MOF-based catalytic systems for PFAS degradation: mechanisms, composite design and future challenges

1 Introduction

Per- and polyfluoroalkyl substances (PFAS) are large and structurally diverse class of synthetic organic compounds that have emerged as persistent global pollutants. They are synthetic compounds featuring a hydrophobic fluorinated hydrocarbon backbone of variable length in which hydrogen atoms are partially or fully substituted with fluorine, forming one of the strongest chemical bonds with a dissociation energy of approximately 488 kJ mol⁻¹. This exceptional bond strength supports their remarkable thermal stability and chemical resistance including both hydrophobic and lipophobic behavior. Due to these properties PFAS have been incorporated into a wide range of industrial and consumer applications such as non-stick cookware, food packaging, carpets, textiles, upholstery and fire-fighting foams. Their structural versatility has led to the identification of more than 14,000 individual compounds typically grouped by chain length and terminal functionality into subclasses such as perfluoroalkyl acids, perfluoroalkyl iodides, and polymeric fluoropolymers reflecting its chemical diversification in manufacturing sectors. The global production of PFAS has gradually increased steadily over recent decades, exceeding one million tonnes per year. In 2024, production is estimated at approximately 0.7 million tonnes, with projections reaching around 1.35 million tonnes by 2034 at an annual growth rate of 5 to 6 percent, indicating a rapidly expanding trend (Evich et al., 2022). Moreover, they are highly resistant to natural degradation, with their half-lives ranging from 50 years in water to more than 200 years in soil. Their persistence and continuous release into the environment have led to widespread contamination which poses serious threat to aquatic ecosystems and human health.

PFAS exposure occurs through multiple pathways including ingestion of contaminated food and water, inhalation of polluted air and dermal contact with PFAS-containing materials (Figure 1). The United States Food and Drug Administration’s (US FDA’s) seafood survey in 2022 concluded that dietary intake particularly from water and seafood has been identified as a major exposure route (74%) for the general population (FDA, 2022). In addition, occupational exposure is significant in professions such as firefighting and fluorochemical manufacturing. Li et al. demonstrated that the persistence of PFAS in humans is largely attributed to their strong binding affinity to plasma proteins (Li Y. et al., 2018). Both perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS) exhibit high protein binding in humans and animal models, leading to prolonged retention. PFOS shows a strong tendency to accumulate in the liver and serum where it can disrupt normal hepatic function, with its elevated levels partly explained by enterohepatic recirculation. PFAS are widely distributed in aquatic environments originating from sources such as industrial discharge, landfill leachate, and wastewater treatment plants. High concentrations have been reported near industrial regions, with surface water levels reaching up to 13.5 μg/L in China and 11.34 μg/L of PFOA in Alabama, USA (Zhu et al., 2023). Even in less impacted regions across Europe, Asia, and North America, concentrations frequently exceed 70 ng/L. Exposure to PFAS has been linked to biological effects in aquatic organisms, including altered growth and physiological changes. Schultes et al. reported correlations between PFAS accumulation and changes in liver somatic index and body size in Baltic cod (Schultes et al., 2018). Hanssen et al. observed bioaccumulation of multiple PFAS compounds in mussels (Hanssen et al., 2010). Additionally, exposure to specific PFAS such as 6:2 Cl-PFAES, HFPO-DA, and PFECHS has been shown to reduce the growth of marine microalgae such as Chlorella, demonstrating broad-spectrum toxicological effects across environmental and biological systems. The surfactant nature of PFAS enhances their mobility in groundwater, enabling transport far beyond their original sources while resisting natural degradation. Among existing technologies, granular activated carbon (GAC), anion exchange resins (AIX), and membrane-based systems are the most widely applied. GAC removes PFAS mainly through hydrophobic and electrostatic interactions while AIX operates through ion exchange mechanisms. Membrane processes such as nanofiltration and reverse osmosis rely on pressure-driven separation, effectively rejecting PFAS based on size exclusion. However, these approaches require high energy input, complex operation, and elevated costs, and they do not achieve mineralization producing instead concentrated waste streams that require high-temperature incineration often exceeding 1100 °C (Kuchar et al., 2017).

Metal-organic frameworks (MOFs) have recently gained attention as promising materials for PFAS remediation due to their well-defined porous structures, large surface areas, and tunable chemical properties. Constructed through coordination of metal nodes with organic linkers, MOFs allow precise control over pore size, surface functionality, and electronic structure (Li et al., 2024). Currently, more than 20,000 distinct MOFs have been reported, each engineered for applications including gas storage, catalysis, and environmental remediation (Furukawa et al., 2013). In the context of PFAS, MOFs offer potential advantages over traditional materials by improving adsorption efficiency and enabling degradation through photocatalytic and redox-mediated pathways. Although several reviews have addressed PFAS adsorption on porous materials or the general photocatalytic destruction of PFAS, a focused critical analysis of multifunctional composite and polymetallic MOF systems specifically engineered for PFAS mineralization remains lacking. The central idea of this review is that the decisive advances in MOF-based PFAS treatment are emerging not from pristine single-metal frameworks but from their rational integration into multi-component architectures MOF@metal-oxide heterojunctions, MOF@carbon hybrids and bi-/tri-metallic frameworks in which complementary functions (broadened light absorption, suppressed charge recombination, coupled oxidative–reductive radical chemistry and enhanced hydrolytic stability) operate cooperatively within a single platform. Accordingly, this review is organized around three questions that define its focus first what specific structural and electronic advantages composite and polymetallic MOFs confer for C-F bond cleavage and complete defluorination, relative to single MOFs and conventional adsorbents, second how large the quantitative performance gap is measured by defluorination and mineralization rather than parent-compound removal between these advanced systems and both pristine MOFs and commercial granular activated carbon (GAC) and anion-exchange (AIX) media and third which design strategies offer the most credible route toward the still-unsolved short-chain PFAS bottleneck. By foregrounding mechanism structure performance relationships and benchmarking against commercial technologies, this work moves beyond a descriptive catalogue of materials toward an actionable design framework for next-generation MOF catalysts targeting complete PFAS mineralization.

2 Fundamental properties of PFAS

Because the physicochemical behavior of PFAS, C-F bond stability, amphiphilicity and chain-length-dependent persistence directly governs how these molecules interact with MOF active sites, only the features relevant to catalytic degradation are discussed here. The degree of fluorination fundamentally influences the physicochemical behavior of these compounds, with fully fluorinated substances exhibiting greater persistence compared to their partially fluorinated counterparts.

According to the length of the carbon chains, PFAS are classified into two types: long-chain PFAS and short-chain PFAS. Long-chain PFAS refer to perfluoroalkyl sulfonic acids (PFSAs) with at least six carbons and perfluoroalkyl carboxylic acids (PFCAs) with at least eight carbons; short-chain variants refer to PFCAs with at most seven carbons and PFSAs with at most five carbons. Due to the increasing chemical inertness for long-chain PFAS, it is more challenging for them to be degraded. Consequently, researchers have placed emphasis on developing systems targeting long-chain PFAS over the years. However, with international regulations targeting long-chain PFAS being proposed, global production has shifted towards short-chain PFAS, resulting in a notable rise in studies concerning their removal (Cousins et al., 2022).

2.1 C-F bond strength and thermodynamic recalcitrance

The exceptional resistance of PFAS to degradation is primarily governed by the intrinsic strength and electronic characteristics of the carbon fluorine (C-F) bond. With bond dissociation energies typically in the range of 480–500 kJ mol⁻¹, the C-F bond represents one of the most stable linkages in organic chemistry (O'Hagan, 2008). This high stability originates from the significant electronegativity difference between carbon and fluorine (Δχ ≈ 1.57 on the Pauling scale), which results in strong polarization and tight electron localization toward the fluorine atoms. Consequently, the bond exhibits a short bond length (∼1.35 Å) and high bond energy, limiting its susceptibility to cleavage under conventional environmental conditions. Furthermore, the dense arrangement of fluorine atoms along the carbon backbone creates a steric and electronic shielding effect that protects the chain from external reactive species, a phenomenon termed the fluorine umbrella effect. Quantum chemical calculations have demonstrated that the activation energy for homolytic C–F bond cleavage in perfluoroalkyl chains exceeds 450 kJ mol⁻¹ under thermal conditions, far beyond what is achievable through conventional oxidative processes (Bentel et al., 2019). Reductive pathways, by contrast, can lower this activation barrier through outer-sphere electron transfer mechanisms, where photoexcited electrons in semiconductor or metal-node systems are transferred directly to the antibonding σ* orbital of the C-F bond, facilitating defluorination. This mechanistic insight underpins the growing interest in photocatalytic and reductive approaches to PFAS degradation, where electronic excitation-rather than brute thermal energy drives C-F bond cleavage. Xiao et al. demonstrated that under UV irradiation, indirectly photolytic defluorination of PFOA could achieve defluorination efficiencies exceeding 60% within 3 h through a stepwise chain-shortening mechanism (Xiao et al., 2013).

2.2 Influence of carbon chain length

Carbon chain length is a fundamental structural factor that influences the degradation behavior of PFAS. As the perfluoroalkyl chain becomes longer, molecular size and electron density distribution increase, which strengthens intermolecular interactions and enhances overall chemical stability. This is mainly due to the cumulative effect of multiple C-F bonds, which collectively reinforce resistance to bond cleavage. In contrast, short-chain PFAS possess fewer carbon–fluorine units and generally show slightly lower persistence, although they still remain highly resistant to conventional degradation pathways. From a degradation perspective, chain length directly affects how PFAS molecules interact with reactive sites on catalysts and advanced materials such as MOFs.

Long-chain PFAS tend to exhibit stronger hydrophobic interactions and are more likely to accumulate at hydrophobic or partially hydrophobic catalytic interfaces, with this local enrichment near active sites enhancing degradation probability under suitable catalytic conditions. However, their increased steric bulk can also limit diffusion into confined catalytic pores if the framework is not appropriately engineered. Wang et al. demonstrated that MOFs with tailored pore environments could accommodate PFCA molecules of different chain lengths and promote their degradation through enhanced host guest interactions (Wang et al., 2017). In such systems, confinement within MOF channels increases local concentration and improves interaction probability with catalytic metal nodes, facilitating C-F bond activation under appropriate conditions. Short-chain PFAS, being more hydrophilic and mobile, may diffuse away more rapidly, reducing interaction time and making their degradation particularly challenging a critical limitation that current MOF-based systems have yet to fully resolve.

