Synergistic Effect of NiFe-LDH and PES/SPSf Matrix on Metal Ion Rejection Efficiency from Surface Water

TL;DR

This study creates a new membrane combining PES/SPSf and NiFe-LDH to efficiently remove heavy metals and proteins from water.

Contribution

A novel membrane with enhanced metal ion rejection and antifouling properties through a synergistic polymer-LDH combination.

Findings

The M3 membrane achieved a pure water flux of 218 L.m−2h−1, three times higher than the base membrane.

M3 removed 92.4% of BSA and over 90% of Cd2+, Pb2+, and Cu2+ from water.

The membrane maintained over 65.9% flux recovery after three fouling–cleaning cycles.

Abstract

Clean water remains a pressing global challenge and developing membranes that are both efficient and durable is critical. This study combined two polymers, polyethersulfone (PES) and sulfone-modified polysulfone (SPSf), with NiFe-layered double hydroxides (LDHs) to create a new class of multifunctional membranes. The membranes were characterized using FTIR, SEM, water contact angle, and zeta potential. The addition of NiFe-LDH fillers improved the hydrophilicity and surface structure of the membranes and enhanced the separation performance of the resulting membranes. The best-performing membrane (M3, with 2 wt.% NiFe-LDH) delivered pure water flux of about 218 L.m−2h−1, which was nearly three times higher than that of the pristine PES/SPSf membrane. Furthermore, M3 removed approximately 92.4% of bovine serum albumin (BSA), attributed to the synergistic combination of size exclusion, electrostatic repulsion, and hydrophilicity. The membrane also showed excellent antifouling properties, maintaining over 65.9% and 71.2% flux recovery after three fouling–cleaning cycles for BSA solution and surface water, respectively. Importantly, the M3 membrane achieved high removal efficiencies for heavy metals, rejecting 91% of Cd2+, 93% of Pb2+, and 88% of Cu2+. These results highlight how the synergy between PES/SPSf and NiFe-LDH can overcome the common challenges of fouling and low metal ion rejection, offering a promising route toward practical and sustainable water treatment solutions.

Figures

Full text

1. Introduction

In recent years, escalating concerns over water scarcity and contamination have driven demand for advanced water treatment and purification technologies [1,2]. Although traditional methods achieve varying degrees of success, they often fall short in addressing complex pollutants or achieving high purification levels [3,4,5]. Consequently, there is a pressing need for innovative materials and approaches to enhance wastewater treatment [6]. Polyethersulfone (PES) and sulfonated polyethersulfone (SPSf) membranes have emerged as highly promising materials in advanced filtration processes due to their excellent mechanical strength, thermal stability, availability of tunable pendant groups, and chemical resistance. In addition, they exhibit remarkable oxidative stability, good film-forming ability, and resistance to hydrolysis and microbial attack. Their intrinsic rigidity and amorphous nature contribute to superior dimensional stability and pressure tolerance, while their tunable surface properties and compatibility with various modifiers make them suitable for applications ranging from water purification to biomedical separations [7,8,9]. These properties make PES/SPSf membranes suitable for a wide range of water treatment applications, including the removal of suspended solids, organic pollutants, and microbial contaminants [10,11,12]. However, while these membranes provide effective filtration, they often face challenges such as fouling, reduced flux, and limited capability in treating specific contaminants [13,14].

On the other hand, layered double hydroxides (LDHs) have gained significant attention for their catalytic and adsorption properties [15,16,17]. These materials, also known as hydrotalcite-like compounds, are an emerging class of materials with remarkable properties that make them highly effective for water treatment applications [18,19]. LDHs consist of positively charged brucite-like layers and charge-balancing anions intercalated between these layers. The general formula for LDHs is [M1^2+−x^Mx^3+^(OH)2]x[An−]x/n·mH_2_O, where M^2+^ and M^3+^ are metal cations, and An− represents the interlayer anions [20,21,22,23]. The unique structural features of LDH include high surface area, tunable interlayer spacing, and rich surface groups that endow them with exceptional catalytic [7,24], adsorptive, and ion-exchange capabilities [25]. When integrated with a membrane, like sulfonated polyethersulfone (SPSf), NiFe-LDHs benefit from the membrane’s enhanced hydrophilicity and high surface area [26]. The combination of PES and SPSf provides a robust support matrix that enhances the dispersion of NiFe-LDHs and promotes intimate interfacial contact between the photocatalyst and the contaminants, thereby improving photocatalytic performance. These properties render the composite membranes promising candidates for advanced water purification applications. The synergistic integration of PES/SPSf membranes with NiFe-LDHs offers a novel strategy to overcome the inherent drawbacks of conventional purification processes, such as low selectivity, high reagent consumption, and the generation of secondary waste.

