Understanding PFAS Adsorption: How Molecular Structure Affects Sustainable Water Treatment

Abstract

Per- and polyfluoroalkyl substances (PFASs) are a broad group of synthetic chemicals characterized by strong carbon–fluorine bonds, making them highly persistent and widely distributed in the environment. Their chain length and functional head groups, such as sulfonate and carboxylate groups, determine key molecular properties like hydrophobicity, acidity, and sorption behavior. These properties significantly impact the effectiveness of PFAS removal from water systems. This review provides a structural classification of PFASs and explores removal strategies, with a particular emphasis on adsorption. It examines sustainable adsorbents, including both natural materials (e.g., cellulose, chitosan, lignin, and cyclodextrins) and engineered synthetic materials (e.g., covalent organic frameworks, metal–organic frameworks, and molecularly imprinted polymers). The discussion highlights important parameters such as chain length and functional chemistry, as these can greatly influence removal efficiency. Furthermore, the discussion addresses the adsorption mechanisms-such as electrostatic attraction, hydrophobic interaction, and fluorophilic interactions-to show how they contribute in different systems. By combining structural insights with adsorption performance data, this review aims to help design and select sustainable, high-performance adsorbents for efficiently reducing PFASs in contaminated water.

1. Introduction

Per- and polyfluoroalkyl substances (PFASs) encompass a remarkably broad class of synthetic organofluorine chemicals characterized by carbon chains that are fully or partially substituted with fluorine atoms. These compounds are known for the strength of the carbon–fluorine bond, which is among the strongest in chemistry. Because of this bond’s stability, PFASs exhibit significant chemical and thermal persistence, resisting degradation in natural environments and within living organisms [1,2]. The class is extraordinarily extensive: thousands of individual PFASs are cataloged by regulatory bodies and chemical databases such as PubChem and Chemical Abstract Service (CAS) [3]. The Organization for Economic Co-operation and Development (OECD) registry suggests that over seven million PFAS structures could exist [4]. In essence, PFASs represent a family of man-made “forever chemicals” that resist typical environmental or biological breakdown processes. These substances can be broadly categorized based on chain length, functional group chemistry, and whether they are polymeric or non-polymeric [5]. Polyfluoroalkyl substances act as precursors in industrial processes, with numerous minor variants existing that differ in chain length, head group chemistry, and branching [6].

The unique physiochemical properties of PFASs, including being lipophobic, hydrophobic, surface-active, and resistant to thermal and chemical degradation, have led to their use across various industries. Notably, PFASs have been employed in aqueous film-forming foams (AFFFs) for firefighting [7], surface treatments for textiles [8], paper [9], and food contact materials [10], non-stick cookware coatings [11], stain-resistant treatments for carpets and upholstery [12], as well as in automotive [5], aerospace [13], electronic [14], and medical devices [15]. Their use is so widespread that PFASs have been found in rainwater, groundwater, surface and drinking water, soils, sediments, indoor air, dust, and wildlife at locations ranging from heavily industrialized areas to remote ecosystems [15,16]. Widespread environmental presence has prompted extensive regulatory, scientific, and public response concern.

From a chemical perspective, the strength of the carbon–fluorine bond drives PFASs’ environmental persistence and buildup. This high-energy bond strongly resists hydrolysis, oxidation, and biological breakdown. Environmental transformation mainly involves breaking down functional groups or precursors into persistent PFAAs; fully mineralizing the fluorinated chain remains extremely difficult, even over geological timescales [17]. As a result, many PFAS have half-lives of several years to decades in water or soil, and some also have human serum half-lives of three to nine years [18] depending on chain length and head group chemistry [19,20]. Biomonitoring consistently detects PFASs in both human and wildlife populations worldwide [21]. Nearly every adult in serum studies exhibited measurable levels of PFOS, PFOA, or their replacements, even after regulatory measures reduced the production and use of long-chain PFASs in many areas [22]. For example, urine and serum samples often contain several nanograms per milliliter of PFOS and PFOA, while shorter-chain analogs and GenX derivatives are also widespread. In food systems, PFASs bioaccumulate through the food chain; fish, shellfish, and agricultural produce frequently serve as significant sources of exposure depending on local conditions contamination.

Human exposure to PFASs happens through several linked pathways. The most researched route is ingestion via contaminated water, food, or consumer products [23]. Dermal absorption and inhalation (from dust, indoor air, or spray products) add to the overall exposure load [24]. Occupational contact with and use of AFFFs in firefighting operations are recognized as high-risk scenarios.

Once absorbed, PFASs spread through the circulatory system and bind to serum proteins like albumin and enzymes involved in fatty acid transport [25]. Tissue accumulation is notable in the liver, kidney, and, to a lesser extent, in lipid-rich area tissues. PFASs easily cross the placenta and are found in breast milk, posing potential exposure risks to fetuses and nursing infants [26,27]. Epidemiological and toxicological studies link PFAS exposure at both low and high levels to various adverse health effects [28]. These include elevated liver enzymes, dysregulated lipid metabolism, a higher risk of kidney and testicular cancer, endocrine disruption (especially thyroid function), reproductive effects (such as altered fetal growth and developmental delays), and immunotoxicity (including reduced vaccine effectiveness) [29,30].

A growing scientific consensus states that all PFASs should be regulated together as one chemical class. Leading among these voices is Kwiatkowski et al., whose landmark article in Environmental Science & Technology Letters argues that persistence, mobility, and toxicity support a comprehensive regulatory approach [31]. They warned that addressing individual PFAS compounds separately will only lead to ongoing replacements with untested analogs. Meanwhile, Cousins et al. have compared PFAS contamination to breaching a planetary boundary, noting that environmental levels of PFOS, PFOA, PFHxS, and PFNA already exceed guideline limits in rainwater and soils worldwide [32]. Cousins advocates for regulatory commitment to industrial decarbonization, a class-based phase-out of non-essential uses, and funding for safer alternatives. In 2025, Cousins and colleagues created a database that cataloged over 500 environmentally friendly alternatives, showing steady progress toward a shift to green chemistry [33]. Relying on this scientific foundation, many jurisdictions have started implementing decisive regulatory steps. In the United States, the Environmental Protection Agency (EPA) enforced landmark drinking water regulations (National Primary Drinking Water Regulations, NPDWR) effective April 10, 2024. Legally enforceable maximum contaminant levels (MCLs) were set at 4 parts per trillion (ppt) for PFOA and PFOS, and at 10 ppt for PFNA, PFHxS, GenX (HFPO-DA), and PFBS; additionally, a hazard index limit of 1 was imposed for mixtures [34].

Across North America, individual U.S. states have enacted new legislation. For example, Wisconsin has expanded its authority to require PFAS cleanup even without confirmed discharges [35]; New York has limited certain PFASs in personal care and menstrual products [36]. New Mexico has gone further by banning added PFASs in consumer products, except for fluoropolymers [37]. Canadian provinces are also considering limits and initial steps toward class-based PFAS regulation [38]. Due to these regulatory efforts, there has been significant innovation in remediation technologies. Traditional groundwater and drinking-water treatments, such as granular activated carbon (GAC), ion exchange resins, and membrane filtration, effectively remove PFASs but do not destroy the chemicals, creating hazardous residuals that pose long-term disposal challenges [39]. Recent research is shifting toward destructive technologies that break the carbon–fluorine bond. Laboratory and pilot successes have been reported for plasma-based oxidation, electrochemical systems (e.g., low-cost graphene sponge electrodes that achieve partial defluorination), and supercritical water oxidation, reaching destruction efficiencies of over 99%. New strategies involve combining capture methods (such as foam fractionation) with advanced oxidation, thermal mineralization, or SCWO destruction [40].

