The goal of this study is to develop a multistage membrane filtration cascade using reverse osmosis (RO), nanofiltration (NF), and ultrafiltration (UF) to produce skim milk concentrate. Using the cascade or evaporation, skim milk concentrates with a dry matter of 40 %, 45 %, and 50 % were produced before being spray dried to produce skim milk powder. Ceramic rotating disc membranes were used to process the highly viscous skim milk concentrate in the UF stage of the cascade. By recycling the permeates of the NF and UF stages of the cascade, the composition of the final UF retentate was close to that of regular skim milk concentrate. The flux of the membrane cascade ranged from 0.9–13 kg m -2 h -1 . The concentrates showed some differences in composition and viscosity, resulting in only minor variations in the functional properties of the skim milk powders. Membrane cascades are an emerging technology for concentrating skim milk. Compared to evaporators, membrane cascades mostly require electrical energy. This means that their use could reduce the environmental impact of the dairy sector. • A membrane cascade comprising reverse osmosis (RO), nanofiltration (NF) and ultrafiltration (UF) was developed to concentrate skim milk up to a dry matter of over 48 %. • Rotating ceramic membrane discs were essential to reach this dry matter. • Skim milk powders produced by membrane cascade had good functional properties. • Recycling of the NF and UF permeates produced an UF retentate composition comparable to skim milk.

A novel high-flux polyethersulfone (PES)-based nanofiltration (NF) membrane was developed by embedding amino-functionalized N-doped hollow porous carbon spheres (N-HPCS/NH₂) for efficient treatment of beverage-industrial wastewater. The N-HPCS/NH₂ nanofiller was synthesized via a hard-templating method and surface-functionalized using 3-aminopropyltriethoxysilane (APTES). The modified nanofiller exhibited a porous hollow structure with a high specific surface area (630.24 m2/g) and abundant amino groups, contributing to improved hydrophilicity and pollutant adsorption. The incorporation of low concentrations (0.1 wt%) of N-HPCS/NH₂ into the PES matrix resulted in membranes (M1) with superior structural, mechanical, and surface charge characteristics. Compared to the bare membrane, M1 exhibited a ~ 4.5-fold increase in pure water flux (230 L/m2·h), enhanced anti-fouling properties (FRR > 90%), and significantly higher removal efficiencies for salts (up to 84.6% for Na₂SO₄), dyes (> 99% for RR195, MB, RhB), and heavy metals (up to 99.81% for Pb2⁺). The efficiency of the M1 membrane was also studied in a salt-dye binary filtration system containing Na₂SO₄ at different concentrations (100-5000 ppm) and 200 ppm RR195. It was observed that increasing Na₂SO₄ concentration reduced RR195 removal slightly from 99.82% to 98.8%, suggesting minor interference due to ionic shielding. The membrane performance remained stable under continuous operation for 72 h and showed negligible decline in real beverage-industry wastewater treatment. The synergistic effects of enhanced hydrophilicity, surface charge, and active adsorption sites enabled the M1 membrane to achieve both high permeability and selectivity. These results demonstrate the potential of PES/N-HPCS/NH₂ membranes as robust candidates for sustainable industrial wastewater treatment.