The treatment of heavy metal-contaminated water is of critical importance due to the persistence, bioaccumulation, and toxicity of these pollutants [27]. Heavy metals such as Cd, Pb, Ni, and Cr are non-biodegradable and therefore tend to accumulate in aquatic organisms, thereby disrupting ecosystems and entering the food chain [28]. Their presence in water sources poses serious health risks to humans, including renal dysfunction, neurological disorders, and carcinogenic effects [29,30]. Membrane-based technologies have emerged as efficient and sustainable alternatives to conventional methods, such as chemical precipitation, ion exchange, and adsorption. Unlike these approaches, membrane systems offer high removal efficiency, tunable selectivity, and minimal chemical or energy demand. Furthermore, their ability to integrate with photocatalytic or adsorption mechanisms enables the simultaneous separation and degradation of heavy metals and organic pollutants [31,32,33]. Several studies have demonstrated the superior potential of polymer-based and composite membranes, particularly those incorporating sulfone polymers and layered double hydroxides, for the effective and selective removal of heavy metals from aqueous media [34,35]. This combination improves the dispersion of NiFe-LDHs and ensures better contact between the photocatalyst and the contaminants, further boosting photocatalytic efficiency. These characteristics make them an attractive option for enhancing water treatment technologies. The synergistic integration of PES/SPSf membranes with NiFe-LDHs represents a novel approach to addressing the limitations of conventional water purification methods.

This approach aims to develop a composite system that not only improves the efficiency of water purification but addresses challenges such as membrane fouling and selective contaminant removal. This study explores the potential of PES/SPSf membranes incorporated with LDH for advanced water purification. The research focuses on evaluating the performance of the composite membrane system in terms of filtration efficiency and overall water treatment effectiveness. The objectives are to investigate the advantages of this integrated approach, assess the impact on membrane performance and stability, and provide insights into its practical applications in addressing water quality challenges. By bridging the gap between membrane technology and catalytic materials, this work aims to contribute to the development of more effective and sustainable water treatment solutions. Through a comprehensive analysis of the synergistic effects of PES/SPSf membranes and NiFe-LDH fillers on solvent transport and solute rejection, the work investigates the feasibility of mixed matrix membranes in wastewater treatment.

2. Materials and Methods

2.1. Materials

Unless stated otherwise, all chemicals used are analytical reagent grade, and ultrapure water was used throughout the experiments. Urea (ACE, Durban North, South Africa), FeCl_3_ (97%), Ni(NO_3_)2·6H_2_O (CAS No.: 13520-61-1), ethanol (99.5%), polyvinyl pyrrolidone K10 (PVP) (CAS No.: 9003-39-8), and N,N-dimethylacetamide (DMac) (99.5%) were supplied by Sigma Aldrich (Johannesburg, South Africa). Ammonia (25%) was obtained from Merck (Johannesburg, South Africa) and was used without further modification. Deionized water (DI) was obtained from a Millipore Milli-Q system (Merck, Darmstadt, Germany). Polyethersulfone (PES) powder of average molecular weight 62,000 g/mol^−1^ was purchased from Solvay Advanced Polymer (Oudenaarde, Belgium), while sulfonated polysulfone (SPSf) with 25% sulfonation was purchased from Shandong Jinlan Special Polymer Co., Ltd. (Linyi, China).

2.2. NiFe-Layered Double Hydroxide Nanomaterial (LDH) Synthesis

With a few modest adjustments, NiFe-LDH was produced using the hydrothermal technique described in [36]. In 80 mL of DI water, 1.33 g of Fe(NO_3_)2·9H_2_O and 3.90 g of Ni(NO_3_)2·6H_2_O were dissolved, and the mixture was agitated. After 30 min of stirring, 1.94 g urea was added, and the mixture was transferred into a Teflon-lined autoclave and heated in an oven at 150 °C for 21 h. The autoclave was then allowed to cool to ambient temperature, and the product was cleaned with water using a centrifuge. The product was then pulverized into a fine powder after being dried for 12 h at 100 °C in an oven.

2.3. Fabrication of the PES/SPSf Membranes

Fabrication of the polymeric membranes was carried out using a non-solvent induced phase separation (NIPS) procedure, adapted from [8], where the base polymers (PES and SPSf) were dissolved in DMac such that the base polymer consisted of 26 wt.% of the dope solution. The pore former, PVP 10 KDa, DI water, and the LDH wt.% were calculated according to the plus method. The wt.% of PVP was kept at 8 wt.%, that of DI at 5 wt.%, while that of the LDH was varied between 0, 0.1, 0.4, and 2 wt.%. The resulting membranes were named M0, M1, M2, and M3, respectively (Table 1). The LDH was dispersed in the DI water via ultrasonication and subsequently added to each dope solution after 6 h of stirring at ambient temperature. Each dope solution was stirred for a further 6 h at ambient temperature to enable dispersion of the LDH and homogenization. The dope solutions were degassed for 12 h, and flat-sheet membranes were fabricated. A film of each dope solution was cast onto a smooth quartz glass plate using an Elcometer 4340 Motorised/Automatic Film Applicator (Manchester, United Kingdom) with the casting knife set at 200 µm, a casting speed of 3 m·min^−1^, and the stage temperature set at 25 °C. The freshly coated glass plate was immediately immersed in a coagulation bath consisting of DI at 25 °C and, after complete precipitation, was stored in DI for a minimum of 48 h.