2.3 Functional group chemistry

PFAS molecules contain at least one ionizable functional group at their terminus, predominantly carboxylate (–COOH) or sulfonate (–SO₃H) moieties. Based on these terminal functionalities, PFAS are categorized into two principal groups: perfluoroalkyl carboxylic acids (PFCAs) and perfluoroalkyl sulfonic acids (PFSAs) (Figure 2) (Li et al., 2023). The substantial electronegativity of fluorine atoms combined with dissociable end groups imparts a net negative surface charge to PFAS molecules under most environmental conditions, with charge density concentrated at the functional head region. This charge distribution significantly influences degradation pathways and efficiency. PFOS exhibits a pKa of approximately 3.72, whereas PFOA demonstrates a pKa near 0.2, indicating stronger acidity. Consequently, PFOS maintains higher ionization levels across broader pH ranges compared to PFOA, affecting its reactivity with degradation agents. Strategic pH manipulation can modulate the protonation state of PFAS molecules, thereby influencing their susceptibility to oxidative, reductive, or photocatalytic degradation mechanisms (Gao et al., 2017).

3 Architectural principles of metal organic frameworks (MOFs)

Metal organic frameworks (MOFs) are crystalline hybrid materials constructed from inorganic metal nodes or clusters interconnected by polytopic organic linkers through strong coordination bonds, yielding periodic three-dimensional networks with exceptional porosity and structural programmability (Figure 3). This modular architecture provides a uniquely versatile platform for addressing the formidable challenges posed by PFAS. The metal nodes typically high-valent transition metals or zirconium clusters—serve a dual functional role: they act as strong Lewis acid sites capable of coordinating directly to the hydrophilic head groups of PFAS molecules (carboxylate, sulfonate, or phosphonate moieties), and they can function as catalytic centers for subsequent C–F bond activation (Li J. et al., 2018). The development of MOFs began in the late 1990s, with a major milestone achieved in 1999 when Omar M. Yaghi and colleagues reported MOF-5, establishing the foundation for modern MOF chemistry. The organic linkers, generally aromatic carboxylates, provide hydrophobic pore environments that facilitate fluorophilic interactions with the perfluorinated carbon tails of PFAS through van der Waals forces and, in certain frameworks, π–π stacking with fluorinated aromatic linkers engineered for enhanced affinity. Pore dimensions can be systematically tuned through linker elongation or topology selection to achieve size-selective capture of specific PFAS chain lengths (Wu et al., 2023). The band gap and light absorption characteristics of photoactive MOFs can also be modulated through mixed-metal node strategies or linker amination, enabling visible-light-driven generation of reactive species capable of initiating C–F bond cleavage. Cheng et al. demonstrated that the incorporation of -NH₂ functionalization into UiO-66 red-shifted light absorption from ∼350 nm to ∼500 nm, dramatically enhancing photocatalytic activity under simulated solar irradiation (Cheng et al., 2018). A critical conceptual distinction in MOF-based PFAS remediation lies between passive adsorptive removal and active catalytic destruction. Adsorption-driven by a synergistic combination of electrostatic attraction to the PFAS head group, fluorophilic partitioning of the perfluorinated tail into hydrophobic MOF cavities, Lewis acid–base coordination at metal nodes, and in some cases anion-exchange interactions represents the initial capture phase and is often remarkably efficient. MOFs such as MIL-101(Cr) and UiO-66 have demonstrated exceptional PFAS uptake capacities exceeding those of conventional activated carbons and ion-exchange resins, attributable to their ultrahigh surface areas (up to 5900 m² g⁻¹ for MIL-101) and precisely tailored pore environments (Férey et al., 2005). However, adsorption alone merely transfers PFAS from the aqueous phase to a solid matrix, producing a concentrated toxic waste stream that necessitates energy-intensive thermal destruction or indefinite landfill storage—neither of which guarantees permanent elimination. Catalytic degradation, by contrast, targets the chemical Achilles’ heel of PFAS: the carbon–fluorine bond. In hydrolytic and thermal catalytic pathways, Lewis acidic metal nodes coordinate to the carboxylate or sulfonate head group, polarizing the adjacent C-C and C-F bonds and lowering the activation energy for defluorination. In photo-oxidative and photocatalytic mechanisms, MOFs generate reactive oxygen species capable of initiating reductive defluorination. The ultimate measure of success is complete mineralization conversion of organic fluorine to inorganic fluoride ions, CO₂, and SO₄ ²⁻ rather than mere transformation to shorter-chain perfluoroalkyl acids that remain environmentally persistent (Maimaiti et al., 2018).

4 Mechanisms of PFAS degradation using MOFs

The degradation of PFAS by MOF-based systems proceeds through several mechanistically distinct yet often synergistic pathways. These include adsorption-coupled surface reactions, photocatalytic oxidation and reduction, Fenton and Fenton-like radical processes, and direct reductive defluorination facilitated by electron-rich MOF structures or photogenerated electrons. The dominant mechanism in any given system is governed by the metal node identity, organic linker properties, pore architecture, solution chemistry, and the presence of co-oxidants or light sources. A comprehensive mechanistic understanding is essential for rational design of high-performance MOF catalysts capable of achieving complete PFAS mineralization under practical treatment conditions.

4.1 Adsorption-coupled degradation

4.1.1 Electrostatic and coordinative interactions

The initial step in most MOF-based PFAS degradation systems involves the accumulation of PFAS molecules at or within the framework a prerequisite for effective catalytic contact. Under environmentally relevant pH conditions (pH 6–8), PFAS exist predominantly in their deprotonated anionic form, owing to their low pKa values (PFOA: ∼0.2; PFOS: ∼3.72). This negative charge creates favorable electrostatic attraction toward positively charged or Lewis acidic metal nodes within the MOF structure. Coordinatively unsaturated metal sites such as Cr³⁺ in MIL-101, Zr⁴⁺ in UiO-66, and Fe³⁺ in MIL-53 directly coordinate with the carboxylate or sulfonate head groups of PFAS, forming inner-sphere surface complexes that not only concentrate the pollutant at catalytic sites but also polarize the adjacent C–F bonds, reducing the activation energy required for subsequent bond cleavage (Serre et al., 2002). Du et al. employed density functional theory (DFT) calculations to elucidate the binding geometry of PFOA at Cr³⁺ Lewis acid sites in MIL-101(Cr), demonstrating that the terminal carboxylate group coordinates in a bidentate bridging mode, inducing significant electron density withdrawal from the α-carbon and weakening the adjacent C-F bond (Du et al., 2014). This coordinative activation represents a fundamentally different mechanism from simple physisorption, as it directly contributes to the thermodynamic accessibility of C-F bond cleavage pathways. Such Lewis acid-mediated activation has been shown to reduce the C-F bond dissociation energy by 15–25 kJ mol⁻¹ in computationally modeled systems, a reduction sufficient to facilitate catalytic defluorination at ambient temperatures under appropriate light or oxidant conditions.

4.1.2 Hydrophobic and fluorophilic interactions

Beyond coordinative interactions at the head group, the perfluorinated carbon tail of PFAS engages in powerful hydrophobic partitioning within the nonpolar pore environment characteristic of many MOFs. Frameworks constructed from aromatic linkers such as terephthalate (BDC), 2-aminoterephthalate (BDC-NH₂) or trimesate (BTC) create pore environments with substantial hydrophobic character that exhibit preferential affinity for the fluorocarbon backbone through London dispersion forces and fluorine–fluorine van der Waals interactions. Sörengård et al. demonstrated that MOFs with highly hydrophobic pore surfaces showed PFAS distribution coefficients (Kd) one to two orders of magnitude higher than conventional activated carbons, attributable to this enhanced fluorophilic partitioning (Sörengård et al., 2020). Furthermore, the rigid pore geometry of MOFs creates a confinement effect that increases the contact time between PFAS molecules and catalytic sites an important advantage over homogeneous catalytic systems where diffusion-limited kinetics constrain performance. Huang et al. reported that PFAS molecules confined within UiO-66 mesopores exhibited first-order degradation kinetics with rate constants approximately 3.8-fold higher than those observed for surface-adsorbed molecules on amorphous silica supports, attributing this enhancement to pore confinement-induced increase in reactive collision frequency (Huang et al., 2017). This spatial integration of adsorptive preconcentration and catalytic activation represents the defining mechanistic advantage of MOF-based systems over conventional granular materials. Su et al. demonstrated the successful application of ZIF-67@C₃N₄ and MIL-100(Fe)@C₃N₄ heterostructures for the elimination of perfluorooctanoic acid (PFOA) from water (Su et al., 2023). These composites were fabricated through an in situ hydrothermal process and combined the adsorption capability of MOFs with the photocatalytic properties of graphitic carbon nitride (g-C₃N₄) (Figure 4). The introduction of MOF components not only increased the available surface area and adsorption sites but also promoted visible light harvesting and more efficient separation of photogenerated electron hole pairs. Consequently, enhanced production of reactive oxygen species contributed to improved PFOA degradation. Under optimized experimental conditions, the ZIF-67@C₃N₄ and MIL-100(Fe)@C₃N₄ composites achieved removal efficiencies of 79.2% and 60.5%, respectively, demonstrating the effectiveness of coupling adsorption and photocatalytic degradation within a single platform. Modification of MIL based frameworks has also emerged as an effective strategy for improving PFAS adsorption performance. Azmi et al. prepared a hydrolyzed polyacrylamide functionalized MIL-96 material (MIL-96-RHPAM2) through an in situ synthesis route (Mohd Azmi et al., 2021). Although polymer incorporation reduced pore accessibility and specific surface area, the modified adsorbent displayed a remarkable PFOA uptake capacity of 340 mg g⁻¹, surpassing many conventional carbon based adsorbents. The enhanced performance was primarily attributed to the presence of abundant positively charged and polar functional groups, including Al³⁺, –COOH, –OH, –NH₂, and NH₃ ⁺, which promoted strong interactions with PFOA molecules. However, the decrease in porosity led to slower adsorption rates. Mechanistic analysis indicated that electrostatic attraction served as the dominant adsorption pathway, while hydrogen bonding, hydrophobic interactions, and van der Waals forces also contributed to pollutant capture. In another study, Yang et al. evaluated the adsorption behavior of Fe-BTC, MIL-100-Fe, and MIL-101-Fe toward PFOA (Figure 4). Among the investigated materials, Fe-BTC exhibited the greatest adsorption capacity of 404.4 mg g⁻¹ despite possessing the smallest surface area. This superior performance was linked to the abundance of accessible Lewis acidic sites that facilitated strong acid base interactions with PFOA molecules. Adsorption data were best described by the Langmuir isotherm model, suggesting monolayer adsorption on homogeneous active sites. Rapid pollutant uptake occurred during the first hour of contact, while equilibrium was generally established within approximately 3 hours (Yang et al., 2020). Increasing solution pH negatively affected adsorption efficiency, which was attributed to reduced electrostatic attraction between the adsorbent surface and PFOA species.

