Comparative Analysis of Commercial and Novel High‐Pressure Membranes for Perfluoroalkyl and Polyfluoroalkyl Substances (PFAS) Removal

ABSTRACT

High‐pressure membrane technologies can be effective in mitigating perfluoroalkyl and polyfluoroalkyl substances (PFAS) contamination in water matrices. This review explores recent developments in both commercial (e.g., NF and RO) and novel membrane technologies, focusing on their removal mechanisms, influential factors, and challenges. Key determinants, including solution pH, PFAS molecular structure, co‐contaminants, and natural organic matter, are summarized for their impacts on PFAS removal efficiency. Novel membranes incorporating materials like graphene oxide, quaternary ammonium compounds, and metal–organic frameworks are highlighted for their potential to enhance PFAS removal, particularly the removal of short‐chain PFAS. Despite promising developments, challenges such as fouling, energy demands, and scalability necessitate further research. This review highlights the significance of lab‐scale studies and innovative designs in bridging the gap between laboratory findings and practical applications, thereby paving the way for sustainable, large‐scale PFAS treatment.

Summary

• RO membranes achieved over 99% PFAS removal, outperforming NF in many conditions.

• NF membranes show inconsistent performance, significantly affected by water chemistry factors such as pH, co‐contaminants, and natural organic matter.

• Novel membranes with nanomaterials and adsorptive fillers enhance PFAS selectivity and reduce fouling.

• Machine learning and renewable energy integration offer promising pathways for sustainable, large‐scale PFAS remediation.

This study highlights membrane filtration as an effective method to remove PFAS from consumer or industrial sources. It emphasizes key mechanisms such as size exclusion, electrostatic interactions, and affinity, along with critical influencing factors, including membrane characteristics, solution chemistry, and PFAS properties that enhance removal efficiency.

1. Introduction

Since the 1940s, thousands of perfluoroalkyl and polyfluoroalkyl substances (PFAS), including both conventional (e.g., perfluorooctane sulfonic acid [PFOS] and perfluorooctanoic acid [PFOA]) and alternative short‐chain or newly recognized fluorinated substitutes, have been synthesized, differing in chain length, structure, and charge (Crone et al. 2019; Hara‐Yamamura et al. 2022; Lenka et al. 2021; Saleh et al. 2019; Zhang et al. 2022). The exceptional surface activity, stability, and resilience of PFAS under challenging conditions attributable to their unique carbon–fluorine (C–F) bond have led to their widespread use in everyday products, including cookware and shampoos, as well as in industrial surfactants and firefighting foams (Das and Ronen 2022; Saleh et al. 2019; Xue et al. 2025; Zhi et al. 2022). This C–F bond is also responsible for making PFAS highly resistant to natural degradation, causing them to persist and accumulate in the ecosystems (Jin, Peydayesh, Joerss, et al. 2021; Ma et al. 2022). PFAS have been detected in marine organisms and in various water sources, including lakes, rivers, tributaries, and groundwater, often at concentrations as low as nanograms per liter (ng/L) (Jin, Peydayesh, Joerss, et al. 2021; Johnson et al. 2022). Still, they could cause numerous health issues like hormone disruption, high cholesterol, pregnancy‐related issues, cancerous disease, and immune toxicity, turning PFAS pollution into a major concern (Chen et al. 2023; Crone et al. 2019; Das and Ronen 2022; Hara‐Yamamura et al. 2022; Jin, Peydayesh, Joerss, et al. 2021; Li et al. 2023; Saleh et al. 2019; Tow et al. 2021). It may also be adsorbed onto microplastic/nanoplastic, resulting in more severe adverse effects (Samaei et al. 2025). Since 2009, under the Stockholm Convention, PFAS have been progressively listed as persistent organic pollutants (POPs), beginning with PFOS in 2009, followed by PFOA (and related compounds) in 2019, and PFHxS (and related compounds) in 2022 (Jin, Peydayesh, Joerss, et al. 2021; Li et al. 2023; Ma et al. 2022). Moreover, the long‐chain perfluorocarboxylic acids (PFCAs) will be considered by the Conference of the Parties in May 2025.

To mitigate the risk associated with PFAS exposure, the US Environmental Protection Agency (USEPA) established a lifetime health advisory level of 70 ppt for PFOA and PFOS (individual/combined) in drinking water in 2016 (Murray et al. 2019; Wan et al. 2022). In April 2024, USEPA finalized the standard, setting the limits at 4 ppt for PFAS and PFOA, and 10 ppt for PFNA, PFHxS, and HFPO‐DA (USEPA 2024). Moreover, other countries have also established their own limits based on local regulations. For instance, in 2020, Japan set a limit of 50 ppt for PFAS in drinking water (Botta and Yamasaki 2020; Wang et al. 2015; Zeng, Tanaka, Suzuki, Yukioka, and Fujii 2017). Additionally, the United Kingdom has implemented a stricter limit of 10 ppt, while countries such as Germany, Italy, the Netherlands, and Sweden have limits ranging from 200 to 500 ppt, depending on the specific compound and application.

The increasing stringency of regulations has driven a new wave of research into PFAS removal technologies. Conventional treatment methods, such as coagulation, adsorption, biodegradation, micro/ultrafiltration (MF/UF), media filtration, ozonation, and chlorination, have proven ineffective at removing (Lenka et al. 2021; Ma et al. 2022; Mastropietro et al. 2021; Wang et al. 2020). In response, the techniques such as ion exchange resin and granular activated carbon treatment, as well as chemical degradation processes, have shown relatively promising results, particularly for long‐chain PFAS (i.e., PFOA and PFOS) (Deng et al. 2010; Kucharzyk et al. 2017; Lee et al. 2022). The effectiveness of these methods depends on various factors, including the type of GAC or resin used, bed depth, water flow rate, temperature, and the presence of co‐contaminants, and they can achieve up to 100% removal of long‐chain PFAS for a limited duration (USEPA 2018). However, they are generally ineffective for removing short‐chain PFAS, given by their unique nature, such as smaller molecular size, high solubility, low hydrophobicity, and weaker ionic interactions (Glover et al. 2018; Li et al. 2020; Murray et al. 2019; Rahman et al. 2014; Wei et al. 2019). Also, the disposal of exhaust adsorbents presents an additional challenge, requiring careful management to avoid secondary environmental contamination (Lee et al. 2022; Saleh et al. 2019; Zhang et al. 2023; Zhi et al. 2022). In contrast, high‐pressure membranes can remove both long and short‐chain PFAS from water. RO demonstrates superior efficiency, achieving over 99% removal of nearly all PFAS types, while NF is highly efficient in removing long‐chain PFAS (> 95%). Although the removal efficiency for short‐chain PFAS is lower than that for long‐chain PFAS, the results are still satisfactory (Liu et al. 2021).

Compared with conventional methods, pressure‐driven membrane processes like nanofiltration (NF) and reverse osmosis (RO), have emerged as highly effective approaches for removing PFAS of all chain lengths. These high‐pressure membranes typically achieve PFAS rejection rates above 90% for long‐chain PFAS (Chen et al. 2023; Ma et al. 2022; Saleh et al. 2019; Tow et al. 2021), and even for short‐chain PFAS, rejection rates greater than 80% have been reported (Li et al. 2021; Zhi et al. 2022). However, this efficiency depends on several factors, including solution concentration, pH, and the presence of co‐contaminants. Variations in these factors as well as differences in the type and structure of PFAS can significantly impact removal performance. Despite their effectiveness, pressure‐driven membrane processes face several challenges, such as high operating costs, reduced efficiency for short‐chain PFAS, membrane fouling, concentrate disposal, and management of spent membranes (Murray et al. 2019). These drawbacks have spurred efforts to develop tailored membranes with enhanced performance. As a result, an increasing number of studies have focused on innovative membrane design and their effectiveness in eliminating short‐chain PFAS species. Figure 1 illustrates the number of publications on pressure‐driven membrane technology for PFAS removal from wastewater published annually over the last decade.

FIGURE 1.

Numbers of studies on pressure‐driven membrane technologies for PFAS removal over the last 10 years. The data were generated by searching Keywords “PFAS removal” and “membrane” in the Web of Science database, with only those papers focusing on high‐pressure membrane technologies for PFAS removal included.

This review aims to summarize recent findings on high‐pressure membrane processes for PFAS removal, emphasizing the performance of both commercially available NF and RO membranes alongside novel innovations. The removal mechanisms related to PFAS removal, including adsorption, electrostatic interactions, and hydrophobic effects, were discussed. Also, we explored key influencing factors such as solution pH, PFAS concentration, coexisting ions, and organic matter. Challenges like fouling, energy demand, and scalability are also discussed, providing insights into future research directions for optimizing membrane technologies and translating lab‐scale innovations into practical applications.

2. Overview of Commercial NF and RO in PFAS Removal

Microfiltration (MF) and ultrafiltration (UF) are low‐pressure membranes (10–100 nm), which are generally ineffective for PFAS removal due to their large pore sizes, resulting in low rejection rates (Abbasian Chaleshtari and Foudazi 2022), seen in Figure 2. For example, the UF/MF membrane system failed to remove the PFAS due to its vast pores, including PFOS, PFDOA, PFHxA, and FOSA (Zahmatkesh et al. 2024). Conversely, commercial NF and RO membranes are distinct pressure‐driven technologies with differing pore size ranges. RO membranes have smaller pore sizes (less than 1 nm), while NF membranes have larger ones (1–10 nm), both demonstrating high effectiveness in removing PFAS and producing high‐quality water (Abbasian Chaleshtari and Foudazi 2022). Table 1 summarizes the progress on the application of NF and RO membranes for PFAS treatment in aqueous environments.

FIGURE 2.

Schematic representation of membrane filtration capabilities for pollutants in μm.

TABLE 1.

Summary of studies on PFAS removal using commercial NF and RO membranes.

Membrane type	Membrane model	Target PFAS species	Removal efficiency (%)	Reference
RO	BW30	C4–C11 PFAS (i.e., PFBA, PFBS, PFPeA, PFHxS, PFHpA, PFOA, PFOS, PFNA, PFDA, and PFUnDA)	> 99.0	(Ma, Lei, et al. 2024)
NF	NF90		90.0–99.0
RO	BW30	PFOA	> 99.0	(Ma, Zhang, et al. 2024)
NF	NF90	PFBA	97.0–99.0
NF	NF90 and NF270	PFBA	> 97.0	(Griffin et al. 2024)
		PFBS
		PFOA
		PFOS
		TFMS
NF	NF90 and NF270	C3–C8 PFAAs (i.e., PFPrS, PFBS, PFPeA, PFPeS, PFHxA, PFHxS, PFHpS, PFOA, and PFOS)	97.0	(Safulko et al. 2023)
RO	CR100 and SW30
NF	DK, NF90, XN45, NF270, and DL	PFBA	81.0–99.9	(Zhi et al. 2022)
		PFPeA
		PFHxA
		PFHpA
		PFOA
		PFBS
		PFOS
		4:2 FTS
		6:2 FTS
		PFMOPrA
		PFMOBA
		HFPO‐DA
NF	NF270	42 PFASs (10 PFAAs and 32 PFASs)	> 98.0	(Liu et al. 2021)
RO	ESPA		> 99.0
NF	NF90	PFBS	90.0–97.4	(Li et al. 2021)
		PFBA	79.7–86.0
		PFHxS	98.3–99.3
		PFHxA	98.8–99.6
		PFOS	95.0–96.4
		PFOA	92.2–97.4
	NFG	PFBS	38.9–64.1
		PFBA	16.5–18.7
		PFHxS	20.8–24.6
		PFHxA	24.7–38.6
		PFOS	27.9–33.1
		PFOA	19.6–55.6
NF	ESNA1‐K1	PFOS	92.0–99.0	(Zhao et al. 2020)
NF	NF90	PFHxA	> 99.0	(Kim, Chernysheva, et al. 2024)
	NF270		86.9–96.2
RO	XLE, BW30, and SW30		> 99.0
NF	ESNA1‐K1	PFOS	92.7–97.9	(Zhao et al. 2018)
UF	UH030	PFHxA	68.9	(Zeng, Tanaka, Suzuki, Yukioka, and Fujii 2017)
NF	NTR‐7410		95.3
NF	NTR‐7450		96.3
RO	NTR‐759		99.2
NF	NF270	PFOS	94.3–99.5	(Zhao et al. 2016)
NF	NF90	PFOA	99.3–99.9	(Hang et al. 2015)
	NF270		92.0–99.0
NF	NF270	PFOS	94.0–99.3	(Zhao et al. 2013)
RO	SG, LFC1, LFC3, BW30, and ESPA3	PFOS	> 90.0	(Tang et al. 2007)
NF	DK, NF90, and NF270		90.0–99.0

