Distribution, advanced separation and degradation technologies for per- and polyfluoroalkyl substances removal in water: Efficiency, challenges, and future perspectives

3.3.1. Surface water

PFASs are highly persistent and mobile from one environmental media to another, such as surface water, groundwater, soil, air and biota. Their fate in aquatic environments is controlled by physicochemical, hydrological and geomorphological processes which influence their mobility and spatio-temporal distribution. A summary of PFAS incidence in surface water, groundwater, sediments and atmosphere is presented in Table 1 [17-27] that summarizes PFAS concentrations, sources and other parameters globally. Studies on surface water in China showed that industrial and domestic sources account for the majority of PFAS pollution. Extensive PFAS pollution was identified in the Jiulong River-Xiamen Bay catchment with all 25 principal PFAS compounds recorded at 56 sampling locations [17]. Similarly, extensive PFAS pollution was identified in the Jiulong River-Xiamen Bay catchment with all 25 principal PFAS compounds recorded at 56 sampling locations [17]. A study by Liu et al. [18] found several perfluorohexanoic acid (PFHxA) and PFOA hotspots among 28 samples acquired when water was diverted from the Yangtze River to Taihu Lake, where 10 of the 18 targeted PFAS compounds were present. Total perfluoroalkyl acid (∑PFAA) levels ranged from 117.77 to 543.3 ng L⁻¹ during diversion conditions in January and they remained lower ranging from 19.13 to 231.35 ng L⁻¹ under extreme summer flooding conditions in early 2020. The PFSAs were also found to have increased contributions downstream of entering tributaries and in the headwater region, while elevated PFAS contamination within the lake indicated water diversion projects as a dominant PFAS source in the area [18]. Short-chain PFASs were predominant, while emerging compounds including F53B were found in 90% of samples, and paper/packaging industry, machinery manufacturer, WWTP discharge and airport/port activities were identified as major sources.

WWTPs are specifically recognized as major point source of PFAS in surface waters. A regional study throughout Southern and Central California found PFOA, PFOS, PFBS, and additional PFAS in effluent of 16 WWTPs headed into downstream waters, some compounds exceeding from experimentally derived lowest observed effect levels [19]. A fate and transport model was used to estimate the daily PFAS loads which enter these WWTPs. The analysis showed that population size and urban land use were the main drivers of PFAS input. Model-predicted sum of selected PFAS compounds annual influent load to California was estimated to be approximately 61,000 ± 40,000 kg/year. It is further observed that the sum of annual PFAS concentrations in the analyzed cities strongly correlated with total pollution burden, and hence their ephemeral ubiquity. Similar patterns of PFAS distribution in rivers and surface waters have been observed across Asia and North America [20,27].

3.3.2. Groundwater

In the environment, PFAS can soak into soil, rivers and groundwater. It is well documented that PFAS pollution of groundwater may arise from AFFFs, fluorochemical plant discharges or migration from polluted soils and water systems [21]. This is a key concern, given that groundwater is the main drinking water source for at least half of the world’s population [18]. Communities served by private wells are also especially at risk as they can be directly exposed to PFAS through their drinking water. It is however well-documented that concentrations of PFAS in groundwater can be found at high level, as Gobelius et al. [21] investigated the groundwater near a fire training area at Stockholm Arlanda Airport, Sweden and reported high levels of ∑26PFAS, between 16–160 nwg/g (dry wt.) in soil and 1200−34,000 ng/L in groundwater. Fire training facilities and other locations are particularly at risk for heavy PFAS contamination from the routine use of AFFFs containing PFAS during training and in response to incidents. Studies have found that there is PFAS contamination is linked to urbanization and industrialization and is more prominent in developed countries particularly US and Europe. In a study by Dauchy et al. [22], 17 groundwater samples were collected from a 30-year-old French firefighter training ground. The purpose of the study was to deduce and delineate mechanisms associated with PFAS entry into the aquifer. The highest observed sum of PFAS, 300-8,300 ng/L, was detected in monitoring wells located on the training site and in a spring located down gradient of groundwater flow. Furthermore, the fluorotelomer 6:2 FTAB was detected in six monitoring wells suggesting this compound could reach groundwater located as deep as 20 m below ground surface. High perfluoroalkyl substances (PFASs) levels were observed in rural areas near a fluorochemical industrial park (FCIP) located in the Yangtze River Delta [23]. Total PFASs (∑=PFAS) were detected in groundwater between 4.8 and 614.6 ng/L, and concentrations were significantly higher in rainwater and groundwater when compared to surface water. PFAS penetration into ground aquifers may extend up to 6 km, at a maximum distance nonetheless being still predominantly controlled by advective water transport. Atmospheric deposition also expands the influence region of the FCIP to 60 km.

3.3.3. Sediments

PFAS from paper and packaging industries, machinery manufacture, WWTP discharges, and airport and dock activities are transported out from surface waters and deposit in sediments where their presence is affected by environmental gradient such as water depth or salinity. An et al. [17] explored PFAS pollution in various environmental media of Jiulong River and Xiamen Bay, China, and reported it as widespread pollutant in the entire catchment. According to their data, perfluorosulfonates (PFSAs) showed a higher adsorption potential to sediment when compared with perfluorocarboxylates (PFCAs), while sediment–water distribution coefficients (log Kₑ) for PFCAs increased with the number of -CF₂- groups for PFAAs, demonstrating partitioning behavior depending on length of chain. Similar findings have been observed in other studies whereby Khan et al. [20] performed the initial study on PFAAs in sediments of the Indus Drainage System from Pakistan, reporting 13 PFCAs and 4 PFSAs. The riverine concentrations of ΣPFAAs varied between 0.78 and 29.19 ng g⁻¹ dw in the particulate matter with short-chain PFAAs (PFBA, PFPeA, PFHxA) dominating after phasing out long-chain PFOA and PFOS. The average ΣPFAA mass loadings from municipal, industrial and agricultural sources were 0.34, 0.64 and 0.79 ng g⁻¹ dw, respectively and the mass flux of short-chain PFCAs was greater than PFOS suggesting a trend toward short-chain substitutes. In Europe, Reinikainen et al. [25] identified 23 PFAS in a variety of environmental compartments including sediments, at firefighting training and industrial sites in Finland with the drivers for their site specific human- and eco-relevance being relatively low despite exceeding established regulatory thresholds.