4.2 Photocatalytic degradation

4.2.1 Band gap engineering and light absorption

Photoactive MOFs function as semiconductor analogs in which the organic linkers serve as light-harvesting antennae and the metal nodes provide electron-accepting conduction band states. The energy gap between the highest occupied molecular orbital (HOMO) of the linker and the lowest unoccupied molecular orbital (LUMO) of the metal cluster determines the optical band gap (Eg) and, consequently, the portion of the solar spectrum that can drive photocatalytic reactions. Pristine UiO-66(Zr) possesses a wide band gap of approximately 4.0 eV, restricting its activity to UV irradiation (<310 nm). Zhang et al. demonstrated that amino-functionalization of the BDC linker (UiO-66-NH₂) narrows the band gap to approximately 2.75 eV through the introduction of a new HOMO state from the–NH₂ lone pair, enabling significant light absorption in the visible range up to ∼450 nm (Zhang et al., 2020). This strategy—linker functionalization for band gap reduction—has become a central design principle in photoactive MOF engineering for PFAS applications.

Mixed-metal or bimetallic node strategies offer an additional dimension of band gap control. The incorporation of Fe³⁺ into otherwise UV-active Zr-MOF structures creates ligand-to-metal charge transfer (LMCT) transitions that extend absorption into the visible region. Wang et al. synthesized a series of bimetallic Fe/Zr-MOFs and demonstrated a systematic red-shift in absorption onset from 380 nm (pure Zr-MOF) to 520 nm (20 mol% Fe substitution), with photocatalytic PFOA degradation efficiency under visible light increasing proportionally with Fe content up to an optimal loading of 15 mol% (Wang et al., 2016). Beyond this threshold, competitive recombination at excess Fe sites diminished photocatalytic performance, highlighting the importance of optimizing metal ratios in mixed-node architectures.

4.2.2 Reactive oxygen species generation and C-F bond attack

Upon photoexcitation, electrons promoted to the conduction band of the MOF react rapidly with dissolved molecular oxygen to generate superoxide radical anions (O₂•^-, E° = −0.33 V vs. NHE), while the corresponding valence band holes react with water or surface hydroxyl groups to produce hydroxyl radicals (•OH, E° = +2.80 V vs. NHE). These highly reactive oxygen species (ROS) attack PFAS through fundamentally different mechanisms. Hydroxyl radicals, possessing the highest oxidation potential among commonly generated ROS, primarily attack the hydrophilic carboxylate or sulfonate head group of PFAS, triggering an oxidative decarboxylation step analogous to the Kolbe electrolysis mechanism. This generates a perfluoroalkyl radical (C_nF_2n+1•) that undergoes sequential defluorination, producing shorter-chain PFCA homologs and ultimately releasing inorganic fluoride (Hori et al., 2005).

Superoxide radical anions, by contrast, mediate reductive defluorination through direct electron injection into the C–F antibonding orbital (Figure 5). Lin et al. employed electron paramagnetic resonance (EPR) spectroscopy combined with radical scavenging experiments to demonstrate that O₂•^- was the predominant ROS responsible for C-F bond cleavage in a MIL-101/rGO photocatalytic system, achieving a defluorination efficiency of 72.6% after 150 min under visible light irradiation (Lin et al., 2015). The presence of both oxidative (•OH) and reductive (O₂•^-, photogenerated e⁻) species in well-designed MOF photocatalysts enables a dual-pathway attack on PFAS oxidative decarboxylation from the functional head combined with reductive defluorination along the fluorocarbon chain synergistically accelerating the overall mineralization rate.

4.3 Fenton and fenton-like reactions

Iron-containing MOFs represent the most extensively investigated class of Fenton-active catalytic platforms for PFAS degradation, owing to the natural abundance of iron, its well-established redox chemistry, and the structural flexibility of Fe-carboxylate frameworks. The Fenton mechanism in MOF systems proceeds through the generation of hydroxyl radicals via reduction of hydrogen peroxide (H₂O₂) by structural Fe²⁺ sites: Fe²⁺ + H₂O₂ → Fe³⁺ + •OH + OH⁻ (k ≈ 76 M⁻¹ s⁻¹). Subsequent regeneration of Fe²⁺ from Fe³⁺ by reaction with H₂O₂ or organic reductants completes the catalytic cycle: Fe³⁺ + H₂O₂ → Fe²⁺ + HO₂• + H⁺. Cao et al. reported that MIL-53(Fe) activated H₂O₂ at pH 3.5 to achieve 79.4% PFOA removal within 90 min, with EPR analysis confirming •OH as the primary reactive species (Cao et al., 2010). The confined pore environment of the MOF was proposed to enhance •OH radical lifetime by reducing diffusive losses, contributing to the superior performance compared to homogeneous Fe²⁺/H₂O₂ systems.

Extended Fenton-like systems employing persulfate (PS, S₂O₈ ^2-) or peroxymonosulfate (PMS, HSO₅ ⁻) as oxidants have attracted considerable attention due to the generation of sulfate radicals (SO₄•^-, E° = +2.60 V vs. NHE) alongside •OH. Sulfate radicals are particularly effective for PFAS degradation because their longer half-life (∼30–40 μs, versus ∼20 ns for •OH in environmental matrices) and selectivity toward electron-rich functional groups including the carboxylate terminus of PFAS enhance steady-state radical concentrations at catalytic interfaces. Liu et al. demonstrated that Fe-BTC activated PMS at pH 5.0 to generate a SO₄•⁻-dominated reactive environment, achieving 83.5% PFOS removal and 48.2% defluorination within 60 min—performance substantially superior to that of homogeneous Fe³⁺/PMS systems at equivalent Fe loadings (Liu et al., 2012). The heterogeneous nature of MOF-based Fenton catalysts provides additional stability advantages as structural Fe sites are less susceptible to precipitation or complex formation than free aqueous Fe ions. Elucidating the contributions of individual radical species in complex MOF-Fenton systems requires careful application of radical scavenging and kinetic modeling approaches. The discriminant scavenger approach employing methanol (•OH + SO₄•⁻ scavenger), tert-butanol (selective •OH scavenger), and p-benzoquinone (O₂•^- scavenger) has been widely applied to deconvolution radical contributions. Li et al. employed this approach in characterizing the PFOA degradation mechanism in a MIL-88B(Fe)/H₂O₂ system, demonstrating that •OH accounted for approximately 68% of the total oxidative capacity while Fe-oxo species contributed the remaining 32%, with direct C-F bond cleavage occurring at the α-carbon adjacent to the carboxylate terminus (Li et al., 2014). Isotopic labeling studies using ¹⁸O-labeled H₂O₂ have further confirmed that the oxygen incorporated into carboxylate degradation products originates predominantly from the peroxide rather than from dissolved water, providing mechanistic evidence for the Haber-Weiss radical cycle operating within MOF catalytic pockets. The photo-electro-Fenton system was then further advanced into the SPEF system by utilizing a more sustainable and readily available light source sunlight. Following the design principles of SPEF, Wang et al. introduced a dual-function MOF/carbon nanofiber (MOF/CNF) composite membrane (Figure 6a) for efficient solar photo-electro-Fenton degradation of PFOA (Wang et al., 2022). The bifunctional cathode was fabricated by solvothermal growth of Fe/Co bimetallic MOFs onto electrospinning PAN derived CNFs, exhibiting both photo and electrocatalytic activity. EPR analysis confirmed enhanced OH generation under solar irradiation. The system achieved 99% PFOA removal within 120 min (Figure 6b). XPS analysis revealed valence changes of Fe and Co, and a corresponding mineralization mechanism was proposed (Figure 6c). In 2022, the same group advanced this design by integrating a glucose fuel cell (GFC) with the SPEF system to create a sustainable biomass-powered platform for PFOA degradation (Figure 6d). Oxygen-deficient CoFe alloy nanoparticles were anchored onto CNFs (CoFe-OVs@CNF), enabling dual-function cathodes (Figure 6e). The toxicity evolution of degradation intermediates was evaluated, confirming the system’s potential to degrade PFOA and other persistent pollutants while mitigating toxic risks (Figure 6f). To further expand the application scope of the SPEF system for PFAS degradation, Hou et al. developed a Cu-based peroxidase-mimicking colorimetric sensor integrated with an SPEF system for PFOA detection and removal (Figure 6g). The flexible, freestanding CNF-Cu/C membrane was synthesized via solvothermal processing, secondary seeding, and in situ thermal reduction (Figure 6h). Derived from MOF/PAN precursors, the resulting 3D carbon network exhibited excellent conductivity, dispersion, and cycling stability (Figure 6i). The CNF-Cu/C membrane showed strong peroxidase-like activity and enabled rapid PFOA detection via inhibition of TMB chromogenic reactions (Figure 6j), with a detection limit of 0.133 mM. Under optimized conditions, the Cu-SPEF system achieved 98% PFOA removal within 180 min. This work illustrates a successful integration of SPEF degradation and real-time detection, with significant practical potential.