ABSTRACT Nanofiltration (NF) is critically important for removing toxic heavy metals from wastewater, providing a high‐efficiency method that safeguards public health and protects ecosystems from irreversible contamination. The main purpose of this research was to fabricate thin film nanocomposite (TFN) membranes modified with composite titanium dioxide (TiO 2 )/silica (SiO 2 ) nanoparticles for removing heavy metals. FT‐IR and XRD tests confirmed the chemical structure of the fabricated silica‐modified titania (TiO 2 /SiO 2 ). Furthermore, the Brunauer–Emmett–Teller (BET) method revealed that the composite particles possess a high specific surface area (150–250 m 2 /g). The addition of TiO 2 /SiO 2 nanoparticles up to 2 wt% induced key structural changes for the TFN membranes: water contact angle measurements indicated a sharp increase in hydrophilicity (from 68.5° to 40.4°), and AFM analysis confirmed a rise in surface roughness. The combined effect of these properties led to an obviously superior pure water flux in the TFN membranes. The membrane incorporating 1 wt% TiO 2 /SiO 2 nanoparticles (NF‐1) demonstrated exceptional performance, achieving high heavy metal rejection rates (98.84% for Pb 2+ , 91.25% for H 2 AsO 4 − , and 96.01% for Co 2+ ) alongside outstanding antifouling properties, as indicated by a 97.4% flux recovery ratio and a minimal irreversible fouling ratio. Therefore, these properties make the NF‐1 membrane a highly efficient and sustainable option for water treatment applications.

Adsorption of per- and polyfluoroalkyl substances (PFAS) by granular activated carbon (GAC) and ion exchange resins (IX) is negatively impacted by elevated concentrations of effluent organic matter (EfOM) and other background water constituents (e.g., coadsorbing inorganic ions) in complex matrices such as wastewater effluent. Here, we evaluated a hybrid system comprised of nanofiltration (NF) as an initial treatment step to reduce concentrations of PFAS, EfOM, and select inorganic ions, followed by either GAC or IX treatment of membrane permeate. A pilot membrane system utilizing a loose nanofilter (NF270) was operated continuously for >45 days, treating wastewater effluent with PFAS periodically added to the feed to evaluate rejection in high recovery (90%) batch experiments. Experimental results demonstrated >92% rejection of C ≥ 4 perfluoroalkyl acids (PFAAs), >98% rejection of hexafluoropropylene oxide dimer acid (Gen-X), and lower rejection (63-92%) of ultrashort chain PFAS and perfluorobutane sulfonamide (FBSA). Without NF pretreatment of wastewater effluent, rapid small scale column tests (RSSCTs) of adsorbents showed that PFAS maximum contaminant level (MCL) criteria were exceeded within 81 bed volumes (BVs) for GAC and 9,000 BVs for IX. In comparison, GAC treated >5,000 BVs of permeate from the NF experiments before exceeding MCLs, while IX-treated NF permeate never exceeded MCLs for the duration of the experiment (450,000 BVs). The proposed NF-adsorbent treatment train represents a promising strategy for PFAS removal from complex wastewater matrices, preventing point source discharges into the environment.

Size exclusion is commonly believed to be the major mechanism governing the rejection of antibiotics by nanofiltration (NF), while both size exclusion and Coulombic interactions affect the rejection of negatively charged polyfluoroalkyl substances (PFASs). Nevertheless, the potential interactions between polar antibiotics (e.g., zwitterionic fluoroquinolones and nonionized sulfonamides) and the membrane are often overlooked. Herein, this study identifies the key impacts of Coulombic and dipole-surface interactions in the rejection of polar antibiotics by polyamide NF membranes with various pore sizes and charges. Membranes with identical pore sizes yet different charges showed distinguished rejection behavior of such antibiotics. Further analysis reveals a strong relevance between the molecular polarity and the passage of antibiotics. This result indicates the non-negligible impact of Coulombic and dipole-surface attraction for the passage of polar zwitterionic fluoroquinolones and nonionized sulfonamides, respectively. In comparison, the dipole-surface interaction had a negligible impact on the rejection of low polar and charged PFASs compared to the dominant size exclusion and Coulombic interaction. Leveraging on these results, a mechanistic framework integrating with size exclusion, Coulombic interaction, and dipole-surface interaction toward efficient rejection of various contaminants were developed. Such improved fundamental understanding promotes the rational design of NF membranes for highly effective water decontamination.

Ternary network derived from polyphenol-inspired sticky nanoparticle: nanofiltration separation efficiency and end-of-life membrane regeneration potential.