Further evidence demonstrates that discharges from AFFFs have contributed to sediment contamination. An inadvertent AFFF release at a site near an airport in Michigan (USA) was reported as one of the major sources of PFOS to sediments and WWTP effluents yielding estimated concentrations between 20.05 and 2100.85 ng g⁻¹; rapid decrease in PFOS concentrations within seven days after monitoring influent input confirms that monitored concentration levels could be controlled [27]. In Taiwan, Shiu et al. [26], reported that Σ8 PFAS concentrations in the sediments and suspended particulate matter (SPM) were 0.05–0.13 ng g⁻¹ and 0.54–9.08 ng g⁻¹, respectively, with low upstream riverine sources affected by long-range transport via sea spray aerosol deposition of these compounds. Their work also offered the first proof that transparent exopolymer particles (TEPs) promoted PFAS accumulation in SPM. Likewise, there was also a report of 35 PFAS compounds in sediment samples from Poyang Lake China with concentrations of 0.26–2.9 ng g⁻¹ and sources apportioned to food packaging and textile treatments as well as fluoropolymer manufacturing [28]. Together, they indicate that sediments are key sinks for PFAS in water bodies and point to management strategies that need to address legacy as well as emerging PFAS species.

4.1.1. Adsorption

Adsorption is commonly employed for the removal of PFAS as it is an operationally simple, co-contaminant resistant process which has a relatively high treatment efficiency. granular activated carbon (GAC), powdered activated carbon (PAC), ion-exchange resins remain traditional adsorbent that has been widely used, with great interest devoted to engineered alternatives for improved performance. These adsorbents include non-bio-based materials such as silica, zeolites, and graphitized carbon nitride (g-C₃N₄), and bio-based counterparts like aminated rice husk or modified chitosan that provide an adjustable surface chemistry and enhanced PFAS affinity [29].

The adsorption of PFAS is mainly controlled by electrostatic attraction and hydrophobic interaction as illustrated in Figure 5, which is due to their amphiphilic nature with an anionic hydrophilic head group and hydrophobic fluorinated carbon chain. Under environmentally relevant pH conditions, PFAS species largely exist in the anionic forms because of their low pKa values, thus, electrostatic interactions between the positively charged adsorbent surface are the primary removal mechanism; hence, adsorption is significantly pH sensitive [30]. π–π interactions are generally negligible due to the absence of π-electrons in PFAS molecules; however, hydrogen bonding may occur between oxygen atoms in PFAS anions and surface functional groups such as carboxylic (–COOH) and hydroxyl (–OH) groups. The contribution of hydrogen bonding remains debated, as competition with water molecules and reduced surface hydrophobicity on oxygen-rich adsorbents can limit its effectiveness [31]. Adsorption efficiency is influenced by PFAS chain length, solution chemistry (pH and ionic strength), and adsorbent properties, although large-scale application is constrained by high material demand, costly regeneration processes, performance loss after repeated cycles, and limited removal efficiency for short-chain PFAS (<C6), particularly with activated carbon-based systems [32].

Close

Figure 5.

Adsorption mechanisms of adsorbents used for PFAS removal [30]. “Reprinted (adapted) with permission from the American Chemical Society.

Export to PPT

(i) MIL MOF for PFAS removal

MIL-type metal–organic frameworks have been extensively investigated for the removal of PFAS from aqueous systems, both in their pristine form and after chemical modification. Using MIL-101(Cr) as a model adsorbent, Guo et al. [37] explored how PFAS molecular structure governs adsorption behavior. Their results showed that short-chain, branched, and sulfonated PFAS exhibited higher uptake than long-chain, linear, or carboxylated analogues. Specific functional motifs further influenced performance: ether linkages, such as those present in GenX, enhanced adsorption affinity, whereas partial fluorine substitution by hydrogen (e.g., 6:2 FTS) weakened interactions with the framework. In contrast to adsorption capacity, uptake rates were higher for long-chain and carboxylate PFAS. PFAS retention on MIL-101(Cr) occur via combination of electrostatic attraction, coordination interactions, hydrogen bonding, and π–π interactions, all of which were sensitive to PFAS molecular characteristics and electronegativity (Figure 6a).

Close

Figure 6.

