. Confocal laser scanning microscopy (CLSM) analysis revealed an average biofilm thickness of 700 μm in HF-0, HF-1, and HF-3, with a 100 μm decrease in HF-2, even though Escherichia coli growth was observed in HF-3 fibers. Regardless of nZVI dosage, a significant decline in COD removal and nitrification rates occurred at low HRTs and gas pressures.

  • The incorporation of nZVI into PVDF gas transfer membranes changed the membrane's properties.

  • The effect of nZVI on biofilm thickness was insignificant.

  • HRT and gas pressure rather than nZVI dosage played a key role in COD removal and nitrification rate.

Due to bubble-free aeration and high oxygen utilization efficiency, membrane biofilm reactors (MBfRs) provide significant advantages over conventional biofilm processes regarding aeration costs in wastewater treatment. Moreover, high effluent quality, (Sunner et al. 2018; Sathyamoorthy et al. 2019), up to 100% oxygen transfer efficiency (Heffernan et al. 2017; Bicudo et al. 2019), design of compact reactor (Sunner et al. 2018), and combining the sludge retention time (SRT) with the hydraulic retention time (HRT) which enables slow-growing microorganisms such as ammonium oxidizing bacteria are among the advantages of this technology. On the other hand, since the stripping off volatile organic compounds and microorganisms embedded in the biofilm are protected by external extracellular polymeric substances, efficient and continuous substrate degradation rates can be achieved for the treatment of high-strength industrial wastewater and xenobiotics (Abdelfattah et al. 2020). MBfR technology has been investigated for the removal of many contaminants including fluorinated organics (Heffernan et al. 2009; Misiak et al. 2011); phenolic compounds (Hanay et al. 2014; Mei et al. 2019; Tian et al. 2019); dye (Wang et al. 2012); pharmaceutical compounds such as cefalexin (CFX) and sulfadiazine (SDZ) (Wang et al. 2021); tetracyclines (TCs) (Salman et al. 2022), and organonitrile compounds from water and wastewater (Li & Liu 2019).

For this innovative technology, the membrane module configuration, aeration mode, type of membrane material, and control of biofilm thickness are important challenges. Among these, control of biofilm thickness has been investigated with different approaches since the accumulation of excess biofilm causes some adverse effects on MBfR performance (Syron & Casey 2008). Excessive biofilm growth reduces the HRT and limits the diffusion of substrate and electron donors through the biofilm. Various strategies such as operating the membrane biofilm system under thermophilic conditions (Liao & Liss 2007); periodic air stripping (Côté et al. 2015); biological scouring of the biofilm attached to the membrane with the addition of higher organisms into the biofilm structure (Aybar et al. 2019) and quorum quenching (QQ) technique (Taşkan et al. 2020) have been applied for excessive biofilm control. On the other hand, a few studies focusing on the significance of the membrane materials used in MBfR are available in the literature although high oxygen permeability and low water permeability are critical for membrane materials being used in this technology (Terada et al. 2009; Nisola et al. 2013; Wu et al. 2019; Kobayashi et al. 2022). Polymeric membranes are widely used due to their low cost, ease of manufacture, easy modification, and good separation efficiency (Acarer 2023). Kobayashi et al. (2022) manufactured polystyrene elastomer (PS) and polyurethane (PU) composite membranes to evaluate the oxygen transfer rates and vapor permeation. They concluded that PS composite membrane resulted in high oxygen permeability, rapid start-up, and high carbon removal as well. In another study, the performance of polyvinylidene fluoride (PVDF) and polypropylene (PP) membrane hollow fiber (HF) membranes in terms of nitrogen and carbon removal, biomass properties on the membrane surface, and membrane pore-blocking properties were investigated. The PVDF membrane exhibited more rough and hydrophilic properties and a better microbial affinity compared to PP (Wu et al. 2019). Additionally, Penboon et al. (2019) also stated that the usage of PVDF in manufacturing polymeric membranes has offered various advantages due to its chemical resistance, superior mechanical properties, and thermal stability.

The nano-sized zero-valent iron (nZVI) is used as an alternative material in environmental remediation, drinking water filters, and disinfection due to its low cost and environmentally friendly (Wang & Zhang 1997; Hanay & Türk 2015; Ling et al. 2018) and it has been extensively studied in the removal of various organic and inorganic compounds from water and wastewater (Wang & Zhang 1997; Hanay & Türk 2015; Ling et al. 2018). Moreover, by considering the catalytic properties of nZVI, the usage of a composite membrane with nZVI has been increased. For example, Li et al. (2021) investigated the use of nZVI/PVDF composite membrane in chlorophenol removal in a non-biofilm-based membrane bioreactor, and high chlorophenol removals were achieved due to the high interaction of nZVI with the model pollutant. Hou et al. (2020) observed an increased trichloroethylene decomposition, good stability, and reliable operating capability in the Fe/Pd nanoparticle doped PVDF composite membrane. In another study, the ZVI/cellulose acetate ultrafiltration membrane exhibited higher thermal stability and roughness with the increase in the mass ratio of ZVI in the composite material (Saranya et al. 2015). Silva et al. (2021) also reported that the PVDF/nZVI catalytic membrane exhibited excellent mechanical stability and high permeability and the high removal efficiency of bisphenol-A (52% ± 0.5%) under low permeate flux (50 LMH) in the presence of 10 mM H2O2. On the other hand, the influence of nZVI on microorganisms is still controversial even though the antibacterial effect of nZVI on membrane fouling has been investigated. Aljohny et al. (2020) stated that the composite membrane synthesized with cellulose acetate and nZVI nanoparticle showed an antibacterial effect against Salmonella typhi. Dizge et al. (2017) manufactured the polyethersulfone (PES) composite membrane containing ZVI and determined that the membrane exhibited antibacterial properties against Escherichia coli bacteria. In another study, nZVI was added to the laboratory-scale anoxic-oxic submerged membrane bioreactor throughout the operation and it was observed that the addition of nZVI reduced the membrane fouling layer and the amount of biomass and extracellular polymeric substance due to the oxidative stress of the bacteria (Zhou et al. 2017). Moreover, nZVI nanoparticles can also potentially act as a source of Fe due to the joining of the enzymes. With these aspects, combining nZVI with microorganisms is a robust tool in the field of bioremediation, biotransformation, and bioenergy (Liu et al. 2022). It has been reported that FeS2 nanoparticle provides iron and sulfur as nutrients for protein construction or energy production (Jørgensen et al. 2019). To our knowledge, only one study on the use of nZVI in HF membrane manufacturing via wet spinning and thermal calcination for flow-through heterogeneous Fenton process was reported by Sun et al. (2022). From this point, how nZVI addition to gas permeable membrane structure can affect the antibacterial properties and membrane characteristics including surface morphology, porosity, and permeability is the main research topic in this study. For this purpose, nZVI-based with different amounts of nZVI gas permeable membranes were synthesized and the performance in membrane biofilm reactor in terms of oxidation of COD and NH3-N was investigated.

