Phosphate-rich municipal wastewater treatment using magnetite nanoparticles dispersed oakwood biochar-assisted pretreatment followed by air-gap membrane distillation

An integrated approach combining Fe 3 O 4 nanoparticle-loaded biochar-assisted pretreatment and air-gap membrane desalination (AGMD) was investigated for treating phosphate-rich wastewater. This study addresses key limitations of conventional coagulants, such as residual chemical load and limited removal efficiency. Fe 3 O 4 nanoparticles were co-precipitated on oakwood biochar (Fe 3 O 4 /BC). The Fe 3 O 4 /BC composite acts as a coagulant that can be rapidly separated under an applied magnetic field. This study systematically evaluated the pretreatment performance for turbidity and phosphate removal in phosphate-rich wastewater. A 94% phosphate and 77% turbidity reduction was obtained with a 2.5 g L -1 Fe 3 O 4 /BC dosage. In parallel, hydrophilic polyethersulfone (PES) and hydrophobic polytetrafluoroethylene (PTFE) membrane configurations were tested in AGMD with and without Fe 3 O 4 /BC-assisted pretreatment. The AGMD permeate flux, overall energy efficiency, and membrane fouling were compared in PES/AGMD, PT + PES/AGMD, PTFE/AGMD, and PT + PTFE/AGMD processes. A 21 times greater initial flux was obtained with the PES membrane in AGMD at 70 °C and 27.3 °C hot feed and coolant temperatures than with the PTFE membrane. However, higher flux reduction and higher salt deposition were observed on the hydrophilic PES membrane without pretreatment. The energy consumption for the PES/AGMD, PT + PES/AGMD, PTFE/AGMD, and PT + PTFE/AGMD processes were 1.516, 1.492, 4.559, and 4.245 kWh L -1, respectively. Overall, the Fe 3 O 4 /BC-assisted pretreatment combined with AGMD presents a promising strategy for phosphate-rich municipal wastewater treatment. This enables effective contaminant removal while protecting membrane performance and reducing operational costs. Further work will focus on the reusability of Fe 3 O 4 /BC as a coagulant, long-term membrane performance, and pilot-scale validation.

Water purification is critical to protecting the health of living organisms and ensuring ecosystem safety. Wastewater can have a wide range of chemical contaminants and physical impurities. Current treatments include physical, chemical, and biological approaches. Wastewater remediation is important for removing chemicals, pathogens, and other contaminants to reduce the risk of waterborne diseases during conversion to drinking water.

Current water purification methods include micro- and ultrafiltration (Fatima et al., 2023; Pinto et al., 2017; Tchobanoglous et al., 1998), reverse osmosis (Bodalo-Santoyo et al., 2003), ion exchange (Marin et al., 2024), adsorption (Olabode et al., 2026; Rodrigo and Kommalapati, 2025), UV disinfection (Das, 2001; González et al., 2023), ozonation (Ikehata et al., 2008; Li et al., 2023), and advanced oxidation processes (Ikehata et al., 2008) used to purify chemical, physical, and biological contaminants from wastewater. Emerging technologies such as membrane filtration offer high removal efficiency. However, membrane filtration requires significant energy and is prone to membrane fouling. Therefore, a low-cost pretreatment method is required to improve water quality prior to membrane filtration.