Numerous studies have highlighted the effectiveness of both RO and NF in removing PFAS from water. RO, with its smaller pore size and enhanced electrostatic repulsion between the solute and membrane surface, consistently achieves higher efficiency compared with NF (Ma, Lei, et al. 2024; Soriano et al. 2019; Zeng, Tanaka, Suzuki, Yukioka, and Fujii 2017). While RO performance remains stable across varying conditions, the efficiency of NF fluctuates significantly depending on factors such as water chemistry and PFAS natures (Ma, Zhang, et al. 2024). Ma, Lei, et al. (2024) evaluated the performance of commercial RO (BW30‐2540) and NF (NF90‐2540) membranes under varying conditions, focusing on the effects of pH, ionic strength, NOM, and PFAS structure (i.e., chain length) on removal performance. Their findings revealed that RO membrane consistently achieved over 99% removal efficiency across all tested conditions, while the NF membrane exhibited more variable performance, ranging from 90% to 99%, depending on the specific factors. The study also found that at transmembrane pressures (TMP) above 41.37 bars, the permeate flux increase began to decline due to pore compression, reducing water transfer channels. However, higher pressures decreased PFOA and PFBA concentrations in the permeate, improving the rejection rate. This was attributed to the dilution effect of RO, where increased water flux enhances recovery, while PFAS flux remains pressure‐independent, consistent with the solution–diffusion mechanism. Moreover, at a fixed ionic strength, PFOA was more effectively removed than short‐chain PFAS, such as PFBA, due to differences in molecular size and electrostatic interactions. High ionic strengths (i.e., 1000 mM) negatively impacted removal performance by causing membrane pore swelling, which reduced PFAS rejection rates. Interestingly, the presence of NOM enhanced PFAS removal, including humic acid (HA), sodium alginate (SA), and bovine serum albumin (BSA), by forming complexes and fouling layers. Furthermore, PFAS retention improved at higher pH levels, where the increased anionic nature of PFAS molecules enhanced electrostatic repulsion with the negatively charged membrane surface. These findings highlight the superior and consistent performance of RO membranes under complex conditions and underscore the influence of environmental factors on NF membrane performance. In another study, Ma et al. (2024) investigated the effects of surfactants, ion valency, and solution temperature on the filtration of PFOA and PFBA using the same types of RO and NF mentioned in the previous study (Ma, Zhang, et al. 2024). Positively charged surfactants like cetyltrimethylammonium bromide (CTAB) and high‐valency ions (e.g., Al³⁺ and PO₄ ³⁻) improved PFAS removal by binding with the negatively charged PFAS, accordingly increasing effective molecular size and enhancing electronegativity. Also, anionic surfactants like sodium dodecyl sulfate (SDS) enhanced PFAS removal by increasing the overall electronegativity. However, the high temperature (i.e., 45°C) reduced efficiency by causing thermal expansion of membrane pores. While surfactants and ionic variations had minimal impact on RO membranes, given their reliance on size exclusion, the factors enhanced NF membrane rejection of PFOA and PFBA by 2%–3%. In short, RO membranes maintained high rejection regardless of coexisting ions, while NF membranes showed improved rejection efficiency under varying ionic conditions.

In the study by Soriano et al. (2019), the removal efficiency of the PFHxA (i.e., short‐chain PFCA) present in aqueous solutions was evaluated using various RO (i.e., XLE, BW30, and SW30) and NF (i.e., NF270 and NF90) membranes (Soriano et al. 2019). RO membranes consistently outperformed NF membranes, even in the presence of salt (achieved by adding 36 mg/L of NaCl, 575 mg/L of CaSO₄, and 98 mg/L of NaHCO₃). Notably, the impact of salt is particularly pronounced in membranes with larger pores, such as the NF270. NF90 demonstrated a better balance between flux and removal efficiency, while NF 270 achieved the highest removal efficiency of 96.2% at neutral pH (7.1). However, its efficiency decreased to 86.9% under acidic conditions (pH 3.5). In contrast, RO membranes maintained rejection rates above 99% regardless of pH, with XLE membranes achieving the highest rejection rates from 99.0% to 99.4%. Moreover, Zeng, Tanaka, Suzuki, Yukioka, and Fujii (2017) clearly reported that PFHxA rejection was not dependent on membrane MWCO, as no clear correlation was observed between MWCO and rejection rate (Zeng, Tanaka, Suzuki, Yukioka, and Fujii 2017). A distinctive rejection rate of 99.2% was observed for RO (R1, NTR‐759 HR, MWCO = N/A), while the two NF membranes achieved rejection rates of 95.3% (N1, NTR‐7410, MWCO = 3000 Da) and 96.3% (N2, NTR‐7450, MWCO = 1000 Da), respectively. The UF (U2, UH030, MWCO = 30,000 Da) membrane exhibited the lowest rejection rate of 68.92%. These results indicate that PFHxA rejection was not dependent on membrane MWCO, as no clear correlation was observed between MWCO and rejection rate. Instead, surface potential played a more dominant role, with lower zeta potential corresponding to decreased rejection rates. The overall performance followed the hierarchy of RO > NF > NF. Although RO showed the highest rejection rate, NF membranes were deemed more suitable because of their optimal balance of PFAS and NaCl rejection rates with superior water permeability.

In another comparative study, five commercial RO (i.e., SG, LFC1, LFC3, BW 30, and ESPA3) and three commercial NF membranes (i.e., DK, NF90, and NF270) were used to remove PFOS (Tang et al. 2007). Similar findings were reported that RO membranes consistently achieved high performance (> 99%), while the performance of NF membranes varied between 90% and 99% depending on the membrane type. Over time, NF membranes demonstrated improved rejection of PFOS (16%), with reduced fluctuation. It could be attributed to the increased PFOS accumulation on the membrane surface and entrapment of PFOS molecules within the polyamide layer. Furthermore, RO systems have proven highly effective for point‐of‐use PFAS treatment, as reported, no PFAS was detected in water treated by household‐scale RO systems (Griffin et al. 2024). Similarly, Zhi et al. (2022) conducted a comprehensive evaluation of five commercially available NF membranes (NF90, NF270, DK, DL, and XN45) to assess their effectiveness in treating groundwater contaminated with 12 legacy and emerging PFAS, including five PFCAs, two PFSAs, two FTSs, and three PFEAs (Zhi et al. 2022). Despite the membranes having similar structural characteristics, their rejection rates varied significantly, ranging from 66% to > 99.9%. The removal efficiency followed the order: DK > NF90 > XN45 > NF270 > DL. Notably, the DK and NF90 membranes excelled in removing short‐chain PFAS (C3 and C4), such as PFBA, PFBS, 4:2 FTS, PFMOPrA, and PFMOBA. This superior performance was attributed to the tightly knit membrane structure, which, however, resulted in reduced water permeability of DK membrane. In contrast, the DL membrane, with the loosest structure, exhibited higher water permeability but lower rejection performance, further validating the relationship between membrane structure and PFAS removal performance (Tang et al. 2007; Zhi et al. 2022). They also observed that membrane intrinsic structural characteristics, such as MWCO, permeability, and salt selectivity, had a statistically more significant impact on PFAS removal than surface properties like hydrophilicity, charge, and roughness. This finding suggests that steric hindrance plays a dominant role in PFAS removal. Additionally, the mass of PFAS adsorbed on the membrane was positively correlated with molecular parameters such as molecular weight (MW) and logK_ow but showed a weak correlation with membrane properties. This indicates that both adsorption and rejection of PFAS are driven by similar mechanisms, primarily size exclusion and hydrophobic interactions.

So far, the discussion has primarily focused on laboratory‐scale studies; however, real‐world situations often differ due to the presence of various co‐components in the water. A recent study investigated PFAS‐contaminated wastewater from the semiconductor industry, using both flat‐sheet filtration (NF90 and NF270 membranes) and a spiral‐wound NF90 membrane in a pilot‐scale system (Griffin et al. 2024). In single‐cell tests, NF90 demonstrated superior performance, achieving a 97% rejection rate for all PFAS, including PFBS, PFOA, and PFOS (500 μg/L each), and the ultrashort‐chain trifluoromethane sulfonic acid (TFMS, C1 PFSA, 6.81 mg/L). Larger PFAS molecules like PFOS and PFOA were predominantly removed via size exclusion, while smaller ones like PFBA and TFMS were removed through electrostatic repulsion, given their molecular weights are close to the membrane's MWCO. The pilot system exhibited even better performance, likely due to handling‐related imperfections in the flat sheet membranes. In a closed‐circuit semi‐batch process, similar results were observed, with rejection rates exceeding 99% for larger PFAS, such as PFOA, PFOS, 6.2 TFS, and 8.2 FTS. Short‐chain PFAS showed slightly lower rejection rates, including TFA (93.7%), TFMS (95.0%), PFBS (96.5%), and PFHxA (97.6%). In the treatment of fab wastewater, all measured PFAS were removed with an efficiency exceeding 99.6%, except for the ultrashort‐chain compounds TFA, which showed a rejection rate of 92.0 ± 1.0%, and TFMS, with a rejection rate of 98.3 ± 0.4%. The slight differences were likely due to compositional differences between the two water sources. These batch rejection values reflected a similar chain‐length dependence as observed in the flat‐sheet and pilot experiments, highlighting the effectiveness of NF membranes in treating PFAS‐contaminated industrial wastewater. Moreover, Liu et al. (2021) evaluate the performance of pilot‐scale spiral‐wound NF (NF270) and RO (ESPA) membranes in treating diluted mixed‐source Aqueous Film Forming Foam (AFFF) stock (AFFF spiked) and AFFF‐impacted groundwater (AFFF GW), analyzing 10 PFAAs and 32 PFASs (Liu et al. 2021). The rejection of all PFAAs in the AFFF spiked water matrix was consistently > 98% for NF and > 99% for RO. However, lower rejection rates (92%–98%) were observed for AFFF GW, with the most significant decreases seen for shorter‐chain PFSAs. This reduction may be attributed to certain groundwater constituents interacting with the sulfonate headgroups of PFSAs, consequently hindering effective rejection. Interestingly, the inhibitory effect was not observed for shorter‐chain PFCAs (i.e., PFPeA, PFHxA, and PFHpA). This study also highlighted that the role of elevated calcium (Ca²⁺) levels in AFFF GW, which were found to cause greater membrane fouling through the formation of DOM‐Ca²⁺ complexes. Furthermore, hydrophobic PFAAs, particularly PFOS and PFHpS, exhibited adsorptive losses onto the spiral‐wound membrane elements, indicating that adsorption plays a role in removal. They also emphasized that, in addition to membrane structure, the complex nature of PFAS (i.e., chain length) can contribute to the removal performance and removal mechanisms.

In summary, both NF and RO membranes effectively remove PFAS under specific conditions, but challenges remain due to the variety of PFAS, diverse water matrices, and operational complexities. RO offers high removal efficiency but is costly and susceptible to fouling, while NF membranes, particularly NF90, perform better than NF270 for PFAS removal. However, NF membranes are effective for long‐chain PFAS but less so for short‐chain ones due to larger pore sizes. Both technologies face additional challenges, including a trade‐off between retention efficiency and water permeability, a lack of selectivity, potential fouling, and high energy demands, which reduce their feasibility for large‐scale PFAS treatment. A further concern is the generation of significant volumes of PFAS‐concentrated wastewater, which poses environmental risks and is challenging to manage. Closed‐circuit membrane filtration offers a potential solution by reducing waste volume while maintaining high removal efficiency as single‐pass filtration (> 97% PFAA rejection) (Safulko et al. 2023). However, this approach is more complex and expensive than the conventional membrane filtration process. To overcome these limitations, researchers are actively developing novel membranes with improved performance characteristics. The recent advancements are discussed in the following section.