Chemicals

Iron(III) chloride (FeCI3) (CAS No. 10025-77-1) and sodium borohydride (NaBH4) (CAS No. 16940-66-2) were purchased from Merck (Germany). PVDF as a used polymer material (Kynar 301F) was obtained from Solvay Specialty Polymers. 1-Methyl-2-pyrrolidinone (NMP, ≥99.5%) used as a solvent in the composition of the polymer solution and internal coagulant solution was purchased from Isık Chemical Materials LTD (Turkey). The two-component epoxy adhesive (Jalasanj Company) was used to adhere to the produced membranes in preparation for the module.

The synthesis of nZVI and nZVI-based gas transfer membranes

nZVI particle was prepared by using the borohydride reduction method according to the study of Hwang et al. (2011). The four open-necked 500 mL bottle was used for the synthesis procedure. The solution was mixed strongly with a mechanical stirrer at 250 rpm. To reduce ferric ions to nZVI, a borohydride solution (358.5 mM ) of 250 mL was added with a constant delivery rate of 20 mL/min into ferric ion (Fe3+) solution (71.7 mM Fe3+) of 250 mL from one of the necks on side of the flask. Nitrogen gas was used to hinder iron oxidation, and the inlet and outlet of gas were provided by the other two necks of the flask. This reaction (R.1) is described as follows:
formula
(R.1)

The obtained particles were washed with ethanol and then dried in an oven at 100 °C for 2 h.

To produce HF membranes via the wet phase inversion method, a similar fabrication approach was followed as described in the study of Aksoy & Hasar (2021). The pilot-scale HF manufacturing membrane system (MEMTEK) was employed for this purpose (Aksoy & Hasar 2021). First, to prepare pristine PVDF dope solution, PVDF was dried for 24 h at 60°C. The dried PVDF was added slowly to the solvent and mixed for 48 h at 70°C. The dope solution was first held at room temperature, then transferred into the polymer dope tank, and kept overnight under a vacuum to remove air bubbles resulting from stirring stages. For the preparation of nZVI blended PVDF dope solution, four different concentrations of nZVI (0.25, 0.5, 0.75, and 1 w/w) were homogeneously dispersed in NMP using Bandelin-Sonopuls homogenization. The dried PVDF was added to this solution as previously described. The dope solutions were sonicated for 30 min in an ultrasonication bath to remove the air bubbles and then sealed under a vacuum (Turken et al. 2015). The NMP/water 70%/30% (w/w) was used for the bore liquid. The composition of the membrane fibers is shown in Table 1. Dope solution and bore liquid was delivered to the spinneret with a flow rate of 6.0 and 3.0 mL/min, respectively, at the same temperature of 25°C after pressurizing with 2 atm using nitrogen gas through a spinneret. Then, fibers moved directly from the spinneret to a coagulation bath filled with tap water at a temperature of 25°C, passed through the washing tank, and collected properly by a wind-up drum with a speed of 0.1 m/s. The wrapped fibers were stored in distilled water for 3 days to completely remove the remaining solvent and additive. Finally, the collected fibers were vertically hung to dry at room temperature for 1 d before characterizing membrane fibers.

Table 1

The content of polymer solutions in manufacturing hollow fiber membranes

Hollow fiber membranesPVDF (wt.%)NMP (wt.%)nZVI (wt.%)
HF-0 20 80 – 
HF-1 20 79.75 0.25 
HF-2 20 79.5 0.5 
HF-3 20 79.25 0.75 
HF-4 20 79 
Hollow fiber membranesPVDF (wt.%)NMP (wt.%)nZVI (wt.%)
HF-0 20 80 – 
HF-1 20 79.75 0.25 
HF-2 20 79.5 0.5 
HF-3 20 79.25 0.75 
HF-4 20 79 

Characterization studies of nZVI and nZVI-based gas transfer membranes

The size and morphology of the nZVI particle were determined using scanning electron microscopy (SEM) LEO-EVO 40 (Cambridge, England) and an attached X-ray energy-dispersive spectrometer (EDX). Images of the samples were recorded at different magnifications at an operating voltage of 20 kV. The surface area of nZVI was determined by A Micromeritics TriStar II 3020 analyzer using the BET (Brunauer–Emmett–Teller) single-point method. The cross-section, outer, and inner surface morphologies of each composite membrane and pristine membrane were examined using SEM (Quanta FEG 250) after coating with gold/palladium using a sputter coater (Quorum SC7620). The coating procedure was carried out within 60 s and at a voltage in the range of 10 and 20 mV. Surface functional groups of membranes were performed using a Fourier Transform Infrared (FTIR) spectrometer (Perkin Elmer Spectrum 100). The contact angle between the water drop and membrane surfaces was measured by a high-resolution digital camera and a contact angle goniometer (Attension T200 Theta) after placing a liquid drop on the top layer of each fiber's sample. The measurements were done at different points of a fiber to ensure a reliable value. The mechanical property of HF membranes was evaluated in terms of Young's modulus by a dynamic mechanical analyzer (DMS 6100 Exstar). In this procedure, the fiber samples were first fastened after being mounted between the grips. The measurements were carried out once 3 s in force increments of 250 N over 20 steps for a total load of 5,000 N. A stress–strain curve was provided for each fiber at the end of the procedure. Finally, average values of Young's modulus were determined by the slope of stress–strain curve in elastic regions after measurements. The coefficients of gas diffusion for the fabricated HF membranes were determined by using U-shaped test modules in the reactor. Distilled water was filled after the prepared U-shaped modules were installed into the reactor. Reactor content was deoxygenated with the addition of sodium sulfite and cobalt chloride and mixed with a magnetic bar at the bottom of the reactor until the dissolved oxygen (DO) stabilized at a concentration of about zero. The reactor was properly sealed to prevent oxygen inflow from the atmosphere. Pure oxygen was supplied from a gas cylinder at a constant gas pressure of 0.14 atm until reached 6.0 mg/L. DO concentration in the reactor was measured using a DO meter (YSI Model 55, OH). The graphs were plotted based on the data collected by measuring DO concentrations over time. Finally, coefficients of oxygen transfer were found as calculated by Aksoy & Hasar (2021). The porosity was determined by the gravimetric method (Li et al. 2009), according to Equation (1)
formula
(1)
where Ɛ is the porosity; m1 is the weight of the wet membrane; m2 is the weight of the dry membrane; ρw is the water density (0.998 g cm−3); A is the effective area of the membrane (m2), and l is the membrane thickness (m).