In general, FeCl 3 (Amuda and Amoo, 2007), FeSO 4 (Rodrigues et al., 2013), polyaluminium chloride (Zouboulis and Tzoupanos, 2010), and polyacrylamide polymers are commonly used as coagulants in wastewater treatment. Even though these coagulants improve solids removal, they can also create downstream challenges, such as residual excess chemicals in treated water. To overcome this problem, low-cost and efficient natural materials need to be explored. Biochar is a carbon-rich product which is produced under a low-oxygen environment through a process called pyrolysis. Biochar is mainly used as an adsorbent material (Navarathna et al., 2020) in wastewater treatment facilities, as well as for carbon sequestration and soil improvement in agriculture (Abeysinghe et al., 2026; Rodrigo, Varco, et al., 2025). Biochar has a high porous structure and high surface area, rich in surface functional groups such as -OH, -COOH, and -C=O groups that can adsorb pollutants (Navarathna et al., 2022). These sites can serve to facilitate bridging among colloidal particles during coagulation. However, to improve the coagulant properties, less soluble Fe 3 O 4 nanoparticles (Rodrigo et al., 2024) can be loaded on the biochar porous structure. The combination of a porous biochar structure and Fe 3 O 4 nanoparticles enables Fe 3 O 4 /BC to destabilize colloidal suspensions more effectively. During coagulation, positively charged sites on Fe 3 O 4 /BC can neutralize the negatively charged sites on colloidal particles, thereby reducing electrostatic repulsion between colloids. Further, Fe 3 O 4 /BC can lead to the formation of larger flocs, and the porous structure allows more to adsorb within its pores. Meanwhile, the porous biochar surface and Fe 3 O 4 nanoparticles provide a high surface area and functional groups that enable adsorption of dissolved and colloidal contaminants onto the solid surface. Additionally, on an industrial scale, the magnetic property of Fe 3 O 4 /BC is useful for the rapid separation of chemical contaminants and adsorbed flocs in a magnetic field prior to downstream processing. Together, these mechanisms improve overall contaminant removal, and the magnetite phase may also allow easier separation under magnetic field.

In most wastewater treatment processes, air-gap membrane distillation is used as the final desalination step before discharge. AGMD is a thermal separation technique that uses a porous membrane to transfer water vapor from a hot feed into an air gap, where it condenses upon contact with the condensation plate (Alklaibi and Lior, 2005). AGMD has the potential for high rejection of salt and non-volatile contaminants, with relatively reduced wetting and scaling risks. However, AGMD typically experiences lower permeate flux and potential efficiency losses due to air-gap thermal resistance and heat losses. Also, AGMD performance can still degrade if membrane wetting or scaling occurs. Therefore, the membrane selection is important, where the hydrophobic nature, porosity, and thickness determine the rate of water vapor diffusion towards the air gap. Hydrophobicity is important when selecting a membrane for AGMD. The membrane wetting resistance is important for maintaining high selectivity and high-quality permeate water, and for preventing membrane fouling. Polyethersulfone (PES) is a hydrophilic membrane that is highly prone to fouling when the water matrix contains high levels of foulants and high total dissolved solids. Therefore, pretreatment is crucial before using a hydrophilic PES membrane in the AGMD process. However, a polytetrafluoroethylene (PTFE) membrane is hydrophobic, chemically inert, minimizes wetting, and reduces fouling. In the literature, a few studies have shown improvements in water quality and membrane performance in membrane distillation when coagulation is used as a pretreatment (Abdel-Karim et al., 2019; Essalhi and Khayet, 2012; Rodrigo et al., 2026). In this study, the use of a hydrophilic PES membrane and a hydrophobic PTFE membrane was compared, along with and without Fe 3 O 4 /BC assisted pretreatment process.

This study investigates the potential of Fe 3 O 4 /BC as a coagulant for pretreatment and the use of air-gap membrane distillation as the final stage of the treatment process before discharge. Further, this study compares the use of hydrophilic polyethersulfone and hydrophobic polytetrafluoroethylene membranes, with and without pretreatment. This study highlights and compares each treatment process to overcome its limitations. Moreover, this study enables optimization of energy consumption for each treatment, thereby supporting water management in wastewater treatment facilities.

Chemicals and materials

All chemicals [FeCl 3 (162.2 g mol -1, 98%), FeSO 4. 7H 2 O (278.01 g mol -1, 98%), and KOH (56.01 g mol -1)] were purchased from ThermoFisher Scientific. Oakwood chips were purchased from WebstaurantStore® (Item #: 688CHIPOAK). The wastewater sample was collected from the Prairie View A&M University (PVAMU) wastewater treatment plant (30.1018 N, −95.9952 W) on 10 November 2025. The sample was collected at a surface depth of 0.5 m.