2.4. The Layered Double Hydroxide (LDH) Nanofiller and MMMs Characterizations

The surface functional groups for the LDH and the membranes were determined via Fourier transform spectroscopic analysis on the Frontier Perkin Elmer L280104 Perkin Elmer SP8000 model (Shelton, CT, USA). Measurements were carried out in the diamond/ZnSe attenuated total reflectance (ATR) analysis mode. Scanning electron microscopy (SEM) imaging was carried out on a JEOL SEM IT 300 (Boston, MA, USA). Each sample was mounted onto metallic stubs using carbon tape and sputter-coated with gold nanoparticles using a Quorum Q150T series coater (East Sussex, United Kingdom). The water contact angle was measured at room temperature using the DataPhysics Optical Contact Angle (OCA) 15 EC (G10, KRUSS, Hamburg, Germany) equipped with video capture for surface energy evaluation. Trace metal determination was performed using an Agilent 7850 inductively coupled plasma mass spectrometer (ICP-MS). The instrument was operated under standard plasma conditions with an RF power of 1550 W to ensure robust plasma generation. Argon was supplied as plasma gas at a flow rate of 15 L/min^−1^, while the auxiliary gas was maintained at 0.9 L/min^−1^ to stabilize plasma shape. The nebulizer gas flow was optimized at 0.99 L/min^−1^, and the peristaltic pump speed was set to 0.1 rps to deliver samples steadily into the plasma.

The zeta potential of the membrane’s active layer was measured using a SurPASS Electrokinetic Analyzer (Anton Paar GmbH, Graz, Austria) at a background electrolyte of 10 mM KCl. Zeta potential determination was carried out in the pH range from 3.0 to 12.0. The measured streaming potential data were converted to the zeta potential with the Helmholtz–Smoluchowski Equation (1), using the instrument’s software, as follows:

$$[eqn]$$

where η is the viscosity, ΔV is the electro-osmotic velocity, δ is the electrolyte conductivity, ΔP is the applied pressure, and ε is the permittivity of water.

2.5. Membrane Filtration Experiments

The wet and dry technique was used to measure the membranes’ porosity and water absorption. A membrane support ring (A = 0.00126 m^2^) was used to cut the membranes, and they were subsequently submerged in DI water for 24 h before being weighed (M_wet_). To determine M_dry_, the same membranes were allowed to air dry before being weighed. Equations (2) and (3) were utilized to determine the water absorption capacities and porosity based on the M_wet_ and M_dry_ values as follows:

$$[eqn]$$ $$[eqn]$$

where W_wet_ is the membrane’s wet weight, W_dry_ is the membrane’s dry weight, ρ is the density of DI water (0.998 g·mL^−1^), and δ is the thickness of the membrane measured from cross-sectional SEM images using ImageJ software 1.5e.

The average pore size of the membranes was calculated using the Guerout–Elford–Ferry Equation (4), as follows:

$$[eqn]$$

where ε is the membrane porosity, η is the water viscosity ( $[eqn]$ 10^−4^ Pa s), r_m_ is the mean pore radius (nm), Q is the volume of permeated water per unit time (m^3^/s), A is the effective membrane area (m^2^), and $[eqn]$ is the operating pressure (MPa).

A Sterlitech dead-end filtration cell (Seatle, WA, USA) was used to evaluate the membranes’ water filtration and antifouling assessment capabilities. Before the assessments, the membranes were pre-compacted at a higher trans-membrane pressure (200 kPa) than the one utilized for evaluation (150 kPa). Equations (5) and (6) were used to determine the membranes’ flux and permeance as follows:

$$[eqn]$$ $$[eqn]$$

where P is the trans-membrane pressure (bar), t is the operating time (h), and V is the volume (L).

BSA and metal-contaminated wastewater were used to evaluate the membranes’ rejection capacity. A total of 1 g of BSA was dissolved in 1000 mL of DI water to make the solution. The UV-Vis. was used to quantify the concentration of BSA in the feed and permeate. Equation (7) was used to determine their rejection as follows:

$$[eqn]$$

BSA solution and metal-contaminated wastewater were also used to evaluate the membranes’ antifouling capabilities. The tests were conducted at 150 kPa after the membranes were pre-compacted for 30 min at 200 kPa. The flux recovery ratio (FRR), total fouling ratio (R_t_), reversible fouling ratio (R_r_), and irreversible fouling ratio (R_ir_) were computed using Equations (8)–(11) to quantify the antifouling behavior, as follows:

$$[eqn]$$ $$[eqn]$$ $$[eqn]$$ $$[eqn]$$

Pure water was passed through the membrane for 30 min to acquire J_w_1, and readings were collected every 5 min. After that, the BSA solution or metal-contaminated wastewater was passed for a further 30 min, and readings were taken every 5 min to obtain Jp. The membranes were then cleaned by backwashing with deionized water for 10 min to remove foulants from the surface and pores. This was achieved by first reversing the membrane, topping the cell with DI water, and passing it through the membrane for 10 min. Thereafter, it was re-inserted into the cell in the correct orientation, topping the cell with the water of interest, and obtaining the measurements again. Starting with pure water to obtain J_w_2, pure water was filtered through the membrane for 30 min, and the permeate volume was recorded at 5-min measurement intervals. For reusability, pure water was substituted with the water of interest, BSA or metal-containing water.