(a) Schematic diagram of PFAS removal [37]; (b) The hetero structure of ZIF67@C₃N₄ and MIL-100(Fe)@C₃N₄ [38]; (c) Bar graph comparing the PFOA adsorption performance of various MOFs (conditions: 20 mg adsorbent, 20 mL of 1000 mg/L PFOA solution, pH 3.3 ± 0.1) and Langmuir and Freundlich isotherms models for the adsorption of PFOA by Fe-based MOFs (Initial PFOA concentration 50–1000 mg/L, Fe-based MOFs concentration 1000 mg/L, pH = 3.3 ± 0.1) [40]. (Reprinted with permission from Elsevier)

Export to PPT

Beyond adsorption, coupling MIL-based MOFs with photocatalytic materials has emerged as an effective strategy for PFAS degradation. Su et al. [38] reported enhanced PFOA removal using ZIF-67@C₃N₄ and MIL-100(Fe)@C₃N₄ composites synthesized via in situ hydrothermal growth (Figure 6b). Incorporation of MOFs increased surface area and adsorption capacity while simultaneously improving visible-light absorption and charge carrier separation in C₃N₄, leading to greater reactive oxygen species generation. Under optimized conditions, PFOA removal efficiencies of 79.2% for ZIF-67@C₃N₄ and 60.5% for MIL-100(Fe)@C₃N₄ were achieved, highlighting the benefits of integrating adsorption and photocatalysis within a single material system.

Chemical modification is frequently applied to MIL-based frameworks to further enhance PFAS uptake. Azmi et al. [39] synthesized a hydrolyzed polyacrylamide-modified MIL-96 (MIL-96-RHPAM2) using an in situ approach. Although polymer incorporation substantially reduced surface area and pore accessibility, the modified MOF exhibited a high PFOA adsorption capacity (340 mg g⁻¹), exceeding that of conventional activated carbon. This improvement was attributed to the abundance of positively charged and polar functional groups (Al³⁺, –COOH, –OH, –NH₂, –NH₃⁺), which strengthened electrostatic interactions with PFOA. The reduced porosity, however, resulted in slower adsorption kinetics. Overall, PFOA removal was primarily driven by electrostatic attraction, with additional contributions from hydrophobic interactions, hydrogen bonding, and van der Waals forces. Similarly, Yang et al. [40] compared Fe-BTC, MIL-100-Fe, and MIL-101-Fe for PFOA adsorption (Figures 6c and d). Fe-BTC exhibited the highest adsorption capacity (404.4 mg g⁻¹) despite the lowest surface area, attributed to abundant Lewis acid sites enabling strong acid–base complexation. Adsorption followed the Langmuir model, indicating monolayer coverage, with rapid uptake within 1 h and equilibrium reached in ∼3 h for all MOFs. Adsorption decreased with increasing pH due to weakened electrostatic interactions.

(ii) UIO for PFAS removal

Zr-based MOFs, particularly UiO-66 derivatives, have gained attention for PFAS removal due to their tunable porosity, surface chemistry, and catalytic functionality. A range of UiO-66-based composites including GO@UiO-66, polymer UiO-66 hybrids, and ethylenediamine tetramethylene phosphonic (EDTMP)-modified UiO-66 have demonstrated enhanced PFOA removal through combined adsorption and heterogeneous catalytic pathways [35]. Incorporation of polymeric or carbonaceous components improves mass transfer and increases access to active sites, thereby strengthening overall removal performance. In line with this, Nemati et al. [41] reported that the PFOA adsorption on UiO-66 is highly pH-dependent, with an acidic condition being favorable for its adsorption. The removal exceeded 90% under optimal conditions (pH 3, 0.1 g L⁻¹ UiO-66 and 25 mg L⁻¹ PFOA) within 60 min with adsorption capacity of the UiO-66 of ∼470 mg g⁻¹.

A deeper insight into the effects of pore structure on PFAS adsorption was provided by Sini et al. [42] in a comparative study on UiO-66 and UiO-67. Monolayer adsorption of UiO-66 agreed with the Langmuir model, whereas the larger pores of UiO-67 promoted multilayer adsorption and intrapore diffusion described by Freundlich equation. Equilibrium was established in 60 min, and the data fitted a pseudo-second-order kinetics model (R² > 0.99) for both materials but UiO-67 achieved much higher adsorption capacities (PFOS: 350 mg g⁻¹; PFOA: 470 mg g⁻¹) compared to UiO-66. The superior enhancement reveals that hydrophobic interactions increased with pore size are the predominant factors regulating PFAS adsorption, which could better validate the proposed method of combining UiO-66 and carbon-based nanomaterials to improve uptake.

In recent years, efforts have been dedicated to handling those well-known limitation such as competitive adsorption, slow kinetics and low capacity by surface functionalization and defect engineering. Ilic et al. [43] prepared Zr-based MOFs and polymer MOF composites with fine-tunability on the pore access, surface chemistry, and water tolerance by variation of synthetic parameters to achieve fast PFAS removal for environmental concentrations. The defect-rich Zr(IV)-MOF combined with an organosilicone shell presented a synergistic hydrophobic and electrostatic interaction, which further improve the PFAS adsorption process (Figure 7), as evidenced by spectroscopy, microscopy and molecular dynamics. Organosilicone coatings facilitated adsorption of long-chain PFAS such as PFOS, whereas over-functionalization in UiO-66-NH₂ and UiO-66-NO₂ resulted in decreased removal efficiency due to the steric hindrance that limited access to hydrogen-bonding sites. In contrast, in the present study both hydrophobicities displayed balanced synergies of a hydrophobic coating that promoted overall PFAS adsorption but kept open access to active Zr(IV) sites and thus favors rather strong binding even for various PFAS.

Close

Figure 7.