3. Membrane Process

The development of novel membranes focuses on the fabrication of the active separation layer to enhance specific separation performance. These modified membranes provide advantages such as optimized pore size, improved PFAS selectivity, greater fouling resistance, reduced energy consumption, and extended lifespan, making them a more cost‐effective and sustainable solution (Table 2).

TABLE 2.

Summary of studies on PFAS removal using novel membranes.

Membrane type	Fabrication process	Material used for fabrication	Pore size	Initial PFAS concentration	Permeability (LMH/bar)	Membrane performance	Reference
Polyacrylonitrile ultrafiltration membrane (PAN‐Q)	Molecular modification and wet‐phase inversion	Polyacrylonitrile (PAN), N,N‐dimethyl‐1,2‐ethanediamine (DMEN), and N,N‐dimethylformamide (DMF)	10.9 nm	0.1 mg/L	21.3	> 98% PFOS removal > 96% removal with coexisting NOM > 90% efficiency after regeneration Improved antifouling performance Four regeneration cycles (99.1%–90.6%)	(Fang et al. 2024)
SFAC activated adsorptive TFC membrane	Interfacial polymerization (IP)	Super fine activated carbon (SFAC)	5.5 nm	2 mg/L	21	94% PFOS removal High regeneration (90% PFOS removal after four cycles)	(Kasula et al. 2024)
PDVF‐g‐QA & PDVF‐g‐TA membrane	Thermal induced graft polymerization		36.9 nm	0.24 μmol/L	165.6	> 90% PFOA removal pH swing regeneration 84.7% removal capacity after three adsorption/desorption regeneration cycles.	(Wan et al. 2024a)
Dual‐functional adsorptive membranes, with hydrophobic backbone and quaternary ammonium (QA) moieties,	Nonsolvent induced phase separation (NIPS)	QA moieties	14.8–22. 8 nm	0.24 μmol/L	34.6	Selective interception of organic competitors like HA 96.8% PFOA removal > 90% removal efficiency maintained across five reuse cycles	(Wan et al. 2024b)
Pore functionalized MF membrane with anion exchange moiety	Vacuum filtration and polymerization	Quaternary/tertiary amine	N.A.	150 μg/L	45–131	90% PFOA rejection 50%–80% PFBA rejection at 30% permeance recovery Regeneration performance up to three cycles (over 90%)	(Thompson et al. 2024)
Mixed matrix composite nanofiltration (MMCNF)	Phase inversion	β‐Cyclodextrin polymer	346 Da	285.5 μg/L	8.7	Near complete (> 99.9%) PFOA rejection Recovery of 99% removal capacity after regeneration	(Chaudhary et al. 2023)
Polysulfone–graphene oxide hollow fiber membranes (PSU‐GO HFs)	Phase inversion extrusion	Graphene oxide	13 nm	0.5 μg/L	N.A.	Higher selectivity for short‐chain PFAS.	(Zambianchi et al. 2022)
NF‐PNIPAm‐PVDF membranes	In‐pore polymerization and IP	Poly‐N‐isopropylacrylamide (PNIPAm) and iperazine‐trimesoyl chloride	N.A.	160 mg/L	7–24	Over 80% removal of PFOA and PFOS rejection	(Léniz‐Pizarro et al. 2022)
Amyloid–carbon hybrid membrane	Commercial	Activated carbon	N.A.	400 ng/L	1738	100% removal of both long and short‐chain PFAS > 96% removal of PFBA removal 38% more sustainable than commercial NF membrane	(Jin, Peydayesh, Joerss, et al. 2021)
ZIF‐L incorporated TFN membrane	IP	ZIF‐L	400 Da	200 μg/L	47.56	2.3 times high water flux > 97% retention of representative PFASs 80% short‐chain PFAS removal (C6) Over 88% flux recovery ratio after humic acid contamination–cleaning	(Bi et al. 2024)
Hyaluronic acid (HA) interlayered TFC membrane	IP	HA	N.A.	100 μg/L	29.18–30.12	Improved rejection of short‐chain PFASs (i.e., 44.96%–99.43%, C4 to C6) Improved reversible fouling property	(Chen et al. 2023)
MXene interlayered TFCi membrane	IP	MXene	N.A.	10–200 μg/L	12.16	Improved rejection of short‐chain PFAS (C4, 93.35% of PFHxA) Improved water permeance	(Ma et al. 2022)
MXene/CNT membrane	Vacuum filtration	MXene and CNT	N.A.	10–200 μg/L	1243.7	Improved PFOA rejection (96.56%) Improved flux	(Xu et al. 2022)
MXene‐PA thin film nanocomposite (TFN) membrane	IP	MXene	N.A.	N.A.	29.26	Better PFOS rejection (96%) Improved permeability	(Le et al. 2022)
Amine functionalized‐boron nitride BN (NH2) nanosheet–decorated membranes	IP	Amine functionalized‐boron nitride BN (NH2) nanosheet	N.A.	N.A.	8.83	1.04 times higher permeate flux than PPA membrane 93% rejection of target PFAS	(Abdikheibari et al. 2022)
MOF membrane	IP	Copper‐tetrakis (4‐carboxyphenyl) porphyrin (Cu‐TCPP) MOF nanosheets	N.A.	50 μg/L	21.4	Effectively removes five short‐chain and six long‐chain PFAS Over 84% rejection of short‐chain PFAS Superior anti fouling performance	(Zhang et al. 2024)
Piperazine (PIP) + Bipiperidine (BP) membranes	IP	PIP and BP	1.2 nm	N.A.	Approximately 12	Improved water flux ∼90% rejection of PFOA Improved fouling properties through salt passage	(Boo et al. 2018)
Asymmetric membrane	Self‐assembly and non–solvent‐induced phase‐separation process (SNIPS)	Cobaltocenium‐containing block copolymers (BCP)	17–23 nm	N.A.	521	Surface and pore customization 96.6%–99.3% PFOA retention A stable permeance of 48 to 171 LMH/bar	(Rittner et al. 2024)
Anionic crystalline ionic (COF) membrane	Counter diffusion IP method	TpPa‐SO3H COF	1.46 nm	1 mg/L	19.9–37.5	Highly negative surface charge, ordered pore channels with suitable pore sizes 99% rejection of PFOS 90%–95% rejection of PFOA and short‐chain PFAS High salt/PFAS selectivity Constantly over 90% rejection for both PFOA and PFOS in a 4‐day test	(Nguyen et al. 2024)
Organically modified montmorillonite (OMMT)‐PMIA membrane	Wet‐phase inversion and solution dispersion	Poly(m‐phenylene isophthalamide) (PMIA) and organically modified montmorillonite (OMMT)	N.A.	100 μg/L	N.A.	Improved thermal stability and hydrophilicity Improved permeation and rejection Over 95% of PFOS removal when adding 2% OMMT	(Luo et al. 2016)
Electrospun polyacrylonitrile (PAN) and amidoxime surface functionalized nanofibrous membrane	Electrospinning	Polyacrylonitrile (PAN)	N.A.	100 mg/L	N.A.	High GenX removal capacity ∼0.6 mmol/g with ASFPAN at pH = 4	(Mantripragada, Obare, and Zhang 2023)
PAN/Algae bicomponent nanofibrous membrane (ES (PAN/Algae))	Electrospinning	PAN and chlorella	N.A.	100 mg/L	N.A.	Improved GenX removal efficiency (0.9 mmol/g at pH 6)	(Mantripragada, Deng, and Zhang 2023)
PVDF‐g‐QA membrane	Partial defluorination, QA grafting and electrospinning	PVDF and QA	67 nm	0.12 μmol/L	32.3	Low‐pressure membrane Removal efficiency as commercial NF Easy membrane regeneration with mild and minimal solution Consistent PFOA and PFOS removal (> 90%) during an 8‐h test of 258 L per m2 of membrane	(Wan et al. 2022)
Multiwalled carbon nanotubes (MWCNT)–filled electrospun nanofibrous membrane	Electrospinning	MWCNT	N.A.	1–100,000 μg/L	N.A.	Improved sorption of PFOS (16.29 ± 0.26 μmol/g vs. 0.92 ± 0.06 μmol/g for pristine membrane)	(Dai et al. 2013)
Poly(diallyl dimethyl ammonium chloride) (PDDA)‐GO membrane and poly (sodium 4‐styrenesulfonate) (PSS)‐GO membrane	Vacuum‐assisted self‐assembly (VASA)	PDDA and PSS	0.7 nm	1–100 μg/L	PSS‐GO = 83.1 PDDA‐GO = 80.5	High PFOA rejection (over 90%) High water flux	(Khorramdel et al. 2022)
GO‐polyethyleneimine (PEI)	Surface modification	PEI	0.2 μm	50–200 μg/L	12.6	93.7% PFOA rejection at 50 μg/L	(El Meragawi et al. 2020)
Poly(m‐phenylene isophthalamide) (PMIA) hollow fiber membrane	Dry‐jet wet spinning	PMIA	0.404 nm	0.1 mg/L	N.A.	> 99% PFOS rejection	(Wang et al. 2015)

3.1. Adsorption‐Enhanced Functional Composite Membrane

As adsorption plays a crucial role in PFAS removal by membranes, enhancing the adsorptive property significantly improves membrane performance. For example, Johnson et al. (2019) functionalized aluminum oxide hydroxide membranes with fluorinated silane to enhance the adsorption of PFOA (0.39 ng/L) and PFOS (0.86 ng/L) (Johnson et al. 2019). A 90% rejection rate was achieved as the perfluorinated side chains adsorbed onto the membrane surface through C–F interactions. Under optimized conditions (pH 7.5, 30 min of filtration, and operated at 0.317 bar), the removal efficiency was further improved to 99.9% by incorporating polyethylene glycol. Another innovative approach involves thin‐film nanocomposite (TFN) membranes incorporating superfine activated carbon (SFAC) to enhance the removal of PFOS, creating additional adsorption sites and altering surface properties (Kasula et al. 2024). The SFAC‐TFN‐0.5 membrane demonstrated the highest adsorption capacity, with ~90% of PFOS being adsorbed and only 5% rejected, achieving a total removal efficiency of 95% (initial concentration = 2 ppm). Remarkably, the membrane maintained consistent PFOS removal efficiency across three adsorption cycles, with only a 5% decrease in efficiency observed after the fourth cycle. This strategic functionalization improved PFOS rejection beyond conventional filtration mechanisms, leveraging both size‐exclusion and adsorption effects. In another study, the addition of multiwalled carbon nanotubes (MWCNT) enhanced membrane performance by doubling the specific surface area with just 0.5 wt%, increasing sorption capacity for PFOS by 18 times compared with the pristine membrane. This modification overcame initial electrostatic repulsion through strong hydrophobic interactions between the MWCNTs and PFOS (Dai et al. 2013).

The hydrophilic surfaces can enhance antifouling capabilities, reducing the accumulation of contaminants on the membrane. Fang et al. (2024) developed an adsorptive polyacrylonitrile UF membrane with an average pore size of 10.9 nm, grafted with tertiary amino groups and subsequently quaternized, to treat PFOS at an initial concentration of 0.20 μM (Fang et al. 2024). The membrane achieved over 98% removal at neutral pH, with over 90% removal efficiency recovered through eluent regeneration. PFOS removal remained consistent (> 96%) even in the presence of coexisting NOM, attributed to the synergistic effects of electrostatic, hydrophobic and dipole–dipole interactions. Notably, the hydrophilic modification enhanced the membrane's antifouling performance, achieving a remarkable flux recovery rate of 99.0%, even in the presence of surface electrostatic attraction. The addition of organically modified montmorillonite (OMMT) also offers a simple and effective approach with enhanced hydrophilicity and thermal resistance (Luo et al. 2016). It resulted in denser, thinner skin layers, higher porosity, and superior PFAS removal performance, as well as improved permeation efficiency.