Nanofiltration (NF) is increasingly applied for advanced drinking water treatment, but achieving stable operation at high recovery rates remains challenging for surface waters with high scaling potential. This pilot study investigated the performance and optimization of a three-stage NF270 system (4:2:1 tapered array) for treating coagulated surface water in northern Jiangsu, China, aiming to identify sustainable operating conditions for high-recovery applications. The NF system was operated at recoveries of 80-90% with a feed flux of 20-23 LMH, and the effects of forward flushing frequency, acid dosing location, and concentrate recirculation on fouling behavior were evaluated. The NF270 membrane achieved consistent removal of organic matter (effluent chemical oxygen demand (CODMn) < 0.5 mg/L), hardness (40-60% rejection), and alkalinity (~20% rejection), meeting Jiangsu Province drinking water standards. However, operation at 90% recovery resulted in rapid third-stage fouling, with permeate flow declining by >60% within 2.5 h. Osmotic pressure analysis (local concentrate osmotic pressure: 3.8-4.2 bar; net driving pressure: 0.8-2.2 bar) confirmed physical scaling rather than hydraulic limitation as the dominant mechanism. Stage-wise concentration factor calculations (CF1 = 1.6, CF2 = 2.9, CF3 = 4.4) revealed local Langelier Saturation Index (LSI) values of 1.8-2.2 in the third stage, identifying CaCO3 supersaturation as the primary scaling cause. Reducing recovery to 85% and flux to 20 LMH with 2 h forward flushing extended stable operation. Acid addition effectively mitigated scaling, but dosing location was critical: first-stage addition (pH 8.1 → 7.6) reduced third-stage LSI to 0.7-0.9 and stabilized performance, whereas third-stage addition (pH 8.0 → 7.3) inadvertently promoted Al(OH)3 precipitation from residual coagulant (feed Al: 0.07-0.11 mg/L). Concentrate recirculation (90% ratio) did not alleviate fouling. These findings demonstrate that for aluminum-rich coagulated surface waters, optimizing recovery, flushing frequency, and acid dosing location is essential for sustainable NF operation, and provide engineering guidance for full-scale applications.

ABSTRACT Efficient lithium recovery from alkaline salt lake brines and acidic battery leachates demands nanofiltration (NF) membranes with high pH‐resistance. Poly(quaternary ammonium) (PQA) has arisen as an ideal alternative. However, low monomer reactivity and uncontrolled interfacial monomer diffusion usually lead to thick PQA layer with broad pore size distribution, restricting the separation performance. Herein, a branched tertiary amine monomer, tris(2‐dimethylaminoethyl)amine, which possesses active sites with higher N‐exposure factor and homogeneous spatial distribution, was designed to accelerate the monomer reaction. Simultaneously, sodium dodecyl sulfate (SDS) was introduced to endow homogeneous monomer distribution and retarded monomer diffusion. Accordingly, the resultant PQA‐T membrane exhibits ultralow thickness (∼15 nm) and narrow pore size distribution, exhibiting exceptional water permeance of 30.8 L m −2 h −1 bar −1 and a high MgCl 2 rejection of 99.2 ± 0.4% (S Li/Mg up to 136), overperforming all reported acid/alkali‐resistant membranes. Notably, it shows high separation factors under acidic and alkaline conditions (S Li/Mg = 81.8 at pH 10; S Li/Co = 28.7 at pH 2), showing great potential for lithium extraction under harsh pH conditions, which is further confirmed by two‐stage NF for both real salt lake brine and battery leachates. Additionally, we demonstrate large‐area fabrication of the PQA membrane (∼625 cm 2 ) while maintaining consistent separation performance.