E. coli culture was employed for antibacterial tests. E. coli was filtered through HF modules for 10 min. These modules were then put on an agar medium and cultured for 3 days at 37 °C. The growth of the bacteria was seen visually (Yavuz et al. 2019). Fiber samples from each reactor were collected at the end of the study to quantify the antibacterial activity (biofilm thickness) on the fiber surface by confocal laser scanning microscopy (CLSM) (C1 plus, Nikon, Japan). Bacterial cells were seen by staining biofilms on membrane fibers with 100 μL of SYTO9 (Molecular Probes) for 20 min in the dark (Taşkan et al. 2020). Using CLSM, biofilm formations were identified following staining. The biofilm thickness on the membrane surface was measured at several points to calculate the average biofilm thickness.

The membrane biofilm reactor configuration and operation

The outer diameter of each membrane is given in Figure 2. Four reactors containing two modules were designed with 30 membrane fibers (one fiber length of 11 cm) which provided a total membrane surface area of 118.3, 109.84, 118.54, and 111.81 cm2 for HF-0, HF-1, HF-2, and HF-3, respectively. Pure oxygen gas was supplied at set pressures of 0.14 and 0.27 atm to the inside of the HF through the manifold at the base. The complete stirring inside the reactors was provided with a magnetic stirrer placed at the bottom of the reactors. A peristaltic pump (Watson Marlow 205S) was used to feed the influent solution to the reactor. The schematic diagram of the MBfR was described in detail by Hasar & Ipek (2010).

The start-up of the reactor was initiated by seeding with 250 mL of inoculum which was the supernatant of precipitated activated sludge from the municipal wastewater treatment plant of Malatya, Turkey. The inoculation liquid was mixed by the magnetic stirrer in the reactor for 72 h to establish the biofilm on the membrane surface under a pure oxygen pressure of 0.27 atm before the acclimation study began. After the biofilm started to develop on the hydrophobic membrane, acetate, and ammonium were fed to the reactors. The synthetic solution was prepared daily and 2 mL of mineral solution was added. The contents of synthetic influent and mineral solution are shown in Table 2. The reactor was operated in a temperature-controlled room at 25 ± 2 °C. Two O2 pressures (0.14, 0.27 atm) were tried at each HRT. The pH was kept between 7.40 and 7.60 in the influent. The operation conditions of MBfR are given in Table 3.

Table 2

The contents of influent and mineral solution

Influent solution (g/L)Mineral solution (μg/L)
NaC2H3O2: 0.225 ZnSO4·7H2O: 100 
NH4Cl: 0.113 MnCI2·4H2O: 30 
NaHCO3: 0.125 H3BO3: 300 
KH2PO4: 0.05 CoCI2·6H2O: 200 
MgSO4·7H2O: 0.05 CuCI2·2H2O: 10 
NaC2H3O2: 0.385 NiCI2·6H2O: 10 
Influent solution (g/L)Mineral solution (μg/L)
NaC2H3O2: 0.225 ZnSO4·7H2O: 100 
NH4Cl: 0.113 MnCI2·4H2O: 30 
NaHCO3: 0.125 H3BO3: 300 
KH2PO4: 0.05 CoCI2·6H2O: 200 
MgSO4·7H2O: 0.05 CuCI2·2H2O: 10 
NaC2H3O2: 0.385 NiCI2·6H2O: 10 
Table 3

The operation conditions of membrane biofilm reactor

PeriodsHydraulic retention time, hOxygen pressure, atmInfluent COD concentration, mg/LInfluent NH4-N concentration, mg/LNH4-N Loading mg/m2·dThe flux of NH4-N, mg/m2·d
HF-0HF-1HF-2HF-3
P1 24 0.27 140–160 25–30 407.04 387.6 387.12 384 399.84 
P2 24 0.14 140–160 25–30 419.28 359.9 388.11 368.64 385.2 
P3 12 0.27 140–160 25–30 865.54 691.44 725.28 725.28 692.64 
P4 12 0.14 140–160 25–30 852.48 688.56 633.84 708.72 657.36 
P5 0.27 140–160 25–30 1,675.44 863.52 608.88 654 1,182 
P6 0.14 140–160 25–30 1,639.92 938.88 689.76 842.64 850.8 
P7 0.27 140–160 25–30 3,384.17 799.92 1,044 850.1 1,556.6 
P8 0.14 140–160 25–30 3,499.68 847.2 458.4 473 413.28 
PeriodsHydraulic retention time, hOxygen pressure, atmInfluent COD concentration, mg/LInfluent NH4-N concentration, mg/LNH4-N Loading mg/m2·dThe flux of NH4-N, mg/m2·d
HF-0HF-1HF-2HF-3
P1 24 0.27 140–160 25–30 407.04 387.6 387.12 384 399.84 
P2 24 0.14 140–160 25–30 419.28 359.9 388.11 368.64 385.2 
P3 12 0.27 140–160 25–30 865.54 691.44 725.28 725.28 692.64 
P4 12 0.14 140–160 25–30 852.48 688.56 633.84 708.72 657.36 
P5 0.27 140–160 25–30 1,675.44 863.52 608.88 654 1,182 
P6 0.14 140–160 25–30 1,639.92 938.88 689.76 842.64 850.8 
P7 0.27 140–160 25–30 3,384.17 799.92 1,044 850.1 1,556.6 
P8 0.14 140–160 25–30 3,499.68 847.2 458.4 473 413.28 

Analytical methods

The samples of influent and effluent from MBfRs were periodically collected and analyzed for the measurement of pH, NH4-N, NO3-N, NO2-N, and COD. The pH values were analyzed by pH-meter (Thermo Scientific) and the concentrations of NH4-N, NO3-N, and NO2-N were determined by spectrophotometer (PerkinElmer). The COD analysis was performed according to Standard Methods (Greenberg et al. 1992). The concentration of total Fe ion and Fe+2 ion in the sample solution was measured by the Ferrozine method (Stookey 1970). The samples were filtered by 0.22-μm disposable syringe filters, and 1 mL of 3.6 M H2SO4, 1 mL of 4.9 mM Ferrozine solution, and 1 mL of acetate buffer were added to the samples before the UV-spectrophotometer analysis. The total aqueous Fe(III) concentration could be evaluated by subtracting the dissolved total Fe ion concentration from the aqueous Fe(II) concentration.

surface loading and fluxes of NH4-N

The computation for surface loadings and fluxes for NH4-N was calculated in our previous studies (Çelik et al. 2018). The oxygen equivalent flux for NH4-N was calculated from Equation (2) by assuming a complete oxidation of NH4-N according to R.2.
formula
(R.2)
formula
(2)
where is the oxygen equivalent flux (mg O2/m2 day) and J JNH4-N is JNH4-N flux (mg JNH4-N /m2 day).