2.6. Heavy Metal Wastewater Sampling

The heavy metal contaminated water was collected from Florida Lake, Roodepoort, South Africa, as indicated in Figure 1. Glass bottles were used for water collection and transported to the lab in a cooler with ice packs. This was performed to prevent any physicochemical changes in the water. The pH, turbidity, electrical conductivity (EC), and total dissolved solids (TDS) were measured before and after treatment using a multimeter.

3. Results and Discussion

3.1. Characteristics of LDH NMs Nanofiller and PES/SPSf/LDH MMMs

3.1.1. FTIR

The PES/SPSf-LDH membrane surface was examined for functional groups, and FTIR analysis was performed to determine whether SPSf and NiFe-LDH had been successfully blended into the PES membrane. Figure 2 illustrates that all spectra had characteristic peaks at 3095 cm^−1^ for the –C–H stretching vibration, 2359 cm^−1^ for the –S–stretch, 1236 and 1103 cm^−1^ for the –SO_3_H of the SPSf, and 1147 cm^−1^ for the stretching of –SO_2_. Moreover, all FTIR spectra of PES/SPSF membranes showed a –S– functional group peak in the same region (2359 cm^−1^), which was not present from the spectrum of the pristine PES membrane [38,39]. These ranges correspond to the asymmetric and symmetric stretching of the –SO_3_H and O=S=O groups of the SPSf polymer and are 1235–1250 cm^−1^, 1000–1072 cm^−1^, and 1119–1150 cm^−1^. This confirmed the successful incorporation of SPSf and the interaction between SPSf and the PES sheet membranes. However, the FTIR spectra of the NiFe-LDH/PES/SPSf composite showed some novel peaks compared to the PES/SPSf membrane. The antisymmetric stretching vibration of the intercalating $[eqn]$ anion from the breakdown of urea during hydrothermal synthesis is responsible for the band at 1383 cm^−1^ [40]. The peak at 3000 cm^−1^ in the NiFe-LDH/PES/SPSf composite membrane spectra was caused by either the covering of hydrogen-bonded surface OH groups or the stretching vibration mode of the OH group. Overall, the FTIR verified the presence of PES/SPSf and the successful coating of NiFe-LDH.

The NiFe-LDH nanofiller was characterized in our previous work [41].

3.1.2. SEM

Figure 3 presents the membrane surface morphology (top surface, bottom surface, and cross-sectional images) analyzed using SEM. The top surface images (M0–M3) present a clean surface with no visible pores. This is attributed to the formation of a selective layer driven by structure polymer–polymer interactions, rapid surface gelation during phase inversion process. The bottom surface (M0’–M3’) presents clearly visible pores fairly distributed across the surfaces. The pores on the bottom surfaces appear because the membranes undergo delayed demixing, which allows polymer-lean phases to nucleate, thereby growing into visible voids. The M0 membranes depict bigger pores than the modified membranes and this can be attributed to the fact that the NiFe-LDH nanoparticles increase the viscosity of the solutions, with a slower solvent–nonsolvent exchange rate, resulting in delayed but controlled demixing, which causes macrovoids. A further decrease in the number of pores was observed with increasing nanofiller content.

The cross-sectional morphology of the membranes (M0*–M3*) is presented in Figure 3c. This shows that the membranes exhibited a spongy morphology in the top layer and a relatively smaller underlying substructure with the macrovoids. This feature has been observed by previous researchers who have reported on the PES/SPSf^2^ membranes of a similar PES/SPSf wt.% composition [7,37,42,43]. The insets indicate that the spongy matrix consists of an increasingly porous network as the loading of the nanofibers was increased (M2 and M3). These nanoporous networks are polymeric gel sub-units formed through pre-crystallization, which is induced by the introduction of water as a non-solvent, which is consistent with other PES/SPSf MMMs [43]. These are attributed to the increased loading of hydrophilic nanoparticles, such as LDH nanocomposites. Additionally, the increasing spongy matrix is credited to the relatively equal outflow of the solvent and inflow of the non-solvent at the polymer crystallization stage. This equilibrium is facilitated by strong interactions among PES, SPSf, and LDH within the polymer matrix, which influence the dope solutions structural stability during phase inversion, thereby preventing the formation of the finger-like and macrovoid sublayers observed by Ly et al. (2020) in the PES/SPSf/LDH mixed matrix membrane [43].

3.1.3. WCA

The membranes’ hydrophobicity or hydrophilicity was tested using the sessile drop method. The membrane contact angles obtained are presented in Figure 4. The M0 membrane exhibited the highest contact angle of 73°, which is moderately hydrophilic. The PES polymer is known to be hydrophobic [8], and SPSf polymer introduces -SO_3_H groups, which are hydrophilic, resulting in moderate wettability [31,44]. The wettability for the M1–M3 membranes further decreased from 61° to 45°, and this enhancement was attributed to the introduction of abundant hydroxyl (-OH) groups from the NiFe-LDH surface and edges of the nanosheets [45,46]. Additionally, the improved hydrophilicity was associated with the increased content of the interlayer spacing of the NiFe-LDH. As the nanofiller content was increased, the water affinity improved.