Efficient lithium recovery from alkaline salt lake brines and acidic battery leachates demands nanofiltration (NF) membranes with high pH-resistance. Poly(quaternary ammonium) (PQA) has arisen as an ideal alternative. However, low monomer reactivity and uncontrolled interfacial monomer diffusion usually lead to thick PQA layer with broad pore size distribution, restricting the separation performance. Herein, a branched tertiary amine monomer, tris(2-dimethylaminoethyl)amine, which possesses active sites with higher N-exposure factor and homogeneous spatial distribution, was designed to accelerate the monomer reaction. Simultaneously, sodium dodecyl sulfate (SDS) was introduced to endow homogeneous monomer distribution and retarded monomer diffusion. Accordingly, the resultant PQA-T membrane exhibits ultralow thickness (∼15nm) and narrow pore size distribution, exhibiting exceptional water permeance of 30.8 L m-2 h-1 bar-1 and a high MgCl2 rejection of 99.2 ± 0.4% (SLi/Mg up to 136), overperforming all reported acid/alkali-resistant membranes. Notably, it shows high separation factors under acidic and alkaline conditions (SLi/Mg = 81.8 at pH 10; SLi/Co = 28.7 at pH 2), showing great potential for lithium extraction under harsh pH conditions, which is further confirmed by two-stage NF for both real salt lake brine and battery leachates. Additionally, we demonstrate large-area fabrication of the PQA membrane (∼625 cm2) while maintaining consistent separation performance.

Graphene oxide (GO) nanofiltration (NF) membranes have emerged as promising candidates for processing of aqueous streams, due to fast water transport characteristics and excellent chemical stability in harsh operating environments. Their practical applications under realistic high salt/solute concentration conditions are challenged by a lack of systematic approaches to control selectivity, as well as membrane swelling and mechanical stability. Here, we show that intercalation of rGO (reduced GO) membranes with polyaromatic conjugated molecules of different structural classes is an effective strategy to enhance and tune their performance. Specifically, these membranes allow entry into a "deep NF" regime with remarkably high inorganic monovalent salt rejections (NaCl, 75-85%) and near-total divalent salt rejections (Na2SO4, ∼98%) in a large concentration range (0.01-0.1 M) while maintaining high water permeability.

The growing demand for critical materials such as lithium, nickel, cobalt, and manganese in electric vehicles, renewable energy systems, and advanced electronics has intensified the need for sustainable recovery strategies. Spent lithium-ion batteries, mineral tailings, and industrial by-products represent valuable secondary resources that can support circular economy objectives. However, conventional hydrometallurgical and pyrometallurgical processes are energy-intensive, chemically demanding, and often generate significant secondary waste, including sludge and saline effluents. Membrane-based separation technologies have emerged as promising alternatives due to their modular design, lower energy requirements, and potential for selective metal recovery. Pressure-driven processes, including ultrafiltration (UF), nanofiltration (NF), and reverse osmosis (RO), as well as electro-driven systems such as electrodialysis (ED), enable target separations (e.g., Li/Mg, Co/Ni, Li/Co-Ni-Mn) through size exclusion, charge-based selectivity, and ion-membrane interactions. Nevertheless, membrane performance in realistic leachates characterized by low pH, high ionic strength, oxidants, and complex metal speciation is strongly governed by chemistry-driven failure modes. These include inorganic scaling (e.g., gypsum and silica), metal hydroxide precipitation, colloidal and organic fouling, redox-induced instability, and polymer degradation, which collectively contribute to flux decline, selectivity loss, and reduced membrane lifetime. This review critically evaluates membrane applications for battery and mineral waste valorization, linking dominant failure mechanisms to solution chemistry and long-term stability. Techno-economic considerations are discussed using normalized metrics (kWh/m 3 , kWh/kg metal, $/kg product), highlighting conditions under which membranes can become cost-competitive with conventional extraction routes.