Sorbent structures and characterization. a) Molecular building blocks of UiO-66-X; b) Polyhedral representation of the sorbent framework; c) An 11 Å octahedral cage featuring a 6 Å pore size; d) A tetrahedral cage with a 7.5 Å diameter; e) Schematic of the polymer (PDMS/OS) coating strategy used to enhance surface hydrophobicity; f) Powder X-ray diffractograms of UiO-66-X; g) Reduction in BET surface areas after OS functionalization; h) H₂O vapor adsorption isotherms (298 K) for four UiO-66-X variants, comparing untreated and OS-modified samples; i) Static water contact angles for OS-treated UiO-66 derivatives; j) FE-SEM images depicting the morphologies of UiO-66, UiO-66-PDMS, and UiO-66-OS crystals [43].

Export to PPT

(iii) Zeolitic imidazolate frameworks (ZIFs)

ZIF-8, composed of Zn²⁺ and imidazole, is widely employed as a self-sacrificing template due to its unique thermal properties. When heated above 907°C (the boiling point of Zn), ZIF-8 completely vaporizes, yielding a non-metallic porous carbon material characterized by a high specific surface area, well-defined porosity, and uniform heteroatom distribution. These structural characteristics can increase the active sites and accelerate electron transfer. Extending prior work on ZIF-8-derived carbons for organic contaminant adsorption, Yan et al. [36] produced N-doped porous carbons (NPCs) through the controlled pyrolysis of ZIF-8 (Figure 8) and evaluated their performance for PFOA removal. Adjustment of the carbonization temperature enabled systematic tuning of pore architecture and surface chemistry. Among the materials tested, NPC-1100 exhibited the best performance, achieving nearly complete removal (99.7%) of 10 mg L⁻¹ PFOA within 10 min at an adsorbent dosage of 0.1 g L⁻¹. The adsorption capability could be enhanced for longer perfluoroalkyl chains up to 72.6% (C4), 97.3% (C7) and 99.8% (C9) owing to the strong hydrophobic interaction, and it was pH-independent and resistant towards inorganic ions or organic matter. In addition, the NPC-1100 exhibited good recyclability with 85.9% removal efficiency retained after five cycles and its practical feasibility was well demonstrated in real water samples indicating that it has a potential significance in water treatment devices.

Close

Figure 8.

(a) Schematic synthesis of NPCs. SEM images of (b) ZIF-8 and (c) NPC-1100. (d) HRTEM image and SAED pattern and (e) TEM elemental mapping of NPC-1100. (f) N₂ adsorption–desorption isotherms, (g) XRD patterns and (h) Raman spectra of NPCs [36]. Reprinted with permission from Elsevier.

Export to PPT

Inspired by the progress of porous carbons derived from ZIF, Ilango et al. [1] prepared two sets of ZIF-8 derived hierarchical carbons, i.e., ultra-microporous ZIF-8 carbon (ZFC) and hyper-microporous activated ZFC (AZFC), for the treatment of 11 PFASs in ultrapure water. The AZFC exhibited >99% removal of most PFAS and 94% for GenX within 24 h, demonstrating a synergistic role of tuned porosity to facilitate fast uptake and nitrogen-rich sites with high affinity for short-chain PFAS. The adsorbent exhibited excellent regenerability and anti-interference ability (≥4 regeneration cycles, different water qualities), indicating potential real-world application. Building upon these findings, Zhang et al. [44] transformed ZIF-67 into cathodes for electrochemical PFAS removal, and could perform simultaneous adsorption and defluorination of short-chain 2- (Trifluoromethyl)acrylic acid (TFMAA) from groundwater. In ion-adsorption, only 10.46% of Pt-G was adsorbed in 48 h; meanwhile, the introduction of a direct electron transfer and atomic hydrogen-mediated processes along with an enhanced H generation over carbonization-induced nitrogen loss led to a 99.66% removal and 97.16% defluorination on applying the applied potential (1.2 V vs Ag/AgCl). Similarly, Konno et al. [45] have reported that decreasing ZIF-67 crystal size improves adsorption velocity: nanocrystals (0.15 μm) achieved equilibrium four times faster than macrocrystals (2.95 μm) although with comparable capacities of 734.7 and 727.3 mg g⁻¹, respectively, highlighting the role of intraparticle diffusion. Together, these studies demonstrate that structural tuning and electrochemical integration significantly enhance PFAS removal efficiency and degradation.

4.1.2. Ion-exchange resins

Ion-exchange (IX) resins have recently been advanced as an attractive option for PFAS treatment, with much research examining the potential for these materials in laboratory and pilot-scale studies. These resins, which are composed of polystyrene or polyacrylic beads carrying charged functional groups and counter ions, act as PFAS capture media mainly by means of an electrostatic binding mechanism and ion exchange, with the hydrophobic effect also playing a role [46] (Figure 5). The capacities of these resins depend largely on their functional groups, porosity or chemical composition and the structural polymer backbone of the matrix, with anion exchangers generally being better. For instance, Liu et al. [47] have shown these polystyrene-divinylbenzene (PS-DVB) beads to exhibit >90% removal efficiency for 35 PFAS compounds, demonstrating their broad-spectrum potential, while the polymethacrylic and polyacrylic resins were less effective overall and may offer a greater chance of selective targeting. Commercial resins like IRA 958 and IRA 67 have not only demonstrated that IX technology has a practically potential (high adsorption capacities of up to 5 mmol/g were obtained for PFOA and PFOS), but also show good removal efficiency [46]. In recent developments such as magnetic IX resins, operationally feasible for > 90% PFOA removal with ease of separation and regeneration have been reported [48]. Likewise, strong base AnX resins are able to readily sequester PFEAS from natural waters [49], indicating that Across the range of PFAS chemistries IX resins can be useful.