Given that PFAS are negatively charged, the deprotonation of amine groups enhances the membrane's surface charge, thereby increasing electrostatic repulsion between the membrane and PFAS molecules. Thompson et al. (2024) developed and evaluated the separation performances of primary and quaternary amine–based membranes, using polyvinylidene difluoride (PVDF) and polyether sulfone (PES) membranes as the platform, for removing short‐ and long‐chain perfluorinated carboxylic acids (i.e., PFBA and PFOA, respectively) (Thompson et al. 2024). The membranes exhibited water permeances ranging from 45 to 131 LMH/bar and achieved up to 90% rejection of PFOA and 50%–80% rejection of PFBA. Regenerated membranes retained their capture performance for three cycles of continuous operation. However, allylamine functionalization significantly reduced flux, not primarily due to fouling, but likely because of extensive amine loading, which obstructed active sites and reduced water permeance within the polymer matrix. Although modified membranes showed effective removal of PFBA and PFOA and maintained performance across multiple regeneration cycles, optimizing amine loading is critical to balance rejection efficiency and water permeance. In another study, Wan et al. (2024a) enhanced PDVF material by incorporating tertiary amine (TA) and quaternary amine (QA) to individually remove PFOA and PFOS (0.24 μM). QA‐modified membranes outperformed TA‐grafted membranes across all pH levels and PFAS types, particularly for PFOS, achieving removal rates of over 90%. The isoelectric point of PDVF‐g‐QA membrane shifted to pH 12.5, enabling a broad pH operating range and facilitating a pH swing adsorption–desorption process. The PVDF‐g‐QA membrane achieved 84.7% PFOA removal after three cycles of regeneration using the pH swing method (treatment capacity of 2930 L/m²). Furthermore, the adsorption capacities for PFOA and PFOS were 0.44 and 0.65 mmol/g, respectively, which can be attributed to a synergistic mechanism of electrostatic attraction and hydrophobic interactions. However, high ionic strength and divalent ions (e.g., Ca²⁺ forming Ca‐PFOA complexes) reduced PFOA removal by 12.4%, while negatively charged NOM decreased removal by 15%–20% through competitive adsorption. Latterly, Wan et al. (2024) developed a QA‐grafted adsorptive membrane with PVDF hydrophobic backbones to selectively concentrate PFAS among competing substances (Wan et al. 2024b). The membrane achieved a PFOA removal efficiency of 96.8% while selectively filtering out organic competitors. Pore adjustment, achieved by controlling PVDF and QA content, allowed selective removal of NOM, with larger molecules like HA excluded by size while PFOA was adsorbed. QA modification shifted the surface potential from negative to positive (i.e., −39.1 to 28.2 mV) under neutral pH conditions (pH 7.8), enabling electrostatic attraction, which, combined with hydrophobic interactions, resulted in a concentration factor of 18.5–19.1 over HA and an enrichment selectivity of 9.6–12.5, outperforming typical NF processes (i.e., 4–10). Also, the membrane maintained high water permeability of 34.6 LMH/bar and over 90% rejection performance for five reuse cycles, demonstrating its stability and effectiveness.

Graphene oxide (GO) and other organic materials have also been utilized to develop advanced membranes with broad applications in various filtration processes. Among these innovations, GO‐based/modified membranes have garnered significant attention for PFAS removal from wastewater. El Meragawi et al. (2020) developed GO‐nanofiltration membranes for removing PFAS from wastewater. These membranes achieved 74.3% efficiency for 50 ppm PFOA at a TMP of 1 bar and a permeate flow rate of 10 ± 2.1 LMH/bar, which was lower compared with standard NF membranes (Appleman et al. 2013; El Meragawi et al. 2020). The reduced performance was linked to the extended interlayer spacing of GO in aqueous environments, caused by water molecules clustering around its oxidized functional groups. This expanded spacing allowed water to pass through the membrane while blocking PFOA molecules via size exclusion. Similarly, Meragawi et al. (2020) achieved significant results with GO membranes modified with polyethyleneimine (PEI), attaining a water permeance of 15.9 ± 1.3 LMH/bar and a high PFOA removal efficiency of 96.5% (El Meragawi et al. 2020). By reducing interlayer spacing (< 7 Å) and enhancing surface hydrophilicity, the modified GO membranes maintained a retention rate of over 90% across a wide concentration range (100 ppb to 100 ppm). It demonstrated their enhanced functionality and reliability for PFAS removal. Moreover, Khorramdel et al. (2022) developed GO membranes modified with poly diallyl dimethyl ammonium chloride (PDDA) and polystyrene sulfonate (PSS), achieving an unprecedented PFOA rejection efficiency of 98.4% and an impressive water flux of 83.1 LMH/bar (Khorramdel et al. 2022). Compared with regular GO membranes, these modifications resulted in a 300% increase in water flux and a 70% improvement in rejection efficiency. Such improvements were attributed to the highly charged and electron‐rich GO nanosheets, which reduced the interlayer spacing in the modified membrane while simultaneously improving surface hydrophilicity. This dual effect facilitated PFOA retention and significantly enhanced water permeation. Further highlighting the versatility of GO‐based membranes, the PSU‐GO hollow fiber membrane developed by Zambianchi et al. (2022), proved to be seven times more efficient than the pristine PSU membrane in removing C3–C13 PFAS (0.5 μg/L each) from tap water, with a notable selectivity for short‐chain PFAS (Zambianchi et al. 2022). Both PSU and PSU‐GO modules showed higher removal rates for long‐chain PFAS (i.e., C8–C13). However, the PSU‐GO membrane exhibited superior performance for sulfonated PFAS compared with their carboxylate ones of the same chain length (e.g., C6: 99% for PFHxS vs. 79% for PFHpA; and C4: 35% for PFBS vs. 4% for PFPeA). The removal process was governed by a delicate balance of hydrophobic and electrostatic interactions, with hydrophilicity being slightly more dominant. Also, the PSU‐GO membrane surpassed the pristine PSU membrane in filtration capacity for other contaminants, including ciprofloxacin, Pb, Cu, and Cr(III), showcasing its multifunctional capabilities. Pervez, Jiang, Mahato, et al. (2024) employed surfactant modification to alter the surface chemistry of GO adsorbents, enhancing their affinity for PFAS capture. GO modified with cetyltrimethylammonium bromide (CTAB), a cationic surfactant, achieved nearly 100% removal of all tested PFAS within 1 h (e.g., PFBS and PFPeA), except for PFBA, which required 8 h to reach a 95% removal efficiency. Similarly, GO modified with cetyltrimethylammonium chloride (CTAC), a structurally similar surfactant, removed almost 100% of both short‐ and long‐chain PFAS within 1 h. Moreover, the effectiveness of the GO‐CTAC adsorbent (i.e., highly positive surface charge), except for PFBA, was not significantly impacted by variations in pH, the concentration of NOM, or ionic strength. The effectiveness of CTAC was attributed to its ability to reduce repulsion between PFAS molecules, enabling micelle formation at concentrations below the critical micelle concentration and enhancing PFAS adsorption via micelle aggregation. This innovative approach could potentially be applied to GO‐based membranes to further improve short‐chain PFAS capture efficiency. However, their mechanical integrity remains a critical factor for real‐world applications, especially when GO is combined with other nanomaterials. Tailoring physicochemical features (e.g., surface functional groups and interlayer spacing) can further enhance membrane performance, yet the intrinsic swelling behavior of GO can introduce additional permeation pathways and may reduce overall PFAS rejection (Wei et al. 2018). Consequently, continued innovation in membrane synthesis and structural design is essential to develop robust, high‐performing GO‐based membranes suitable for large‐scale PFAS remediation.

The achievement in adsorptive membranes demonstrates significant potential for efficient PFAS removal, utilizing surface modifications like amine grafting, hydrophilic coatings, and surfactant incorporation to enhance adsorption, size exclusion, and electrostatic interactions. The functionalized PVDF materials and GO‐based/modified membranes offer high rejection rates for both short‐ and long‐chain PFAS, with improved regeneration capabilities and selective removal even in challenging conditions (e.g., high ionic strength and NOM presence). However, challenges such as reduced flux, trade‐offs between adsorption efficiency and water permeability, and fabrication complexity remain. Future efforts should prioritize scalable designs, improved fouling resistance, and cost‐effective solutions to maximize their practical applicability for PFAS‐contaminated water treatment.

3.2. Mixed Matrix Membrane

Mixed matrix membranes, composed of both inorganic and organic components, offer a promising alternative to monolayer membranes for PFAS removal, addressing challenges such as fouling, low mechanical stability, and high capital costs (Barhoum et al. 2023; Zahmatkesh et al. 2022). Metal–organic frameworks (MOFs), known for their exceptional porosity, well‐defined pore structures, and compatibility with polymeric materials, have shown great potential in fabricating high‐performance TFN membranes (Zhao et al. 2021). Among them, zeolitic imidazolate frameworks (ZIFs) stand out as key members of the MOF family (Han et al. 2020). In a recent study, a high‐performance TFN membrane was fabricated using an interfacial polymerization (IP) approach, incorporating 5 wt% ZIF‐L (relative to the weight of ethylene imine polymer [PEI]) on a polysulfone (PSF) selective layer (Bi et al. 2024). The modified membrane exhibits a 2.3‐flod increase in water flux (up to 47.56 LMH/bar) compared with the pristine TFN membrane (20.46 LMH/bar), and achieved 98.47% removal for PFOS and 95.85% removal for PFOA. Also, the retention rates exceeded 80% for both PFHxA and PFHxS (i.e., short‐chain PFAS, C6), attributed to enhanced electrostatic interactions and hydrogen bonding. High PFAS concentration, ionic strength and cationic valence (i.e., Ca²⁺ and Mg²⁺) further improved PFAS retention. Moreover, experimental data revealed the membrane's effectiveness in removing PFAS at ambient and elevated concentrations in real‐world scenarios, with rejection rates of 89.50% for PFHxS in seawater and 86.77% in municipal water.

Another approach involves using amine‐functionalized boron nitride (BN(NH₂)) nanosheets into untreated poly piperazine amide (PPA) membranes, which enhances membrane surface charge and surface wettability. This modification achieved a remarkable 93% rejection rate for the short‐chain PFAS (i.e., potassium nonafluoro‐1‐butanesulfonate, C₄F₉SO₃K), along with a 1.04‐fold increase (53 L/m²·h) in permeation rate compared with unmodified PPA membranes (26 L/m²·h) (Abdikheibari et al. 2022). The potential of such nanomaterials can be further advanced by utilizing them into lamellar membranes, offering a promising direction for improved PFAS removal technologies. For example, Zhang et al. (2024) developed a lamellar MOF membrane by incorporating copper‐tetrakis (4‐carboxyphenyl) porphyrin (Cu‐TCPP) MOF nanosheets modified with polyamide (Zhang et al. 2024). This membrane effectively removed 11 types of PFAS (50 μg/L each), including five short‐chain and six long‐chain variants with molecular weights ranging from 214.0 to 514.1 Da. It demonstrated high water permeance (21.4 LMH/bar) and achieved an overall removal rate exceeding 90%, even rejecting 84.2% of the smallest short‐chain PFBA (214 Da). The integration of MOF nanosheets with polyamide segments enhanced the membrane adsorption capacity, imparted stronger surface negative charges, and improved hydrophilicity, which strengthened interactions with PFAS molecules. Furthermore, the polyamide modification created a denser membrane structure, improving its stability and antifouling performance by preventing foulants from penetrating the membrane. The stronger negative charge and increased hydrophilicity also repelled negatively charged organic foulants, further enhancing its filtration efficiency. Hence, stacking MOF nanosheets creates a lamellar structure, accordingly, providing enhanced water transport channels and abundant functional groups that facilitate favorable interactions with solutes and enhance PFAS rejection (Ge et al. 2024).