3.1.4. Zeta Potential of the Membranes

The zeta potential of the PES/SPSf and PES/SPSf/NiFe-LDH membranes (Figure 5) was measured across a wide pH range to study the effect on membrane performance. The PES/SPSf membrane showed negative zeta potential values across the entire pH range, becoming increasingly negative as pH increased. This behavior was attributed to the deprotonation of sulfonic and ether groups in the polymer chains. The negative surface charge promotes electrostatic repulsion toward anionic species (including negatively charged heavy metals), thereby improving solute rejection efficiency. Incorporation of NiFe-LDH into the PES/SPSf membrane matrix significantly altered the surface charge properties of the modified membranes, as evidenced by a progressive increase in zeta potential (i.e., less negative or even positive values) with increasing LDH loading. At a low pH, the zeta potential of the modified membranes shifted from negative to positive. This shift indicates successful surface modification and charge compensation by the positively charged hydroxide layers of the LDH. At a moderate FeNi-LDH nanofiller content, partial neutralization improved membrane hydrophilicity, which is good at facilitating reduced fouling propensity, since the more hydrophilic LDH domains attract hydration layers that hinder the adsorption of pollutants. By contrast, at a higher filler content, the membrane surface became increasingly positive, especially at neutral-to-basic conditions. This positive charge enhances electrostatic attraction toward anionic contaminants, potentially increasing fouling and reducing rejection at neutral-to-basic pH conditions.

3.2. Membrane Performance Assessment

3.2.1. Pure Water Flux, Water Uptake, and Porosity

The performance of the PES/SPSf and NiFe-LDH/PES/SPSf membranes were evaluated in terms of key parameters, such as pure water flux (PWF), water uptake, pore size, thickness, and porosity, which are crucial for assessing the suitability of these membranes in various filtration applications (Figure 6). The porosity and water uptake of the membranes were evaluated because they are critical properties of polymeric membranes. These interdependent properties significantly influence water permeance and separation processes. The pore size and porosity also play important roles in determining water uptake capabilities. Generally, bigger pore sizes correspond to higher porosity, as they provide a greater void volume within the membrane matrix. The porosity and water uptake results are presented in Figure 6b,c. The pristine M0 (PES/SPSf) membrane exhibited a porosity and water uptake of 40.7% and 19.7%, respectively. It is well known that PES is inherently hydrophobic, which limits its water adsorption capability [47]; however, the incorporation of SPSf, which contains hydrophilic sulfonic acid groups [48], enhanced the water adsorption of the blended membrane. The water uptake and porosity of the modified membranes (M1–M3) improved from 19.7% to 27.9% and from 40.7% to 78.8%, respectively. This improvement can be attributed to the favorable characteristics of the NiFe-LDH nanofiller. NiFe-LDH is highly hydrophilic due to its abundant polar hydroxyl groups and interlayer anions that facilitate hydration, as well as positively charged LDH layers that attract water molecules [49]. Additionally, a synergistic interaction between the LDH layers and the sulfonic acid groups (-SO_3_H) of SPSf forms strong hydrophilic domains. This ultimately leads to the greater attraction and retention of water molecules, a higher degree of swelling, and an improved overall hydration capability.

Figure 6b present the average pore radii of the fabricated membranes. The pristine M0 membrane exhibited a relatively larger average pore radius of 82.4 nm, whereas the modified membranes showed smaller pore sizes decreasing from 70.2 nm for M1 to 67.1 nm for M3. This trend can be attributed to several factors related to the interactions between the LDH nanofillers, the casting solution, and the phase inversion process. Incorporating LDH nanofillers into the casting solution generally increases its viscosity, which has been well established in membrane science [50]. A more viscous solution slows down the solvent–nonsolvent exchange during the phase inversion process, and a slower demixing rate typically leads to the formation of smaller pores. Another plausible explanation is that LDH nanoparticles can restrict polymer chain mobility, promoting the formation of a more interconnected polymer–nanoparticle network. As the nanofiller content increases in the casting solution, a greater number of nanoparticles occupy the microvoids, thereby contributing to pore shrinkage.

The thickness of the membranes was measured using ImageJ software on the cross-sectional images of SEM. The M0 membrane exhibited a thickness of 111.67 μm, which represented the baseline structural dimension of the unmodified polymer matrix. Upon incorporating the nanofiller into the membrane casting solution, slight variations were noted across the modified membranes (M1–M3). The membrane with the lowest nanofiller loading (M1) exhibited a thickness of 117.67 μm. The slight increase was associated with the incorporation of a small amount of nanofiller, which can potentially increase the casting solution viscosity, which then slows down the demixing rate during phase inversion [51]. Slower demixing usually leads to the formation of a slightly thicker membrane. By contrast, the membrane containing medium nanofiller loading (M2) had a thickness of 112.33 μm. This suggests that, at this intermediate loading, the nanofiller influence on the casting solution and precipitation kinetics was less pronounced. The nanofiller content was sufficient to modify the membrane’s surface and structural characteristics but not high enough to significantly influence its kinetics. At the highest loading (M3), the measured thickness was 113.67 μm, which is slightly higher than M0 and M2 membranes. In the literature, it is reported that a higher nanofiller content generally increases solution viscosity and can limit solvent–nonsolvent exchange, which then results in marginally thicker membranes [51]. However, because the observed increase is small, it indicates that, even at a higher loading, the polymer matrix retained stable phase inversion behavior. Thus, the modified membrane thickness remained within a narrow range, demonstrating that NiFe-LDH incorporation has minimal influence on the final membrane thickness.