Pilot studies have also shown the benefits of IX resins over common GAC. For instance, SorbixA3F resins showed the most-effective performance for PFOA and PFOS removal from groundwater (GAC’s adsorption capacity was exceeded by more than 4 times), presumably due to their simultaneous adsorption processes as well as ion exchange; however, on-site regeneration generates need of proper treatment for the PFAS-loaded waste. Conte et al. [50] emphasized the effect of resin chemistry and proved that non-ionic resins were generally less efficient, since electrostatic interactions dominated the adsorption of short-chain PFAS, while hydrophobic interactions controlled long-chain PFAS. Du et al. [51] have reinforced that removal efficiency is a function of the resin functional > groups and that IX resins provide several advantages compared to GAC like greater adsorption efficiency, shorter contact times, smaller equipment footprints and regeneration. In mixed-matrix media such as those present in fire training area groundwater composed of anionic, cationic, and zwitterionic PFAS, complete removal may require the strategic combination of resins capable of capturing specific charged components. Overall, these results imply that IX resins offer an adaptable and effective tool for removal long- and short-chain PFAS in water treatment operations that can be integrated with existing remediation practices (as indicated by enriching the contents of Table S3), combining synergetic benefits as well as continuing operation efficiency.

4.1.3. Membrane technology for PFAS removal

Membrane filtration has demonstrated as a promising method for the removal of PFAS from water, and its selectivity was mainly dominated by the surface characteristics including membrane material, porosity, pore size and zeta potential. Membranes are less affected by salts, dissolved organic matter or other co-contaminants in the water compared to ion exchange resins and activated carbon. However, the presence of fouling layers (deposits organic and/or inorganic foulants on membrane surface) can greatly affect the removal efficiency. It has been found that fouling, especially of alginate by clogging can reduce water flux acting as a physical obstacle while potentially enhancing the PFAS removal by sorption. Nevertheless, the bound PFAS might be gradually leached and releasing back from it can complicate accurate measurement and secondary contamination [31,52]. Microbial fouling may also impact PFAS adsorption. For the fouling effect, it can be alleviated by an adequately membrane cleaning or washing; thus demanding for a knowledge and control of fouling layers in order to achieve sustainable PFAS removal.

Membranes are generally classified as dense or porous. Porous membranes (with a broader size distribution) such as ultrafiltration (UF) and microfiltration (MF) have larger pore sizes and are not efficient for PFAS rejection. For instance, UF membranes like UP020 and UH030 exhibited approximately 68.9-83.7% rejection for PFHxA because of relatively large pore size [53]. UF membrane surface modification has been demonstrated as a can enhance the performance; for instance, a modified layerby-layer deposition of polyelectrolytes (polyallylamine hydrochloride and polyacrylic acid) treatment with UA60 reduced its molecular weight cut off (MWCO) from 2263 Da to 1411 Da and brought down porosity by 9.2%, rendering the PFOA and PFOS removal efficiency enhanced by about 30% [54]. Another alternative method is membrane distillation (MD), which uses hydrophobic membranes with pore sizes ranging between 0.1 and 0.45 μm in combination with a thermal gradient to separate water vapor from the feed solution, leaving PFAS behind [55].

Dense membranes like NF and reverse osmosis (RO) provide a promising route for water and wastewater treatment due to its high efficiency in removing PFAS from waters up to trace concentrations (Figure 9) [52]. RO is the predominant process used in commercial and industrial applications because of its high level of removal capabilities usually higher than that obtained for NF membranes, while complying with effluent regulations such as discharges in the US, Canada, and Australia. PFAS rejection becomes relatively high for both RO and NF due to the small pore size, i.e., 0.0001–0.001 μm for RO and 0.001–0.01 μm for NF. In contrast, MF and UF membranes showed less removal because of their larger pores. While polymeric membranes can be engineered with appropriate RO and NF pore sizes, ceramic membranes face fabrication challenges at such fine scales. NF membranes are particularly effective due to “physical sieving,” which blocks larger contaminants such as colloids and organic macromolecules while allowing water and PFAS to pass through because of their narrow pore size, low molecular weight cut-off, and negative surface charge. Studies demonstrate NF membranes can achieve removal rates exceeding 93% for diverse PFAS compounds. For example, Franke et al. [56] reported 99% removal of 15 PFAS species with chain lengths from C4 to C12 (PFHxS, PFBS, PFOS, PFHxA, etc.) at concentrations of 6–110 ng/L and molecular weights of 213–500 g/mol, using NF270 at a feed flow rate of 2.3 m³ h⁻¹. Similarly, Liu et al. [57] treated aqueous film-forming foam (AFFF) containing PFAS using a spiral-wound NF/RO thin-film composite membrane, achieving rejection rates over 97% across flux rates of 7–50 LMH, feed flows of 5.7–13.2 Lpm, and feed pressure of 60 psi, for both laboratory (≈60,000 ng/L) and groundwater (≈6,000 ng/L) matrices.

Close

Figure 9.

PFAS removal by NF membrane, reverse osmosis and adsorption process [52].

Export to PPT

Nonetheless, although membrane have the high efficiency to eliminate wide range of pollutants compared to adsorption, it still face some operational challenges like fouling, energy demand and brine management. Low PFAS removal efficiencies are attributed to inorganic, organic and microbial fouling, and the resulting brine concentrate with high contaminant concentration that need proper disposal or side treatment. Furthermore, as an alternative, the membrane process is more energy intensive than adsorption or ion exchange and treatment cost for about $0.12/m³ of water has been calculated. RO is particularly suitable for rather clean low suspended solids and low-organic C- water, emphasizing the necessity of good quality in guaranteeing the sustainable operation over time. In general, NF and RO membranes offer well-performing and robust PFAS removal; however, this is highly dependent on suitable fouling management, brine application and energy demand differences are shown in Table S3.