This review provides a detailed look at per- and polyfluoroalkyl substances (PFASs) from both chemical and environmental cleanup perspectives. It starts by explaining the basic chemistry, sources, and environmental behavior of PFASs, emphasizing their unique physicochemical properties, durability, and tendency to bioaccumulate, which earns them the name “forever chemicals.” Then, it presents a systematic classification of PFASs, dividing them into polymeric and non-polymeric groups. Within the non-polymeric group, it further differentiates between long-chain and short-chain PFASs based on the number of perfluorinated carbon atoms and their functional head groups, including carboxylates, sulfonates, and others. This classification system is crucial for understanding the environmental fate, transport mechanisms, and reactivity of different PFAS species.

This review centers on current and emerging PFAS remediation strategies, which are mainly divided into destructive and non-destructive methods. Particular focus is placed on adsorption-based removal technologies, especially those using sustainable and synthetic polymeric adsorbents. We carefully evaluate the design, functionalization, and performance of these polymers in capturing PFASs from contaminated environments. Key factors affecting PFAS adsorption, such as chain length, head group chemistry, and the interaction between these molecular features and the adsorption mechanisms, are thoroughly discussed. We explain how the combination of chain length and head group polarity influences the affinity of PFAS molecules for different adsorbent surfaces, ultimately impacting removal efficiency. There are three reasons to focus on adsorption: First, it is a non-destructive method that does not produce secondary pollutants. In contrast, destructive technologies can convert primary pollutants into more hazardous secondary byproducts, creating additional challenges. Second, adsorption is simple in both the synthesis of adsorbents and their use. While many existing reviews highlight the development of synthetic adsorbents, our focus is on how the length of the adsorbent chain and its functional groups influence PFAS removal effectiveness. Third, adsorption is a more cost-effective and sustainable method compared to alternative removal technologies, as it requires little to no energy input. Finally, we offer a forward-looking perspective, providing guidance for researchers and stakeholders working on next-generation PFAS remediation technologies. We highlight future directions, including the development of multifunctional polymeric materials, mechanistic understanding of PFAS–polymer interactions, and the application of green chemistry principles, to advance sustainable management of PFAS contamination.

2. Classification of PFASs

PFASs are broadly classified into polymeric and non-polymeric categories [41,42] based on their chemical structures, molecular arrangements, and the number of carbon–fluorine (C-F) bonds. These two main classes are further divided into subclasses and groups based on attached functional groups, degree of fluorination, and characteristic behaviors. Polymeric substances are subdivided into three categories; fluoropolymer PFASs are broadly classified into polymeric and non-polymeric groups. These include mers, perfluoroethers, and side-chain fluorinated polymers, which differ in their chemical structures, backbone composition, and fluorination patterns. Non-polymeric substances are divided into per- and polyfluoroalkyl substances. Perfluoroalkyl substances are further classified by carbon chain length (short- or long-chain) and by attached functional head groups (perfluoroalkyl sulfonic acids and carboxylic acids), as shown in Figure 1. In this section, we review the classification of PFASs based on carbon chain length and attached functional groups. This classification enables a more detailed discussion of PFASs. Figure 2 presents a flow diagram summarizing the classification of PFASs discussed in this section.

2.1. Polymeric and Non-Polymeric

Polymeric PFASs are large molecules composed of small repeating units with high molecular weights, such as polytetrafluoroethylene (PTFE), which has 10,000–180,000 carbons linked together to form a high molecular weight structure. Due to their large molecular weights, these polymeric PFASs exhibit low mobility in the environment, resulting in reduced ubiquity. In contrast, non-polymeric substances consist of 4 to 12 carbon atoms with attached fluorine atoms [41]. They possess smaller molecular sizes and lower molecular weights, which makes them highly mobile in water systems, thus increasing their potential for bioavailability and entry into the food chain. Due to the significant differences in their properties, polymeric fluorocarbons are evaluated separately from non-polymeric fluorocarbons. Non-polymeric PFASs are commonly used in both industrial and consumer products, leading to greater environmental and human exposure. As a result, they have been studied more extensively than their polymeric counterparts, especially regarding their mobility, persistence, and toxicity.

Molecular weight is crucial in evaluating PFAS toxicity. High molecular weight compounds, especially polymeric PFASs (generally > 1000–10,000 Da), usually cannot cross cell membranes, which limits their interaction with internal organs and decreases their potential for bioaccumulation. Conversely, non-polymeric PFASs, with much lower molecular weights, can more easily penetrate biological barriers and are more likely to build up in tissues, leading to toxic effects. Therefore, as the degree of polymerization (Mn) increases in PFASs, the associated health risks tend to decrease [43].

2.2. Per- and Polyfluoroalkyl Substances

Per- and polyfluoroalkyl substances are non-polymeric fluorocarbons. Perfluoroalkyl substances are characterized by having each carbon atom in the backbone fully substituted with fluorine atoms, creating highly stable carbon–fluorine bonds [44]. These perfluoroalkyl substances are further classified into two main categories: perfluoroalkyl acids and perfluoroalkane sulfonamides. Polyfluoroalkyl substances have carbon chains in which some—but not all—carbon atoms are bonded to fluorine atoms [45]. Although the name “polyfluoroalkyl” suggests a structure of repeating carbon units like true polymers, these compounds do not actually have such structures. Instead, the prefix “poly” indicates the presence of multiple fluorine atoms attached to the molecule. Fluorotelomers are a well-known class of polyfluoroalkyl substances that are commonly used for their water- and oil-repellent properties in consumer products such as non-stick cookware, carpets, stain-resistant clothing, and food packaging [46].

Perfluoroalkyl substances are subdivided into short-chain and long-chain types based on the length of the carbon chain. The carbon chain is directly connected to functional groups (alkyl acid groups) [47]. This functional group, often called the “head,” determines many of the chemical, environmental, and biological properties of each PFAS compound. Perfluoroalkyl acids are the most environmentally relevant and well-studied class of PFAS, as Perfluoroalkyl carboxylic acids (PFCAs) and Perfluoroalkyl sulphonic acids (PFSAs) are very dominant in the environment [31]. They exist in an anionic form in the environment, impacting their mobility, toxicity, adsorption, and bioaccumulation.

2.2.1. Long-Chain

The term “long-chain PFAS” refers to perfluoroalkyl acids with n ≥ 7. These long-chain PFASs are widely produced, highly toxic chemicals that have been well studied. The legacy PFAS molecules, perfluorooctanoic acid (PFOA) and perfluorosulfonic acid (PFOS), belong to this subclass of PFAS [48]. They are highly toxic even at low doses because they are more persistent and less mobile than short-chain compounds. Therefore, they can bioaccumulate easily [49] and can cause severe health problems like kidney and liver cancer, and disrupt the reproductive system in humans [50,51].

2.2.2. Short-Chain

With the phase-out of long-chain PFASs, short-chain PFASs were developed as an alternative. The term “short-chain PFAS” generally refers to perfluoroalkyl substances with n < 7 [49], where n is the number of carbon atoms present in the chain. The increasingly used short-chain PFASs were assumed to have lower toxicity and bioaccumulation potential compared to longer chains; however, they are more mobile and have less adsorption potential [51]. Because of this property, degrading and removing short-chain PFASs from the environment is challenging [52].

2.2.3. Sulphonic Acids

Perfluoroalkyl substances in which a carbon chain is linked to a sulphonic acid group (-SO₃H) are classified as perfluoroalkyl sulphonic acids (PFSAs). They are strong acids with very low dissociation constant (pKa) values and are highly stable because resonance stabilizes the structure through three oxygen-bonded atoms sharing the negative charge. Perfluorooctane sulfonic acid (PFOS) is a well-studied compound in this group, known for its excellent water- and oil-repellent properties.

2.2.4. Carboxylic Acids

Perfluoroalkyl carboxylic acids (PFCAs) consist of a perfluorinated alkyl chain attached to a carboxylic acid (-COOH) functional group at the other end. They have lower pKa values than sulfonic acids, making them less acidic. The negative charge on the carboxylate is localized on two oxygen atoms attached to a carbon atom, which gives them unique structural and physical properties. Perfluorooctanoic acid (PFOA) is the representative compound of the PFCA class. They are named by replacing the “S” suffix in PFOS with “A”.