An acrylic air-gap membrane distillation cell was purchased from Sterlitech Corporation in Kent, WA, United States. The AGMD cell has a 5.6 mm air-gap thickness, a 1.9 × 10 −4 m feed-channel depth, a 0.146 m feed-channel width, and a 0.014 m 2 effective membrane surface area. A 0.1 μm pore size hydrophilic polyethersulfone (PES) (product no. 012005) and hydrophobic polytetrafluoroethylene (PTFE) membranes (product no. 1121841) were also purchased from Sterlitech Corporation, Kent, WA, United States. We used the same pumps, flow meters, hot feed generators, and chiller that were used in our previous study (Rodrigo et al., 2025).

Preparation of oakwood biochar (BC) and synthesis of Fe 3 O 4 nanoparticles on BC (Fe 3 O 4 /BC)

A 100.0 g portion of oakwood biomass was oven-dried at 105 °C for 12 h, and the moisture content was calculated using Equation 1. Then, a 25.00 g portions of dried oakwood biomass were placed in a bomb reactor and pyrolyzed at 400 °C for 1 h in a muffle furnace under reduced oxygen levels. The yield percentage of BC was calculated using Equation 2.

Magnetite, Fe 3 O 4, deposition onto BC was described previously in our study (Rodrigo et al., 2022; Rodrigo et al., 2024). Briefly, a 1:2 M ratio of Fe 2+ (5.56 g of FeSO 4. 7 H 2 O) and Fe 3+ (6.62 g of FeCl 3) solution was prepared in 350 mL of deionized water. Then, 25.0 g of dried oakwood biochar (BC) was mixed (500 rpm) into this solution to form a slurry. The mixture was kept for 3 h to adsorb Fe 2+ and Fe 3+ ions onto the biochar surface. The pH was raised to 9.09 by adding 100 mL of 2 mol L -1 KOH solution at a rate of 5 mL min -1. The pH change during the addition of KOH is shown in Supplementary Figure S1. The resulting mixture was aged 24 h at 25 °C to promote further nucleation and deposition of Fe 3 O 4 nanoparticles on the BC surface. An ethanol wash was performed to remove any organic and inorganic residues, and Fe 3 O 4 /BC was dried at 50 °C for 6 h.

Water quality

Gravimetric analysis

The Prairie View A&M wastewater was characterized by measuring total solids (TS) (total solids method 2540B), total dissolved solids (TDS) (total dissolved solids method 2540C), total suspended solids (TSS) (total suspended solids method 2540D), total volatile solids (TVS) (total volatile solids method 2540E), total fixed solids (TFS), and other water quality parameters (pH, conductivity, etc.). The total solids (TS) were measured as follows. Briefly, 200.00 mL portions of untreated or treated wastewater samples were measured into preweighed and dried crucibles and kept in an electric oven for 24 h at 95 °C. Then, the temperature was raised to 105 °C for an additional 4 h. The crucibles were weighed after reaching room temperature. The TFS were measured by further heating the crucibles for 4 h at 550 °C, and then measuring the final weights after they reached room temperature.

Chemical oxygen demand (COD), total organic carbon (TOC), and phosphate analysis

The chemical oxygen demands of the wastewater and treated samples were measured using the HACH Method 8000. 2 mL of each sample was mixed with 2 mL of HACH reagent and placed in the DRB 200 digester for 2 h at 150 °C. Similarly, 2 mL of DI water was used as the blank. The digestate’s chemical oxygen demand was measured using the 435 COD HR method in the DR 390 instrument at 620 nm.

The TOC content of the initial and final wastewater samples was measured using the HACH Method 10,129. Initially, 0.4 mL of pH two buffer solution was mixed with 10 mL of the sample in a 50 mL Erlenmeyer flask. The flask was then placed on a magnetic stirrer for 10 min. One TOC persulfate powder pillow was added to each acid-digested sample. After that, 3 mL of organic-free water and 3 mL of the prepared sample were transferred to the reagent blank and prepared sample vials, respectively. Then, blue indicator ampules were placed inside vials and kept in the reactor for 2 h at 105 °C. Finally, the TOC content was measured using the DR 390 instrument at 600 nm.