The properties of nZVI and nZVI-based gas transfer membranes

Figure 1 shows the SEM images and EDX results of the synthesized nZVI. SEM images show that nZVI has a spherical structure and spherical formations with different particle sizes are agglomerated. Some chain and spherical aggregations are caused by magnetic interaction between particles (Hanay & Türk 2015). According to EDX results, the contents of Fe, C, and O were 91.37, 5.23, and 3.41 wt.%, respectively. The low content of oxygen indicates that the synthesized iron is not oxidized and iron corrosion products are not formed (Xu & Wang 2011). The specific surface area of nZVI was found to be 5.99 m2/g according to BET analysis. The SEM images are shown in Figure 2 and HF-0 and HF-1 HF membranes exhibit similar cross-sectional images. It is seen that finger-like structures extend from the outer part of the membranes to the inner part, and the parts closer to the inner part have a spongy structure. The finger-like structure on the outside formed due to rapid solidification during wet phase inversion. However, using a high amount of solvent as the internal coagulant may have resulted in a spongy structure in the inner parts of the membranes as a result of the delay of phase inversion by the internal coagulant, which was also stated by Hosseini & Mansourizadeh (2017). Additionally, there are macro cavities in the cross-section of the membrane. HF-2 and HF-3 HF membranes exhibit similar cross-sectional images. Unlike HF-0 and HF-1, the macro cavities are completely removed and, the spongy structure dominates from the outer part to the inner part. There are also thin and short finger-like structures on the outer part of the membranes. The inner space of the HF-4 HF membrane is not formed properly due to the addition of a high concentration of nZVI.
Figure 1

SEM images and EDX patterns of nZVI.

Figure 1

SEM images and EDX patterns of nZVI.

Close modal
Figure 2

SEM images of HF membranes.

Figure 2

SEM images of HF membranes.

Close modal
Figure 3 shows the FTIR spectrum of HF membranes. In this study, 761; 875; 1,071; and 1,172 cm−1 peaks (α-phase) and 839; 1,275; and 1,401 cm−1 peaks (β-phase) were monitored. Nine peaks belonging to pure PVDF HF membrane (HF-0) were also detected from the FTIR spectrum of nZVI-doped HF membranes, and it was concluded that nZVI did not form different functional groups on the membrane surface. Hashim et al. (2011) in their study showed that the characteristic peaks in the FTIR graphic results of pure Solef 6010 PVDF polymer were at 765, 796, 859, 874, 976, 1,070, 1,148, 1,180, 1,206, 1,383, and 1,423 cm−1 and these peaks were found in the α-phase while the other observed values of 840; 1,275; and 1,404 cm−1 were in the β-phase. Moreover, as illustrated in Figure 3, the peaks at 1,071, 1,172, 1,275, and 1,401 cm−1 are related to the C–F bond stretching oscillation of PVDF, and the peaks at 700–900 cm−1 are related to the vibration of PVDF. This is in agreement with a recent study by Gao et al. (2022), which reported similar peaks.
Figure 3

FTIR spectra of HF membranes.

Figure 3

FTIR spectra of HF membranes.

Close modal
The measurement of the contact angle is a fundamental criterion utilized to categorize the wetting characteristics of a fluid as either hydrophilic or hydrophobic. This parameter, known as the surface contact angle, represents the angle formed at the point where the liquid–solid and liquid–vapor interfaces intersect (Ahmad et al. 2018). It is customary to regard θ = 90° as a demarcation point between hydrophilicity and hydrophobicity. Values below 90° signify hydrophilic surfaces of membranes, while those exceeding 90° are observed on hydrophobic surfaces of membranes (Gugliuzza 2015). In this study, the variation of water contact angle values at different amounts of nZVI was measured as 79.05°, 69.49°, 62.59°, 60.12°, and 60.86° for HF-0, HF-1, HF-2, HF-3, and HF-4, respectively (Figure 4(a)). It indicates that the incorporation of nZVI into the PVDF polymer solution leads to an increase in the surface hydrophilicity of HF membranes. This phenomenon may be explained by the fact that nZVI has a high surface area and high reactivity, which can increase the production of free radicals during the synthesis of HF membranes. These free radicals can react with PVDF polymer chains, leading to the formation of functional groups such as hydroxyl, carboxyl, and amino groups on the membrane surface. These functional groups can increase the surface hydrophilicity of the membrane by increasing the interaction between the membrane surface and water molecules, thereby reducing the value of the water contact angle. This observation is consistent with previous studies reported by Yang et al. (2022) and He et al. (2020). They stated that the incorporation of micro and nano ZVI into the PVDF membrane led to a decrease in the water contact angle, which was related to the formation of hydrophilic functional groups on the membrane surface.
Figure 4

The properties of HF membranes: (a) water contact angle, (b) Young's Modulus, (c) oxygen transfer coefficient, (d) porosity.

Figure 4

The properties of HF membranes: (a) water contact angle, (b) Young's Modulus, (c) oxygen transfer coefficient, (d) porosity.

Close modal

Young's Modulus is a measure of the hardness of a material and represents the ratio of stress to strain under tension or compression loading conditions (Rabuni 2014). As depicted in Figure 4(b), a significant decrease was observed in the Young's Modulus values of HF membranes as the nZVI was used in the polymer solution. Moreover, a slight difference among Young's Modulus values of membranes containing the different nZVI amounts was determined. For instance, HF-0 containing 20 wt.% of PVDF has the highest Young's Modulus value of 90.304 MPa, while HF-1 containing 20 wt.% of PVDF and 0.25 wt.% of nZVI has the lowest Young's Modulus value of 28.491 MPa. The decrease in Young's Modulus with increasing nZVI content can be attributed to the fact that nZVI particles may result in some defects in the polymer matrix, leading to a decrease in the effective cross-link density of the polymer structure. As a result, polymer chains become more mobile and the overall stiffness of the material decreases. These results are consistent with the findings of previous studies investigating the effect of nanoparticle incorporation on the mechanical properties of polymer composites (Demirsoy et al. 2015; Kamal et al. 2019).

High oxygen permeability and low water permeability are critical for membrane materials being used in MABR (Kobayashi et al. 2022). The gas transfer performance of membranes was examined by calculating the oxygen diffusion coefficient and the results are shown in Figure 4(c). The addition of nZVI into the PVDF polymer solution resulted in decreasing the oxygen diffusion coefficient values in all studied membranes. A possible explanation for this variation is that the formation of aggregates of nZVI particles can be able to clog the pores of membranes and reduce the oxygen permeability. Additionally, the morphology or surface chemistry of membranes could be altered by the presence of nZVI, which can affect the interaction between membranes and oxygen molecules. For example, Najafi et al. (2018) suggested that the incorporation of silica particles into the polymer matrix increases the number of polar OH groups as well as the morphological changes induced at the CA/silica interface, resulting in increased gas solubility in the membrane, while the diffusivity of gases decreases due to blockage by impermeable silica particles. Takase et al. (2021) reported that high hydrophilicity suppressed oxygen diffusion. Unexpectedly, during the analysis of the oxygen transfer test, the HF-4 HF membrane, containing 1% wt of nZVI ruptured. Therefore, further studies are needed to elucidate the mechanisms underlying this negative effect and to optimize the preparation of HF membranes for efficient gas transfer performance.