As shown in Figure 3, Liang et al. analyzed 107 PFAS compounds in wastewater and classified them based on their structural features. The identified PFASs were divided into five groups, covering fifteen classes: 9-PFCA (C2-C10), 32-H-PFAA (C2-C16), 52-Ether-PFAA (C3-C16), 9-Cl-PFAA (C2-C9), and 5-I-PFAA (C2-C8). The total reported concentration of all PFASs was 36 mg/L [53].

3. Methods for PFAS Removal

Different methods for PFAS removal are discussed in the literature. Primarily, two types of methods are used: non-destructive and destructive. Since many reviews have already been published specifically on treatment methods [54,55,56,57,58,59], this section review will briefly examine the treatment approaches available so far, their effectiveness, cost-efficiency, and what remains to be performed to remove PFASs from our environment effectively.

3.1. Non-Destructive Methods

Non-destructive techniques for PFAS removal include traditional coagulation, flocculation, adsorption, ion exchange resins (IX), and membrane filtration [57,60,61]. Coagulation and flocculation are traditional methods that do not effectively remove PFASs [60,62,63]. Membrane filtration, adsorption, and anion exchange resins have become more effective and extensively researched approaches in recent years [64,65]. Figure 4 illustrates the fundamental mechanisms of PFAS removal through these non-destructive techniques.

For filtration, high-pressure membranes such as reverse osmosis (RO) and nanofiltration (NF) are very effective at removing PFASs from water [66]. The effectiveness of these membranes in removing PFASs depends on several factors, such as the molecular weight of PFASs, chain length, functionalities on the membrane surface, pore size, pore distribution, pH, and filtrate concentration [66,67]. Membranes with negatively charged and hydrophilic surfaces can potentially reject PFAS molecules. PFAS removal by RO and NF membranes improves with longer chain length due to increased size exclusion and hydrophobic interactions [68]. In a study by Fujioka et al., the effect of functional groups on PFAS removal by NF is discussed. They note that PFASs with carboxylate groups show higher rejection rates than sulfonate because PFOA (−0.83 electron units) has a greater partial negative charge compared to PFOS (−0.66 electron units) [69].

Adsorption is a sustainable, cost-effective, and eco-friendly method commonly used to remove contaminants from water [70], including PFASs [57]. Destructive approaches, such as advanced oxidation [71] or thermal treatment [72], often require high energy input and may generate harmful secondary byproducts that complicate remediation [73]. In contrast, adsorption is a non-destructive, simple, and scalable process that can be applied under mild conditions without producing secondary pollutants [74]. But disposal of spent adsorbent can produce secondary pollution [75]. Moreover, adsorption is particularly well-suited for PFASs due to their amphiphilic nature, which allows for tailored interactions such as hydrophobic, electrostatic, and hydrogen bonding with functionalized polymeric adsorbents [76]. These features make adsorption both a scientifically relevant and practically viable technology for PFAS removal. Different adsorbent materials have been developed based on the type of contaminant, each with its advantages and disadvantages [77]. For PFASs, positively charged adsorbent materials have demonstrated higher efficiency because PFASs exist in an anionic form under environmentally relevant conditions [78]. When synthesizing adsorbents, parameters such as surface area, pore size distribution, pore volume, surface charge, and surface morphology are carefully controlled. Various adsorbent materials, including activated carbon [79], biochar, metal–organic frameworks (MOFs), aerogel, hydrogel, and carbon nanotubes have been used to mitigate PFASs [70,80].

Activated carbon (AC) is a well-studied class of adsorbents that has been employed for PFAS removal, specifically for long-chain PFASs [81]. AC sorbents are efficient, simple, and cost-effective for PFAS removal [82]. In full-scale wastewater treatment plants, granular activated carbon (GAC) and powdered activated carbon (PAC) have been in use for effective removal of PFASs [83,84]. Numerous researchers have studied PFAS removal using GAC [81,85,86,87,88,89] and PAC [90,91]. They concluded that long-chain PFASs are removed more effectively by AC than short-chain analogues, and that sulfonic acids show higher removal than carboxylic acids. Comprehensive reviews already exist on AC-driven PFAS removal [92], and on regeneration of spent AC [93]; therefore, we focus here on the key research gap that AC shows less adsorption for short chains that remains to be addressed. One potential strategy to overcome this limitation is the development of narrow pore structures that can restrict access to long-chain PFASs or larger co-contaminants, thereby enhancing the overall efficiency for short-chain PFAS removal [93].

Natural and synthetic polymers are widely used for removing PFASs from water. Studies indicate that porous organic polymers, which are structurally adjustable and chemically stable, can achieve high PFAS removal efficiency under environmentally relevant conditions [94]. Natural polymers, especially those modified with fluorinated or amine groups, have shown excellent ability in PFAS adsorption because of hydrophobic, electrostatic, and fluorous interactions [44]. Studies also show that adding both fluorinated and cationic groups to polymeric sorbents increases removal efficiency for PFASs, mainly through rapid fluorine–fluorine and electrostatic interactions that bind both long- and short-chain PFASs. Long-chain PFASs can interact via hydrophobic tails through hydrophobic interactions, and the chain length and orientation of PFAS molecules significantly influence the adsorption process. Since short-chain PFASs have fewer fluorine moieties, they exhibit less hydrophobic interaction, but electrostatic interaction becomes the dominant mechanism in this type of adsorption.

Ion exchange resins, particularly anion exchange resins (gel type and microporous), have demonstrated effectiveness in removing PFASs [95]. In a study by Appleman et al., anion exchange resins showed minimal to no removal of short-chain PFASs but achieved removal rates of 92% and 75% for long-chain PFOS and PFOA, respectively [96]. A clear distinction exists in how chain length and alkyl acid type influence PFASs in this study. Another research examined polystyrene and polyacrylic resins for PFAS removal, with polystyrene showing superior performance over the latter. In the presence of sulphonate, the removal of PFCA decreases because the more extensive charge distribution of sulphonate, compared to carboxylate, outcompetes carboxylate in interactions [97].

3.2. Destructive Method

Destructive methods involve breaking part or all of the C-F bonds in PFAS molecules, resulting in mineralization, which is often gauged by the amount of fluoride released [98]. Several destructive methods have been reported, including electrochemical oxidation, thermal degradation, advanced oxidation/reduction processes (AOPs and ARPs), and plasma treatment. In destructive methods, the C-F bond breaks down either by applying heat or through radical attack. The detailed cyclic mechanism for PFAS degradation by different species is shown in the destructive section of Figure 4. Electrochemical oxidation of PFASs begins when current passes through the contaminant solution between the cathode and anode. Indirect and direct oxidation reactions start the conversion of the precursor PFAS molecule through a series of reactions into an intermediate. But this treatment method has a drawback: it can increase the concentration of short-chain intermediates by breaking the long PFAS chain [99]. Thermal degradation of PFASs is an energy-intensive process because it involves burning spent media, such as adsorbents and filters, at very high temperatures (600–1000 °C). During this intense heat, PFAS molecules can be transformed into greenhouse fluoromethane gas (CF4) [73].

ARPs and AOPs generally utilize ozone, persulfates, peroxides, and UV radiation to initiate oxidation and reduction reactions. Persulfates and metal oxides are used to generate reactive species, such as sulfite radicals, hydroxyl radicals, and hydride electrons, which initiate the reaction mechanism. Several researchers have published a comprehensive review of the procedures and mechanisms for PFOS and PFOA degradation using ARPs and AOPs, including Cardoso et al. [100], Alalm and Boffito [101], and Umar [102]. Two types of catalytic systems are used for the degradation of PFASs, namely homogeneous and heterogeneous [103]. In heterogeneous catalytic systems, catalysts containing functional groups can interact with the carboxylate or sulfonate head of PFOA, thereby weakening the bond at the head group and promoting its breakdown. In contrast, homogeneous systems depend on reactive species like radicals, holes, or electrons to break C-C and C-F bonds [101].