The phosphate concentration was determined by the UV-visible molybdenum blue method. Briefly, to 4.0 mL of the water sample, 1.0 mL of ammonium molybdate and 0.5 mL of hydrazine sulfate were added, and the mixture was made up to 10.0 mL. The resultant mixture was kept for 30 min at 60 °C. After developing the blue color phosphomolybdate complex, the absorbance was measured at 860 nm. DI water was used as the blank solution.

Characterization of BC, Fe 3 O 4 /BC, and PTFE membrane

Fourier transform infrared spectroscopy

Fourier transform infrared spectroscopic (FTIR) analysis was performed using a Nicolet iS5 FT-IR spectrometer (Thermo Scientific, Inc.) to characterize the surface functional groups present on BC and Fe 3 O 4 /BC materials and PES and PTFE membranes. All FTIR spectra were acquired over the 4000–400 cm -1 wavenumber range using 32 scans, with a resolution of 4 cm -1.

Scanning electron microscopy and X-ray diffraction

Scanning electron microscopy (SEM, JSM-6010LA, JEOL) was employed to investigate the morphology of the BC and Fe 3 O 4 /BC before and after pretreatment. Briefly, the samples were mounted on carbon tape, gold-coated by sputtering, and analyzed at an operational voltage of 10–15 kV. The X-ray powder diffraction patterns of BC and Fe 3 O 4 /BC were obtained from a Rigaku Ultima III instrument using Cu Kα radiation, where λ is equal to 0.154 nm. Diffractions were obtained from 5°–80°.

Pretreatment

A jar test was performed to pretreat wastewater from Prairie View A&M University using BC and Fe 3 O 4 /BC as coagulants. Fe 3 O 4 /BC was mixed with wastewater to Fe 3 O 4 /BC concentrations of 0.625, 1.250, 2.500, 3.750, 5.000, 6.250, and 7.500 g L -1, and a 7.500 g L -1 dose of BC was used to compare with Fe 3 O 4 /BC. A 200 rpm initial mixing rate was applied for 3 min, then reduced to 20 rpm over a 3–30 min interval. Then, the treated solutions were allowed to settle for another 30 min before the supernatant was separated. Turbidity and remaining phosphate concentrations were measured. The optimized Fe 3 O 4 /BC dose of 2.5 g L -1 was used to pretreat the wastewater prior to air-gap membrane distillation.

Air gap membrane distillation

A hot feed temperature of 70 °C and a coolant (DI water) temperature of 27.3 °C were used in all experimental runs that were carried out for 2.5 h for the PES membrane and 5 h for the PTFE membrane. A 2 L min -1 (0.12 m s -1 cross-flow velocity) flow rate was used for both hot feed and coolant. The investigations were conducted with both PES and PTFE membranes. The investigation involved recording the real-time flux at 1-min intervals and subsequently monitoring the variations for 2.5 and 5 h for the PES and PTFE membranes, respectively.

The real-time permeate flux (J) was calculated as follows (Equation 3).

J is the permeate flux, ΔW is the weight change of water collected per minute (g min -1), d H 2 O is the density of water at 25 °C (0.997 kg L -1), and A is the effective membrane area (0.014 m 2).