On the other hand, the increase in the amount of nZVI in PVDF solution increases the porosity of membranes (Figure 4(d)). This behavior can be explained by the fact that nZVI nanoparticles have a high surface area and can form voids within the membrane structure during the wet phase transformation process (Al Harby et al. 2022). These findings are consistent with previous studies investigating the effects of various nanoparticles on the porosity of polymer membranes. For example, Ghaemi et al. (2015) prepared a mixed matrix PES nanofiltration membrane by the addition of various concentrations of modified nanoparticles based on n-Fe3O4. They concluded that the total porosity increased with the addition of nanoparticles as compared to pure PES membranes. Li et al. (2009) prepared PES–TiO2 hybrid membranes by a phase transformation method and reported that the composite membrane had a top surface with high porosity at a low loading amount of TiO2 whereas the skin layer became much looser for a significant aggregation of TiO2 nanoparticles at a high loading amount of TiO2.

The effect of nZVI addition on antibacterial properties was examined by the visible observation of E. coli growth on agar. After the E. coli filtration process, the HFs placed in the agars were left to incubate for 72 h, and the images of E. coli growth on the fiber surfaces obtained at the end of the incubation are shown in Figure 5. In contrast to our expectations, E. coli growth occurred on nZVI-added membranes. As can be seen from the images, E. coli growth close to the surfaces of HF-2 and HF-3 fibers was observed while a small part of HF-0 fiber exhibited a similar behavior which may result from contamination during the analysis. The E. coli growth was more evident in HF-3 fiber while no growth was observed in the HF-1 fiber. The bactericidal effects of nZVI nanoparticles are still controversial. It has been determined that Fe(II) and Fe(III) released from nZVI can directly participate in the metabolism of microorganisms by binding to specific proteins, although excessive Fe2+ intake can cause oxidative damage in cells (Zhao et al. 2021). On the other hand, Fe precipitation may promote the survival of microorganisms by preventing the accumulation of toxic metabolic products (e.g. H2S) and the passage of nanoparticles into the cell (Harouaka et al. 2016). In this study, the concentration of total Fe ion and Fe2+ ion was not determined in the effluent sampled at the end of each period. It can be inferred that nZVI can be embedded into the membrane fibers without any Fe leaching and it can lead to the growth of E .coli bacteria.
Figure 5

The images of Escherichia coli growth on HFs at the end of incubation.

Figure 5

The images of Escherichia coli growth on HFs at the end of incubation.

Close modal

Performances of COD removal and nitrification and fluxes of ammonium

The MBfRs were operated to evaluate the simultaneous performance of COD removal and nitrification at different HRTs and O2 gas pressures throughout 120 days after the acclimation time of about 60 days. The influent COD concentration was between 140 and 150 mg/L whereas NH3-N concentration was kept to be in the range of 25 and 30 mg/L throughout the study.

At P1, COD removal efficiency did not change significantly and the average COD removal efficiency was approximately 95% for all reactors. However, the average COD removal efficiencies at P2 were calculated as 88, 94, 95, and 97% for HF-0, HF-1, HF-2, HF-3, respectively (Figure 6(a)). No significant differences in COD removal were observed between membrane biofilm reactors using the different nZVI dosages in membrane manufacturing. Moreover, compared to the HF-0, the contribution of nZVI improved the COD removal efficiency by approximately 9% as compared to HF-3. In these periods, the nitrification rate occurred greater than 86% at both gas pressures (Figure 6(b)). The average COD removal efficiency was determined as 90% at P3 and P4. However, a slight decrease in nitrification rate was observed at P3 and P4 regardless of the applied oxygen gas pressure. The highest nitrification rate was 81% for HF-0 at P4. The further decrease in HRT (from 12 to 6 h) did not any change in COD removal at P5 and P6 whereas the nitrification rate significantly slowed down. For instance, the average NH4-N removal efficiencies were 52, 34, 39, and 47% for HF-0, HF-1, HF-2, HF-3, respectively. For all studied membrane biofilm reactors, the effect of HRT on COD removal seemed to be important at P7 and P8 which HRT was 3 h and oxygen gas pressure was 0.27 and 0.14 atm, respectively. The COD removal efficiencies averaged 77, 77.5, 79, and 86% for HF-0, HF-1, HF-2, and HF-3 reactors at P7 while those were 71, 74, 79, and 77% at P8, respectively. Similarly, nitrification deteriorated by reducing the HRT when a significant decrease in nitrification rate was determined for all membrane reactors. As can be expected that low HRT led to less contact between the biofilm and the substrate, thus reducing the amount of substrate diffusing into the inner layer of the biofilm (Wang et al. 2021). In contrast to our results, Wang et al. (2021) stated that reducing the HRT from 4 to 3 h in the MBfR decreased the removal efficiency of the antibiotic species as cefalexin and sulfadiazine compounds from 86 to 61.5% while the nitrification performance and microbial community structure did not change. In another study, the effects of HRT, O2 gas pressures, and electron donors (ammonium and chlortetracycline (CTC)) on ammonium and CTC removal were investigated in an MBfR (Aydın et al. 2021). They found that decreasing O2 gas pressure and HRT did not affect the removal of the primary electron donor (NH4-N) whereas the removal of the secondary electron donor (CTC) slowed down. A possible explanation for this is the type of carbon resource used in membrane biofilm reactor operation. In this study, acetate was used as a carbon resource as it is more readily bioavailable for heterotrophic bacteria than a pharmaceutical compound. On the other hand, the concentrations of removed NH4-N were similar to the concentration of formed NO3-N in all periods and the the concentration of NO2-N was as low as 0.1 mg/L which implies that denitrification did not occur during operation. It is well known that the COD/N ratio of the influent wastewater is important for simultaneous nitrification and denitrification (SND) and it typically determines the potential degree of denitrification for wastewater. This issue has been reported experimentally or mathematically in many studies. For example, mathematical MBfR simulations have suggested that optimum SND occurs at a COD:N ratio of 3.75, and total nitrogen removal will not exceed 30% when the COD:N ratio is less than 1 or greater than 7 (Matsumoto et al. 2007). In another study, based on stoichiometric relationships and model simulations, optimum SND could be achieved when the influent COD:N ratio was approximately 4 (Shanahan & Semmens 2004). As the influent of COD/N ratio was kept to be about 6 in this study total nitrogen removal can be expectedly limited.
Figure 6

The concentrations of COD (a) and NH4-N+ (b); NH4-N+ fluxes (c); O2 fluxes for NH4-N+ removal (d) during the operation of MBFRs. P1; HRT: 24 h, , P2; HRT: 24 h, , P3; HRT: 12 h, , P4; HRT: 12 h, , P5; HRT: 6 h, ; P6; HRT: 6 h, , P7; HRT: 3 h, , P8; HRT: 3 h, .