Figure 6a demonstrates the water fluxes of the pristine PES/SPSf and NiFe-LDH modified membranes. The PES/SPSf membrane (M0) demonstrated moderate flux rates, with flux values increasing in the range 38.8–143.7 L/m^2^·h^1^ withing the pressure range 100–300 kPa. As expected, the water flux of the mixed matrix membranes increased with applied pressure in the range 44.8–218 L/m^−2^·h^−1^ between M1 and M3, with the pure water permeance increasing from 52.1 to 90.5 L/m^−2^·h^−1^Bar^−1^ under the same pressure conditions, demonstrating a higher increase with pressure for the mixed matrix membranes. The lower increase with M0 can be attributed to the membrane’s semi-hydrophilic nature and the relatively dense PES matrix, which somewhat restricts water flow [52,53]. The PES/SPSf blend thus forms a semi-permeable membrane well-suited to microfiltration (MF) processes requiring moderate filtration rates. Incorporating SPSf into the PES matrix improved the membrane’s hydrophilicity (Figure 4 for CA), which contributed to enhanced water permeability. Their enhanced permeation arises from the unique layered architecture of NiFe-LDH, combined with anion incorporation, which provides effective water channels for greater water throughput. This reflects the porosity of the internal structure. This improvement stems from the increased surface area and adjustable interlayer spacing of the LDH structure, which allows for easier water passage through the membrane [54,55]. LDH modifications, such as exfoliation, yield a larger effective open framework, further boosting permeability. Ionic interactions within the LDH layers and the adjustable interlayer spacing further amplified permeability. These findings align with the existing literature [7,56], where optimized LDH interlayer spacing has been shown to promote high flux. The hydrophilic characteristics of NiFe-LDH/PES/SPSf membranes also contributed to enhanced water flux, as these polar functional groups have high affinity for water molecules.

3.2.2. BSA Rejection and Flux

BSA flux and rejection tests (Figure 7) were performed to assess the separation mechanism of the fabricated membranes. For the PES/SPSf membrane (M0), BSA rejection was primarily governed by size exclusion, moderate hydrophilicity (arising from sulfonation), and electrostatic repulsion. BSA has a hydrodynamic diameter of approximately 7 nm [57], whereas the membranes exhibited relatively larger pore sizes, decreasing from 82.3 nm for M0 to 67.1 nm for M3 (all within the UF range, Figure 6b). Despite this size discrepancy, the M0 membrane achieved 81.3% BSA rejection, which can be attributed to the effective physical sieving of BSA molecules. Additionally, the dense sponge-like structure of the PES/SPSf membrane suppressed macrovoid formation and promoted a more uniform pore distribution. The incorporation of SPSF introduced sulfonic acid groups (-SO_3_H) into the membrane matrix, imparting a negative surface charge (as seen in Figure 5). At a neutral pH, BSA is also negatively charged, resulting in electrostatic repulsion between the protein and the membrane surface. The moderate hydrophilicity of the membrane further contributed to improved rejection. For the NiFe-LDH modified membranes, BSA rejection ranged between 85.4% and 92.4%, and was governed by a synergistic combination of size exclusion, electrostatic repulsion, and hydrophilicity. These modified membranes displayed a narrower pore-size distribution due to better control phase separation during fabrication. The layered morphology of NiFe-LDH also acted as an additional barrier, helping to prevent pore blockage and foulant deposition while providing further physical sieving of BSA rejection. NiFe-LDH nanosheets are typically positively charged under neutral-to-mildly-basic pH conditions (Figure 5) because of surface -OH groups and protonated Ni^2+^/Fe^2+^ cations, leading to a net positive surface charge. Consequently, the electrostatic repulsion mechanism in the modified membranes were complex, as the PES/SPSf matrix remained negatively charged, while the nanofiller was positively charged. Despite these opposing charges on the nanofiller, the results show that negative charges from SPSf dominated the protein rejection. Increasing the NiFe-LDH nanofiller content improved the rejection further, primarily due to enhanced hydrophilicity and smoother membrane surfaces. For BSA permeation flux, the modified membranes exhibited higher values than the control membrane. This improvement is attributed to the abundant hydroxyl groups on NiFe-LDH, which significantly enhanced water permeability. The combination of strong a negative charge from SPSF and the hydrophilic nature of LDH also reduced BSA adsorption on the membrane surface, thereby mitigating fouling.