Wastewater quality

Prairie View A&M University wastewater was characterized by analyzing pH, conductivity, TS, TDS, TFS, TSS, COD, TOC, Turbidity, and phosphate. Table 1 reports the initial wastewater quality. The wastewater influent exhibited a slightly acidic pH (6.47 ± 0.03) and a high amount of dissolved solids (831 ± 10 mg L -1) compared to suspended solids (40 ± 2 mg L -1). Further, a high phosphate concentration (18.4 ± 0.3 mg L -1) was observed, which may be influenced by higher turbidity (13 ± 1 NTU). Our previous study reported an initial pH of 6.80 ± 0.30, conductivity of 626 ± 1 μS cm -1, TS of 405 ± 7 mg L -1, TFS of 232 ± 18 mg L -1, TDS of 391 ± 4 mg L -1, TSS of 14 ± 3 mg L -1, and COD of 36 ± 3 mg L -1 (Rodrigo et al., 2025). The water quality values reported in this study were slightly higher than those in our previous study. Understanding the initial water quality is required to select the treatment process, determine the membrane, and optimize operational conditions.

Characteristics of BC, Fe 3 O 4 /BC, PES membrane, and PTFE membrane

X-ray diffraction pattern of BC and Fe 3 O 4 /BC

Figure 1 displays the XRD patterns of pristine BC and Fe 3 O 4 /BC. A peak around 23° indicates the formation of graphitized biochar (Figure 1A). In the Fe 3 O 4 /BC composite (Figure 1B), in addition to the characteristic peak of BC, several other peaks were observed at 18.5, 30.9, 35.9, 37.3, 43.5, 54, 57.5, 63, 66.7, and 74°, which can be assigned to (111), (202), (311), (222), (400), (331), (511), (404), (531) and (533) planes of Fe 3 O 4 (JCPDS card no. 01-1111). Fe 3 O 4 nanoparticles on Fe 3 O 4 /BC exhibit a face-centered cubic structure on the BC surface and an average crystallite size of ∼24 nm (Rodrigo et al., 2022; Rodrigo et al., 2024). studies have reported a similar structure with characteristic X-ray diffraction patterns, coprecipitation of Fe 3 O 4 on biochar surfaces. The results from both XRD and FTIR analyses confirmed the successful incorporation of Fe 3 O 4 particles into the biochar.

Elemental composition and scanning electron microscopy

Oak wood biomass had a moisture content of 8.2% ± 0.3%. Biochar yield was 62.4% ± 0.7% at a pyrolysis temperature of 400 °C. After depositing Fe 3 O 4 nanoparticles onto 25.0 g of BC, the final dried weight of Fe 3 O 4 /BC increased to 29.33 g, which had a 17.3% loading of Fe 3 O 4.

Elemental combustion analysis of BC found 59.9% ± 1.0% C, 5.1% ± 0.1% H, 33.9% ± 1.0% O, 0.8% ± 0.1% N, 0.0% ± 0.0% S, and 0.3% ± 0.0% ash (Table 2). After the deposition of Fe 3 O 4 nanoparticles, the resulting Fe 3 O 4 /BC C% dropped to 52.2% ± 0.2%, while ash% increased to 14.8% ± 0.6%. The morphologies of BC and Fe 3 O 4 /BC are shown in Figure 2. Pristine BC exhibited an irregular, flake-like morphology. On the other hand, the Fe 3 O 4 -NP-loaded BC exhibited a compact morphology with interconnected flakes. Compared to the BC, the composite exhibited lower porosity and fewer voids, which was due to the deposition of Fe 3 O 4 particles.

Fourier transform infrared spectroscopy

The chemical composition of pristine BC, Fe 3 O 4 /BC composite, PES, and PTFE was studied by FTIR, and the results are depicted in Figure 3. The BC demonstrated a broad O-H stretching band in the region of 3600–3200 cm -1 (Figure 3A). A weak peak at 2902 cm -1 corresponds to the C-H stretching vibration. The peaks around 1703 and 1606 cm -1 are attributed to C=O and C=C groups, respectively. The band near 1000 cm -1 is due to the C-O stretching vibration. After loading Fe 3 O 4 on BC (Figure 3A), the resulting composite preserved all the characteristic vibrations of BC, with additional peaks appearing at 560 and 620 cm -1 corresponding to the Fe-O vibration of Fe 3 O 4. These findings are similar to our previous studies (Pant et al., 2017; Rodrigo et al., 2024). The PES depicted several characteristic bands attributed to the benzene ring, ether, and sulfone functional groups. The C-H stretching peak originating from the benzene ring was observed at 3075 cm -1. The peaks between 1600 and 1400 cm -1 were due to C=C vibration. The characteristic sulfone absorptions (S=O) were present at 1151 and 1105 cm -1. The ether stretch (C-O-C) was observed at around 1250–1230 cm -1. The absorption bands presented at 837, 780, and 720 cm -1 were attributed to the para-substituted aromatic C-H bending, aromatic ring deformation modes, and C-S stretching, respectively (Alenazi et al., 2018; Koloti et al., 2018). On the other hand, the PTFE exhibited characteristic C-F stretching peaks at 1206, 1153, 636, 553, and 503 cm -1, indicating its fluorinated background (Rodrigo, Fatima, et al., 2025).