Figure 6

The concentrations of COD (a) and NH4-N+ (b); NH4-N+ fluxes (c); O2 fluxes for NH4-N+ removal (d) during the operation of MBFRs. P1; HRT: 24 h, , P2; HRT: 24 h, , P3; HRT: 12 h, , P4; HRT: 12 h, , P5; HRT: 6 h, ; P6; HRT: 6 h, , P7; HRT: 3 h, , P8; HRT: 3 h, .

Close modal

Results from the calculation of fluxes demonstrated that the fluxes of ammonium increased significantly with increasing loading of ammonium at the first four periods regardless of nZVI dosages (Figure 6(c)). For instance; at P3, the ammonium loading of 36.1 mg/m2 d creates a flux of 28.8 mg/m2 d, 26.9 in HF-1, 30.22, and 28.9 mg/m2 d for HF-0, HF-1, HF-2, and HF-3, respectively. However, the ammonium loading increased almost 2-fold, approximately 69.8 mg/m2 d when HRT was reduced to 6 h (P5), and the flux was 36, 25.4, 27.3, and 49.3 mg/m2 d for HF-0, HF-1, HF-2, and HF-3, respectively. With the further decrease in oxygen gas pressure at HRT of 6 h (P6), a similar behavior within P5 was observed. The highest flux was calculated for HF-3, which had the highest nZVI dosage. When the ammonium loading was increased from 68.32 mg/m2 d to 141 mg/m2 d (P6), the ammonium fluxes formed in HF-0, HF-1, HF-2, and HF-3 were 33.3, 43.5, 35.4, and 64.9 mg/m2 d, respectively, as HRT was reduced to 3 h and ammonium loading was 141 mg/m2 d. At P8, the ammonium load applied to the system was 145.8 mg/m2 d, while the ammonium fluxes in HF-0, HF-1, HF-2, and HF-3 were 35.3, 19.1, 19.7, and 17.2 mg/m2 d, respectivel, additionally, as illustrated in Figure 6(d). The equivalent oxygen fluxes of NH4-N followed a similar pattern to NH4-N fluxes.

Biofilm thickness

CLSM images of fibers of HF-0, HF-1, HF-2, and HF-3 membranes are shown in Figure 7. Three samples were used for CLSM from fibers taken from each membrane fiber. Accordingly, the average biofilm thickness was determined as 700 μm in HF-0, HF-1, and HF-3 whereas a decrease of 100 μm was observed for HF-2. As mentioned above, although the growth of E .coli varied according to nZVI dosage, no clear relationship was found between the nZVI dosage and biofilm thickness. This may be related to membrane characteristics including surface morphology, porosity, and permeability as they affect microbial affinity for the membrane and oxygen transfer; hence they impact biofilm characteristics (Syron & Casey 2008; Lu et al. 2021). The membranes of HF-2 and HF-3 exhibit higher membrane porosity than that of HF-0 and HF-1 which can result in greater microbial detachment. However, HF-2 and HF-3 have a higher oxygen transfer coefficient than that of HF-1. It can be inferred that more oxygen can be supplied to the biofilm and in consequence enhance the growth of microorganisms on fibers. On the other hand, total Fe and Fe(II) analysis shows that nZVI can embedded in all membrane fibers which probably promotes the iron bacteria which should be investigated in further studies. The results from the previous studies with the aim of simultaneous COD and total nitrogen removal, the values of biofilm thickness reported in the literature vary considerably and were determined as 1,6 mm, 600 μm, 2,100 μm, and 270 μm (Semmens et al. 2003; Satoh et al. 2004; Li et al. 2008; Hwang et al. 2010). Wang et al. (2021) also determined the biofilm thickness in the range of 200–400 μm during the operation of MBfR, where nitrified bacteria and antibiotic species cephalexin and sulfadiazine compounds were removed. Combined with the results of membrane characteristics, it can be inferred that the evaluation of the influence of nZVI on biofilm thickness is still sophisticated. Consequently, further analysis of the microbial community, including iron bacteria should be performed to elucidate the linkage of nZVI with biofilm thickness.
Figure 7

CLSM observations of biofilms on membrane fibers.

Figure 7

CLSM observations of biofilms on membrane fibers.

Close modal

It can be summarized that the usage of nZVI in manufacturing the gas transfer membrane caused changes in membrane properties, and it could not provide the expected antibacterial effect on biofilm thickness. Additionally, HRT and oxygen gas pressures influenced the performance of the MBfR in terms of COD and ammonium removal rather than nZVI addition. From the results of measurements of biofilm thickness, it remains an open question why nZVI did not have an antibacterial effect on biofilm. Therefore, further research is needed to examine the microbial community structure including the iron bacteria.