Retaining essential minerals (e.g., Ca2+, Mg2+) is critical for generating healthy, palatable water. However, conventional polyamide nanofiltration (NF) membranes suffer from excessive mineral salt rejection, chlorine susceptibility, and scaling. Although polyester NF membranes are inherently chlorine-resistant, they lack the precise selectivity required for safe, mineral-rich potable water, limiting their application in drinking water treatment. Herein, we developed a high-selectivity dense polyester NF membrane using 3,5-dihydroxybenzoic acid (DHBA) as an aqueous monomer and incorporated sodium dodecyl sulfate (SDS) to assist interfacial polymerization by enhancing trans-interface reactant diffusion, creating membranes with a uniform pore size distribution. By tuning the SDS concentration, the optimal polyester membrane (SAIP-DHBA-0.125) achieved a water permeance of 8.3 L m-2 h-1 bar-1, superior water/Na2SO4 and CaCl2/Na2SO4 selectivities (42.7 and 177.7, respectively), and enhanced antiscaling and chlorine resistance, surpassing reported polyester membranes, most lab-made polyamide NF membranes, and a commercial nanofiltration membrane (NF270, Dupont). When applied to real tap water with high hardness, the SAIP-DHBA-0.125 membrane produced a high-quality permeate that complies with the Standards for Healthy Drinking Water Quality (T/BJWA 001-2021), while retaining essential minerals (Ca2+: 37.6 mg L-1, Mg2+: 56.3 mg L-1), approximately 55% higher than the NF270 membrane. Moreover, it maintained ∼96.0% of its initial flux and stable total dissolved solids rejection during 96 h of operation, underscoring its potential for the sustainable production of health-oriented potable water.

Pesticide contamination from agricultural activities has become a growing environmental concern since pesticides can migrate across environmental compartments and accumulate on undesirable surfaces and in water bodies.Given their high toxicity to living organisms and resistance to degradation, developing effective removal strategies is essential.This study investigates the removal of pesticide atrazine from a binary solution using commercially available nanofiltration (NF) and reverse osmosis (RO) membranes, with molecular weight cut-offs of 150-300 Da and 100-200 Da, respectively.The experimental study was conducted in a laboratory-scale RO/NF system with six cells connected in parallel over a duration of 3 h.Removal efficiency was determined by analysing all samples (feed and permeate) using liquid chromatography-tandem mass spectrometry.The results showed that the atrazine removal efficiency ranged from 16.0 to 84.9 % with NF membranes, and from 64.0 to 93.1 % with RO membranes, indicating that size exclusion was the main removal mechanism.

Practical Bayes filters often assume the state distribution of each time step to be Gaussian for computational tractability, resulting in the so-called Gaussian filters. When facing nonlinear systems, Gaussian filters such as extended Kalman filter (EKF) or unscented Kalman filter (UKF) typically rely on certain linearization techniques, which can introduce large estimation errors. To address this issue, this paper reconstructs the prediction and update steps of Gaussian filtering as solutions to two distinct optimization problems, whose optimal conditions are found to have analytical forms from Stein's lemma. It is observed that the stationary point for the prediction step requires calculating the first two moments of the prior distribution, which is equivalent to that step in existing moment-matching filters. In the update step, instead of linearizing the model to approximate the stationary points, we propose an iterative approach to directly minimize the update step's objective to avoid linearization errors. For the purpose of performing the steepest descent on the Gaussian manifold, we derive its natural gradient that leverages Fisher information matrix to adjust the gradient direction, accounting for the curvature of the parameter space. Combining this update step with moment matching in the prediction step, we introduce a new iterative filter for nonlinear systems called Natural Gradient Gaussian Approximation filter, or NANOfilter for short. We prove that NANO filter locally converges to the optimal Gaussian approximation at each time step. Furthermore, the estimation error is proven exponentially bounded for nearly linear measurement equation and low noise levels through constructing a supermartingale-like property across consecutive time steps. Real-world experiments demonstrate that, compared to popular Gaussian filters such as EKF, UKF, iterated EKF, and posterior linearization filter, NANO filter reduces the average root mean square error by approximately 45% while maintaining a comparable computational burden.