3.2.3. Fouling Resistance and Reversibility Using BSA Solution and Surface Water

The membranes were also evaluated for fouling resistance performance and reusability using both a BSA solution and real surface water samples. These tests aimed to evaluate the effect of incorporating a positively charged and hydrophilic NiFe-LDH nanofiller into the PES/SPSF matrix on fouling behavior, flux decline, and flux recovery after cleaning. The flux recovery ratio (FRR), total fouling ratio (R_t_), reversible fouling ratio (R_r_), and irreversible fouling ratio (R_ir_) were calculated using Equations (7)–(10), based on initial pure water flux (J_w_1), permeate flux during BSA or surface water filtration (J_p_), and pure water flux after cleaning (J_w_2).

BSA is predominantly hydrophobic, and surface water typically contains partially hydrophobic natural organic matter (NOM) [58,59]. The control M0 membrane exhibited a total fouling ratio and irreversible fouling ratio of 76.3% and 59.6% when filtering the BSA solution, and 50% and 35.9% when filtering surface water, respectively (Figure 8a,b), consistent with the inherent hydrophobic characteristics of the PES/SPSf membrane. Fouling resistance using these membranes was achieved through enhanced hydrophilicity by the introduction of sulfonic acid groups and a negatively charged surface. The resulting hydration layer and electrostatic repulsion delayed foulant deposition and adhesion, yielding FRR of 40.4% for the BSA solution and 64.0% (Figure 8c,d) for the surface water. The NiFe-LDH modified membranes exhibited an improved fouling resistance performance, with further enhancement as nanofiller loading was increased. This improvement stemmed from the synergistic contributions of sulfonic acid groups from SPSf and abundant hydroxyl groups from NiFe-LDH, which dramatically improved hydrophilicity. For BSA filtration, R_t_ decreased progressively from 51.1% (M1) to 47.8% (M2) and 46.0% (M3), respectively. For surface water, Rt values were 42.1%, 37.5%, and 35.6% for the M1, M2, and M3 membranes, respectively. The high reversible fouling in the modified membranes indicated that most foulants could be easily dislodged from the membrane surface and pores by simple water backwashing, confirming excellent reversibility and suitability for long-term operation. Additionally, in Figure 8a–d, the modified membranes exhibited considerable higher FRR, indicating lower R_ir_, good surface characteristics, as well as consistent flux over time.

As shown in Figure 8c,d, all the membranes experience flux decline during filtration of the BSA solution or surface water (J_p_), indicating an accumulation of foulants on the membrane surface and/or pores over time. After hydraulic cleaning of the membranes, permeate flux (J_w_2) was partially recovered, with the degree of recovery strongly dependent on the surface chemistry. Membranes with superior characteristics consistently exhibited better fouling resistance performance. Notably, the BSA solution caused more severe and irreversible fouling than the surface water, likely due to its higher concentration, stickiness, and stronger hydrophobic/hydrogen-bonding interactions with the membrane surface, even on moderately hydrophilic SPSf-based materials. By contrast, the surface water contained lower concentrations of diverse organic and inorganic foulants, leading to milder and more reversible fouling.

Reusability studies of the membranes were further assessed through three consecutive filtration–cleaning cycles using both the BSA solution and surface water, as shown in Figure 8e,f. The study was performed to evaluate the long-term stability, fouling resistance, cleanability, and the overall environmental benefits of these membranes. The flux was expressed in percentages (normalized flux) for a clearer comparison of the fouling and cleaning behavior of M0–M3, and enables an accurate calculation of the total, reversible, and irreversible fouling ratios. The pristine M0 membrane (Figure 8e,f) showed the highest fouling tendency among all membranes for both feed solutions. Its FRR after three cycles was only ~25.3% for the BSA solution and 48.6% for the surface water after three cycles, indicating limited hydrophilicity and considerable irreversible foulant adhesion (Figure 8e,f). The hydration layer assumed to be formed by hydrophilic SPSf moieties is believed to delay foulant adhesion on the membranes. Surface water foulants were generally easier to reverse than BSA foulants, as evident from the higher FRR values in Figure 8f compared to Figure 8e. Incorporation of NiFe-LDH significantly enhanced the reusability of the modified membranes. For the BSA solution, the M1 and M2 membranes achieved FRR values of approximately 55.3% and 51.7%, respectively, during the first fouling cycle, indicating better reversibility and reduced irreversible fouling. The optimized M3 membrane exhibited the best performance, maintaining an FRR above 65.9% after three cycles with the lowest irreversible fouling ratio (R_ir_ < 10%). For surface water, the M1, M2, and M3 membranes achieved FRR values of 60.2%, 65.1%, and 71.2%, respectively, after the third cycle. These improvements are attributed to the increased surface hydrophilicity and negatively charged surface imparted by NiFe-LDH, which effectively repel protein molecules and minimize pore blockage. Thus, NiFe-LDH modification not only enhances permeability and rejection performance but prolongs operational stability by mitigating fouling.

It is worth emphasizing the environmental advantages of membrane-based wastewater treatment systems, which include high retention of pollutants, water reuse potential, and low usage of chemicals compared to conventional treatment processes. Among the tested membranes, the NiFe-LDH modified membranes offer better environmental benefits because their enhanced antifouling properties and cleanability allow for extended operation with milder cleaning protocols and less frequent replacement, thereby reducing operational costs and chemical usage. Nevertheless, the non-biodegradability of the polymeric base materials remains a challenge that requires further attention in future research.