Covalent organic frameworks (COFs) have also gained interest in advanced membrane technologies. Incorporating thin TpPa‐SO₃H into membrane selective layer enhances its negative charge, strengthening the Donnan effect (Nguyen et al. 2024). The TpPa‐SO₃H membrane achieved high rejection rates of over 99% for PFOS and ~95% for PFOA (1.0 mg/L each), while maintaining high water permeance ranging from 19.9 to 37.5 LMH/bar. Notably, it allowed scale‐forming salts to pass, achieving PFAS, achieving a high salt/PFAS selectivity. The high rejection rates for PFBS (~90%) and PFBA (~90%) are primarily attributed to the Donnan exclusion mechanism, which underscores the importance of the highly negative charge imparted by the sulfonic acid groups in the TpPa‐SO3H membrane. These head groups create a highly electronegative environment generated by the distribution of negative charge across sulfur and oxygen atoms, resulting in effective electrostatic repulsion. Further functionalization of COFs with amine groups has shown significant potential for enhancing PFAS removal, particularly for GenX (Ji et al. 2018). A COF with 28% amine loading demonstrated the removal of approximately 90% of 12 out of 13 PFAS tested (i.e., 63% of PFBA, C4 PFCA). Notably, the amine‐loaded COF outperformed both GAC and PAC adsorbents, achieving 91% removal of GenX, compared with 0% and 72% removal by GAC and PAC, respectively. The superior adsorption performance of the COF was attributed to the combined effects of amine groups and hydrophobic interactions, which facilitated efficient PFAS adsorption onto the COF surface.

Advanced polymer‐based material is another promising alternative for PFAS removal, with recent studies demonstrating innovative approaches to enhance membrane performance and selectivity through tailored functionalization. For instance, Rittner et al. (2024) developed an asymmetric membrane using cobaltocenium‐containing block copolymers (BCPs), enabling the customization of surface properties and pore sizes to meet specific filtration requirements (Rittner et al. 2024). These cobaltocenium‐based polymers exhibited exceptional adsorption and desorption capabilities for PFAS removal, achieving PFOA retention rates of 99.3% at low water flux (48 ± 2 LMH/bar) and 96.6% at high water flux (171 ± 26 LMH/bar). Additionally, it demonstrated efficient ion‐capturing performance, achieving 46.4% for anionic pollutants and 99.8% for cationic metallic pollutants. Moreover, the selective solubility of the cobaltocenium BCPs allowed for effective recycling of the porous membranes, enabling the formation of new membranes from recovered materials. Furthermore, upcycling of used membranes through calcination under reductive or oxidative conditions produced cobalt‐containing ceramics with tunable compositions and well‐defined porous structures, offering a sustainable approach to material reuse and ceramic production. To develop solute‐selective, chemically stable, and cost‐effective hollow fiber NF membranes, Wang et al. (2015) employed dry‐jet wet spinning technology to fabricate a novel poly(m‐phenylene isophthalamide) (PMIA) NF membrane for PFOS removal (Wang et al. 2015). The PMIA membrane achieved enhanced PFOS rejection, improving from 91.17% to 97.49% as the pH increased from 3.2 to 9.5, owing to its isoelectric point at 3.54. Additionally, increasing Ca²⁺ concentrations from 0.1 to 2 mM further improved rejection from 97.10% to 99.40% at a transmembrane pressure of 4 × 10⁵ Pa, attributed to pore blockage and bridging interactions between Ca²⁺ and the sulfonate group of PFOS. Sorption/desorption experiments revealed that PFOS adsorption on the membrane surface was five times higher in the presence of Ca²⁺. The membrane's high performance was driven by a combination of steric hindrance and Donnan exclusion effect. Notably, the PMIA NF membrane demonstrated superior packing density and a larger membrane surface area compared with flat sheet membranes, rendering it economically viable for mass production. Still, short‐chain PFAS, particularly C4 compounds, require further investigation, as these (ultra)short‐chain PFAS are increasingly prevalent in aqueous environments.

Bio‐incorporated/based polymeric membranes are being explored as an environmentally friendly alternative for membrane fabrication, offering promising potential for PFAS removal. Some researchers developed a novel mixed‐matrix composite NF (MMCNF) membrane incorporating β‐cyclodextrin microparticles into a polyethersulfonate (PES) membrane, significantly enhancing its adsorption capacity, up to 1000 times that of the unmodified PES membrane (Chaudhary et al. 2023). This membrane achieved near‐complete removal of PFOA (> 99.9%), with adsorption contributing 30% of the removal. Higher β‐cyclodextrin loading (i.e., 8%) further improved removal efficiency, whereas the membrane without β‐cyclodextrin achieved only 73% PFOA removal. Additionally, adsorption delayed membrane saturation, reducing the frequency of regeneration needed. As proved, the removal performance equaled RO membranes, while offering the extra benefit of producing a lower salinity retentate, which reduces both energy consumption and brine treatment costs. Meanwhile, approximately 99% of the removal capacity was recovered after three regeneration cycles, ensuring the membrane's long lifetime.

Another study conducted by Jin, Peydayesh, Joerss, et al. (2021), they developed an amyloid fibril membrane and an amyloid–carbon hybrid membrane and tested their performance in removing 31 target PFAS (400 ng/L each), including long‐ and short‐chain PFCAs and PFSAs and their precursors (Jin, Peydayesh, Joerss, et al. 2021). The β‐lactoglobulin amyloid fibril membrane exhibited exceptional adsorption capabilities for long‐chain PFAS, due to their stronger hydrophobic interactions (more C–F bonds) with amyloid fibrils. The removal efficiency increased with chain length for both PFCAs and PFSAs, aligning with previous studies on other membrane types. When tested with PFAS‐contaminated water from Xiaoqing River basin, the membrane effectively removed both high (> μg/L) and trace (ng/L) levels of long‐chain PFAS, achieving 73.70% removal of PFOA (C8 PFCA) from 327 to 241 μg/L, 93.52% removal of HFPO‐TrA (C7 PFECA) from 33.93 to 2.20 μg/L, and 72.29% removal of PFNA (C9 PFCA) from 51.65 to 14.31 ng/L. However, the removal of short‐chain PFAS was significantly lower, such as PFBA (3.66%, C4 PFCA), PFPeA (4.14%, C5 PFCA), PFHxA (4.23%, C6 PFCA), PFHpA (20.62%, C7 PFCA), and HFPO‐DA (9.81%, C5 PFECA, GenX parent acid). Surprisingly, the amyloid–carbon hybrid membrane yielded highly promising results, eliminating both long‐ and short‐chain PFAS to levels within the detection limits. While trace levels of PFBA remained, the membrane still achieved a significant removal efficiency of 96.31% (from 102 to 3.76 ng/L). The removal mechanisms were attributed to hydrophobic interactions between the C–F bonds of PFAS and the hydrophobic amino groups of amyloid fibrils, as well as the enhanced adsorption capacity provided by the porous carbon. Furthermore, the hybrid membrane was reported to be 38% more environmentally sustainable compared with commercial NF membranes, while also offering cost‐effective performance.

In short, the development of mixed‐matrix membranes has shown substantial progress in PFAS removal. Innovative approaches, such as incorporating MOFs, COFs, and functional materials, have significantly improved filtration capacity, surface properties, and rejection rates. Further enhancements could be optimizing the embedded sorbents within the membranes, such as using nano‐sized particles. Despite these advancements, challenges remain, particularly in addressing the removal of (ultra)short‐chain PFAS and ensuring long‐term stability under real‐world conditions. Future research should also focus on refining material design, improving regeneration capabilities, and scaling up these technologies to deliver cost‐effective and environmentally sustainable solutions for PFAS‐contaminated aqueous environments. Bio‐incorporated/based polymeric membranes stand out as an excellent example.

3.3. Nanofiber and (Catalytic) Nanomaterial Membrane

Nanomaterials integrated with membrane technology offer unique advantages over bulk materials, significantly enhancing PFAS removal performance. Nanofiber membranes, produced through various spinning methods, feature a high surface‐to‐volume ratio and controlled porosity, enabling multifunctional separation mechanisms such as size exclusion, adsorption, and depth filtration. Moreover, given their high porosity, nanofiber membranes can operate at low pressure and are well‐suited for gravity‐driven filtration applications. Also, functionalization with coatings or fillers further enhances their potential for water and wastewater treatment applications (Liao et al. 2018; Sanaeepur et al. 2022).

Electrospun nanofiber membranes offer a novel approach for PFAS removal, leveraging an electrostatic attraction mechanism rather than conventional electrostatic repulsion, along with tailored surface properties to enhance adsorption efficiency (Pervez, Jiang, and Liang 2024). For example, Mantripragada, Obare, and Zhang (2023) evaluated the GenX removal efficiency of electrospun polyacrylonitrile nanofibrous membrane (ESPAN) and amidoxime‐functionalized ESPAN (ASFPAN) nanofiber membranes (Mantripragada, Obare, and Zhang 2023). Both membranes performed better at pH 4, with ASFPAN membrane outperforming ESPAN membrane by 88% at a 0.24 g/L loading and 100 mg/L GenX concentration, achieving a GenX removal efficiency of ~35% and a filtration capacity of ~0.6 mmol/g (more than doubled than that at pH 6). GenX adsorption onto the pristine PAN membrane was primarily driven by hydrophobic and dipole–dipole interactions between the C≡N groups of PAN and the C–F groups of GenX. In contrast, the ASFPAN membrane exhibited enhanced adsorption through Coulombic attraction between its positively charged C=N + (OH)–H groups and the negatively charged –COO⁻ groups of GenX. Additionally, the super hydrophilic surface of ASFPAN facilitated water penetration into the nanofibrous structure, further improving GenX removal efficiency. Later, Mantripragada, Obare, and Zhang (2023) developed an electrospun PAN/Algae bicomponent nanofibers membrane (ES (PAN/Algae)) and achieved a maximum GenX removal capacity of ~0.9 mmol/g at pH 6, with approximately 72% GenX removal under conditions of 100 mg/L GenX and 50 wt.% Chlorella with PAN (Mantripragada, Deng, and Zhang 2023). The removal mechanism was attributed to non‐Coulombic interactions, including hydrophobic interactions, dipole–dipole interactions, and hydrogen bonding. Hydrophobic GenX molecules were drawn toward the membrane surface, followed by dipole–dipole interactions between the C–F bonds in GenX and the C≡N groups of PAN, as well as between the C–F bonds in GenX and the C=O groups of algae. Additionally, hydrogen bonding occurred between the C=O groups of GenX and the N‐H/O‐H groups of algae, further facilitating GenX adsorption.