4.4 Reductive defluorination

Reductive defluorination represents a fundamentally distinct mechanistic pathway in which electrons, rather than oxidizing radicals, serve as the primary agents of C-F bond cleavage. This approach exploits the thermodynamic accessibility of dissociative electron attachment (DEA) to C-F bonds, where an incoming electron is captured into the σ* antibonding orbital, causing bond elongation and eventual homolytic cleavage to generate a carbon-centered radical and a fluoride anion: C_nF_2n+1COOH + e⁻ → C_nF_2nCOOH• + F⁻. The carbon-centered radical subsequently undergoes further reduction, hydrogen atom transfer, or β-scission reactions that collectively achieve chain shortening and defluorination. Unlike oxidative approaches that preferentially attack the carboxylate terminus, reductive defluorination can proceed at any C-F bond along the chain, potentially enabling direct cleavage of the highly stable internal C-F bonds that represent the thermodynamic bottleneck for complete mineralization (Bentel et al., 2020). Photo excited electrons in semiconductor-like MOFs are well-positioned to mediate reductive defluorination, as their quasi-Fermi levels can be engineered to fall well negative of the C-F bond reduction potential (approximately −1.0 to −1.4 V vs. NHE for perfluoroalkyl chains). Qian et al. demonstrated visible-light-driven reductive defluorination of PFOA by a copper-doped Zr-MOF (Cu@UiO-66), achieving 71.3% defluorination efficiency after 3 h under simulated solar light (Qian et al., 2017). Mechanistic investigations using in situ X-ray absorption spectroscopy (XAS) revealed that Cu(I)/Cu(II) redox cycling at the Cu dopant sites generated strongly reducing Cu(I) intermediates capable of outer-sphere electron transfer to PFOA establishing a conceptual framework for designing Lewis acidic MOF nodes with appropriate reduction potentials for direct C-F bond activation. The integration of both oxidative and reductive capabilities within a single MOF platform remains a key design objective for achieving complete mineralization.

4.5 Synergistic coupling of degradation pathways in composite and polymetallic MOFs

The preceding subsections treat adsorption-coupled activation, photocatalytic oxidation, Fenton-like radical generation, and reductive defluorination as distinct pathways. In practice, the defining advantage of composite and polymetallic MOFs is that these pathways operate cooperatively, and this coupling accounts for the markedly higher defluorination efficiencies summarized in Tables 1, 2. Three levels of synergy can be distinguished. First, adsorption preconcentration is mechanistically continuous with catalytic defluorination: Lewis-acidic metal nodes that bind the carboxylate or sulfonate head group simultaneously concentrate PFAS at the catalytic centre and polarize the adjacent Cα–CF bond, lowering the barrier for the first defluorination step, so that reactive species generated at the node attack a pollutant already held in place; composite design amplifies this because the high-surface-area MOF performs capture while the coupled carbon or oxide component supplies or separates the charge carriers that drive destruction. Second, oxidative and reductive attack are coupled within a single platform: oxidative routes (•OH, SO₄•^-) preferentially attack the terminal head group, whereas the thermodynamically stubborn internal C-F bonds are more amenable to reductive attack and Z-scheme MOF/g-C₃N₄ configurations preserve strongly oxidizing holes and strongly reducing electrons at spatially separated sites, allowing oxidative decarboxylation and reductive defluorination to proceed in parallel as exemplified by the MIL-88B(Fe)/g-C₃N₄ (Zhang et al., 2019) and MIL-101/rGO (Lin et al., 2015) systems. Third, polymetallic nodes integrate complementary redox functions at the atomic scale: metals of differing reduction potential establish internal electron-transfer cascades that sustain turnover, so that in PCN-250(Fe-Ni-Cu) the Fe³⁺/Fe²⁺ couple supplies Fenton oxidation, Ni²⁺ extends visible-light sensitization, and the Cu²⁺/Cu⁺ couple provides reductive defluorination within the same pocket (Khan et al., 2026). This atomic-scale integration is the structural basis for the cooperative oxidation-reduction chain-shortening mechanism invoked to rationalize near-complete mineralization, and it has been demonstrated explicitly in bifunctional single-atom electrocatalysts that combine reductive PFOA activation with in-situ Fenton oxidation to reach simultaneous high removal and defluorination.

TABLE 1. Comparative performance of selected MOF-based systems for PFAS degradation.

TABLE 2. Tiered comparison of PFAS treatment materials by destruction capability.

5 Role of metal nodes in MOF performance

The identity of the metal node exerts a profound influence on the PFAS degradation performance of MOF-based systems, governing adsorption affinity, catalytic mechanism, structural stability, and the kinetics of reactive species generation. Three principal categories of metal nodes have been extensively investigated: iron-based frameworks, which are prized for their Fenton-like activity and visible-light photocatalytic properties; zirconium-based frameworks, which offer exceptional hydrothermal stability and strong Lewis acidity; and zinc-based frameworks, which provide tailored hydrophobic pore environments and zinc oxide-like photocatalytic properties in their calcined derivatives. The following sections systematically examine the mechanistic contributions and structural design implications of each metal node class.

5.1 Fe-based MOFs: redox activity and fenton performance

Iron-containing MOFs, including MIL-53(Fe), MIL-88B(Fe), MIL-101(Fe), and the mixed-valence Fe-BTC framework, represent the most catalytically active class for PFAS degradation in oxidative environments. The ready cycling between Fe²⁺ and Fe³⁺ oxidation states under environmentally relevant conditions endows these materials with inherent Fenton activity, enabling H₂O₂ activation to generate •OH radicals at pH ranges of 3–6. Cao et al. demonstrated that MIL-53(Fe) achieved 79.4% PFOA removal with 38.7% defluorination in the presence of H₂O₂ at pH 3.5 over 90 min, with the observed pseudo-first-order rate constant (kobs = 0.021 min⁻¹) approximately 4.2-fold higher than that of free Fe³⁺ at equivalent concentrations, attributing the enhancement to MOF-mediated confinement of reactive intermediates near PFAS molecules pre-adsorbed in the pores (Cao et al., 2010).

The photoactive properties of Fe-carboxylate frameworks further extend their utility to solar-driven PFAS degradation. The broad visible-light absorption of Fe-MOFs arises from d–d transitions of Fe³⁺ and LMCT transitions between the carboxylate linkers and Fe³⁺ nodes, with absorption extending to approximately 600 nm. PCN-250(Fe₂M) frameworks, in which two Fe³⁺ nodes are bridged by a paddlewheel-like Fe₂ (μ-OH) unit, have attracted particular interest for PFAS photocatalysis. Wang et al. reported that PCN-250(Fe₂Ni) achieved 88.9% PFOA removal and 55.2% defluorination under simulated sunlight irradiation (100 mW cm⁻²) within 120 min, with the Ni substitution improving both structural stability and electron-transfer kinetics at the bimetallic node (Wang et al., 2021). The Fe₂ (μ-OH) cluster was identified as the primary photocatalytic active site, generating •OH via photo-Fenton chemistry upon irradiation, while the Ni²⁺ substituent modulated the Fe³⁺ reduction potential to optimize the Fe³⁺ → Fe²⁺ photoreduction step.

5.2 Zr-based MOFs: stability and lewis acid catalysis

Zirconium-based MOFs, exemplified by the UiO series (UiO-66, UiO-67, UiO-68) and NU-1000, combine outstanding chemical and hydrothermal stability with high-density Lewis acidic Zr₆O₄(OH)₄ clusters that exhibit exceptional affinity for phosphonate and carboxylate groups. The exceptional stability of Zr-MOFs attributed to the high charge density of Zr⁴⁺ and the consequent strength of Zr–O coordination bonds (dissociation energy ∼760 kJ mol⁻¹) enables their application across a wide pH range (pH 1–12) without structural collapse, a critical requirement for real wastewater treatment scenarios. Huang et al. demonstrated that UiO-66 maintained over 90% of its crystallinity after 30 days of immersion in aqueous solution at pH 3–10, outperforming conventional Fe-MOFs that undergo significant structural degradation below pH 5 (Huang et al., 2017).

The PFAS degradation mechanism in Zr-MOFs proceeds predominantly through adsorption-coupled Lewis acid activation rather than radical generation. The Zr₆O₄(OH)₄ cluster presents both Lewis acidic Zr⁴⁺ sites and Brønsted acidic μ₃-OH groups, creating a bifunctional catalytic environment for PFAS head group coordination. Defect engineering deliberately introducing missing-linker defects in UiO-66 through modulated synthesis with monocarboxylic acid modulators (formic acid, acetic acid, trifluoroacetic acid)—has been shown to dramatically increase the density of coordinatively unsaturated Zr sites accessible to PFAS molecules. Gu et al. demonstrated that defect-rich UiO-66 (synthesized with 50 equivalents of formic acid modulator) achieved PFOA adsorption capacities of 312 mg g⁻¹, approximately 2.8-fold higher than pristine UiO-66, with subsequent photocatalytic degradation under UV irradiation yielding 42.1% defluorination significantly exceeding the performance of defect-free UiO-66 (Gu et al., 2017).

5.3 Zn-based MOFs: photocatalytic potential and derived oxides

Zinc-based MOFs, most notably ZIF-8 [Zn (MeIM)₂, MeIM = 2-methylimidazolate], combine a sodalite-like topology with exceptional hydrophobic pore character arising from the organic imidazolate linkers, making them particularly suited for the adsorptive removal of long-chain PFAS through hydrophobic partitioning. ZIF-8 possesses a BET surface area of approximately 1630 m² g⁻¹ and uniform micropores (diameter ∼3.4 Å, accessible window ∼11.6 Å) that provide selective size-based capture of fluorocarbon chains (Park et al., 2006). Sörengård et al. reported that ZIF-8 achieved Langmuir maximum adsorption capacities of 195 mg g⁻¹ for PFOS comparable to commercial anion exchange resins through combined hydrophobic and electrostatic mechanisms, with regeneration studies demonstrating maintained capacity over five consecutive adsorption cycles (Sörengård et al., 2020).

Beyond direct adsorption applications, Zn-MOFs serve as valuable sacrificial templates for the synthesis of zinc oxide (ZnO)-based photocatalysts with engineered porosity. Controlled thermal decomposition (calcination at 350 °C–450 °C in air) converts the organic linker framework into porous ZnO architectures that retain the morphology of the parent MOF while gaining the semiconductor photocatalytic properties characteristic of ZnO (Eg ≈ 3.37 eV). Chen et al. demonstrated that MOF-derived mesoporous ZnO nanostructures, synthesized by calcination of ZIF-8 precursors, exhibited approximately 2.7-fold higher PFOA photodegradation rates under UV irradiation compared to commercial ZnO, attributable to the hierarchical pore structure facilitating improved PFAS mass transfer and the high density of surface hydroxyl groups serving as •OH radical precursors (Chen et al., 2016). This MOF-template strategy bridges the disciplines of framework chemistry and semiconductor photocatalysis, opening new avenues for the rational design of porous oxide catalysts with inherited structural advantages. Considered critically, the three metal-node classes embody a clear set of trade-offs rather than a ranking. Fe-based frameworks offer the strongest intrinsic redox and Fenton activity and good visible-light response, but are the most vulnerable to hydrolytic instability and metal leaching, especially under the acidic conditions their Fenton chemistry favours. Zr-based frameworks invert this profile: outstanding stability across pH 1–12 and strong Lewis acidity, but limited intrinsic radical generation, so that their best performance is realized only when coupled to a photoactive or reductive partner. Zn-based frameworks provide hydrophobic, size-selective capture and useful derived-oxide photocatalysis, but the lowest standalone defluorination.