3.2.4. Heavy Metal Removal Performance

The removal of heavy metals ions (Cd^2+^, Pb^2+^, and Cu^2+^) was investigated to evaluate the practical applicability of the PES/SPSf/NiFe-LDH membranes in real-world water purification scenarios, as presented in Figure 9. The pristine M0 membrane exhibited the highest rejection of Pb^2+^ (52%) and the lowest rejection for Cu^2+^ (39%), which follows the decreasing ionic radii of the ions (Cd^2+^ – 1.12 Å, Pb^2+^ – 0.97 Å, and Cu^2+^ – 0.73 Å), thereby suggesting rejection according to ionic size. This rejection of metal ions by ultrafiltration membranes despite their size being small is consistent with the observations by other researchers where up to 50% Cu^2+^ was rejected by hollow fiber PES membranes [60]. It is also noted that adsorption plays a critical role in metal rejection in mixed matrix membranes that incorporate high surface fillers, such as LDHs, as alluded herein below. Although divalent cations are positively charged, and thus electrostatically attracted to the negatively charged surface, a portion of the ions was still rejected, possibly due to steric hindrance and the hydration layer effect. Additionally, adsorbed metal ions possibly bonded to sulfonate and other functional groups on the membrane surface, contributing to removal through adsorption and/or complexation mechanisms. By contrast, incorporation of NiFe-LDH significantly enhanced heavy metal removal performance. The M3 membrane achieved rejection exceeding 90% for Cd^2+^ and Pb^2+^, and above 85% for Cu^2+^. This substantial improvement can be attributed to the synergistic effect of the modified membrane matrix: (i) increased surface charge and improved hydrophilicity, which reduced free ion passage; (ii) abundant active adsorption and ion exchange sites provided by the NiFe-LDH layers; and (iii) narrower and more uniformly distributed pores, which strengthened size exclusion effects. These findings clearly demonstrate that NiFe-LDH plays a dual role, simultaneously enhancing both the separation efficiency and the pollutant adsorption capacity of the membranes.

3.2.5. Characteristics of Surface Water

Surface water usually contains NOMs, turbidity, dissolved ionic species, and microorganisms, which tend to reduce its quality. In this work, UF membranes were used to provide physical and physicochemical separation to improve the effluent. Based on Table 2, the fabricated PES/SPSf and NiFe-LDH/PES/SPSf membranes enhanced all the measured water quality parameters, including pH, total dissolved solids (TDS), electrical conductivity (EC), turbidity, and dissolved oxygen (DO), to meet the South Africa National Standards for drinking water (SANS:241). The M0 membrane improved turbidity to 1.30 NTU which can be attributed to its well-defined pore structure, and the introduction of NiFe-LDH layers further improved it. These layers imparted high hydrophilicity which reduced pore blockage and enabled more uniform filtration. The positively charged LDH layers enhanced the capture of negatively charged NOMs and colloids through electrostatic attraction. Additionally, its adsorption sites helped trap fine suspended particles that would normally pass through the unmodified UF membranes. Hence, the TDS significantly improved for the modified membranes. These layers also captured colloidal and organic acids that contribute to conductivity, hence the improvement. The LDH has strong affinity for organic species. The DO of the feed is 2.26 mg·L^−1^, and it was improved to 4.30 mg·L^−1^ for the M0 membrane and to 5.24 to 7.24 mg·L^−1^ for the M1 to M3 membranes. Generally, membranes do not produce oxygen, instead they can enhance it through the removal of substances such as biodegradable organic matter, microorganisms, and suspended solids. The pH of the feed may have improved through the removal of acidic organic matter, leading to pH shifting, and this improved neutrality.

4. Conclusions

This study demonstrated the successful development of PES/SPSf/NiFe-LDH composite membranes with enhanced performance for water purification. The incorporation of NiFe-LDH improved membrane hydrophilicity, porosity, and surface charge, which translated into higher pure water flux (up to ~218 L/m^−2^ h^−1^ for M3), superior BSA rejection (92.4%), and excellent antifouling behavior with a flux recovery of 65.9% for the BSA solution and 71.2% for the surface water after multiple cycles. Importantly, the membranes also showed strong heavy metal removal capability, with the M3 membrane achieving rejection efficiencies of 91% for Cd^2+^, 93% for Pb^2+^, and 88% for Cu^2+^, underscoring their potential for treating both organic and inorganic contaminants. This indicated the combined contributions of adsorption, increased hydration layer with increasing surface hydrophilicity, and surface charge. The synergistic combination of PES/SPSf and NiFe-LDH thus addresses the common limitations of polymeric membranes, particularly low rejection and fouling. While future work should explore long-term stability, scalability, and performance with complex real wastewater matrices, the present findings highlight PES/SPSf/NiFe-LDH membranes as promising candidates for sustainable water purification in applications ranging from drinking water production to wastewater remediation and industrial effluent treatment.