4.1.4. Foam fractionation

Among the various technologies used for decontamination of PFAS, foam fractionation has been considered as a prospective treatment method because it operates on simple principles, low cost and adaptability to wide range of water matrices. Due to these merits, the technology is now of increasing interest for researchers and water stakeholders and has been widely investigated in separation and concentration of surface-active substances from aqueous solutions [58]. Laboratory, pilot and full-scale studies have shown successful PFAS removal from groundwater, landfill leachate and AFFF-impacted waters [59]. The faom fractonation process takes advantage of this amphiphilic characteristic of PFAS, which contain a hydrophobic and a hydrophilic moieties that favors absorption to air–liquid interfaces. In a common arrangement, air is injected at the bottom of a vertical column and rising bubbles are passed through, collecting PFAS as they rise to create a layer of foam that can be removed by valves or vacuum. As illustrated in Figure 10(a), foam fractionation is a method of physical separation that enriches PFAS without causing chemical decomposition relying solely on interfacial adsorption based on surface activity [60,61].

Close

Figure 10.

Foam fractionation performed with basic mode for the removal of PFAS (a) PFAS removal via adsorption at the air-liquid interface of bubbles without co-surfactants (b-a), (b) PFAS removal in the presence of a cationic surfactant (b-b) [60].

Export to PPT

Foam fractionation is conceptually similar to froth flotation: it is a separation process whereby dissociated components are attached to gas bubbles within liquid phases, precipitated by foaming. Foam separation is generally not applied for the removal of particulate matter or suspended hydrophobic particles, whereas flotation operates with such insoluble hydrophobic substances. While foam fractionation is most effective for surface-active substances, surface-inactive compounds may also be removed indirectly through complexation with added surfactants that impart surface activity. Therefore, the choice of surfactant is an important factor on PFAS extraction performance. As demonstrated in several studies, cationic surfactants can be very effective because strong electrostatic interaction between the positively charged headgroup and anionic PFAS molecules can improve adsorption, facilitated foam enrichment; whereas anionic surfactants exhibit reduced performance because of electrostatic repulsion [60]. As depicted in Figure 10(b), Buckley et al. [62] achieved >95% removal of PFOA with cetyltrimethylammonium bromide (CTAB) only 20–30 min after foaming, equivalent to removal observed for short-chain PFAS such as PFBS and PFBA following 30 min. In order to increase short-chain PFAS removal, Vo et al. [63] studied mixed cationic–zwitterionic surfactant systems (CTAB with SB-12) and presented evidence that CTAB played the dominant role in demonstrating more than 90% removal of PFBS and perfluoropentanesulfonic acid (PFPeS). Although foam fractionation is effective across a range of PFAS-contaminated waters, enhanced removal of short-chain PFAS often requires co-surfactants, which may increase operational costs and raise concerns regarding secondary environmental impacts due to surfactant toxicity (Table S3).

4.2.1. Biological degradation

Biological process, in particular based on microorganisms and enzymes, is a sustainable and environment friendly way for PFAS degradation without causing secondary pollution [64]. There are two primary enzymatic pathways by which microorganisms transform PFAS compounds: oxidation (Enzymes introduce oxygen atoms across the C-F bond, ultimately breaking it) and reduction (Electrons are added across the C-F bond, leading as well to its cleavage). The high energy of the carbon-fluoride bond in PFAS requires a relatively large amount of energy to activate, thus this reaction generally takes place slower than other treatment mechanisms. A biological approach harnesses the abilities of microorganisms, such as bacteria and fungi, that can cleave the robust C-F bonds of PFAS molecules. Microbial treatment could be economically and environmentally feasible in PFAS remediation due to the use of natural organisms [65]. Despite the fact that the processes of PFASs biodegradation by microorganism are not well characterized, emerging microbiological tools offer good prospects to investigate and elucidate these reactions [64]. For instance, fluorotelomer alcohols (FTOHs) are major sources of PFOA and other PFCAs in the environment. Unlike many persistent PFAS, FTOHs degrade relatively quickly under aerobic conditions, forming PFCAs and aldehydes [66]. As shown in Figure S2, aerobic degradation begins with FTOH oxidation to FTAL, then FTCA, and finally FTUCA, releasing fluoride ions. Anaerobic breakdown of 8:2 FTOH yields intermediates like 7:2 olefin and end products such as PFHxA, PFOA, and PFHpA [67]. PFAS can be effectively treated through microbial remediation using bacteria such as Gordonia and Acidimicrobium, mycoremediation using fungi like Aspergillus niger and P. chrysosporium, and phytoremediation using plants such as Betula pendula and Picea abies. There are limited number of microbes that are capable of breaking down the C-F bond under aerobic or anaerobic conditions. However, the degradation process is either slow (taking anywhere from a few days to 180 days depending on the specific process) or less effective when compared to adsorption or membrane separation methods. In the future, certain bacteria in anaerobic digestion that have the ability to use methane and hydrogen as a source of electrons, while PFAS compounds serve as electron acceptors, could enhance the process of breaking down the C-F bond in the biodegradation of PFAS compounds. The utilization of various bioremediation strategies has the potential to improve the effectiveness of degrading PFAS [64].