Electrospun nanofiber membranes also hold potential for the catalytic degradation and removal of PFAS. For example, an MOF‐derived Cu/Carbon nanofiber membrane was developed as a bifunctional catalyst for PFOA detection and degradation by regulating reactive Cu species (Hou et al. 2024). The CNF‐Cu/C‐800 membrane achieved an optimal PFOA degradation efficiency of 98.22% over 180 min. Increasing the carbonization temperature to 900°C slightly reduced the degradation efficiency to 87%, primarily attributed to a reduction in specific surface area and active sites caused by the aggregation of CuOx nanoparticles and carbon skeleton collapse. The membrane demonstrated excellent stability and reusability, retaining 87% removal efficiency after five cycles while maintaining its structural integrity, highlighting its cyclic degradation capacity and robustness as a photoelectric cathode. The PFOA degradation mechanism in the Cu‐SPEF system primarily involved hydroxyl radicals (·OH) and photocatalyzed holes (h_vb ⁺), with superoxide radicals (·O₂ ⁻) playing a supportive role. The catalytic cycle on the CNF‐Cu/C surface was sustained by the redox interplay between Cu species, where Cu(0) was oxidized to Cu(I) by acid corrosion and oxygen, and Cu(I) activated H₂O₂ to produce ·OH, subsequently regenerating Cu(II). Cu(II) was then reduced back to Cu(I) by H₂O₂ and electrons from the cathode. Under UV or solar irradiation, valence band electrons (e_vb ⁻) transitioned to the conduction band, generating h_vb ⁺ in the valence band. The h_vb ⁺ oxidized PFOA (C₇F₁₅COO⁻) into unstable radicals (C₇F₁₅COO· and C₇F₁₅·), which reacted with ·OH to form intermediates like C₇F₁₅OH, leading to stepwise defluorination and degradation into short‐chain perfluorocarboxylic acids, CO₂, and F⁻. Therefore, h_vb ⁺ and ·OH had a predominant influence on PFOA mineralization. Meanwhile, as reported by other researchers, incorporating g‐C₃N₄/Ag₃PO₄ (0.5 wt%) into polyethersulfone membranes enhances performance by strengthening hydrogen bonds and inducing surface charge, ultimately affecting the membrane's photocatalytic performance. Compared with neat PES membranes and carbon nanomaterial‐incorporated PES membranes, g‐C₃N₄/Ag₃PO₄‐modified membranes exhibit improved flux recovery rates (i.e., 57.5% vs. 39%), higher rejection efficiency, and superior electrostatic interactions. This material outperforms alternatives like TiO₂‐coated MWCNTs due to its enhanced photocatalytic and electrostatic properties, making it a promising choice for advanced membrane modifications (Gokulakrishnan et al. 2021). Another notable example was conducted by Zhang et al. (2020), who demonstrated superior PFOA degradation and defluorination efficiencies using In₂O₃‐based nanomaterials under ideal laboratory conditions. These nanomaterials outperformed other nanophotocatalysts, highlighting their potential for enhanced photocatalytic activity and effectiveness in PFAS removal (Zhang et al. 2020). However, the use of nanomaterials in membrane modification poses challenges such as high operational costs and risks of secondary contamination, highlighting the need for innovative designs to ensure efficient separation and safe post‐treatment handling.

Moreover, nanofiber membranes demonstrate significant advantages in gravity‐driven operations for PFAS removal, offering energy‐efficient solutions with high adsorption capacity and effective separation performance under minimal operational pressure. Wan et al. (2022) developed a cost‐effective electrospun PVDF membrane grafted with quaternary ammonium (QA) groups, achieving PFOA and PFOS removal efficiencies of 97.9 ± 1.4% and 99.1 ± 0.4%, respectively, comparable to commercial NF270 membranes (> 95%) (Wan et al. 2022). QA grafting enhanced electrostatic interactions and permeating flux, while PVDF improved hydrophobicity. An optimal QA dosage (50 wt% of PVDF) balanced hydrophilicity and hydrophobicity. PFOA removal was limited under alkaline conditions and high conductivity, while PFOS removal remained effective because of stronger hydrophobic interactions, consistent with electrostatic and hydrophobic removal mechanisms. The membrane exhibited remarkable regeneration capabilities, with 92.5% PFOA and 94.5% PFOS desorption using a 5% methanol solution at pH 11. Additionally, the gravity‐driven operation led to a low energy demand (213 J/m³ for PFOA and PFOS removal), significantly outperforming NF membranes such as NF270 (2.8 kWh/m³) and NF90 (0.18 kWh/m³) under comparable recovery rates and rejection efficiencies (Aedan et al. 2024). Moreover, Guo et al. (2021) developed a nanofibrous PVA/PDDA membrane for efficient PFAS removal with near‐zero energy consumption, achieving > 99% removal for PFOA and PFOS and > 97% for GenX (10 μg/L for each compound) (Guo et al. 2021). Its high water permeability (2700 LMH/bar) enabled rapid gravity‐driven filtration with a 5 cm water head, corresponding to an estimated energy consumption as low as 2.7 × 10⁻⁴ kWh/m³ (i.e., 972 J/m³). The membrane maintained high performance during mixed PFAS solutions, capturing ≥99% of PFOA, PFOS, and GenX at volumetric loadings ≤ 6000 L/m², although GenX capture decreased slightly at higher loadings due to its lower affinity for amine‐based functional groups. Notably, the membrane demonstrated a high GenX recovery ratio of 94% and a volumetric concentration factor of 40 over 12 cycles. These features, with its low energy demand, excellent capture ratios, and effective reusability, make this membrane a highly advanced and sustainable solution for PFAS‐contaminated water treatment. Thereby, nanofiber membranes provide a sustainable, efficient solution for PFAS removal with minimal energy consumption.

Another alternative method to enhance PFAS rejection in NF membranes is the incorporation of inorganic nanofillers into the membrane structure or surface. This strategy improves contaminant rejection while simultaneously increasing membrane permeability (Tawalbeh et al. 2023). MXenes are a versatile family of 2D transition metal carbides, carbonitrides, and nitrides with the general formula Mn + 1XnTx, where M represents a transition metal, X denotes carbon or nitrogen, n ranges from 1 to 4, and Tx refers to surface functional groups such as –OH, –O, and –F (Alhabeb et al. 2017). A MXene‐reinforced polyamide (PA) membrane, developed through interfacial polymerization, demonstrated high rejection rates for short‐chain PFAS (i.e., 96.85% for PFHxS and 93.35% for PFHxA) and improved water permeance (from 8.63 to 12.15 LMH/bar) (Ma et al. 2022). These improvements were achieved by MXene functionalization, particularly its hydrophilic functional groups, increasing the membrane's morphology and surface charge, and boosting PFAS rejection and water permeability. Additionally, the increased surface electronegativity due to negatively charged MXene facilitated the separation of negatively charged PFAS molecules. Thus, the MXene‐regulated interfacial polymerization offers a promising pathway to optimize PA membranes, addressing the trade‐off between rejection and permeability while effectively tackling the challenge of short‐chain PFAS removal. Similarly, incorporating MXene into a hollow fiber membrane yielded similar results. With 0.025‐wt% MXene loading, water permeability increased from 13.19 to 29.26 LMH/bar while PFOS rejection improved from 72% to 96% (Le et al. 2022).

These findings highlight the advancements in novel membranes for efficient long‐ and short‐chain PFAS removal, emphasizing mechanisms like size exclusion and enhanced surface electronegativity. Improvements in adsorption properties, charge density, and pore optimization have enhanced PFAS selectivity and water permeance, with nanofiber and nanocomposite membranes offering innovative, sustainable alternatives. Despite progress, further research is essential to optimize costs, scalability, and transition these membranes to industrial applications. Furthermore, the potential of integrating machine learning and molecular simulations to unravel key factors influencing PFAS removal by polyamide NF and RO membranes, providing valuable insights for optimizing membrane design and performance (Jeong et al. 2024).

4. Insights Into Membrane Removal Mechanism

Membrane separation is a filtration technique that uses a selective permeable membrane to isolate solutes from a solution. In this process, a pressure gradient drives the solvent through the membrane pores (permeation), while larger solutes and impurities are retained as retentate. The effectiveness of membrane separation in removing PFAS primarily depends on three mechanisms: size exclusion, electrostatic interaction, and solute–membrane affinity (Mastropietro et al. 2021; Osorio et al. 2022). Size exclusion, governed by molecular weight cut‐off (MWCO), blocks molecules based on size, with a lower MWCO membrane achieving higher PFAS rejection rates (Li et al. 2021; Xiong et al. 2021). When the membrane pore size is smaller than the PFAS molecule, size exclusion dominates the removal process. In RO membranes, with pore sizes typically ranging from 0.1 to 1 nm, size exclusion is the primary mechanism for removing PFAS molecules (Mastropietro et al. 2021). Similarly, in NF membrane, size exclusion can account for up to 80% of the total removal efficiency, particularly for PFAS with carbon chain lengths greater than four. For these compounds, the smaller pore size (e.g., 0.36 nm) ensures that size exclusion remains the dominant removal mechanism. Moreover, ions such as Ca²⁺, Na⁺, Mg²⁺, and Fe³⁺ in the feed water can facilitate PFAS to form complexes, which have bigger sizes and can be more effectively removed. Additionally, organic and colloidal fouling during NF/RO filtration can partially block membrane pores, reduce the effective pore diameter and further enhance PFAS rejection (Wu et al. 2020). Conversely, fouling can also lead to foulant‐enhanced concentration polarization, thereby decreasing pollutant rejection (Wang et al. 2015).

Electrostatic interaction between charged solutes and the membrane surface also plays a crucial role. Negatively charged membranes enhance the separation of anionic PFAS through electrostatic repulsion. Generally, a greater negative charge causes greater electrostatic repulsion. At neutral pH, PFAS molecules dissociate into negatively charged ions, which are repelled by the membrane's negative surface charge, leading to improved rejection rates. Moreover, coexisting cations (such as Ca²⁺ or Na⁺) can alter the charge‐shielding effect by associating with PFAS molecules and neutralizing their charges, ultimately increasing their affinity for the membrane surface (Pervez, Jiang, and Liang 2024). As reported, the presence of 2‐mM Ca²⁺ increased PFOS removal by fivefold and simultaneously raised average membrane roughness by 2.5‐fold (from 3.67 to 9.79 nm) owing to charge neutralization (Wang et al. 2015). Furthermore, Kwon et al. (2012) indicated that PFAS with longer carbon chains are less susceptible to the charge‐shielding effect (Kwon et al. 2012). Thereby, rejection of short‐chain PFAS by NF membranes may rely more on electrostatic repulsion than on size exclusion (Zhi et al. 2025). However, when the electrostatic repulsion force is absent (e.g., at the isoelectric point of the membrane) or when the membrane pore size exceeds the size of the PFAS molecules, other mechanisms such as adsorption become more influential.

Adsorption occurs when solute–membrane interactions dominate the removal process, such as hydrophobic interactions, hydrogen bonding, or π–π interactions. Factors influencing adsorption capacity include surface area, pore size, pore density and distribution, surface charge density, and hydrophilicity of the active layer (Li et al. 2021). Nanomaterials such as MOFs, graphene oxide, and MXene can be incorporated into adsorptive membranes (i.e., combining functions of adsorption and filtration) to significantly boost PFAS removal, owing to their precisely tunable chemical and physical properties. It offers several advantages over conventional membranes, including enhanced contaminant‐ion retention, reduced energy requirements, and increased permeate flux (Huang and Cheng 2020).

Hence, the overall efficiency of PFAS removal depends on the interplay among these mechanisms and is influenced by factors such as membrane characteristics (e.g., MWCO, pore size, and surface charge), and solubility within the membrane matrix (Figure 3). Understanding these interactions is crucial for optimizing membrane technologies to achieve higher PFAS rejection rates across varying water treatment conditions. Although numerous studies have evaluated PFAS removal using commercial NF and RO membranes, further investigation is required to identify the most efficient removal mechanism by incorporating its techno‐economic analysis. A deeper understanding of this process could enable the development of membrane optimization to enhance PFAS removal, especially (ultra)short‐chain species, according to the targeted mechanism. The influence of membrane characteristics and water matrices will be discussed in Section 5.

FIGURE 3.

Schematic diagram of PFAS removal mechanism by high‐pressure membrane.

5. Parameters Controlling PFAS Rejection by Membrane System

The efficiency of PFAS removal by membrane filtration is influenced by various parameters, critical for optimizing performance. Membrane material, surface characteristics, and interaction mechanisms with PFAS molecules play pivotal roles, as seen in Table 3. Hydrophilicity or hydrophobicity affects water affinity, while pH impacts electrostatic interactions between PFAS and membranes. Initial PFAS concentrations, often enhancing rejection through adsorption or pore blockage, also affect removal efficiency. Coexisting ions, such as anions or cations, influence performance via electrostatic interactions or ion‐bridging effects. Understanding these parameters enables better membrane design and operation for effective PFAS mitigation (Table 4 and Figure 4). We will discuss various aspects related to membrane technology in the following section.

TABLE 3.

Effect of membrane characteristics on PFAS removal.