Pretreatment

Pretreatment is crucial in wastewater treatment for maintaining membrane lifespan, membrane performance, final water quality, and energy consumption. Therefore, coagulation, flocculation, and sedimentation steps were performed during the pretreatment process to improve water quality and membrane performance and to reduce energy costs. Fe 3 O 4 /BC was used as a coagulant in this study. The remaining phosphate concentration was 18.1 ± 0.2 mg L -1 with pristine BC. Fe 3 O 4 /BC reduced the final phosphate concentrations to 9.2 ± 0.1, 3.4 ± 0.1, 1.1 ± 0.1, 0.9 ± 0.0, 0.7 ± 0.0, 0.4 ± 0.0, and 0.3 ± 0.0 mg L -1, respectively, for 0.625, 1.25, 2.5, 3.75, 5, 6.25, and 7.5 g L -1 Fe 3 O 4 /BC dosages (Figure 4A). Similarly, turbidity declined to 10 ± 1, 7 ± 1, 3 ± 1, 1 ± 1, 0 ± 1, 0 ± 0, and 0 ± 0, respectively, from 0.625, 1.25, 2.5, 3.75, 5, 6.25, and 7.5 g L -1 Fe 3 O 4 /BC dosages (Figure 4B). A 2.5 g L -1 Fe 3 O 4 /BC dose was selected to pre-treat the Prairie View A&M University wastewater before the air gap membrane distillation process, resulting in 94% phosphate and 77% turbidity reductions.

Pretreated water quality is reported in Table 3. The pH of the pretreated water was slightly increased to 6.82 ± 0.04. TS, TDS, and TFS reduction was ∼23%, ∼20%, and ∼25%, respectively. A significant TS, TDS, and TFS reduction was not observed from Fe 3 O 4 /BC because of the incapability of the adsorption of Na +, Mg 2+, K +, Ca 2+, SO 4 2-, NO 3 −, and Cl − ions present in the wastewater. Significant reductions in TSS (70%) and turbidity (77%) were observed due to the presence of a porous structure on BC, which can adsorb suspended solid particles. Also, the reductions in COD (∼42%) and TOC (∼29.4%) were due to the chemisorption capability of Fe 3 O 4 nanoparticles deposited on Fe 3 O 4 /BC. This greater COD reduction could be due to the adsorption of organic matter and suspended solids onto the porous biochar surface.

Post characteristics of BC and Fe 3 O 4 /BC after pretreatment

The morphologies of both BC and Fe 3 O 4 /BC after the pretreatment process were obtained and compared with those from prior treatments. As in Figure 5A, the BC showed clear flake-like structures without the deposition of any additional matter. EDS also confirmed the presence of C and O, which are the main components of biochar. In contrast, significant morphological changes were observed in the Fe 3 O 4 /BC sample after pretreatment (Figures 5C,D). The Fe 3 O 4 /BC exhibited a compact morphology, suggesting additional surface deposition. The EDS analysis revealed the presence of carbon (C), oxygen (O), iron (Fe), sulfur (S), chlorine (Cl), potassium (K), sodium (Na), magnesium (Mg), calcium (Ca), and phosphorus (P). Among these elements, C, O, and Fe were the main constituents of the Fe 3 O 4 /BC, whereas S, Cl, K, Na, Mg, Ca, and P were deposited during the pretreatment process. The elemental composition of the BC and Fe 3 O 4 /BC after the pretreatment is summarized in Supplementary Table S2.