4.2.2. Photocatalytic degradation

The utilization of photocatalysis for eliminating contaminants from wastewater has become increasingly popular in recent decades due to its numerous benefits, including its environmentally friendly nature, straightforward procedure, ability to achieve full destruction, high efficiency, and cost-effectiveness [68]. A photocatalytic system consists of two primary components: a catalyst (such as TiO₂, graphene oxide, ZnO, etc.) and a light source (such as UV, visible, solar, light-emitting diodes (LEDs), etc.) as shown in Figure 11. The catalyst is heterogeneous in nature. The fundamental mechanism of this process involves the creation of electron-hole pairs inside the system when the catalyst surface is exposed to light irradiation. This electron-hole pair makes it possible to produce ̇OH and or Ȯ^2– radicals which will then react with pollutants contained in the wastewater. Photocatalysis possesses proven ability to degrade even very low amounts of pollutants in polluted water. But these efficiencies could be affected by a number of operational parameters including initial pollutant concentration, catalyst surface area, catalyst loading, temperature, light source, pH and reactor geometry [69]. Photocatalysis, is a well studies method, however, much of that research has been conducted on a small scale in laboratory settings. Researchers are currently working on implementing large-scale operations by addressing the limits of the photocatalytic process, such as the use of sunlight instead of UV light and the design of large-scale reactors. The photocatalytic method has a notable and significant characteristic of being able to recover the catalyst after it is used. This allows the catalyst to be reused without needing to be regenerated in most circumstances, making the process economically feasible. Several recent research have demonstrated the use of membrane or adsorption technologies combined with photocatalytic techniques to enhance the efficacy of PFAS removal. The combination of these approaches shows promise for effectively mineralizing PFAS. However, this field of research is currently lacking in thorough investigation and requires quick attention [68].

Close

Figure 11.

Conceptualized illustration of photocatalytic reaction mechanisms of Fe/TNTs@AC. Reprinted with permission from Elsevier [68].

Export to PPT

4.2.3. Sonochemical degradation

As one of the rapidly developing sonochemical techniques, sonolysis is a promising water treatment technology for PFAS decomposition. This approach uses high-frequency ultrasound (100 –1000 kHz) to generate acoustic cavitation wherein the powerful wave vibrations produce microscopic bubbles that implode rapidly, resulting in localized sites of extremely high temperature and pressure [70]. These aggressive microenvironments drive the cleavage of carbon–fluorine (C–F) bond and hence enabling PFAS degradation. The degradation rate typically increases with acoustic power density, and for many studies highest efficiency lies at or near powers where the system is saturated [70]. Apart from energy input, PFAS molecular structure, especially chain length, and operational parameters such as irradiation time and frequency have a decisive influence on degradation efficiency [71]. In terms of mechanism, the sonochemical process for PFAS degradation occurs predominantly through cavitation-induced pyrolysis at elevated transient temperatures and via free-radical oxidization, in which water thermally decomposed generates reactive hydroxyl radicals (•OH), as expressed with the following Eq (1).

However, not all the radicals generated can effectively degrade PFAS due to the strong C–F and C–C bonds. Instead, pyrolysis drives the breakdown of PFOA through decarboxylation Eq (2)

A series of laboratory studies have shown the extraordinary Sonochemical process efficiency for PFAS destruction. At frequencies greater than 16 kHz, the ultrasound produces strong pressure and elevation in temperature values which are good enough to pyrolyze PFAS compounds existed in wastewater. In the course of this process, it also generates strongly reactive hydroxyl radicals (̇OH), which are actively involved in the decomposition of pollutants as an oxidizing agent [71]. Research has indicated that ultrasonic cavitation offers up potential for PFOS removal from water. It is assumed that the high temperatures produced upon collapse of the bubble lead to this degradation, possibly by pyrolysis [70]. Studies have demonstrated varying levels of PFOS degradation using ultrasonic cavitation. Moriwaki et al. found a decrease in the PFOS concentration between 28–60% for air and argon gas at specific frequency and power density [72]. Higher degradation efficiencies were achieved by other studies that reported up to 99% removal at a different frequency [73]. The proposed mechanism for this degradation involves pyrolysis at the hot bubble surface, which could break down the PFOS molecule in two ways: the strong carbon-sulfur (C-S) bond holding the functional group to the perfluoroalkyl chain might be the first to break, releasing sulfate ions as a byproduct. This breakdown would be followed by the mineralization (degradation to simpler compounds) of the remaining perfluoroalkyl chain. Alternatively, some research suggests that the fluorine atoms might be removed from the perfluoroalkyl chain (defluorination) before the C-S bond cleavage occurs [73].

Ultrasonic dissociation may also be effective in eradicating contaminants that readily partition to the air-water interface or have high Henry’s Law constants (favoring volatilization) [74]. As shown in Table 2 [71, 72, 75-79], various sonochemical techniques have been explored for treating PFAS, primarily PFOA and PFOS [80]. Studies by Moriwaki et al. [72] investigated the sonolysis of PFOA and PFOS (both at 10 ppm initial concentration) under argon and oxygen atmospheres. Notably, using argon resulted in promising degradation rates, with pseudo-first-order rate constants of 0.32 min^-1 for PFOA and 0.16 min^-1 for PFOS. Analysis using liquid chromatography-mass spectrometry (LC/MS) suggested that most PFOS and PFOA degradation occurred at the interface between the bulk water and the collapsing cavitation bubbles [71]. Sonochemical degradation was applied to treat landfill groundwater containing PFAS (PFOA and PFOS) along with co-contaminants including volatile organic compounds (VOCs) and dissolved organic matter (DOM) [76]. The presence of organic components reduced the degradation rate of PFOA and PFOS due to competition for sorption sites at the bubble-water interface. However, combining ultrasound with ozonation improved the mineralization (complete breakdown) of PFOA and PFOS during landfill groundwater treatment. The sonochemical treatment can be influenced by several operational parameters, including power densities, ultrasound frequencies, temperature, initial concentration of PFAS, additives, ambient conditions and chemical characteristics of PFAS.