5.4 Ni, Cu and Co-Based MOFs

Beyond the three dominant node chemistries, frameworks built on nickel, copper and cobalt have attracted growing attention for PFAS capture and destruction. Copper nodes are particularly valuable for reductive defluorination, the Cu(I)/Cu(II) couple supplies strongly reducing intermediates capable of outer-sphere electron transfer to the C-F bond, as demonstrated for Cu@UiO-66 (71.3% PFOA defluorination) and exploited in magnetic Cu-based frameworks for recyclable pollutant removal. Cobalt-based frameworks such as ZIF-67 and ZIF-67 derived carbons provide both adsorptive and peroxymonosulfate-activating functions and CoFe bimetallic derivatives have enabled efficient solar photo-electro-Fenton PFOA mineralization (Su et al., 2023; Wang et al., 2022). Nickel substitution as in PCN-250(Fe₂Ni) extends visible-light sensitization and stabilizes the bimetallic node while improving electron-transfer kinetics (Wang et al., 2021). These nodes are most effective as components of bi and tri-metallic or composite architectures, where their complementary redox potentials reinforce the cooperative mechanisms discussed in Section 4.5.

6 MOF-based composites for enhanced PFAS degradation

While pristine MOFs demonstrate promising PFAS degradation capabilities, several intrinsic limitations including restricted visible-light absorption, rapid photogenerated charge recombination, inadequate long-term structural stability, and limited defluorination efficiency—necessitate the development of composite architectures that combine MOFs with complementary functional materials. The strategic coupling of MOFs with metal oxides, carbon-based nanomaterials, and photosensitizing semiconductors has emerged as a powerful approach to address these limitations simultaneously, yielding synergistic improvements in light harvesting, charge carrier dynamics, surface chemistry, and catalytic durability.

6.1 MOF@Metal oxide heterojunctions

The integration of MOFs with semiconductor metal oxides such as TiO₂, ZnO, and Bi₂O₃ yields heterojunction architectures that combine the high surface area and tunable pore chemistry of the MOF with the well-established photocatalytic properties of the oxide semiconductor. In Type-II heterojunctions, the staggered band alignment between the MOF and metal oxide drives directional charge carrier separation: photogenerated electrons migrate from the higher conduction band to the lower one (typically the oxide), while holes transfer in the opposite direction, spatially separating oxidizing and reducing equivalents and dramatically suppressing recombination. Chen et al. demonstrated that a MIL-53(Fe)/ZnO composite achieved 94.1% PFOA removal and 65.8% defluorination under UV-visible irradiation within 120 min—substantially outperforming either MIL-53(Fe) alone (79.4%) or ZnO alone (53.2%)—with the heterojunction interface suppressing charge recombination by approximately 40% as measured by photoluminescence spectroscopy (Chen et al., 2016).

TiO₂-based MOF composites have received particular attention due to TiO₂’s well-documented PFAS photocatalysis. Gu et al. reported that a UiO-66/TiO₂ composite synthesized through an in situ hydrothermal growth strategy achieved a 2.4-fold enhancement in PFOA photodegradation rate compared to P25 TiO₂ under UV irradiation, attributed to the MOF’s role in concentrating PFAS near TiO₂ photocatalytic sites through adsorptive preconcentration and to improved charge transfer across the UiO-66/TiO₂ interface (Gu et al., 2017). The role of interfacial contact area in governing heterojunction performance was demonstrated by Jing et al., who systematically varied the MOF-to-TiO₂ ratio in MIL-101(Cr)/TiO₂ composites and identified an optimal mass ratio of 1:3, yielding the highest defluorination efficiency (61.3%) under simulated solar light—a performance degradation at both higher and lower MOF loadings attributable to competitive light absorption and insufficient interfacial charge transfer, respectively (Jing et al., 2021).

6.2 MOF@Carbon material composites

Carbon-based nanomaterials, including reduced graphene oxide (rGO), carbon nanotubes (CNTs), and g-C₃N₄, serve as versatile coupling partners for MOF-based PFAS degradation systems, contributing electron conductivity, additional adsorption capacity, and in the case of g-C₃N₄, visible-light photocatalytic activity. The integration of MOFs with rGO is particularly effective at enhancing charge carrier mobility, as the high electron conductivity of the graphene network (σ ≈ 10³–10⁴ S m⁻¹) provides efficient channels for transporting photogenerated electrons away from the MOF, suppressing electron-hole recombination. Lin et al. synthesized a MIL-101(Cr)/rGO composite through solvothermal growth of MIL-101 on functionalized rGO sheets, achieving 97.2% PFOA removal and 72.6% defluorination under visible light irradiation—the highest defluorination efficiency reported among non-heterojunction MOF systems—attributable to the rGO-mediated enhancement of electron transfer to dissolved oxygen for O₂•^- generation (Lin et al., 2015).

The emerging class of graphitic carbon nitride (g-C₃N₄)-MOF Z-scheme composites offers a particularly elegant solution to the fundamental trade-off between oxidation power and reduction potential in photocatalytic systems. In Z-scheme configurations, MOFs and g-C₃N₄ are coupled such that photogenerated electrons from the MOF conduction band recombine with holes from the g-C₃N₄ valence band at the interface (the “Z-bridge”), simultaneously maintaining the strongly reducing electrons of g-C₃N₄ and the strongly oxidizing holes of the MOF. This preserves both the oxidative and reductive extremes of the two semiconductors while eliminating the less energetically valuable charge carriers. Zhang et al. reported that a MIL-88B(Fe)/g-C₃N₄ Z-scheme composite achieved 89.7% PFOA mineralization (as measured by total organic fluorine reduction) under visible light—a substantial improvement over either component alone, with •OH radical scavenging experiments confirming that the enhanced oxidation power originated from the preservation of high-potential MIL-88B(Fe) valence band holes in the Z-scheme configuration (Zhang et al., 2019).

6.3 Bi-metallic and tri-metallic MOF architectures

The incorporation of multiple metal nodes within a single MOF structure yielding bimetallic or trimetallic frameworks represents an advanced strategy for engineering synergistic catalytic properties that transcend the capabilities of any single metal species. Bimetallic MOFs such as PCN-250(Fe₂Ni), PCN-250(Fe₂Co), and MIL-101(Fe-Cr) achieve enhanced catalytic performance through several mechanisms: the presence of two or more metal species with different reduction potentials facilitates internal electron transfer cascades that maintain catalytic activity across broader pH and redox conditions; the mixed-metal environment creates structural distortions and electronic perturbations at the catalytic nodes that can activate otherwise inert C–F bonds; and bimetallic compositions can simultaneously optimize both adsorption (through one metal species’ affinity for PFAS head groups) and oxidative catalysis (through the other metal’s radical-generating capacity). Trimetallic MOF variants push this concept further by incorporating three distinct metal species within a single framework topology. Khan et al. synthesized a series of PCN-250-based trimetallic frameworks with systematic variation of the second and third metal species, demonstrating that the Fe-Ni-Cu variant exhibited the highest photocatalytic PFAS degradation performance, attributed to the synergistic interplay between Fe³⁺/Fe²⁺ Fenton activity, Ni²⁺-mediated visible-light sensitization, and Cu²⁺/Cu⁺ reductive defluorination pathways operating cooperatively within the same catalytic pocket (Khan et al., 2026). BET surface area analyses revealed that the trimetallic frameworks maintained surface areas of 1650–1870 m² g⁻¹, comparable to pristine PCN-250(Fe₂M), demonstrating that trimetallic substitution does not compromise pore accessibility a critical prerequisite for effective PFAS mass transfer to internal catalytic sites.

6.4 Addressing the short-chain PFAS bottleneck

Short-chain PFAS constitute the most stubborn target in PFAS remediation, combining weak hydrophobic adsorption with high thermodynamic C-F stability; conventional oxidative routes, along with GAC and AIX media, perform poorly against them. Composite and polymetallic MOFs offer the most credible though not yet complete route to overcoming this bottleneck through three complementary levers. First, reductive defluorination at electron-rich or reductively cycling metal sites attacks internal C-F bonds directly rather than relying on terminal decarboxylation; the Cu-doped Zr-MOF Cu@UiO-66 achieved 71.3% PFOA defluorination via Cu(I)/Cu(II) redox-mediated electron transfer, a mechanism intrinsically less chain-length-dependent than oxidative attack (Qian et al., 2017). Second, polymetallic nodes co-locate oxidative and reductive functions within a single catalytic pocket: the trimetallic PCN-250(Fe-Ni-Cu) framework couples Fe³⁺/Fe²⁺ Fenton activity, Ni²⁺ visible-light sensitization, and Cu²⁺/Cu⁺ reductive defluorination, providing parallel attack pathways better suited to the short-chain regime than any single mechanism (Khan et al., 2026). Third, defect engineering and pore-environment tailoring partially restore the adsorption affinity that short-chain PFAS otherwise lack defect-rich UiO-66 (Gu et al., 2017) and frameworks with chain-length-matched confinement raise local short-chain concentration at catalytic sites. Importantly, no current system achieves complete short-chain mineralization, and the C4→C3 step remains rate-limiting; these strategies should therefore be regarded as the leading research directions rather than a solved problem.

7 Factors affecting PFAS degradation efficiency

The degradation efficiency of PFAS in MOF-based systems is governed by a complex interplay of physicochemical parameters that modulate both the fundamental reaction kinetics and the thermodynamic accessibility of C–F bond cleavage pathways. Understanding and optimizing these operational variables is essential for translating laboratory-scale performance into practically viable treatment systems capable of meeting increasingly stringent regulatory limits for PFAS in drinking water and environmental discharge.

7.1 Solution pH

Solution pH exerts a multifaceted influence on PFAS degradation efficiency by simultaneously modulating the surface charge of the MOF catalyst, the ionization state of PFAS molecules, the speciation of reactive radical species, and the structural integrity of the framework itself. Under acidic conditions (pH 3–5), MOF surfaces tend to carry a net positive charge (below the point of zero charge, PZC), enhancing electrostatic attraction toward the negatively charged PFAS anions and thereby improving adsorptive preconcentration. However, excessively low pH (pH < 3) can compromise the structural integrity of carboxylate-based Fe-MOFs through protonation of the coordination bonds, leading to gradual framework dissolution and metal leaching. He et al. demonstrated that the optimal pH for PFOA degradation by MIL-53(Fe)/H₂O₂ was 3.5, yielding maximum •OH generation from the Fenton reaction while maintaining framework structural integrity over the reaction timescale (He et al., 2017).