This study employs an integrated modeling approach to elucidate the mechanisms of nitrate ion transport through nanofiltration (NF) and reverse osmosis (RO) membranes. The investigation first applied models from irreversible thermodynamics, specifically the Kedem-Katchalsky and Spiegler-Kedem models, to describe convective/diffusive contributions and the impact of the initial nitrate concentration (50-150 mg/L) on phenomenological parameters (reflection coefficient σ, and solute permeability Ps). The results revealed a marked sensitivity of NF membranes to the initial nitrate concentration, in contrast to the stable performance of RO membranes. To deepen this analysis, Response Surface Methodology (RSM) was used as a robust statistical tool to systematically model and quantify the synergistic effects of the initial concentration and other key operational parameters, transmembrane pressure (TMP) and recovery rate (Y) on NF performance. The results highlight the complementarity between transport modelling and statistical approaches for analysing nitrate rejection and permeate flux. The proposed approach provides useful insight into NF membrane-specific behaviour and relative sensitivity to operating conditions, within the scope and limitations of the studied membrane and experimental configurations.

Energy-efficient and environmentally benign organic solvent nanofiltration (OSN)-based thin-film composite (TFC) membranes were fabricated on partially hydrolyzed polyacrylonitrile (HPAN) membranes via in situ oxidative polymerization of polyaniline (PANI). A three-layered PANI coating was applied using the layer-by-layer (LbyL) deposition technique to optimize efficiency. The separation efficiency of these membranes was evaluated for the recovery of n-hexane from vegetable-oil-based micelles using dead-end stirred cells. The membranes' microstructures, surface morphology, structural properties, thermal stability, tensile strength, and hydrophilicity were characterized using SEM, FTIR, TGA, CA, and UTM. Findings indicate that HPAN membranes with a single PANI coating exhibited less than 1% or negative rejection due to inadequate conversion from ultrafiltration (UF) to nanofiltration (NF). Membranes with a double PANI coating demonstrated NF membrane properties with reduced rejection rates. Triple PANI-coated membranes exhibited superior rejection of flaxseed, corn, soybean, sesame, and sunflower oils. It is concluded that the LbyL method for PANI deposition may be effectively utilized to enhance the performance efficiency of OSN-TFC membranes in the oleochemical industry, thereby minimizing solvent recovery waste.

Lithium is a critical element for modern energy storage systems, particularly in batteries powering renewable energy technologies and electric vehicles. With global demand rapidly increasing, attention has shifted toward recovering lithium from unconventional sources such as desalination brine. This byproduct of brackish and seawater desalination contains lithium concentrations higher than those found in seawater, making it a valuable secondary resource. Recent advancements in recovery technologies-including metal-organic frameworks (MOFs), ion-imprinted polymers (IIPs), nanofiltration (NF), and capacitive deionization (CDI) and its variants (MCDI, HCDI, and FCDI)-offer promising pathways for efficient and sustainable lithium extraction. These technologies differ in selectivity, energy efficiency, scalability, and cost. MOFs and IIPs exhibit superior selectivity for lithium ions but are limited by high material costs, whereas CDI-based methods are more energy efficient, regenerative, and environmentally friendly. NF, though well established and scalable, often requires high pressure, increasing energy consumption. This review highlights the potential of hybrid systems that integrate the selectivity of advanced materials like MOFs and IIPs with the operational efficiency of CDI technologies. Such integrated approaches represent a sustainable and cost-effective route for large-scale lithium recovery from desalination brine, addressing both environmental and economic challenges associated with the global lithium supply.

In-situ vacuum-assisted fabrication of highly selective hollow fiber nanofiltration membranes for removing polyfluoroalkyl substances.

Recent advances and future perspectives in ceramic-based nanofiltration membranes: Material innovations, applications, and sustainability challenges.

Cellulose Nanofiber-Based Nanofiltration Membranes for Sustainable Water Purification