Membrane material	Membrane module	Surface charge	Surface wettability	PFAS (concentration)	pH	Other ions/additives	Removal efficiency	Reference
Polyacrylonitrile ultrafiltration membrane (PAN‐Q)	NF	Positively	Hydrophilic	PFOS (0.1 mg/L)	3–12	NOM (HA, FA, and BSA) Surfactants (SDS and CTAB)	> 98% (pH 3)	(Fang et al. 2024)
							> 96% (in presence of NOM)
Polyether sulfonate (PES) ultrafiltration membrane	NF	Negatively	Hydrophilic	PFOS (2 ppm)	2–10	Na+, Mg2+, and Ca2+	94%	(Kasula et al. 2024)
PVDF‐g‐QA &PVDF‐g‐TA membrane	NF	Positively	Hydrophilic	PFOA (100 μg/L)	2–14	HA	> 90%	(Wan et al. 2024a)
PVDF‐QA membrane	NF	Positively	Hydrophilic	PFOA (100 μg/L)	3–8	HA	96.8%	(Wan et al. 2024b)
Primary and quaternary amine functionalized PVDF, PES membrane	MF	Positively	Hydrophilic	PFOA and PFBA (150 ppb)	3–12	HA, Na+, and Cl−	90% PFOA	(Thompson et al. 2024)
							50%–80% PFBA
β‐Cyclodextrin adsorbed PES membrane	NF	Negatively	Hydrophilic	PFOA (161–345 μg/L)	3–10	—	> 99.9% PFOA	(Chaudhary et al. 2023)
Polysulfone–graphene oxide hollow fiber membranes (PSU‐GO HFs)	UF	Positively	Hydrophilic	PFBA, PFPeA, PFBs, PFHxA, PFHpA, PFHxS, PFOA, PFNA, PFOS, PFDA, PFUnDA, PFDoDA, PFTrDA, PFTeDA (0.5 μg/L)	7.5	As(V), Cd, Cr(III), Cu, Ni, Pb, U, and V	99% PFHxS	(Zambianchi et al. 2022)
							79% PFHpA
							35% PFBS
							4% PFPeA
NF‐PNIPAm‐PVDF membranes	NF	Positively	Hydrophilic	PFOA, GenX, and PFBA (70 μg/L)	2–9	Na+	70% PFOA and GenX	(Léniz‐Pizarro et al. 2022)
							50% PFBA
β‐Lactoglobulin amyloid fibril membrane	N.A.	Positively	Hydrophobic	31 PFASs (400 ng/L each)	2–7	—	High efficacy in removing high (> μg/L) and trace (ng/L) levels of PFASs	(Jin, Peydayesh, Joerss, et al. 2021)
ZIF‐L incorporated TFN membrane	NF	Positively	Hydrophilic	PFBA, PFBS, PFHxA, PFHxS, PFOA, PFOS, and PFNA (200 μg/L)	—	K+, Na+, Ca2+, Mg2+, and Cl−	> 97% PFOA and PFOS	(Bi et al. 2024)
							80% short‐chain PFAS removal
Hyaluronic acid interlayered TFC membrane	NF	Negatively	Hydrophilic	PFBA, PFBS, PFHxA, PFHxS, PFOA, and PFOS (100 μg/L)	2–10	HA, SA, Ca2+, Mg2+, and Cl−	> 99% PFHxA rejection	(Chen et al. 2023)
							> 96% PFHxS rejection
MXene interlayered TFC membrane	NF	Negatively	Hydrophilic	PFHxA and PFHxS (10–200 μg/L)	3–12	HA, Ca2+, and Cl−	> 97% PFHxA rejection	(Ma et al. 2022)
							> 99% PFHxS rejection
MXene/CNT membrane	NF	Negatively	Hydrophilic	PFOA (10–200 μg/L)	3–11	Ca2+ and HA	> 90% PFOA retention	(Xu et al. 2022)
MXene‐PA TFN membrane	NF	Negatively	Hydrophilic	PFOS (2 mg/L)	4–10	Mg2+	96% PFOS retention	(Le et al. 2022)
Amine functionalized‐boron nitride BN (NH2) nanosheet–decorated membranes	NF	Negatively	Hydrophilic	Potassium nonafluoro‐1‐butanesulfonate (C4F9 SO3K) (338.19 g/mol)	2–9	—	93% rejection of target PFAS	(Abdikheibari et al. 2022)
PA modified MOF membrane	NF	Negatively	Hydrophilic	Five short‐chain, six long‐chain PFAS (50 μg/L each)	3–10	Na+, Mg2+, Cl−, and SO4 2−	Over 84% rejection of short‐chain PFAS	(Zhang et al. 2024)
Piperazine (PIP) and Bipiperidine (BP) membrane with PA selective layer	NF	Negatively	—	PFOA (1 mg/L)	7–10	Na+, Ca2+, and SO4 2−	90% PFOA retention	(Boo et al. 2018)
Asymmetric membranes based on cobaltocenium‐containing block copolymers (BCP)	—	Positively	Hydrophobic	PFOA (2.556 mg/L)	—	Na+, Pb2+, CrO4 2−, and NO3 −	96.6%–99.3% PFOA retention	(Rittner et al. 2024)
Anionic COF membrane	NF	Negatively	—	PFOS, PFOA, PFBS, and PFBA (1 mg/L each)	2–12	Na+, Ca2+, Mg2+, and Cl−	99% rejection of PFOS	(Nguyen et al. 2024)
							90%–95% rejection of PFOA and short‐chain PFAS
OMTT‐PMIA membrane	NF	Negatively	Hydrophilic	PFOS (100 μg/L)	7	Ca2+, Pb2+, Cl−, SO4 2−, and PO4 3−	(72%–92%) rejection at different pressure & concentration of PFOS (83%–95%) rejection of PFOS at different pressure in the presence of Pb2+	(Luo et al. 2016)
PAN (ESPAN) and amidoxime surface functionalized ESPAN (ASFPAN) nanofibrous membrane	—	ESPAN (negatively)	PAN(ESPAN)—hydrophobic ESPAN (ASFPAN)—hydrophilic	GenX (100 mg/L)	4–6	—	35% GenX removal by ASFPAN membrane at pH 4	(Mantripragada, Obare, and Zhang 2023)
		ASFPAN (positively)
PAN/Algae bicomponent nanofibrous membrane (ES (PAN/Algae))	—	Negatively	Hydrophobic	GenX (100 mg/L)	4–6	—	72% GenX removal at pH 6	(Mantripragada, Deng, and Zhang 2023)
PVDF‐g‐QA membrane	—	Positively	Hydrophobic	PFOS and PFOA (0.12 μM)	3–12	HA, FA, CTAB, and SDS	> 95% rejection	(Wan et al. 2022)
MWCNT‐ENFMs membrane	—	Negatively	Hydrophobic	PFOS (100 μg/L)	2–10	—	Up to 90% removal efficiency	(Dai et al. 2013)
PDDA‐GO membrane and PSS‐GO membrane	—	PSS‐GO (negatively)	Hydrophilic	PFOA (1–100 ppm)	2–11	—	98.4% rejection for 50 ppm PFOA at 0.45 MPa	(Khorramdel et al. 2022)
		PDDA‐DO (positively)
GO‐PEI	NF	Negatively	Hydrophilic	PFOA (50–200 ppm)	4–10	—	96.5% PFOA rejection at 50 ppm	(El Meragawi et al. 2020)
PMIA hollow fiber membrane	NF	Negatively	Hydrophilic	PFOS (50–500 μg/L)	3–10	Na+, Mg2+, SO4 2−, and Cl−	> 99% PFOS rejection	(Wang et al. 2015)

Abbreviations: BSA, bovine serum albumin; CTAB, cetyltrimethylammonium bromide; FA, fulvic acid; HA, humic acid; N.A., not available; NOM, natural organic matter; SA, sodium alginate; SDS, sodium dodecyl sulfonate.

TABLE 4.

Factors affecting PFAS removal by membrane filtration.

Factors	Effects explained
Effect of membrane surface properties	Simplicity and cost: Organic NF membranes are generally preferred over inorganic ones for their easier fabrication, reliability, and lower cost (Zahmatkesh et al. 2024). Surface properties: Membrane hydrophilicity or hydrophobicity is critical: hydrophobic membranes attract bacteria, proteins, and suspended solids more readily, while hydrophilic membranes resist these foulants and maintain higher retention‐permeation efficiencies (Johnson et al. 2022; Rana and Matsuura 2010). Isoelectric point (IEP): Near the IEP, surface charges approach neutrality, weakening electrostatic repulsion and lowering PFAS retention.
Effect of membrane surface potential	Membrane zeta potential: The zeta potential of a membrane is governed by its surface functional groups (e.g., carboxyl and amine groups) and ionic strength of the solution. Effect of ionic strength: Increasing ionic strength compresses electrostatic double layer, thereby reducing the effective surface charge and weakening electrostatic repulsion between similarly charged PFAS and the membrane (Jin, Peydayesh, and Mezzenga 2021), which results in lower rejection efficiency. Negatively charged membranes: Membranes with higher negative surface charges demonstrated stronger repulsion against anionic PFAS, leading to higher removal rates under low ionic strength conditions. Weakened interactions: As ionic strength increases, the electrostatic forces diminish, underscoring the need to maintain adequate surface charge for effective PFAS removal.
Effect of PFAS chain length	Hydrophobicity and chain length: Longer PFAS chains increase hydrophobic interactions with membranes, leading to higher rejection compared with short‐chain PFAS, which are more soluble and interact less with membrane surfaces. Filtration‐Adsorption effects: Hydrophobic membranes perform better by combining size exclusion and adsorption effects. Size exclusion: As chain length increases, the size exclusion becomes more dominant, reducing the rejection difference between neutral solutes and PFAS (Zhang et al. 2024). Short‐chain PFAS challenges: Short‐chain PFAS, with lower hydrophobicity, are difficult to capture, necessitating the development of advanced membrane materials to target these compounds effectively.
Effect of initial PFAS concentration	Influence on flux and rejection: Higher concentrations enhance mass transfer to the membrane surface, improving rejection through filter cake formation, pore clogging, or micelle formation (above the critical micelle concentration). Mechanisms during filtration: Thicker PFAS layers on the membrane surface enhance the sieving effect, while adsorbed PFAS molecules reduce pore size, increasing rejection efficiency. However, excessive concentrations can saturate the membrane, reduce water flux, and lower rejection efficiency over time (El Meragawi et al. 2020). Industrial vs. typical PFAS streams: High PFAS concentrations in industrial waste can be beneficial for removal, but also heighten fouling risks and necessitate frequent regeneration. In typical wastewater or groundwater, achieving critical micelle concentration is unlikely. Operational trade‐offs: While higher PFAS concentrations may enhance retention, they simultaneously lower permeate flux and increase both fouling and operational costs.
Effect of solution pH	Electrostatic interactions: Under high pH levels, both PFAS (negatively charged) and the membrane surface (often negatively charged) exhibit enhanced repulsion, leading to improved rejection (e.g., MXene‐PA membranes achieved higher rejection at pH 10 (Ma et al. 2022)). Low pH conditions: Protonated functional groups impart a positive charge to the membrane, promoting electrostatic attraction and increased PFAS adsorption but reducing overall rejection efficiency. For example, polyamide membranes (e.g., NF 270) benefit from deprotonated carboxyl groups at higher pH levels, enhancing rejection, while polyethersulfonate membranes with sulfonate (‐SO3H) groups maintain stable performance across various pH (Zeng, Tanaka, Suzuki, and Fujii 2017). Design implications: Tailoring membrane functional groups to specific pH conditions is critical for achieving optimal PFAS separation.
Effect of coexisting ions	Ion bridging and complex formation: Multivalent cations (e.g., Ca2+, Mg2+, and Fe3+) and anions (e.g., PO4 3− and SO4 2−) can form larger complexes with PFAS molecules, enhancing sieving effect and thereby improving rejection efficiency (Luo et al. 2016; Zhao et al. 2018; Zhao et al. 2020). Specific ion effects: PFOS rejection rates improved from 94% to 99.5% with increased Ca2+ concentrations (Zhao et al. 2016; Zhao et al. 2013), while PO4 3− showed better performance than SO4 2− and Cl− (Luo et al. 2016). Membrane selection: In some cases (e.g., NF270), the presence of salts (i.e., CaSO4, NaCl, and NaHCO3) may slightly reduce PFHxA rejection. By contrast, NF90 or XLE membranes show minimal sensitivity to these salts (Soriano et al. 2019).
Effect of organic matter	Effect on adsorption and fouling: Organic matter (e.g., humic acid, HA; and sodium alginate, SA), competes for adsorption sites, alters membrane surface characteristics, and promotes fouling. Flux reduction and fouling impact: While HA can marginally improve PFAS rejection (e.g., from 94.1% to 95.1% at 0.4 MPa; Wan et al. 2022; Zhao et al. 2020), it forms dense fouling layers that reduce water flux by up to 61.7% and 58.28% for PFHxA and PFHxS, respectively (Zhao et al. 2018). Additionally, fouling can enhance the sieving effect for short‐chain PFAS but often reduces removal efficiency for long‐chain PFAS by increasing hydrophilicity. Competitive adsorption: NOM decreases removal efficiency (e.g., a 20% drop in PFOA removal with PVDF‐g‐QA membranes) by competing with PFAS (Wan et al. 2024a). Mitigation strategy: Combining various mechanisms, such as electrostatic, hydrophobic, and dipole–dipole interactions, alongside controlled membrane pore sizes, can selectively reject HA and reduce NOM competition (Fang et al. 2024).