Air-gap membrane distillation

Air-gap membrane distillation was performed in the final stage as the final purification step. The AGMD performance was evaluated with and without pretreatment to assess the significance of the pretreatment step prior to air gap membrane distillation. The hot feed and the coolant were maintained at 70 °C and 27.3 °C, respectively, throughout the experiment. Deionized water showed average fluxes of 165.7 ± 2.6 and 7.4 ± 0.3 L m -2 h -1 for PES and PTFE membranes, respectively, at hot feed and coolant flow rates of 2 L min -1. Rodrigo et al. (2025) study has reported 2.3 ± 0.2 L m -2 h -1 water flux at 50 °C of hot feed (DI water), and with similar other conditions (Rodrigo et al., 2025).

Untreated wastewater (TS = 871 ± 10 mg L -1, TDS = 831 ± 10 mg L -1, TSS = 40 ± 2 mg L -1, and Turbidity = 13 ± 1 NTU) was directly exposed to AGMD with hydrophilic PES and hydrophobic PTFE membranes. The flux initiated at 169.7 (Figure 6A) and 7.9 L m -2 h -1 (Figure 6B), respectively, with PES and PTFE membranes, which was a 21 times greater flux observed with the PES membrane compared to the PTFE membrane. However, flux declined to 115.2 and 5.8 L m -2 h -1 at 2.5 and 5 h, respectively, for PES and PTFE membranes. The drastic decrease in flux was due to membrane fouling during distillation. The flux decreased in PES and PTFE to 32.1% and 26.6%, respectively, during the 2.5- and 5-h experimental periods. The higher flux decrease with the PES membrane was due to the large amount of foulant deposition on the hydrophilic PES membrane. However, after the pretreatment with 2.5 g L -1 of Fe 3 O 4 /BC dose, the TSS and turbidity values were reduced by 70% and 77%, respectively (Table 3). A slight decrease in flux was observed during the distillation process for pretreated wastewater. The flux reduction was 155.6 and 7.4 L m -2 h -1 after 2.5 and 5 h, respectively. In summary, direct membrane distillation leads to membrane fouling, which increases energy consumption.

FTIR was performed to analyze the post characteristics on the membrane surface. Salt deposition on the membrane surface can reduce the density of surface functional groups, thereby lowering the %transmittance of significant peaks. The %transmittance of the benzene peaks around 3075 cm -1 and the peaks around 3500 cm -1 in the PES membrane decreased significantly after the AGMD process without the pretreatment process (Figure 7A). However, the FTIR of the used hydrophobic PTFE showed a new peak around 3500 cm -1 without the pretreatment process. This could be due to the deposition of organic contaminants having -OH functional groups (Figure 7B). However, both the PES and PTFE membrane surface morphologies were not much affected after the Fe 3 O 4 /BC assisted pretreatment.

Water quality

Knowing the final water quality is critical before discharge. The final permeate water quality is presented in Table 4. The permeate water quality in AGMD shows a clear improvement in both PES and PTFE membranes. However, pretreatment yielded much higher water quality. The pH of AGMD-treated permeate ranged from 6.80 to 7.23. Conductivity dropped dramatically after the AGMD process, which reflects substantial salt rejection. The reductions in TS, TDS, TFS, and TVS were not significantly improved by the PES membrane, even after the pretreatment process. However, TSS and COD reductions were significantly improved with both PES and PTFE membranes during pretreatment, reflecting the performance of Fe 3 O 4 /BC during pretreatment and the slight incapability of the AGMD process at a 70 °C feed temperature. In summary, the hydrophobic PTFE membrane showed better performance than the hydrophilic PES membrane. However, Fe 3 O 4 /BC-assisted pretreatment improved water quality for both PES and PTFE membranes. Phosphate is effectively removed from the hydrophobic PTFE membrane, and the Fe 3 O 4 /BC-assisted pretreatment reduced residual phosphate in AGMD permeate to 0.1 mg L -1 with the hydrophilic PES membrane.