4.2.4. Electrochemical processes

Electrochemical oxidation (EO) is gaining attention as an effective method for degrading persistent PFAS in water and wastewater, particularly because conventional biological and chemical treatments are largely ineffective against these contaminants [81]. EO makes use of an applied electric current to form powerful oxidizing species at the electrodes, causing direct degradation of organic pollutants (Figure 12a). Several laboratory- and pilot-scale studies reported its efficiency to remove highly recalcitrant organics, including PFAS [82,83]. EO offers several advantages, such as on-site implementation, elimination of added chemical oxidants and operational simplicity; however, challenges remain, including incomplete PFAS degradation under certain conditions, loss in efficiency due to anode scaling, potential formation of harmful byproducts, possible air emissions and high electrode costs. Notwithstanding these limitations, the mineralization of PFAS by EO at lower energy than incineration demonstrated in this study makes EO a potential usable technology for focused PFAS treatment [81].

Close

Figure 12.

(a) Application of various advanced oxidation processes for the degradation of PFASs in diverse aqueous environment [88]. Reprinted with Elsevier permission; b) Proposed degradation pathways for electrochemical oxidation of PFASs: through combination of direct oxidation in the anode surface and indirect oxidation with OH (Cycles I and II), dissolved O₂ (Cycle III) and formation of volatile fluorinated organic compounds (Cycle IV) and c) PFOA degradation via direct and indirect oxidations with SO₄^•− and ·OH [81]. Reprinted with Elsevier permission.

Export to PPT

PFAS mineralization by advanced oxidation process (AOP) or advanced reduction process (ARP) have been extensively studied to minimize the environmental side effects. The efficiency of these procedures is highly related to the formation of reactive species including hydroxyl radicals (·OH) or superoxide radicals (Ȯ₂⁻), which in turn can be determined based on the treatment mode and experimental conditions. AOPs and ARPs include a wide range of methods including UV irradiation, ozonation, photocatalysis, sonolysis, elemental iron reduction, H₂O₂- based treatments, photolysis, electrocatalysis solely or in combination (Figure 12a) [56]. While numerous techniques such as these are capable of breaking down PFAS, their high economic expenditures and tentative long-term applicability frequently confine the use to an adjunct measure. The reported decomposition times for the removal of about 90% PFAS vary from minutes to several days depending on process specific parameters such as light intensity and catalyst type in photocatalysis, ozone dose in ozonation, or pH and catalyst concentration in Fenton-based systems. Accordingly, AOPs and ARPS are often used as a polishing step after separation procedures like adsorption or membrane processes. Notably, perfluoroalkyl sulfonates exhibit strong resistance to oxidative treatments; for example, a 120-min laboratory study using ozone, ozone/UV, ozone/H₂O₂, and Fenton’s reagent failed to degrade PFOS at 20 mg/L [84]. Hori et al. [85] showed that the degradation of fluorotelomer substances under persulfate activation and UV irradiation was orders of magnitude faster than such processes in the dark with near-stoichiometric recovery of fluoride and CO₂ also over longer reaction times. Likewise, oxidation by Fenton was able to eliminate 89% of PFOA after 150 min with H₂O₂ and Fe(III) [86], but fluoride mass balance was not provided. While radical-driven oxidation can mineralize fluorotelomers and perfluoroalkyl carboxylates, in situ application remains challenging due to secondary formation of mobile short-chain PFAS and difficulties in replicating laboratory kinetics under heterogeneous subsurface conditions.

Because fluorine atoms shield the carbon backbone, PFAAs are generally resistant to hydroxyl radical attack, limiting the effectiveness of many oxidative processes. In contrast, reductive pathways may be more favorable, as fluorine atoms are susceptible to nucleophilic and reductive attack due to their high electronegativity. Song et al. [87] and others have demonstrated degradation of PFOA and PFOS using ARPs such as UV-irradiated sulfite, iodide, and dithionite systems. Although laboratory investigations have demonstrated encouraging removal efficiencies, the application of these reagents is accompanied by environmental risks, and operational challenges continue to restrict their deployment in large-scale or in situ remediation scenarios.

To clarify how PFAS are transformed during electrochemical oxidation (EO), several mechanistic pathways have been proposed. Radjenovic et al. [83] delineated four principal routes involving distinct reactive species (Figure 12b). Among these, the direct electron transfer mechanisms (Cycles I and II) involve oxidation of PFAS at the electrode interface, leading to the formation of highly reactive radical intermediates that subsequently undergo degradation. In parallel, oxygen and hydrogen abstraction pathways (Cycles III and IV) involve reactions between perfluoroalkyl radicals and dissolved oxygen or water, contributing to chain scission and mineralization. More recent studies have identified an additional EO pathway involving electrochemically generated sulfate radicals (SO₄•⁻) acting alongside hydroxyl radicals (·OH). As shown in Figure 12(c), PFOA degradation may proceed via either direct electron transfer or SO₄•⁻ attack, generating perfluoroalkyl radicals that undergo sequential “unzipping” reactions, ultimately yielding CO₂ and HF [88]. This pathway offers faster kinetics and reduced energy demand but has so far been confirmed only for PFOA and HFPO-DA, highlighting the need for further investigation across a broader range of PFAS compounds.