The authors gratefully acknowledge the financial support from The Scientific and Technological Research Council of Turkey (TUBITAK) with project number 121Y574.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Abdelfattah
A.
,
Hossain
M. I.
&
Cheng
L.
2020
High-strength wastewater treatment using microbial biofilm reactor: A critical review
.
World Journal of Microbiology and Biotechnology
36
,
1
10
.
Ahmad
D.
,
van den Boogaert
I.
,
Miller
J.
,
Presswell
R.
&
Jouhara
H.
2018
Hydrophilic and hydrophobic materials and their applications
.
Energy Sources, Part A: Recovery, Utilization, and Environmental Effects
40
(
22
),
2686
2725
.
Aksoy
Y.
&
Hasar
H.
2021
Fabrication of PVDF-HF membrane for bubble-free gas transfer via wet phase inversion
.
Journal of Applied Polymer Science
138
(
47
),
51405
.
Aybar
M.
,
Perez-Calleja
P.
,
Li
M.
,
Pavissich
J. P.
&
Nerenberg
R.
2019
Predation creates unique void layer in membrane-aerated biofilms
.
Water Research
149
,
232
242
.
Aydın
E.
,
Erdem
M.
,
Casey
E.
&
Hasar
H.
2021
Oxidation mechanism of chlortetracycline in a membrane aerated biofilm reactor
.
Environmental Technology & Innovation
24
,
101910
.
Bicudo
J.
,
Heffernan
B.
,
Klassen
A.
,
Rao
M.
,
McConomy
J.
,
Syron
E.
&
McDermott
L.
2019
A one year demonstration of nutrient removal with Membrane Aerated Biofilm Reactor (MABR)
. In:
Nutrient Removal and Recovery Symposium 2019
.
Water Environment Federation
.
Côté
P.
,
Peeters
J.
,
Adams
N.
,
Hong
Y.
,
Long
Z.
&
Ireland
J.
2015
A new membrane-aerated biofilm reactor for low energy wastewater treatment: Pilot results
. In
WEFTEC 2015
.
Water Environment Federation
.
Demirsoy
N.
,
Nuray
U.
,
Aysen
O.
&
Kizildag
N.
2015
Nanocomposite nanofibers of polyacrylonitrile (PAN) and silver nanoparticles (AgNPs) electrospun from dimethylsulfoxide
.
Marmara Fen Bilimleri Dergisi
27
,
16
18
.
Dizge
N.
,
Ozay
Y.
,
Simsek
U. B.
,
Gulsen
H. E.
,
Akarsu
C.
,
Turabik
M.
,
Unyayar
A.
&
Ocakoglu
K.
2017
Preparation, characterization and comparison of antibacterial property of polyethersulfone composite membrane containing zerovalent iron or magnetite nanoparticles
.
Membrane Water Treatment
8
(
1
),
51
71
.
Ghaemi
N.
,
Madaeni
S. S.
,
Daraei
P.
,
Rajabi
H.
,
Zinadini
S.
,
Alizadeh
A.
,
Heydari
R.
,
Beygzadeh
M.
&
Ghouzivand
S.
2015
Polyethersulfone membrane enhanced with iron oxide nanoparticles for copper removal from water: Application of new functionalized Fe3O4 nanoparticles
.
Chemical Engineering Journal
263
,
101
112
.
Greenberg
A. E.
,
Clesceri
L. S.
&
Eaton
A. D.
1992
Standard Methods for Examination of Water and Wastewater. American Public Health Association, Washington DC.
Gugliuzza
A.
2015
Membrane Wettability. In: Drioli, E., Giorno, L. (eds) Encyclopedia of Membranes. Springer, Berlin, Heidelberg
.
Hanay
Ö.
,
Taşkan
E.
,
Yıldız
B.
,
Hasar
H.
&
Casey
E.
2014
Gas/substrate fluxes and microbial community in phenol biodegradation using an O2-based membrane biofilm reactor
.
CLEAN–Soil, Air, Water
42
(
1
),
36
42
.
Harouaka
K.
,
Mansor
M.
,
Macalady
J. L.
&
Fantle
M. S.
2016
Calcium isotopic fractionation in microbially mediated gypsum precipitates
.
Geochimica et Cosmochimica Acta
184
,
114
131
.
Hasar
H.
&
Ipek
U.
2010
Gas permeable-membrane for hydrogenotrophic denitrification
.
Clean–Soil, Air, Water
38
(
1
),
23
26
.
Hashim
N. A.
,
Liu
Y.
&
Li
K.
2011
Stability of PVDF hollow fibre membranes in sodium hydroxide aqueous solution
.
Chemical Engineering Science
66
(
8
),
1565
1575
.
He
Z.
,
Mahmud
S.
,
Yang
Y.
,
Zhu
L.
,
Zhao
Y.
,
Zeng
Q.
,
Xiong
Z.
&
Zhao
S.
2020
Polyvinylidene fluoride membrane functionalized with zero valent iron for highly efficient degradation of organic contaminants
.
Separation and Purification Technology
250
,
117266
.
Heffernan
B.
,
Murphy
C. D.
,
Syron
E.
&
Casey
E.
2009
Treatment of fluoroacetate by a Pseudomonas fluorescens biofilm grown in membrane aerated biofilm reactor
.
Environmental Science & Technology
43
(
17
),
6776
6785
.
Heffernan
B.
,
Shrivastava
A.
,
Toniolo
D.
,
Semmens
M.
&
Syron
E.
2017
Operation of a large scale membrane aerated biofilm reactor for the treatment of municipal wastewater
. In:
WEFTEC 2017
.
Water Environment Federation
.
Hou
X.
,
Chen
X.
,
Bi
S.
,
Li
K.
,
Zhang
C.
,
Wang
J.
&
Zhang
W.
2020
Catalytic degradation of TCE by a PVDF membrane with Pd-coated nanoscale zero-valent iron reductant
.
Science of the Total Environment
702
,
135030
.
Hwang
J. H.
,
Cicek
N.
&
Oleszkiewicz
J. A.
2010
Achieving biofilm control in a membrane biofilm reactor removing total nitrogen
.
Water Research
44
(
7
),
2283
2291
.
Hwang
Y.-H.
,
Kim
D.-G.
&
Shin
H.-S.
2011
Effects of synthesis conditions on the characteristics and reactivity of nano scale zero valent iron
.
Applied Catalysis B: Environmental
105
(
1–2
),
144
150
.
Jørgensen
B. B.
,
Findlay
A. J.
&
Pellerin
A.
2019
The biogeochemical sulfur cycle of marine sediments
.
Frontiers in Microbiology
10
,
849
.
Li
T.
,
Liu
J.
,
Bai
R.
&
Wong
F. S.
2008
Membrane-aerated biofilm reactor for the treatment of acetonitrile wastewater
.
Environmental Science & Technology
42
(
6
),
2099
2104
.
Li
J.-F.
,
Xu
Z.-L.
,
Yang
H.
,
Yu
L.-Y.
&
Liu
M.
2009
Effect of TiO2 nanoparticles on the surface morphology and performance of microporous PES membrane
.
Applied Surface Science
255
(
9
),
4725
4732
.
Liao
B. Q.
&
Liss
S. N.
2007
A comparative study between thermophilic and mesophilic membrane aerated biofilm reactors
.
Journal of Environmental Engineering and Science
6
(
2
),
247
252
.
Liu
N.
,