FIGURE 4.

Key factors governing PFAS rejection regarding membrane characteristics and water parameters.

6. How to Improve (Ultra)short‐Chain PFAS Removal?

Membrane filtration can effectively retain a wide range of PFAS, including both long‐ and short‐chain ones. Denser membranes (i.e., RO) exhibit stronger retention capabilities, and short‐chain PFAS typically require higher operating pressures and membranes with smaller pore sizes compared with long‐chain PFAS. The primary approach involved modifying membrane structures to enhance PFAS removal through electrostatic interactions and/or size exclusion mechanisms, as well as incorporating other removal mechanisms (e.g., electrochemical approach).

Tailoring the surface charge of NF is a promising approach to enhance the electrostatic exclusion of (ultra)short‐chain PFAS. Zhi et al. (2025) developed layer‐by‐layer assembled NF membranes with polycation (PDADMAC) and polyanion (PSS) coating solutions. The resulting membrane (i.e., (PDADMAC/PSS)₃) exhibited a negative zeta potential (−76.31 to −49.11 mV) and significantly improved short‐chain PFAS rejection through enhanced electrostatic repulsion (major) and adsorption (minor). It demonstrated the highest removal (86.1%–98.1%) for short‐chain PFAS, including PFBA to PFHpA (C4–C6), PFBS (C4), PFMOPrA (C3), PFMOBA (C4), and GenX (C5, HFPO‐DA), while removing > 99.9% of long‐chain PFAS (≥ C7).

Recent advancements in membrane design have focused on enhancing chemical interactions within membrane pores to improve the removal of short‐chain PFAS. For example, Mahofa et al. (2025) developed a β‐cyclodextrin‐modified graphene oxide (GO‐βCD) membrane featuring asymmetrical nanochannels to enhance the removal of short‐chain PFAS. Through host‐guest interactions, the embedded βCD sites significantly increased intrapore energy barriers and binding affinity. The membrane achieved > 90% of simultaneous rejection of PFBA, PFPeA, and PFHxA, outperforming polyamide membranes, while maintaining high water permeance (22 LMH/bar). Molecular dynamics and transition state theory confirmed that PFAS transport was hindered predominantly by chemical affinity rather than size exclusion or electrostatics, providing a robust strategy for short‐chain PFAS separation.

Furthermore, integrating membrane technology with electrochemical treatment is another promising approach for removing (ultra)short‐chain PFAS. Kim, Elbert, et al. (2024) investigated a redox‐polymer electrodialysis system combining a water‐soluble redox terpolymer with affordable NF membranes (1000 Da MWCO) and activated‐carbon electrodes to drive PFAS ion migration (≤ C4, TFA, PFPrA, and PFBA) and electrosorption (≥ C6, PFHxA and PFOA) in a single step. This approach achieved 86%–98% removal of (ultra)short‐ to long‐chain PFAS alongside continuous desalination to potable‐quality conductivity, even in high‐salt or wastewater matrices. It is also coupled with electrochemical oxidation to mineralize retained PFAS (i.e., 76%–100% defluorination). This study highlighted the viability of integrating the redox‐polymer electrodialysis for the removal of ultrashort‐chain PFAS. Moreover, Liang et al. (2025) reported a highly stabilized interfacial engineering strategy on modifying a reactive electrochemical membrane (REM). They fabricated a N‐doped graphene oxide nanosheet/CeO2@Ti₄O₇ REM, significantly enhancing the interfacial charge transfer and electrode stability. It achieved nearly complete degradation on five ether‐PFAAs (C3–C8) after 250 mins operation, except for C3 (approximately 85%). They also observed the transformation of chloride ions into toxic chlorinated byproducts (e.g., chlorate and perchlorate), which poses a significant concern for this electro‐oxidation process. However, regarding PFCAs degradation (C2–C9), only a slight accumulation of TFA (CF₃COOH, common degradation intermediates) was found.

Previous research on PFAS removal has primarily focused on long‐chain compounds such as PFOA and PFOS. However, emerging ultrashort‐ and short‐chain PFAS, progressively used in industrial and consumer products, also contribute to PFAS contamination in water systems and pose significant challenges. Given the limited studies addressing these compounds, further research is essential to develop advanced membrane technologies capable of selectively removing or simultaneously degrading emerging PFAS from water (Table 5).

TABLE 5.

Summary of pros and cons of membranes for PFAS removal.

Membrane type	Pros	Cons
MF/UF	Suitable for pretreatment	Ineffective for PFAS removal
NF/RO	Effective for PFAS removal Commercial products Pilot‐scale validation	Complex operation and maintenance Energy‐intensive Concentrate stream
Functionalized/modified membrane	High selectivity Fouling control Multimechanisms	Scalability concerns Manufacturing cost Extra chemical waste from production Limited studies on (ultra)short‐chain PFAS removal By‐product formation

7. Future Perspectives

Technical and economic assessments of PFAS removal are strongly affected by regulatory guidelines set by governing authorities. Generally, using NF for PFAS‐containing wastewater costs $0.016–$0.16/m³, while achieving similar removal in drinking water is around $0.11/m³ (Das and Ronen 2022; Franke et al. 2021). Notably, to overcome financial challenges, advancements in membrane design and material development are paramount. Novel designs, such as the incorporation of nanomaterials or adsorptive fillers into polymeric matrices, have demonstrated enhanced PFAS removal efficiency by leveraging adsorption, degradation, and filtration mechanisms. These innovative approaches offer the potential for membranes with higher selectivity, improved flux, reduced fouling propensity, and the ability to operate at lower pressures (i.e., gravity‐driven membrane). It is important to note that machine learning has also been applied to membrane design, effectively integrating AI‐driven fabrication strategies with experimental validation and demonstrating the feasibility of AI‐guided approaches (Yin et al. 2024). Also, the use of renewable energy sources, such as solar power, offers a promising avenue for supporting high‐pressure membrane systems, thereby improving the environmental sustainability, feasibility, and cost‐effectiveness of PFAS remediation, particularly in energy‐limited regions (Shalaby, Kabeel, et al. 2022; Shalaby, Sharshir, et al. 2022). Additionally, the substantial costs associated with manufacturing, installation, operation, maintenance, and concentrating treatment are major constraints in this field. Minimizing the volume of concentrated waste generated during membrane filtration processes can significantly reduce associated treatment costs. Emerging technologies, such as closed‐circuit membrane filtration and foam fractionation, have shown promise in achieving volume reduction (Safulko et al. 2023). Meanwhile, membrane concentrates can be further polished using adsorptive materials (e.g., anion exchange resins or granular activated carbon) to achieve discharge standards (4–90 ng/L total PFAS) at an estimated cost of $0.86–$3.34/m³ (Chen et al. 2024; Franke et al. 2021). However, further research is required to optimize these approaches and evaluate their feasibility for large‐scale applications.

Developing membranes with optimized structural and surface properties tailored for short‐chain PFAS removal is also crucial, since regulatory standards are becoming increasingly stringent. Fine‐tuning parameters such as pore size, surface charge, polymer composition, and operational conditions is essential for achieving optimal performance. Incorporating advanced or hybrid materials (e.g., adsorptive nanoparticles or positively charged groups) into the membrane matrix further boosts removal efficiency, particularly for emerging short‐chain PFAS. Furthermore, by using machine learning (Dangayach et al. 2024) and molecular modeling (de Souza and Meegoda 2024) to predict PFAS‐membrane interactions, next‐generation membranes can tackle emerging challenges and enable sustainable, large‐scale water treatment, and adopting interdisciplinary approaches.

Adsorptive membranes engineered with tailored functional moieties and optimized pore dimensions achieved high removal of long‐chain PFAS despite NOM competition, highlighting the promise of hybrid “interceptive‐adsorptive” mechanisms that exclude interferents while selectively capturing target PFAS. Additionally, short flow‐through residence times and limited adsorbent loading can restrict overall uptake, especially for (ultra)short‐chain PFAS, posing a challenge for adsorptive membrane performance.

Nanomaterials embedded in membranes can degrade, aggregate, or leach out over time, diminishing PFAS removal efficiency. Moreover, the potential toxicity of some nanoparticles poses environmental and health risks if they are released or improperly disposed of. Ensuring consistent and proper nanoparticle alignment within the membrane matrix (Goh et al. 2020), especially at a larger scale, remains a considerable practical challenge.

Functionalized GO‐ and MXene‐based membranes overcome the typical permeability‐rejection trade‐off, yet issues like interlayer delamination and swelling concerns require further optimized fabrication to enhance durability. Moreover, MXenes' inherent properties (e.g., thermal and chemical robustness, high surface area, mechanical flexibility, photothermal responsiveness, tunable band gap, and excellent conductivity) offer numerous potentials for integration with other destructive methods (Xu et al. 2020). However, significant challenges remain, such as high material and production costs, accelerated oxidative degradation, scale‐up feasibility, and the risk of secondary pollutants.

8. Conclusion

High‐pressure membranes (i.e., RO and NF) have shown promising results in addressing PFAS contamination. However, despite their demonstrated effectiveness, these technologies are yet to consistently achieve PFAS levels below the stringent thresholds established for drinking water. Persistent challenges, such as fouling control, high operational costs, and limited long‐term reliability, impede the large‐scale application of these technologies. Furthermore, the complexity of the water matrix, including PFAS concentration, coexisting anions, and NOM, critically impacts the performance and efficiency of membrane systems. Hence, addressing these issues will require ongoing research into cost‐effective treatment solutions for unlocking the full potential of high‐pressure membranes to consistently meet stringent PFAS thresholds nowadays. The development of membranes with optimized structural and surface properties, supported by interdisciplinary approaches, is also critical to overcoming these challenges and achieving reliable, large‐scale PFAS remediation in both water and wastewater treatment.

Author Contributions

Razia Hasan: formal analysis, investigation, visualization, validation, writing – original draft. Jianfei Chen: formal analysis, investigation, visualization, validation, writing – original draft, writing – review and editing. Parnian Mojahednia: validation, visualization, writing – review and editing. Seyed Hesam‐Aldin Samaei: validation, writing – review and editing. Jinkai Xue: conceptualization, funding acquisition, investigation, methodology, project administration, resources, supervision, writing – original draft, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Hasan, R. , Chen J., Mojahednia P., Samaei S.-A., and Xue J.. 2025. “Comparative Analysis of Commercial and Novel High‐Pressure Membranes for Perfluoroalkyl and Polyfluoroalkyl Substances (PFAS) Removal.” Water Environment Research 97, no. 8: e70157. 10.1002/wer.70157.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.