Under neutral to mildly alkaline conditions (pH 6–8), conditions most relevant to natural water matrices, photocatalytic MOF systems generally maintain satisfactory performance provided that the photocatalytic reactive species generation (O₂•⁻, •OH via photochemical rather than Fenton pathways) does not strictly require the acidic conditions that optimize Fenton chemistry. Zr-based MOFs are particularly valuable in this context, as their exceptional stability across a wide pH range (1–12) enables application without pH pre-adjustment—a significant practical advantage. At pH > 9, increased competition from hydroxide ions (OH⁻) as •OH radical scavengers progressively diminishes oxidative efficiency, while simultaneous precipitation of Fe²⁺/Fe³⁺ ions as iron hydroxides in Fe-MOF systems restricts the Fenton cycle and reduces catalytic activity (Nidheesh et al., 2013).

7.2 Light intensity and irradiation wavelength

For photocatalytic MOF systems, light intensity and spectral composition represent primary determinants of reactive oxygen species generation rates and, consequently, PFAS degradation kinetics. Within the linear regime (typically below 80 mW cm⁻²), photocatalytic reaction rates scale proportionally with photon flux, as each absorbed photon generates one electron-hole pair capable of initiating radical chemistry. At higher irradiance (>80–100 mW cm⁻²), saturation effects emerge as the rate of charge carrier generation exceeds the rate of charge carrier consumption, leading to increased recombination and sublinear scaling of degradation rate with intensity. Zhang et al. demonstrated that doubling the irradiance of a UiO-66-NH₂ photocatalytic system from 50 to 100 mW cm⁻² increased the PFOA degradation rate constant by only 1.67-fold (rather than the theoretical 2.0-fold for a strictly linear system), confirming the onset of recombination-limited kinetics at higher intensities (Zhang et al., 2020).

The wavelength distribution of the irradiation source profoundly influences which MOF systems can be effectively activated. Narrow-band-gap MOF composites (Eg < 2.5 eV) can absorb a substantial fraction of the visible solar spectrum, enabling solar-driven PFAS degradation without artificial UV sources. Chen et al. demonstrated that MIL-53(Fe)/ZnO composites under simulated natural sunlight (AM 1.5G filter, 100 mW cm⁻²) achieved 78% of the degradation efficiency obtained under optimized UV-visible irradiation, demonstrating practical solar-driven performance and suggesting viable outdoor deployment scenarios for PFAS-contaminated water bodies (Chen et al., 2016).

7.3 Co-existing ions and natural organic matter

Real water matrices contain a complex assemblage of co-existing ions and dissolved organic molecules that profoundly influence PFAS degradation kinetics through competitive adsorption, radical scavenging, and potential catalytic inhibition. Inorganic anions including Cl⁻, NO₃ ⁻, SO₄ ²⁻, HCO₃ ⁻, and HPO₄ ²⁻ compete with PFAS for active sites on MOF surfaces and consume reactive radicals through chain-terminating reactions. Bicarbonate (HCO₃ ⁻) and carbonate (CO₃ ²⁻) ions are particularly detrimental, reacting rapidly with •OH (k = 8.5 × 10⁶ and 3.9 × 10⁸ M⁻¹ s⁻¹, respectively) to form the much less reactive carbonate radical (CO₃•^-), reducing the steady-state •OH concentration available for PFAS attack. Liu et al. demonstrated that addition of HCO₃ ⁻ at 10 mM reduced the PFOA degradation rate constant in a Fe-BTC/PMS system by 47%, whereas equivalent concentrations of Cl⁻ and NO₃ ⁻ caused more modest reductions of 22% and 18%, respectively (Liu et al., 2012).

Natural organic matter (NOM) exerts dual effects on MOF-based PFAS degradation systems: it competes with PFAS for surface adsorption sites on the MOF (reducing preconcentration efficiency) and simultaneously scavenges •OH and SO₄•^- radicals (reducing degradation kinetics). However, the influence of NOM is modulated by the specific adsorption mechanism of the MOF system—frameworks that rely primarily on electrostatic interactions with PFAS are more susceptible to NOM competition, while those exploiting fluorophilic or size-selective pore interactions are less affected. Incorporating an ion-selective membrane barrier or pH pre-adjustment step into MOF-based treatment systems represents a practical mitigation strategy for NOM interference, reducing the effective NOM load entering the catalytic zone (Dickenson and Higgins, 2016).

7.4 PFAS chain length and concentration

The effect of PFAS carbon chain length on degradation efficiency follows a non-linear relationship that reflects competing influences: longer chains exhibit greater hydrophobic adsorption affinity (enhancing surface concentration near catalytic sites) but also stronger thermodynamic C–F bond stability (impeding catalytic cleavage). For MOF systems with photocatalytic or radical-based mechanisms, the adsorption enhancement effect generally dominates for chain lengths of C6–C10, yielding higher degradation efficiencies for long-chain PFAS compared to short-chain counterparts under equivalent conditions. This trend is reflected in reported degradation efficiencies, which increase from ∼45% for PFBA (C4) to ∼91% for PFOA (C8) before declining modestly for longer chains due to increased steric bulk limiting pore access. Short-chain PFAS (C4–C5 for PFCAs; C3–C4 for PFSAs) present a particularly challenging remediation target, combining relatively weaker adsorption affinity (due to shorter hydrophobic tails) with sufficient thermodynamic stability to resist most radical-based degradation approaches. Initial PFAS concentration also critically affects system performance: at low concentrations (<1 μg L⁻¹, relevant to drinking water applications), reaction kinetics become pseudo-first-order and are often limited by mass transfer to the catalyst surface, whereas at higher concentrations (>100 μg L⁻¹, relevant to contaminated groundwater), competitive adsorption and catalyst saturation effects emerge. Liu et al. demonstrated that increasing PFOA concentration from 10 to 200 μg L⁻¹ in a UiO-66-NH₂ photocatalytic system reduced the degradation rate constant by 68%, attributable to the saturation of Zr₄ ⁺ adsorption sites and the resultant reduction in PFAS surface concentration enhancement (Liu et al., 2021).

7.5 Performance in real water matrices

Almost all the efficiencies in Tables 1, 2 were obtained in deionized or buffered water; reported performance in authentic matrices is markedly lower and more variable. In groundwater from AFFF-impacted sites, elevated sulfate, bicarbonate, and dissolved organic carbon depress radical-mediated defluorination through the scavenging and competitive-adsorption mechanisms discussed above, while the co-occurrence of PFAS mixtures rather than single compounds spreads the available reactive flux across many homologues. In municipal and industrial wastewater, high natural-organic-matter loads and competing anions, together with turbidity that attenuates light penetration, can reduce degradation rate constants by tens of percent relative to clean-water benchmarks. Industrial effluents additionally present extremes of pH and ionic strength that challenge framework stability. The limited pilot- and field-scale data available indicate degradation-efficiency losses of roughly 30%–60% on transition from laboratory to real matrices. Practical mitigation includes an upstream selective pre-concentration or membrane barrier to exclude NOM and particulates, pH conditioning ahead of Fenton-type stages, and the use of hydrolytically robust Zr- or polymetallic frameworks for high-salinity or extreme-pH streams. Systematic benchmarking in standardized real-water matrices remains an outstanding need for credible technology comparison.

8 Performance evaluation and comparison of MOF-Based systems

A rigorous comparative evaluation of MOF-based PFAS degradation systems requires consideration of multiple performance metrics beyond simple removal efficiency, including defluorination efficiency, mineralization rate, reaction kinetics, catalyst stability and recyclability, and performance under realistic water chemistry conditions. Table 1 provides a comprehensive comparison of representative MOF-based systems reported in the literature, encompassing both pristine frameworks and composite architectures, evaluated under standardized or comparable experimental condition.

8.1 Defluorination efficiency as the critical performance metric

Among all performance metrics, defluorination efficiency defined as the molar fraction of organic fluorine converted to inorganic fluoride (F⁻) provides the most unambiguous measure of actual PFAS mineralization and is therefore the most meaningful indicator of treatment success from an environmental protection standpoint. High PFAS removal efficiency (measured as disappearance of the parent compound by chromatographic analysis) can mask the formation of shorter-chain PFAS intermediates that remain environmentally persistent and biologically toxic. Complete defluorination (100%) is thermodynamically equivalent to complete mineralization and represents the ultimate goal of catalytic PFAS treatment.

As evident from Table 1, defluorination efficiencies in current MOF-based systems range from approximately 29.5% (ZIF-8, adsorption-dominated) to 72.6% (MIL-101/rGO composite, photocatalytic), with the highest values consistently achieved by composite systems that combine multiple degradation mechanisms. The gap between removal efficiency and defluorination efficiency which ranges from 20 to 40 percentage points across the systems evaluated represents the residual organic fluorine present in partially degraded intermediates, highlighting the mechanistic challenge of achieving complete C-F bond cleavage in all intermediate chain lengths generated during stepwise degradation. Overcoming this “defluorination gap” remains a central challenge for next-generation MOF-based PFAS treatment systems, requiring both mechanistic innovation and careful optimization of operating conditions. To make the value proposition of advanced MOF architectures explicit, Table 2 benchmarks three tiers of PFAS treatment materials commercial adsorbents, pristine single-metal MOFs and composite/polymetallic MOFs against the metric that ultimately matters: conversion of organic fluorine to inorganic fluoride rather than mere parent-compound removal. Commercial GAC and AIX media achieve high removal of long-chain PFAS but effect no C-F bond cleavage, transferring contamination to a solid phase that still requires high-temperature incineration. Pristine single MOFs improve on this through adsorption-coupled activation but deliver defluorination of only ∼29–51%, dominated by capture rather than destruction. Composite and polymetallic MOFs close much of this gap, raising defluorination to 55%–73% (and mineralization to ∼90% in optimized Z-scheme systems) by combining broadened light absorption, suppressed charge recombination and coupled oxidative reductive pathways within a single material.