Recent studies have demonstrated that plasma treatment can significantly degrade PFASs, with reductions of up to 63.75% for PFOA in laboratory experiments using atmospheric non-thermal plasma (NTP) with air as the working gas [104]. The degradation efficiency depends on operational parameters (such as exposure time, gas composition, and cooling systems) and the water matrix, with greater removal observed for long-chain PFASs compared to short-chain analogs [105]. For example, in deionized water, long-chain PFASs such as PFOS and PFHxA achieved over 90% removal, while short-chain PFASs had lower removal rates. In real water matrices, removal efficiencies ranged from 8% to 50%, depending on the compound and conditions [104]. The degradation mechanism for PFSAs and PFCAs involves electron transfer and bond breaking, while ether-containing PFASs, such as hexafluoropropylene oxide dimer acid (GenX) and 4,8-Dioxa-3H-perfluornonanoic acid (ADONA), are degraded through initial ether-group cleavage followed by additional transformation steps [106].

Destructive technologies used for PFAS remediation often create secondary pollutants. Incomplete mineralization of long-chain PFASs can produce short-chain precursors, which are more mobile and widespread in the environment. Burning PFASs at high temperatures generates ash containing inorganic fluorine and releases a corrosive HF gas. [17,99]. In cases of incomplete incineration, perfluoroisobutylene, fluorocarbons, and fluorinated alkanes can form, which have long half-lives in the atmosphere [107,108]. Complete mineralization of PFAS molecules generally requires very high temperatures, typically around 1000 °C [109]. Electrochemical oxidation generates byproducts, such as Cl₂, HF, ClO₃⁻, and ClO₄⁻, during the unzipping mechanism of CF₂ units and degradation of the working electrode [110]. ARPs also generate short-chain PFAAs as byproducts because of incomplete mineralization [108]. However, plasma causes full dissociation of PFAS molecules into thermodynamically stable species, (HF, CO, CO₂, SO_x), which are then removed by scrubbing [99,108]. Given these risks, careful monitoring of intermediates and byproducts throughout the treatment process is essential to verify complete defluorination.

4. PFAS Removal by Adsorption

Adsorption is a simple yet complex process because designing the adsorbents for effective removal of a diverse range of PFAS molecules is challenging. A good adsorbent should have the following properties: (a) cost-effective, (b) environmentally friendly, and (c) effective for short- and long-chain PFAS removal [79]. Adsorption in complex environments with other co-contaminants has not been thoroughly explored. Most of the adsorbents reported to date mainly focus on long-chain PFASs, such as PFOS and PFOA. Due to strict environmental regulations on these legacy PFAS molecules, short-chain alternatives have been developed. However, they are more toxic and mobile in the environment. This review examines how short- and long-chain PFAS molecules, based on their functional head groups, are adsorbed by natural, synthetic, and composite adsorbents.

4.1. Sustainable Natural Polymers

Biopolymers, such as cellulose, chitosan, lignin, and cyclodextrins, are affordable, sustainable materials with the potential to serve as adsorbents. The chemical structure of biopolymers is illustrated in Figure 5. They possess tunable surface properties. Cellulose and chitosan are the two most abundant natural polymers on Earth. Cellulose is a key structural component of plants and contains a high number of hydroxyl groups (Figure 5). These hydroxyl groups make it highly suitable for chemical modifications such as oxidation, etherification, esterification, and grafting [106,111]. In most studies, cellulose is functionalized with amine groups using compounds such as polyethyleneimine (PEI), triethylamine (TEA), and trimethylammonium chloride to create cationic cellulose. Cellulose can be crosslinked with other polymers, such as chitosan, because they have similar chemical structures, with chitosan containing an amine group at the C2 position. Crosslinking enhances the surface area of the adsorbent and provides more functional sites, which improves the adsorption of PFASs [112]. Recently, Li et al. synthesized amine-functionalized dialdehyde cellulose through an oxidation and amine reduction reaction using 4-, 8-, and 12-carbon alkylamines. These adsorbents were used to remove short- and long-chain PFASs, achieving maximum adsorption capacities of 676, 385, and 230 mg/g for PFOS, PFOA, and PFBA (perfluorobutanoic acid), respectively. The long-chain PFOS and PFOA were removed through hydrophobic interactions between their hydrophobic tails and the alkyl chains attached to amines. Short-chain PFBA showed electrostatic interactions. Sulphonic acid head groups adsorb almost twice as much as carboxylic groups. However, this study did not explain why the acid head groups of PFASs influence the overall adsorption [113]. Cellulose fibers were functionalized with quaternary amines to introduce positive charges in a study by Harris et al. Quaternized cellulose removed 80% of PFOS and PFOA, while less than 30% of short-chain PFAS molecules were removed within a few seconds. Long-chain PFASs have higher octanol/water partition coefficients (Kow), making them more hydrophobic, which leads to increased hydrophobic interactions with the cellulosic materials backbone. They also demonstrated that sulfates have a greater affinity for amines compared to carboxylates [114]. Another study noted that PFAS functional groups are more critical for their adsorption than chain length when using quaternized cellulose functionalized polymer resins, which demonstrated high removal efficiency for both short and long-chain PFASs [115].

Chitin is a plentiful natural polymer often derived from the shells of crustaceans and the cell walls of fungi. It can be partially or fully deacetylated to remove the acetyl group attached to the amine group at the C2 position, transforming it into a cationic biopolymer (chitosan) [116]. The high amount of amine and hydroxyl groups in their structure, as shown in Figure 5, makes them suitable for PFAS removal through electrostatic interactions [117]. Chitosan has been functionalized with nanoparticles or crosslinked with other natural and synthetic polymers to increase its surface area and functional groups, creating more interaction sites for PFAS molecules. Shahrokhi and Park reported the creation of polyethyleneimine-functionalized chitosan beads for PFAS removal. The beads successfully adsorbed PFAS through electrostatic and hydrophobic interactions; however, their adsorption efficiency decreased in the presence of co-contaminants [116]. Wittwer et al. crosslinked chitosan with different concentrations of epichlorohydrin (ECH) and used molecular imprinting with PFAS templates to improve PFOS adsorption [118].

Lignin is an essential chemical component of wood, acting as a binder between cellulose and hemicellulose. Lignin has many environmental applications, and it can be carbonized to serve as an adsorbent for removing emerging contaminants [119]. Mel et al. investigate the partitioning behavior and interaction mechanisms of per- and polyfluoroalkyl substances (PFASs) within pulp and paper (P&P) wastewater systems, focusing on lignin-containing substrates. Their experimental study confirms that electrostatic and hydrophobic interactions primarily drive PFAS sorption to lignin, especially for sulfonate-functional PFASs like PFOS and PFBS, which display higher solid–water distribution coefficients (Kd). Conversely, carboxylate-functional PFASs show lower affinity for lignin and tend to remain more mobile in the water phase [120].

Cyclodextrins are cyclic oligosaccharides derived from starch, composed of a macrocyclic ring of glucose monomers connected by α-1,4 glycosidic bonds (Figure 5). Depending on the number of glucopyranose units—six, seven, or eight—they are classified as α, β, and γ-cyclodextrins, respectively [79]. Wang et al. synthesized an adjustable β-cyclodextrin polymer platform by polymerizing it with anionic and cationic comonomers. They synthesized three adsorbents—two with negative surface charges and one with positive surface charges. The cationic adsorbent worked very well for removing both short- and long-chain PFASs in ultrapure water, but its performance declined especially for short-chain PFASs like PFBA, indicating that these removals heavily rely on electrostatic interactions. In contrast, anionic adsorbents performed better when Na₂SO₄ was added, especially for long-chain PFASs, likely due to stronger hydrophobic interactions and decreased repulsion between the adsorbent and PFAS. PFSAs were removed more effectively than PFCAs, possibly because they are more hydrophobic [121]. Similarly, in another study, β-CD was functionalized with epichlorohydrin (EPI), hexamethylene diisocyanate (HDI), and benzyl chloride (Cl) by Abaie et al. to introduce positive and negative charges on CD. The first two, functionalized with EPI and HDI, showed moderate adsorption of PFASs and PFCAs, while the Cl-functionalized β-CD exhibited significant adsorption. A positive correlation between Kd and Kow values for the two adsorbents (EPI and HDI functionalized) confirmed the hydrophobic relationship between the adsorbent and anionic PFAS. The third adsorbent (Cl-β-CD) displayed comparable adsorption for PFASs [122]. The detailed comparison of the adsorption capacities of functionalized natural polymers, along with the mechanisms involved, is listed in

Table 1. Comparison of efficacy of biopolymers for PFAS removal, along with their mechanism.