Liu
J.
,
Wang
H.
,
Li
S.
&
Zhang
W.
2022
Microbes team with nanoscale zero-valent iron: A robust route for degradation of recalcitrant pollutants
.
Journal of Environmental Sciences
118
,
140
146
.
Lu
D.
,
Bai
H.
,
Kong
F.
,
Liss
S. N.
&
Liao
B.
2021
Recent advances in membrane aerated biofilm reactors
.
Critical Reviews in Environmental Science and Technology
51
(
7
),
649
703
.
Mei
X.
,
Liu
J.
,
Guo
Z.
,
Li
P.
,
Bi
S.
,
Wang
Y.
,
Yang
Y.
,
Shen
W.
,
Wang
Y.
&
Xiao
Y.
2019
Simultaneous p-nitrophenol and nitrogen removal in PNP wastewater treatment: Comparison of two integrated membrane-aerated bioreactor systems
.
Journal of Hazardous Materials
363
,
99
108
.
Najafi
M.
,
Sadeghi
M.
,
Bolverdi
A.
,
Pourafshari Chenar
M.
&
Pakizeh
M.
2018
Gas permeation properties of cellulose acetate/silica nanocomposite membrane
.
Advances in Polymer Technology
37
(
6
),
2043
2052
.
Nisola
G. M.
,
Orata-Flor
J.
,
Oh
S.
,
Yoo
N.
&
Chung
W.-J.
2013
Partial nitrification in a membrane-aerated biofilm reactor with composite PEBA/PVDF hollow fibers
.
Desalination and Water Treatment
51
(
25–27
),
5275
5282
.
Penboon
L.
,
Khrueakham
A.
&
Sairiam
S.
2019
Tio2 coated on PVDF membrane for dye wastewater treatment by a photocatalytic membrane
.
Water Science and Technology
79
(
5
),
958
966
.
Rabuni
M. F.
2014
Chemical and Thermal Stability Studies of Hydrophobic and Hydrophilic Polyvinylidene Fluoride (PVDF) Membranes in Alkaline Environments
.
University of Malaya, Kuala Lumpur. (Malaysia)
.
Salman
M. Y.
,
Taşkan
E.
&
Hasar
H.
2022
Comparative potentials of H2-and O2-MBfRs in removing multiple tetracycline antibiotics
.
Process Safety and Environmental Protection
167
,
184
191
.
Saranya
R.
,
Arthanareeswaran
G.
,
Ismail
A. F.
,
Dionysiou
D. D.
&
Paul
D.
2015
Zero-valent iron impregnated cellulose acetate mixed matrix membranes for the treatment of textile industry effluent
.
RSC Advances
5
(
77
),
62486
62497
.
Sathyamoorthy
S.
,
Tse
Y.
,
Gordon
K.
,
Houweling
D.
&
Coutts
D.
2019
BNR process intensification using membrane aerated biofilm reactors
. In
Nutrient Removal and Recovery Symposium 2019
.
Water Environment Federation
.
Semmens
M. J.
,
Dahm
K.
,
Shanahan
J.
&
Christianson
A.
2003
COD and nitrogen removal by biofilms growing on gas permeable membranes
.
Water Research
37
(
18
),
4343
4350
.
Shanahan
J. W.
&
Semmens
M. J.
2004
Multipopulation model of membrane-aerated biofilms
.
Environmental Science & Technology
38
(
11
),
3176
3183
.
Silva
L. L. S.
,
Abdelraheem
W.
,
Nadagouda
M. N.
,
Rocco
A. M.
,
Dionysiou
D. D.
,
Fonseca
F. V.
&
Borges
C. P.
2021
Novel microwave-driven synthesis of hydrophilic polyvinylidene fluoride/polyacrylic acid (PVDF/PAA) membranes and decoration with nano zero-valent-iron (nZVI) for water treatment applications
.
Journal of Membrane Science
620
,
118817
.
Stookey
L. L.
1970
Ferrozine – a new spectrophotometric reagent for iron
.
Analytical Chemistry
42
(
7
),
779
781
.
Sun
M.
,
Zou
L.
,
Wang
P.
,
Fan
X.
,
Pan
Z.
,
Liu
Y.
&
Song
C.
2022
Nano valent zero iron (NZVI) immobilized CNTs hollow fiber membrane for flow-through heterogeneous Fenton process
.
Journal of Environmental Chemical Engineering
10
(
3
),
107806
.
Sunner
N.
,
Long
Z.
,
Houweling
D.
,
Monti
A.
&
Peeters
J.
2018
MABR as a low-energy compact solution for nutrient removal upgrades – results from a demonstration in the UK
. In:
WEFTEC 2018
.
Water Environment Federation
,
1264
1281
.
Terada
A.
,
Ito
J.
,
Matsumoto
S.
&
Tsuneda
S.
2009
Fibrous support stabilizes nitrification performance of a membrane-aerated biofilm: The effect of liquid flow perturbation
.
Journal of Chemical Engineering of Japan
42
(
8
),
607
615
.
Tian
H.
,
Hu
Y.
,
Xu
X.
,
Hui
M.
,
Hu
Y.
,
Qi
W.
,
Xu
H.
&
Li
B.
2019
Enhanced wastewater treatment with high o-aminophenol concentration by two-stage MABR and its biodegradation mechanism
.
Bioresource Technology
289
,
121649
.
Turken
T.
,
Sengur-Tasdemir
R.
,
Koseoglu-Imer
D. Y.
&
Koyuncu
I.
2015
Determination of filtration performances of nanocomposite hollow fiber membranes with silver nanoparticles
.
Environmental Engineering Science
32
(
8
),
656
665
.
Wang
C.-B.
&
Zhang
W.
1997
Synthesizing nanoscale iron particles for rapid and complete dechlorination of TCE and PCBs
.
Environmental Science & Technology
31
(
7
),
2154
2156
.
Wang
J.
,
Liu
G.-F.
,
Lu
H.
,
Jin
R.-F.
,
Zhou
J.-T.
&
Lei
T.-M.
2012
Biodegradation of Acid Orange 7 and its auto-oxidative decolorization product in membrane-aerated biofilm reactor
.
International Biodeterioration & Biodegradation
67
,
73
77
.
Wu
Y.
,
Wu
Z.
,
Chu
H.
,
Li
J.
,
Ngo
H. H.
,
Guo
W.
,
Zhang
N.
&
Zhang
H.
2019
Comparison study on the performance of two different gas-permeable membranes used in a membrane-aerated biofilm reactor
.
Science of The Total Environment
658
,
1219
1227
.
Yang
C.
,
Li
K.
,
Xu
L.
,
Wang
Z.
,
Yu
L.
&
Wang
J.
2022
Reduction of nitrobenzene by a zero-valent iron microspheres/polyvinylidene fluoride (mZVI/PVDF) membrane
.
Separation and Purification Technology
282
,
120006
.
Yavuz
F. N. S.
,
Sengur Tasdemir
R.
,
Turken
T.
,
Urper
G. M.
&
Koyuncu
I.
2019
Improvement of anti-biofouling properties of hollow fiber membranes with bismuth-BAL chelates (BisBAL)
.
Environmental Technology
40
(
1
),
19
28
.
Zhao
J.
,
Li
F.
,
Cao
Y.
,
Zhang
X.
,
Chen
T.
,
Song
H.
&
Wang
Z.
2021
Microbial extracellular electron transfer and strategies for engineering electroactive microorganisms
.
Biotechnology Advances
53
,
107682
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).