8.2 Kinetic analysis and rate comparisons

The kinetics of PFAS degradation in MOF-based systems generally follow pseudo-first-order rate laws with respect to PFAS concentration at the low concentrations relevant to environmental applications (typically 1–500 μg L⁻¹). Observed rate constants (kobs) reported in the literature span approximately two orders of magnitude, from ∼0.005 min^-1 for adsorption-dominated systems to ∼0.065 min⁻¹ for optimized composite photocatalysts, reflecting the dominant contribution of catalytic mechanism to overall degradation kinetics. Normalization of rate constants to catalyst surface area (kSA = kobs/[catalyst]·BET) enables meaningful cross-system comparison by accounting for differences in catalyst loading and surface area, revealing the intrinsic activity per unit catalytic surface. Lin et al. reported kSA values of 2.8 × 10⁻⁴ L mg⁻¹ m⁻² min⁻¹ for MIL-101/rGO approximately 3.1-fold higher than pristine MIL-101 confirming that the rGO integration genuinely enhances intrinsic catalytic activity rather than merely increasing total surface area (Lin et al., 2015).

9 Challenges and limitations of MOF-Based PFAS degradation systems

Despite the remarkable progress achieved in MOF-based PFAS degradation research, several fundamental and practical challenges must be addressed before these systems can transition from laboratory demonstrations to real-world environmental remediation applications (Ma et al., 2020). These challenges span multiple dimensions from intrinsic material properties (water stability, metal leaching) to system-level limitations (scalability, real-water performance) and fundamental mechanistic gaps (short-chain PFAS removal, incomplete mineralization) and their resolution requires coordinated advances across materials chemistry, process engineering, and environmental science.

9.1 Water and chemical stability of MOF structures

The hydrolytic stability of MOF structures in aqueous treatment environments represents perhaps the most critical practical challenge limiting their deployment at scale. Many catalytically active MOFs, particularly those based on Fe³⁺ and Al³⁺ nodes with carboxylate linkers (e.g., MIL-53, MIL-88 series), exhibit susceptibility to ligand displacement by water molecules at coordinatively unsaturated metal sites, leading to gradual framework amorphization and loss of crystallinity over extended treatment periods. Stock and Biswas systematically evaluated the water stability of 200 MOFs from the Cambridge Structural Database, finding that only approximately 35% exhibited stable crystal structures after 24 h immersion in water at pH 7-a sobering statistic for environmental applications requiring sustained long-term performance (Stock and Biswas, 2012).

The mechanisms of MOF hydrolytic degradation are well-characterized: water molecules coordinate to Lewis acidic metal nodes, displacing carboxylate ligands in a reversible exchange process that leads to progressive linker detachment at elevated temperatures or extreme pH values. Strategies to improve water stability include the use of high-valence metal nodes (Zr⁴⁺, Ti⁴⁺) that form stronger M–O coordination bonds, incorporation of hydrophobic surface coatings (e.g., PDMS, silane functionalization) that restrict water access to the coordination sites, and post-synthetic crosslinking of adjacent MOF layers through covalent bond formation. However, each of these strategies introduces trade-offs: high-valence frameworks often sacrifice photocatalytic activity; hydrophobic coatings reduce accessible surface area; and crosslinking can block pore access. The development of MOF structures that are simultaneously catalytically active, accessible to PFAS, and hydrolytically stable remains an unresolved fundamental challenge (Bedia et al., 2019).

9.2 Metal leaching and secondary contamination

The release of metal ions from MOF-based catalysts during PFAS degradation treatment—termed metal leaching—represents a critical concern for environmental safety, as leached Fe³⁺, Cr³⁺, Zn²⁺, or other metal species may constitute secondary pollutants that exceed regulatory discharge limits. Leaching is particularly acute for Fe-MOFs operated under acidic Fenton conditions (pH 3–5), where proton competition at metal–carboxylate coordination bonds accelerates ligand dissociation and framework dissolution. Studies have documented Fe leaching rates of 0.5–3.2 mg L⁻¹ per treatment cycle for MIL-53(Fe) under typical Fenton conditions—values that exceed the WHO drinking water guideline for iron (0.3 mg L⁻¹) and necessitate post-treatment removal of dissolved iron before water discharge (He et al., 2017).

Chromium-containing MOFs such as MIL-101(Cr), despite their exceptional surface area and catalytic performance, face stringent regulatory scrutiny due to the potential leaching of toxic Cr³⁺ (and its possible oxidation to carcinogenic Cr⁶⁺ under strong oxidative conditions). Several recent studies have documented Cr leaching from MIL-101(Cr) during oxidative treatment, with leached concentrations of 0.02–0.15 mg L⁻¹ reported—below immediate toxicity thresholds but sufficient to trigger regulatory concerns for drinking water applications (Wang et al., 2020). The transition to Cr-free MOF systems (Fe-based, Zr-based, or Al-based frameworks) and the development of post-synthetic metal encapsulation strategies represent necessary directions for environmentally responsible PFAS catalysis development.

9.3 Incomplete mineralization and intermediate accumulation

The incomplete conversion of PFAS to inorganic fluoride, carbon dioxide, and sulfate evidenced by the defluorination efficiency gap discussed in Section 8 is mechanistically attributable to the sequential, stepwise nature of photocatalytic and radical-based chain-shortening processes. Each degradation cycle typically removes one or two carbon units from the PFAS chain, generating a shorter-chain PFCA homolog that must undergo subsequent degradation cycles before complete defluorination is achieved. The accumulation of these shorter-chain intermediates is not merely a kinetic inconvenience—it represents a genuine environmental risk, as short-chain PFCAs (C4–C6) exhibit greater mobility in groundwater, higher bioaccumulation in certain tissues, and increasing environmental ubiquity following regulatory restrictions on longer-chain PFAS.

Kucharzyk et al. identified the C4 step (conversion of PFBA to perfluoropropionic acid, PFPrA) as the rate-limiting degradation step in most photocatalytic systems, attributable to the reduced adsorption affinity of short-chain PFAS at MOF active sites combined with the heightened thermodynamic stability of shorter perfluoroalkyl chains relative to their longer-chain counterparts (Kuchar et al., 2017). This ‘short-chain bottleneck’ implies that achieving complete PFAS mineralization requires either exceptionally long reaction times (potentially hours to days for real-water matrices), elevated photon flux, or the development of catalytic systems with specific affinity for short-chain PFAS a design challenge that has not yet been satisfactorily addressed (Wu et al., 2018). Reductive defluorination pathways (Section 4.4) may offer a more direct route to short-chain PFAS degradation by targeting internal C–F bonds rather than the terminal carboxylate group, representing a promising direction for future mechanistic investigation.

9.4 Scalability and real wastewater performance

The vast majority of MOF-based PFAS degradation studies have been conducted in clean aqueous matrices (deionized or pH-buffered water) at laboratory scale (10–100 mL), using single-compound PFAS solutions at concentrations (10–1000 μg L⁻¹) that are frequently orders of magnitude higher than those encountered in real contaminated groundwater or surface water (typically 0.001–10 μg L⁻¹). The extrapolation of laboratory performance to real treatment scenarios involves substantial uncertainties arising from matrix interference effects (co-existing ions, NOM, turbidity), the presence of PFAS mixtures rather than single compounds, lower target concentrations that increase the diffusion-limited character of mass transfer kinetics, and the engineering challenges of delivering sufficient photon flux to large volumes of turbid water (Gagliano et al., 2020).

The transition from batch laboratory experiments to continuous-flow treatment systems introduces additional engineering complexities, including reactor design for adequate light penetration, catalyst immobilization strategies that prevent filter breakthrough, and long-term catalyst stability under continuous operation conditions. Pilot-scale MOF-based treatment systems for PFAS contaminated water have been demonstrated by only a small number of research groups to date, with most reporting significant performance gaps relative to laboratory benchmarks—typically a 30%–60% reduction in degradation efficiency at pilot scale, attributable primarily to mass transfer limitations and reduced effective irradiance in scaled systems (Bedia et al., 2019). Bridging this scale gap requires interdisciplinary collaboration between materials chemists, environmental engineers, and regulatory scientists to develop MOF-based PFAS treatment modules that can realistically compete with established GAC and ion-exchange technologies at treatment costs acceptable to water utilities.

9.5 Reusability, regeneration and sustainability

Long-term viability depends not only on per-pass efficiency but on retained activity over many cycles and on the ease of catalyst recovery. Reported cycling behaviour varies widely: stable frameworks such as UiO-66 derivatives and several composite photocatalysts retain the majority of their activity over roughly five consecutive cycles, whereas less stable Fe/Al-carboxylate frameworks lose activity more rapidly through partial amorphization and metal leaching. Regeneration is typically achieved by mild solvent washing, thermal treatment, or pH-swing desorption, although each route risks incremental framework damage and must be balanced against cumulative leaching. Magnetic composite designs offer a practical route to rapid catalyst recovery: incorporation of a magnetic core (for example, Fe₃O₄@SiO₂-supported MOFs [NEW-MAG1] or magnetic HKUST-1-type framework enables separation by an external field and has supported ten or more reuse cycles in related adsorbent systems, an approach directly transferable to PFAS catalysts.

9.6 Synthesis cost, catalyst recovery and continuous-flow performance attenuation

Three engineering bottlenecks dominate the transition from laboratory demonstration to deployable technology. (1) Synthesis cost. Laboratory-grade MOFs typically cost on the order of US$20–200 per kg, driven by organic-linker price, large solvent volumes (often DMF), and energy-intensive batch solvothermal protocols. Achieving treatment costs competitive with GAC (∼US$0.10–0.50 per m³) requires roughly an order-of-magnitude cost reduction, attainable through continuous-flow or mechanochemical synthesis, aqueous or solvent-lean routes, solvent recovery, and lower-cost metal precursors and modulators. (2) Catalyst recovery and immobilization. Slurry-phase MOF powders are impractical at scale because of separation cost and breakthrough risk; immobilization onto membranes, monoliths, foams, or 3D-printed supports, or magnetic-core compositing for field-assisted recovery, is a prerequisite for continuous operation. (3) Performance attenuation in continuous-flow reactors. Under sustained flow, efficiency declines through pore blockage and fouling by NOM and particulates, progressive metal leaching and partial framework amorphization, catalyst detachment under hydrodynamic shear, and in photocatalytic configurations attenuation of photon flux in turbid streams and along the reactor depth; reported pilot-scale efficiencies are typically 30%–60% below batch benchmarks for these reasons. Mitigation strategies include membrane- or monolith-integrated catalyst beds, periodic in-situ regeneration and backwashing, hydrolytically robust Zr- or polymetallic frameworks to limit leaching, and annular or LED-illuminated thin-film photoreactors that maintain uniform irradiance.