Biopolymer Adsorbents	PFAS	Initial PFAS Concentration (mg/L)	qmax (mg/g)	Adsorption Mechanism	Ref.
Polyethyleneimine functionalized cellulose microcrystals (PEI-f-CMC)	PFCA (C4-C13), PFSA (C4-C10), ADONA, N-MeFOSAA, N-EtFOSAA	0.002–0.050	2.32 (PFOA)	N/A *	[123]
Alkylamine modified dialdehyde cellulose (DAC)	PFOS PFOA	5–50	576–697 235–346 132	>Electrostatic <Hydrophobic Interaction	[113]
	PFBA	25–150
Quaternized nanocellulose (QNC)	PFOS PFOA PFBS PFBA	60	559 405 319 121	>Hydrophobic <Electrostatic Interaction	[124]
Quaternized wood pulp (QWP)	PFOS PFOA	0.0025	763 605	Hydrophobic and Electrostatic Interaction	[114]
Quaternized cotton	PFOS PFOA	95.02–460.12 78.67–380.95	1650.43 1283.62	Hydrophobic Interaction	[125]
Alkylamine-modified dialdehyde cellulose (DAC)	PFOS PFOA PFBA	5–50 25–150	576–697 235–346 132	Hydrophobic and Electrostatic Interaction	[113]
Polyethyleneimine grafted chitosan beads (GCBs)	PFOS PFOA PFBA PFBS	0.001	500 555.5 1428.6 769.2	Hydrophobic and Electrostatic Interaction	[116]
Crushed chitosan beads (CBs)	PFOS PFOA PFBA PFBS		384.6 312.5 476.2 303.1
MIP-Chitosan beads	PFOS	20–550	3202.1	Electrostatic Interaction	[126]
Kraft/alkali Lignin	PFBA PFBS PFOA PFOS	0.01	N/A	Hydrophobic and Ion-Dipole Interactions	[120]
β-cyclodextrins-HDI	PFDA PFPS PFPA PFOSA PFOS PFOA PFNS PFNA PFHxA PFHxS PFHpA PFHpS PFDS	0.0001–0.001	N/A	Weak Hydrophobic + Electrostatic Interaction	[122]
β-cyclodextrins-EPI				Weak Hydrophobic + Electrostatic Interaction
β-cyclodextrins-Cl				Strong Hydrophobic Interaction
Cyclodextrin polymers (CDPs)	PFAAs	0.01–5	92.54 (PFOA) 93 (PFBA)	Hydrophobic Interaction	[127]

* Not available.

.

4.2. Synthetic Polymers

Synthetic polymers are commonly used for PFAS adsorption and often serve as the base materials for polymer resins. Coordinated polymers such as metal–organic frameworks (MOFs), covalent organic frameworks (COFs), and molecularly imprinted polymers (MIPs) have recently gained attention due to their tunable structures and selective adsorption capabilities. Their adsorption behavior, studied in recent studies, has been summarized in

Table 2. Studies reported for PFAS adsorption by synthetic adsorbents.

Synthetic Adsorbents	PFAS	Initial PFAS Concentration (mg/L)	qmax (mg/g)	Adsorption Mechanism	Ref.
Poly DMAPAA-Q hydrogel a	PFOS PFOA PFBS PFBA GenX ADONA	0.001	N/A *	Hydrophobic and Electrostatic Interaction	[130]
COFs b	PFBS PFBA PFHxS PFHxA PFOS PFOA	21.76 15.52 29.01 22.77 36.26 30.04	N/A *	Hydrogen Bonding, Hydrophobic and Electrostatic Interaction	[131]
QA-COFs c	GenX HFPO-TA	16.50–198.02 20.70–248.42	679.27 894.29	Electrostatic Interaction	[132]
Amine-f-COFs d	GenX	0.2–100	130–200	Electrostatic Interaction	[133]
Nitrogen doped ZIF-8 e	PFSA PFCA	0.01	788.43 1115.11	Hydrophobic and Electrostatic Interaction	[134]
UiO f-67	PFSA PFCA PFASA FTs	N/A *	N/A *	Hydrophobic and Charge-Pairing Interaction	[135]
UiO f-67-F2	PFOA	1000	928	Hydrophobic and Fluorophilic Interaction	[136]
MIL g-53	PFOS	20–80	~220	Electrostatic and Coordination Interaction	[137]

^a—Poly N-[3-(dimethylamino)propyl]acrylamide methyl chloride quaternized hydrogel, ^b—covalent organic frameworks, ^c—quaternary amine-covalent organic frameworks, ^d—amine-functionalized covalent organic frameworks, ^e—zeolitic imidazolate frameworks, ^f—University of Oslo, ^g—Materials Institute Lavoisier, * not available.

. The effectiveness of these polymers depends on how PFAS molecules interact with them, which is influenced by the PFAS chain length and functional groups. The surface charges, pore sizes, and hydrophobicity of the polymers guide the design of materials to enhance PFAS removal from water. Polystyrene and polyacrylic are used as matrices for anion exchange resins, often functionalized with positively charged tertiary and quaternary amines to reduce anionic PFAS contaminants [128]. Ateia et al. reported using a synthetic cationic hydrogel, N-[3-(dimethylamino)propyl]acrylamide methyl chloride quaternary (DMAPAA-Q), for PFAS removal. Their results showed that sulfonated PFASs (PFSAs) were adsorbed more effectively than carboxylated PFASs (PFCAs), regardless of chain length [129]. The study concluded that hydrophobic interactions mainly drive PFSA adsorption, while electrostatic interactions play a more significant role in PFCA binding. DFT calculations revealed that PFSAs exhibited more exergonic binding energies compared to PFCAs, supporting their higher adsorption affinity [130].

Covalent organic frameworks are made up of organic building blocks that are covalently bonded to form porous crystalline structures. COFs are surface functionalized with quaternary amines and fluorine to enhance PFAS adsorption through electrostatic and fluorophilic interactions [128]. The pore size of COFs determines their adsorption ability. A study by Wang et al. mentioned that COFs with a pore size 2.5 to 4 times larger than PFAS molecules can adsorb PFASs. Therefore, the COF structure shown in Figure 6, with a pore size of 28Å, is well suited for adsorbing PFAS molecules containing 7 to 10 carbon atoms. To analyze the influence of COF pore size on PFAS removal, they synthesized five different COFs with varying pore sizes and tested the adsorption of six short- and long-chain PFAS molecules. The adsorption mechanism revealed that excessively large pores reduce the adsorption rate. Long-chain PFASs were more strongly adsorbed by COFs, facilitated by hydrogen bonding and hydrophobic interactions. However, they did not study the impact of acid head groups on the adsorption process [131].

Quaternary amine-functionalized COFs with comparable porosities (including mesoporous and microporous structures) were synthesized by Wang et al. In their study, they found that the oxygen-containing moieties of PFAS molecules formed strong electrostatic interactions with the quaternary amine groups, leading to high adsorption energies. However, they also observed that short-chain GenX molecules were displaced by long-chain PFASs because of differences in exchange energies, indicating a preference for long-chain PFASs to adsorb at the quaternary amine sites [132].