Energy consumption and cost analysis

Electric energy usage during pretreatment and air-gap membrane distillation was considered to calculate energy consumption (kWh) and the average cost per 1 L of purified water. The energy consumption during the pretreatment process was 0.00095 kWh L -1. AGMD requires electric energy to heat the feed from 25 °C to 70 °C, maintaining the hot feed and coolant temperatures of 70 °C and 27.3 °C throughout the experimental time, and circulating the feed and coolant through the AGMD cell. The energy consumptions were 1.52, 1.49, 4.56, and 4.24 kWh L -1 for the PES/AGMD, PT + PES/AGMD, PTFE/AGMD, and PT + PTFE/AGMD processes, respectively (Figure 8). Based on energy consumption, the PTFE membrane-assisted AGMD process requires more energy than the PES membrane-assisted AGMD process. This was due to the higher flux obtained from the PES membrane than the PTFE membrane. Moreover, the Fe 3 O 4 /BC-assisted pretreatment reduced the energy consumption to 0.03 and 0.32 kWh L -1, respectively, for the PES- and PTFE-membrane-based AGMD processes. This was due to the requirement for low energy consumption during the pretreatment process and the improvement in AGMD flux after pretreatment. According to the Texas Utilities Electric Co., the electricity cost (0.118 $/kWh) (Zarnikau et al., 2023). The average costs for the PES/AGMD, PT + PES/AGMD, PTFE/AGMD, and PT + PTFE/AGMD processes were 0.179, 0.176, 0.538, and 0.501 $ L -1, respectively. Overall, the Fe 3 O 4 /BC assisted pretreatment process reduces the energy consumption in the AGMD process.

The Fe 3 O 4 /BC pretreatment can reduce operating costs compared to conventional chemical coagulation such as FeCl 3 and polyacrylamide by reducing chemical residuals. Fe 3 O 4 /BC provides magnetic property which enables rapid separation under an applied magnetic field which helps limit excess chemicals remaining in the treated water. Also, Fe 3 O 4 /BC reduces foulant precursors reaching AGMD, which helps protect membrane performance.

This study highlights the importance of an optimized membrane-based distillation process for phosphate-rich wastewater, where high membrane-fouling conditions necessitate integrated approaches. This study demonstrates effective phosphate (∼94%) and turbidity (∼77%) removal at an optimized Fe 3 O 4 /BC dosage of 2.5 g L -1. Superior permeate quality was achieved through Fe 3 O 4 /BC-assisted pretreatment integrated with AGMD. The descending order of permeate water quality for each treatment is PT + PTFE/AGMD > PTFE/AGMD > PT + PES/AGMD > PES/AGMD. However, the PTFE/AGMD configuration had the highest energy demand compared to the PES/AGMD configurations. Therefore, each configuration can alter the needs of energy cost and water quality. The integrated best performing configuration (PT + PTFE/AGMD) achieved an energy cost of 0.5 $ L -1, which supports the overall potential for cost reduction through combined chemical and energy savings (Figure 9).

Overall, the Fe 3 O 4 /BC-assisted pretreatment demonstrated higher phosphate and turbidity removal, reducing membrane fouling and extending membrane lifespan under the tested conditions. However, this study may not capture the real municipal wastewater, including fluctuating phosphate levels, total solids, and other potential foulants. Future work is required to evaluate the reusability of Fe 3 O 4 /BC, magnetic recovery efficiency, and coagulation performance over multiple cycles to advance the practical deployment of Fe 3 O 4 /BC as a coagulant. Further, the long-term stability and potential leaching of Fe 3 O 4 nanoparticles from the BC matrix have to be investigated under different water matrices. Also, long-term membrane fouling dynamics have to be investigated under continuous operation along with the permeate water quality.