10 Future perspectives

The field of MOF-based PFAS degradation has advanced remarkably over the past decade, transitioning from proof-of-concept adsorption studies to sophisticated photocatalytic and Fenton-like degradation systems capable of achieving meaningful defluorination under realistic conditions. Looking forward, several transformative research directions are poised to overcome the current limitations and establish MOF-based technologies as viable components of next-generation PFAS remediation strategies. These directions span fundamental materials chemistry, applied process engineering, computational design, and regulatory science.

10.1 Trimetallic and multi-metallic MOF engineering

The exploration of trimetallic MOF architectures represents one of the most promising near-term research priorities, offering the potential to synergistically combine the complementary catalytic functions of three metal species within a single, precisely engineered framework (Ali et al., 2026; Meng et al., 2026; Zuhra et al., 2025). As demonstrated in Section 6.3, trimetallic PCN-250 variants have already shown superior PFAS degradation performance relative to their mono- and bi-metallic counterparts. Future work should focus on elucidating the atomic-scale mechanisms underlying the synergistic effects in trimetallic nodes, using advanced characterization techniques including operando X-ray absorption near-edge spectroscopy (XANES), pair distribution function (PDF) analysis, and single-site catalysis approaches to identify the active metal species and their oxidation state dynamics during PFAS degradation (Khan et al., 2026). Systematic combinatorial screening of metal combinations—moving beyond the Fe-M-M′ combinations explored to date to include Ti, V, Mo, and rare earth elements with complementary photocatalytic and Lewis acidic properties—could yield MOF platforms with dramatically enhanced performance (Ali et al., 2025).

Computational screening approaches, employing periodic DFT calculations to evaluate the thermodynamic accessibility of C–F bond activation at candidate trimetallic nodes, can dramatically accelerate experimental discovery by prioritizing the most promising metal combinations for synthesis and evaluation. The growing repository of MOF crystal structures in the Cambridge Structural Database (>12,000 experimentally characterized frameworks) and high-fidelity force field databases (e.g., UFF4MOF) enable molecular dynamics simulations that can predict PFAS diffusion coefficients, pore-accessible volumes, and binding geometries at specific metal sites providing design blueprints for the next-generation of high-performance PFAS catalysts before a single gram of material is synthesized (Raza et al., 2020).

10.1.1 Single-atom catalysts and defect-engineered frameworks

Single-atom catalysts in which isolated metal atoms are anchored on supports or within MOF-derived carbons, maximize atom economy and present uniform, electronically tunable active sites; recent bifunctional SACs that integrate reductive PFOA activation with in-situ Fenton oxidation have achieved near-complete mineralization and >95% defluorination, pointing to MOF-templated SACs as a high-potential direction. Defect engineering deliberate introduction of missing-linker or missing-cluster defects creates coordinatively unsaturated metal sites that simultaneously increase PFAS adsorption affinity and catalytic accessibility, as already demonstrated for defect-rich UiO-66 (Gu et al., 2017).

10.2 Electrochemical coupling and electrocatalytic systems

The integration of MOF-based catalysts with electrochemical systems represents a conceptually elegant strategy for overcoming the light-dependence limitation of photocatalytic approaches and enabling continuous, on-demand PFAS degradation without requiring solar or UV irradiation. In electrocatalytic MOF systems, the MOF catalyst is deposited on a conductive electrode substrate, and an applied electrical potential drives the generation of reactive species (•OH at anodic surfaces, reducing equivalents and H₂ at cathodic surfaces) that attack adsorbed PFAS molecules. The electrochemical driving force can overcome the thermodynamic barriers to C–F bond cleavage that limit purely thermal or photocatalytic approaches, potentially enabling more complete defluorination at ambient temperature. Mao et al. demonstrated that a Ti/BDD (boron-doped diamond) anode functionalized with MIL-101(Cr) achieved 98.7% PFOA removal and 85.3% defluorination at an applied current density of 20 mA cm⁻², with the MOF layer’s adsorptive preconcentration function dramatically reducing the energy requirement compared to direct electrochemical oxidation of dissolved PFAS (Mao et al., 2014).

Coupled photo-electrocatalytic systems, in which MOF-modified electrodes are simultaneously irradiated and electrochemically biased, offer the potential to combine the benefits of both approaches photogenerated charge carriers providing additional reactive species while the applied bias suppresses electron-hole recombination and drives directional electron transfer. The development of MOF-based electrodes with sufficient mechanical robustness to withstand the hydrodynamic stresses of continuous-flow electrochemical reactors, while maintaining high PFAS adsorption capacity and catalytic activity over hundreds of operational hours, represents a key engineering challenge for this emerging approach.

10.3 Artificial intelligence-guided MOF design

The vast chemical design space of MOFs with theoretically millions of possible metal-linker combinations across thousands of known topologies is fundamentally intractable for systematic experimental exploration. Machine learning and artificial intelligence approaches are increasingly being applied to navigate this design space efficiently, training predictive models on existing experimental datasets to identify metal-linker combinations predicted to exhibit optimal PFAS degradation performance. Graph neural networks (GNNs) that encode MOF crystal structures as atom-bond graphs have demonstrated remarkable accuracy in predicting gas adsorption isotherms from crystal structure data alone, with recent extensions to liquid-phase applications showing promise for predicting PFAS adsorption capacities (Chong et al., 2020).

Generative AI models trained on the geometric and chemical features of known high-performance MOF structures can propose entirely novel framework compositions combinations of metals and linkers that have not yet been synthesized optimized for specific target properties such as high PFAS adsorption capacity, visible-light photocatalytic activity, and hydrothermal stability simultaneously. The integration of active learning loops, in which AI predictions guide targeted experimental synthesis and characterization of the most promising candidate frameworks, with experimental data feeding back to improve model accuracy in successive cycles, represents the frontier of computational-experimental co-design for environmental applications (Jablonka et al., 2020). This approach has the potential to dramatically compress the discovery timeline for next-generation PFAS-degrading MOF systems from the current paradigm of years-long empirical optimization to accelerated discovery cycles of months.

10.4 Reactor engineering and scale-up

Translating the impressive laboratory-scale performance of MOF-based PFAS degradation systems into practically deployable treatment modules requires substantial advances in reactor engineering, catalyst immobilization, and process integration. Membrane-integrated MOF systems, in which MOF photocatalysts are incorporated into polymeric or ceramic membrane matrices, offer a compelling solution to the catalyst separation challenge while simultaneously providing size-selective rejection of larger PFAS molecules. Continuous annular photoreactors with LED-illuminated internal surfaces, capable of delivering uniform photon flux to flowing water streams at flow rates of tens to hundreds of liters per hour, represent a more scalable architecture for MOF photocatalytic systems than the batch slurry reactors currently dominating the laboratory literature.

The techno-economic feasibility of MOF-based PFAS treatment systems has received insufficient attention relative to their catalytic performance characterization, yet economic viability is ultimately the gateway to real-world deployment. Preliminary techno-economic analyses suggest that the high synthesis cost of MOF catalysts (typically $20–200 kg⁻¹ for laboratory-grade material) must be reduced by at least one order of magnitude through continuous-flow synthesis, solvent recycling, and use of lower-cost metal precursors to achieve treatment costs competitive with existing GAC-based systems (∼$0.10–0.50 per m³ of treated water) (Woodard et al., 2017). Life cycle assessment (LCA) frameworks that account for the avoided costs of PFAS-related health impacts and regulatory compliance provide a more complete economic picture and may demonstrate superior cost-effectiveness for MOF-based mineralization systems relative to the indefinite waste management requirements of adsorption-only approaches.

11 Conclusion

This review has provided a comprehensive and critical analysis of the rapidly evolving field of MOF-based catalytic systems for PFAS degradation, spanning fundamental C-F bond chemistry, framework architectural principles, mechanistic degradation pathways, composite design strategies, and the current state of performance evaluation. Several overarching conclusions emerge from this analysis that collectively define the current state of knowledge and chart the course for future research priorities. Metal-organic frameworks represent a genuinely transformative platform for PFAS remediation, offering capabilities that fundamentally exceed those of conventional adsorptive materials by integrating adsorptive preconcentration with catalytic C–F bond activation within a single, structurally programmable material. The mechanistic diversity available in MOF-based systems—encompassing photocatalytic, Fenton-like, reductive, and electrochemical pathways, often operating synergistically—provides an unprecedented degree of mechanistic flexibility for targeting the chemical recalcitrance of different PFAS compounds across varying environmental conditions. The composite design strategies reviewed herein—MOF@metal oxide heterojunctions, MOF@carbon hybrids, and trimetallic frameworks—have collectively pushed defluorination efficiencies from the 30%–40% range achievable by pristine MOFs to greater than 70%, demonstrating the power of multi-component system design in overcoming the kinetic and thermodynamic barriers to complete PFAS mineralization.

At the same time, the persistent challenges identified in this review hydrolytic instability, metal leaching, incomplete short-chain PFAS mineralization, and the substantial performance gap between laboratory demonstrations and real-water applications—underscore that MOF-based PFAS treatment technology remains in its early developmental stage. Bridging the transition from promising laboratory results to practically deployable, economically viable treatment systems demands coordinated advances in material stability engineering, mechanistic understanding of short-chain PFAS degradation, electrochemical system integration, and reactor scale-up. The convergence of high-throughput computational screening, machine learning-guided MOF design, and advanced in situ characterization techniques positions the field to accelerate the discovery and optimization of next-generation PFAS-degrading frameworks at a pace not previously achievable through purely empirical approaches. The societal urgency of PFAS contamination with more than 12,000 sites contaminated in the United States alone and regulatory limits for PFAS in drinking water being tightened to the sub-nanogram-per-liter level in multiple jurisdictions provides a compelling mandate for sustained research investment in MOF-based catalytic mineralization technologies. The fundamental promise of these materials to achieve complete, permanent destruction of PFAS at the molecular level rather than merely transferring contamination between environmental compartments justifies the scientific ambition required to overcome the remaining challenges and deliver transformative solutions for one of the defining environmental remediation challenges of our era.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. To improve English language.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Asif Shahzad, University of South Australia, Australia

Reviewed by: Lingxin Chen, Chinese Academy of Sciences (CAS), China

Manavi Yadav, Purdue University Fort Wayne, United States

S. Bagyalakshmi, kgisl Institute of Technology, India