Metal–organic frameworks (MOFs) are regarded as coordination polymers that have a metal atom at the center, along with organic ligands bonded through coordinate covalent bonds [128]. MOFs, such as MIL (Materials Institute Lavoisier), UiO (University of Oslo), and ZIF (Zeolitic Imidazolate Frameworks), and their variants have been extensively explored for PFAS removal. The crystal and unit cell structures of ZIF-8, MIL-53, and UiO-66 MOFs are shown in Figure 7.

Recently, Ilango et al. synthesized nitrogen-doped ZIF-8, which showed efficient removal of both short- and long-chain PFSAs and PFCAs. The material’s mesoporous and hydrophobic structure facilitated the adsorption of long-chain PFASs, while the nitrogen dopants at the ZIF-8 nodes improved electrostatic interactions with short-chain PFASs [134]. Hua et al. studied the dual functionality of UiO-66 through defect engineering. They used UiO-66 to detect PFASs at trace levels, and they mentioned that the accuracy was comparable to LC/MS. UiO-67 removed up to 99% of all PFAS molecules in the water matrix within 30 min [135]. Zhao et al. synthesized a series of MIL-53 metal–organic frameworks with different metal centers (Al, Fe, and Cr) and organic ligands of various lengths to evaluate their performance in PFOS adsorption. The study demonstrated that both the type of metal and the length of the organic linker significantly affected the adsorption capacity. Specifically, longer organic ligands increased pore spacing, which improved the diffusion of PFOS molecules and enhanced adsorption. Additionally, the presence of unsaturated metal sites further increased PFOS removal, likely due to stronger electrostatic and coordination interactions with the PFOS anions [137,138].

Molecularly imprinted polymers (MIPs) are synthetic polymers designed to have specific selectivity for target molecules. They have been utilized as adsorbents for PFAS removal and as sensors for detecting PFASs [139]. Different synthetic methods have been used to create imprints targeting PFAS molecules, with bulk polymerization being the simplest. However, it requires intensive grinding and involves using various molar ratios of the target PFAS molecule as templates, along with crosslinkers and functional monomers, as shown in Figure 8 [139].

Some of the MIP adsorbents used for PFAS removal, including their adsorption capacities and mechanisms for both short- and long-chain PFASs, are listed in

Table 3. Adsorption capacities of molecular imprinted polymers (MIP).

MIPs	Monomers	PFAS Template	Crosslinker	Initial PFAS Conc. (mg/L)	qmax (mg/g)	Adsorption Mechanism	Ref.
Chitosan-based MIP	Chitosan	PFOS	Epichlorohydrin (ECH)	20–550	3202.1	Electrostatic Interaction	[126]
PFOA-MIP	Acrylic acid	PFOA	Ethylene glycol dimethacrylate (EGDMA)	0.02–0.1	5.45	N/A *	[140]
HDI-1	polyurethane and β-cyclodextrin	PFOA	Hexamethylene diisocyanate (HDI)	2.07–2070.35	1089.9	Ion-Dipole and Hydrophobic Interaction	[141]
Bi-functional MIP	2-(trifluoromethyl) acrylic acid, 4-vinyl pyridine	PFOA PFOS	Ethylene glycol dimethacrylate (EGDMA)	0.05–10	6.42 6.27	Hydrophobic Interaction	[142]
Magnetic-MIP-PFOS	Acrylamide + Fe3O4@SiO2NPs	PFOS	Ethylene glycol dimethacrylate (EGDMA)	0.1–0.7	2.401	Hydrophobic, Electrostatic Interaction, and Hydrogen Bonding	[143]

* Not available.

. PFOS and PFOA share a similar linear C8 chain structure. MIP adsorbents, which are templated with PFOS, showed high selectivity only for PFOS due to its sulfonic acid group, which forms strong electrostatic interactions with amino groups in chitosan during imprinting. In contrast, PFOA’s carboxylic group lacked the same interaction strength due to memory selectivity, resulting in significantly lower adsorption and demonstrating the critical role of functional group specificity in selective PFAS recognition [126].

5. Key Parameters Affecting PFAS Adsorption

The amphiphilic nature of PFASs, featuring a hydrophobic fluorinated tail and a polar functional head group [144], makes the adsorption process particularly complex, especially in real water matrices where coexisting contaminants can interfere. The chain length and the nature of the acid head group (e.g., carboxylate vs. sulfonate) play critical roles in the adsorption process [145].

5.1. Effect of PFAS Chain Length

As discussed earlier, PFASs are classified as short- or long-chain based on the number of carbon atoms. This chain length is essential for their adsorption by an adsorbent. Long-chain PFAS molecules have higher molecular weights, making them less mobile and more hydrophobic in water matrices compared to short-chain PFASs [146]. Adsorption of PFOS usually exceeds that of PFOA on different adsorbents [147]. Despite having the same number of carbons, PFOS exhibits greater hydrophobicity due to two main structural differences: first, it has one extra CF₂ unit, and second, there is a sulfonate group in PFOS [147]. Adding a CF₂ unit results in an increase of about 0.55 log units in the Kd (distribution coefficient) value [148], which indicates the tendency of an adsorbate to adhere to the surface of the adsorbent. It is reported that the adsorption of longer-chain PFAAs increases with the length of the carbon chain [149], suggesting that hydrophobicity is the main driving force for sorption [150].

$${∆\mathrm{G}}_{\mathrm{h}\mathrm{y}\mathrm{d}\mathrm{r}\mathrm{o}\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{b}\mathrm{i}\mathrm{c}}=\mathrm{m}.{∆\mathrm{G}}_{{\mathrm{C}\mathrm{F}}_2}$$

(1)

where m is the number of CF₂ units in the chain and ΔG_CF2 is the Gibbs free energy contribution by each CF₂ unit. This equation shows a linear relationship: the longer the chain (more CF₂ units), the more negative the hydrophobic free energy becomes, making adsorption thermodynamically more favorable [151]. The incorporation of additional CF₂ units in the chain increases the overall molecular polarity, making long-chain PFASs more polar [152].

Zhang & Yazadin conducted computational studies to determine the effect of PFAS chain length on adsorption [19]. As the number of fluorinated carbons increases, regardless of the attached functional group, the Gibbs free energy (ΔG) decreases, indicating that longer chain lengths thermodynamically favor the adsorption of PFASs [19]. Short-chain PFASs exhibit high mobility and low adsorption potential [153]. They are more hydrophilic and smaller in size, which makes them harder to adsorb [153]. Their adsorption mechanism primarily depends on electrostatic interactions between their anionic functional groups and the cationic charges of the adsorbent [154]. Moreover, long-chain PFASs are more likely to form bilayers and aggregate in environmental matrices because of their lower critical micelle concentrations (CMCs) compared to short-chain analogs [44,155]. Kang et al. synthesized a PEI-PVC nanofiber adsorbent to improve short-chain PFAS removal, which usually has low adsorption capacity because of high hydrophilicity and dependence on electrostatic interactions. By adding polyethyleneimine, they increased the positive surface charge, resulting in high adsorption capacities for PFBA (84.26 mg/g) and PFBS (214.37 mg/g) [156]. This approach agrees with Zhang and Yazadin’s findings that short-chain PFASs are more difficult to adsorb, unlike long-chain PFASs, which are thermodynamically favored.

5.2. Effect of Functional Head Groups

Perfluoroalkyl acids (PFAAs) are typically classified into two groups based on their functional head groups: perfluoroalkyl sulfonic acids (PFSAs) and perfluoroalkyl carboxylic acids (PFCAs). PFSAs, like PFOS, contain a sulfonate (-SO₃-) group, while PFCAs, such as PFOA, have a carboxylate (-COO-) group [157]. These functional groups play a crucial role in their adsorption behavior.

According to Pearson’s Hard and Soft Acids and Bases (HSAB) principle, hard bases like the carboxylate group in PFCAs have a high charge-to-radius ratio and are less polarizable. In contrast, the sulfonate group in PFSAs is considered a soft base, meaning it is more polarizable. Based on HSAB theory, soft bases tend to interact more with soft acids (or soft cationic adsorbent sites), while hard bases prefer hard acids. Therefore, PFOS (a soft base) tends to interact more strongly with soft cationic adsorbents, whereas PFOA (a hard base) shows stronger binding with hard cationic sites [47,158].

However, adsorption is not only influenced by acid–base interactions. The polarity of the functional group and the hydrophobic nature of the fluorocarbon tail also impact adsorption. Carboxylate groups in PFCAs are less polar than sulfonate groups, making hydrophobic interactions more significant in PFCA adsorption. In contrast, PFSA adsorption is mainly driven by polar interactions due to the higher polarity and charge delocalization of the sulfonate group [19]. As a result, PFSA molecules (such as PFOS) tend to form stronger electrostatic interactions with polar adsorbents, even when they have long fluorocarbon tails. Meanwhile, PFCAs (such as PFOA) are more effectively adsorbed by hydrophobic interactions due to the greater contribution of van der Waals forces along their less polar chain.

5.3. Combined Effect of Chain Length and Functional Groups

Most of the adsorbents worked efficiently at acidic to neutral pH levels because all the PFASs exist in their anionic form, regardless of the attached acid functional group. The pKa values for short- and long-chain PFAS molecules are provided in

Table 4. Structural features, acidity profiles, and thermodynamic properties of PFAS compounds.

Category	Name	Molecular Structures	Molecular Weights a	-CF2 Units	pKa		Log Kow (Neutral)	ΔGhydrophobic (kJ mol−1) d
Short-chain	PFBS		300.09	(-CF2-)4	−3.3 a	−3.3 e	1.98 c	2.6 e	-
	PFBA		214.04	(-CF2-)3	0.4 a	1.1 e	1.05 b	2.3 e	−10.1
	PFPeS		350.10	(-CF2-)5	0.14 a	−3.3 e	-	3.3 e	-
	PFPeA		264.05	(-CF2-)4	0.17 a	0.34 e	3.19 b	3.0 e	−13.4
	PFHxS		400.11	(-CF2-)6	−3.34 a	−3.3 e	3.44 c	4.0 e	-
	PFHxA		314.05	(-CF2-)5	−0.16 a	−0.78 e	3.99 b	3.7 e	−16.8
	PFHpA		364.06	(-CF2-)6	−0.19 a	−2.3 e	4.40 b	4.4 e	−20.1
Long-chain	PFHpS		450.12	(-CF2-)7	−2.29 a	−3.3 e	4.06 c	4.7 e	-
	PFOS		500.13	(-CF2-)8	−3.27 a	−3.3 e	4.05 c	5.4 e	-
	PFOA		414.07	(-CF2-)7	−0.2 a	−0.5 e	4.67 b	5.1 e	−23.5
	PFNA		464	(-CF2-)8	0.52 a	−6.5 e	5.02 b	5.8 e	−26.8
	PFDA		514.80	(-CF2-)9	0.4 a	−5.2 e	5.44 b	6.5 e	−30.2

Log K_ow are predicted average values for PFSA, while values for PFCA are determined experimentally (neutral). Ref. [78] = ^a, Ref. [160] = ^b, Ref. [161] = ^c, Ref. [162] = ^d, Ref. [159] = ^e.

. Since all the PFAS molecules have very low pKa values, they exist in their anionic form under environmentally relevant conditions [159].

As listed in Table 4, we compared the pKa, log Kow (octanol–water partition coefficient), and ΔG_hydrophobic values to understand the adsorption behavior of PFAS molecules. The very low pKa values for long-chain molecules confirm their more polar behavior, which is especially noticeable among long-chain PFSAs. Higher log Kow values indicate greater hydrophobicity. Long-chain PFASs have significantly higher values, supporting the idea of increased hydrophobicity compared to short-chain PFASs; however, PFSA and PFCA show similar patterns. Although PFSAs are more polar, their Kow values are still higher than those of PFCAs. This suggests a complex interplay of hydrophobic and electrostatic forces when comparing the adsorption of both classes based on chain length. The more negative ΔG value supports the hypothesis that longer chains, being more hydrophobic, adsorb more favorably. However, we lack sufficient data to compare the hydrophobicity of PFSAs and PFCAs directly.

6. Adsorption Mechanism

The PFAS adsorption mechanism is complex and depends on both the adsorbent properties and the PFAS chemical structures. Hydrophobic and electrostatic interactions are widely recognized as the primary driving forces. Most reported studies attribute PFAS removal from water to these two types of interactions. Non-covalent interactions are the main factors in how PFAS tails interact [163], as the Sp³ hybridization of C-F bonds [164]. Having three lone pairs on the fluorine atom increases steric and electrostatic hindrance for a nucleophilic attack on the carbon–fluorine tail, thereby promoting hydrophobic character, especially in long-chain PFASs. The non-polar and hydrophobic C-F bonds do not interact with water molecules and tend to adopt energetically unfavorable orientations [163,165,166]. To achieve energetic stability, they interact with hydrophobic moieties from adsorbents that exhibit the same behavior in water, resulting in hydrophobic interactions [167]. Currently, adsorbents made from natural organic materials such as cellulose, lignin, chitosan, and cyclodextrins—including biochars, activated carbons, and aerogels—provide a large surface area with favorable hydrophobic interactions. As the number of -CF2 units increases, the hydrophobic nature of a PFAS increases, as shown in Equation (1), thereby improving adsorption through hydrophobic interactions [47,168].

Electrostatic interactions typically occur between the functional acid head groups of PFAS molecules (anionic) and charged functional sites (cationic) of adsorbents [169]. Although the F atom attached to the carbon chain has three lone pairs, the electrostatic interaction it induces is weak [47,170]. The pH of water greatly affects their adsorption because, at acidic pH, PFAS molecules deprotonate and exist as anionic species. Above pH 7, they are neutral, so electrostatic interactions are unlikely to occur. Hafnium oxide ceramics were used to examine how pH influences the adsorption of POFA by Bronsted acid–base sites. The material’s adsorption capacity decreased from 20.9 mg/g at pH 2.3 to 9.2 mg/g at pH 6.3 due to increased coulomb repulsions [171].

Another method for PFAS adsorption is fluorophilic interaction achieved by introducing fluorine atoms to the adsorbent through chemical modifications [172]. Fluorine atoms are highly electronegative, enabling them to polarize nearby atoms. However, fluorine itself exhibits low polarizability, resulting in weaker interactions with other molecules [173]. Therefore, they tend to cluster together to minimize interactions with non-fluorous compounds and enhance interactions with fluorous compounds. This type of interaction is specific to fluorous adsorbents because of their special design. Fluorophilic interactions are stronger than hydrophobic and electrostatic interactions and are unaffected by the presence of organic matter or co-contaminants [163,174]. Long-chain PFASs with higher fluorophilic characteristics are more fluorophilic, especially PFSA, due to one more CF2 unit compared to PFCA. Ionic flurogels containing abundant quaternary amines and fluorine groups were synthesized through radical copolymerization. One such material, Ionic Fluorogel (IF-20+), showed a 95% removal efficiency for both long- and short-chain PFASs from water collected at the Sweeney Water Treatment Plant in Wilmington, North Carolina. The study emphasized the synergistic effect of fluorophilic and electrostatic interactions as key mechanisms driving the adsorption process [175].

The hydrogen bond between the fluorine atom of PFAS molecules and the hydrogen atom of the adsorbent also contributes to PFAS removal [176]. The strength of these hydrogen bonds usually increases as the PFAS carbon chain lengthens [177]. But the hydrogen bond formed between the PFAS–adsorbent is weaker than the hydrogen bonding in water molecules. So, the role of hydrogen bonding is insignificant in the PFAS adsorption process [176,178]. Zarei et al. reported the combined role of electrostatic interaction and hydrogen bonding using a viologen-based COF for the efficient removal of PFOS and PFOA at very low concentrations (3.8 ng/L) [179].

In conclusion, we can say that PFAS removal from water is governed by a complex interplay of interactions, mainly electrostatic, hydrophobic, fluorophilic interactions, and hydrogen bonding, as shown in Figure 9. The strong hydrophobic interactions arise from the hydrophobic parts of adsorbents and the tails of PFAS molecules. Electrostatic interactions mainly occur between PFAS heads and positively charged functional groups on the adsorbent. Fluorophilic interactions are specific to the adsorbent, while the role of hydrogen bonding remains minimal, although some studies have reported it.

7. Comparative Study of PFAS Removal Technologies

Many PFAS removal technologies have been investigated; however, non-destructive techniques, such as membranes, ion exchange resins (IERs), and activated carbons (ACs), have been commercialized and tested at wastewater treatment plants. Still, no single technology is sufficient for the complete removal of PFASs. In this section, we provide a detailed comparison of PFAS removal technologies, as shown in

Table 5. Comparative cost analysis and scalability of PFAS removal technologies.

Technology		Removal Efficiency	Cost	Scalability	Limitations	References
Non-Destructive	Sustainable polymer adsorbent (aerogel and hydrogel)	>99%	In research	Lab → Pilot scale	Upscaling, disposal of spent adsorbent	-
	Activated carbon (GAC) a	>90%	USD 0.44/m3 WW b	Full-scale TRL c-9	Less removal of SC d, interference of NOM e	[182]
	Ion exchange resins (regenerable)	>90%	USD 0.40/ m3 TW f	Full scale TRL c-9	Regeneration cost	[182]
	Membranes (NF g, MF h and RO i)	>99% RO 90–99% NF g	NF USD 0.016–0.16/m3 TW f	Full scale TRL c-9	Membrane fouling, high energy consumption	[184]
	Coagulation/ flocculation	1–50%	N/A n	DT j	Less removal efficiency	[184]
Destructive	Electrochemical oxidation	60–99%	High cost	Bench → Pilot scale TRL c-7	Less effective for SC d, high cost, electrode stability, and hazardous byproducts	[99]
	Advanced oxidation and reduction process	75% AOP k >90% ARP l	High cost	Full scale TRL c-9 AOP TRL c-5 ARP	Do not fully defluorinate the tail, less efficiency for SC	[184]
	Plasma degradation	>99%, LC m >99%, SC d	-	Lab → Pilot scale	pH sensitive, NOM e interference, long treatment time	[99,185]

^a—granular activated carbon, ^b—wastewater, ^c—technology readiness level, ^d—short-chain, ^e—natural organic matter, ^f—treated water, ^g—nanofiltration, ^h—microfiltration, ⁱ—reverse osmosis, ^j—developing technology, ^k—advanced oxidation process, ^l—advanced reduction process, ^m—long-chain, ⁿ—not available, →: Research is underway and it’s in process of upscaling from lab to pilot scale.

and Figure 10, regarding cost, scalability, and limitations, to give readers a clear understanding of each process. Sustainable polymers, such as cellulose and chitin, have attracted significant attention as affordable and readily available materials for use as sustainable adsorbents. However, these natural materials need to be scaled up into ready-to-use technologies, as many adsorbents are currently tested only in lab-scale batch reactors. Additionally, more research is needed to determine the costs of adsorbent production, regeneration, and disposal of spent materials. Nonetheless, these sustainable polymer adsorbents have the potential to be functionalized to enhance the adsorption of both short- and long-chain PFASs. Extensive research has been conducted on the cost analysis of GAC, IER, and electrochemical oxidation, which we briefly summarized in this section. Liang et al. conducted a pilot study coupling IER with electrochemical oxidation on-site. IER was able to remove PFOA and PFOS to below detectable levels, requiring regeneration after 6–12 months of operation. The concentrated stream was further treated with a Ti₄O₇ electrode, consuming a total of 0.131–0.161 kWh/m³ for PFOA and 0.071–0.094 kWh/m³ for PFOS [180]. A report by the Minnesota Pollution Control Agency (MPCA) states that the total cost for treating one cubic meter of water with regenerable IER is USD 0.40/m³. GAC costs up to USD 0.44 per cubic meter, while nanofiltration membranes have been cost-effective, with costs ranging from about USD 0.016 to USD 0.16 per cubic meter. IER offers a significantly longer service life, approximately 16 times that of GAC, and is at least 20% less expensive across the evaluated price points. GAC can be more cost-effective only if its performance improves by 25% or if its cost drops below USD 2.75/kg, whereas IX performance declines or its cost rises above roughly USD 20.35/kg [181]. Ellis et al. reported that the cost of PFAS removal by single-use and regenerable GAC systems ranges from USD 0.05 to USD 1.00 per cubic meter [182]. In another study, Jiang et al. mention the operating costs for PFAS removal by GAC and AER: USD 0.64/m³ and USD 0.24/m³, respectively [64,183].

When using electrochemical oxidation, pretreatment systems such as IER, RO, and NF can reduce total energy consumption by 50% [186]. However, it still comprises 85% of the total cost of the combined system [187,188]. These are the operating costs of the system after installation. Installation costs are not included. The hybrid system of electrochemical oxidation and ARP removed 90% of PFOS and PFOA from the water by using less than 20 kWh/m³ of energy [189].

8. Conclusions and Future Perspective

This review offers a detailed analysis of how PFAS chain length and functional groups impact their adsorption behavior. Although no single technique has proven to be universally effective for PFAS removal, adsorption remains the most cost-efficient and sustainable method for reducing these “forever chemicals.” Its effectiveness, however, depends on a complex interplay of factors including adsorbent structure, surface charge, pore size, and performance in challenging environments with co-contaminants. Both chain length and functional groups significantly influence PFAS toxicity and their removal efficiency across various adsorbents. Despite notable progress, many of these challenges still need to be fully addressed. This review not only highlights the core difficulties of PFAS adsorption but also offers a new perspective to guide future research toward more effective and sustainable solutions.

8.1. Sustainable Approach

It has been observed that some spent adsorbents are even more difficult to dispose of than PFAS molecules because recycling these adsorbents creates secondary pollutants that are even more toxic. Sustainable solvents and materials should be used to synthesize cost-effective adsorbents. Natural materials such as cellulose, chitin, and lignin have great potential as precursors for PFAS removal. Additionally, adsorbents are often overused to remove more PFASs, which increases costs and leaves some material unused. Regenerating adsorbents over multiple cycles can also reduce their ability to hold PFAS molecules, so these factors should be considered when designing and synthesizing adsorbents.

8.2. Principles for PFAS-Adsorbent Interactions

PFAS adsorption occurs through different primary forces, such as (1) electrostatic interactions between PFAS functional head groups (anionic) and the adsorbent surface charge (cationic), (2) hydrophobic interactions (weak interactions) between the C-F tail of PFASs and hydrophobic regions of the adsorbent, and (3) fluorophilic interactions (F-adsorbent specific) between fluorine moieties of PFASs and the adsorbent. The physicochemical properties of adsorbents determine which type of interaction will be predominant in the adsorption process. Short-chain molecules mainly interact through charge–charge interactions, while hydrophobic van der Waals forces dominate among long chains. Therefore, adsorbents should be designed to effectively remove both short- and long-chain PFASs, regardless of the functional groups of the PFAS molecules.

8.3. Current Research Gaps in PFAS Adsorption Studies

While reporting PFAS concentrations, researchers have used various units, which makes direct comparisons between studies difficult. Most research to date has focused on legacy PFAS compounds like PFOS and PFOA, with little attention paid to mixed matrices containing both long- and short-chain PFASs. Additionally, limited work has been conducted to assess adsorbent performance under realistic environmental conditions; many studies test at unrealistically high concentrations that do not reflect actual contamination levels. Although many laboratory-scale experiments have been conducted, there is a notable lack of pilot- or industrial-scale studies. It is also concerning that there is no comprehensive cost analysis covering the entire process, from adsorbent synthesis and PFAS removal testing to regeneration cycles. Filling these gaps is crucial for turning laboratory success into practical, scalable, and sustainable PFAS remediation solutions.