The scale of the contamination problem is staggering. Surface water concentrations near industrial regions have reached 13.5 micrograms per litre in China and 11.34 micrograms per litre of PFOA in Alabama, while even less impacted regions across Europe, Asia and North America frequently exceed 70 nanograms per litre. Dietary intake, particularly from water and seafood, accounts for roughly 74 percent of general population exposure according to the United States Food and Drug Administration’s 2022 seafood survey. PFOS and PFOA bind strongly to plasma proteins, accumulating in the liver and serum and resisting elimination through enterohepatic recirculation. In aquatic ecosystems, PFAS accumulation has been correlated with altered liver somatic index and body size in Baltic cod, bioaccumulation in mussels, and reduced growth in marine microalgae, demonstrating broad-spectrum toxicological effects across biological systems.

Conventional treatment technologies fall short in a fundamental way. Granular activated carbon, anion exchange resins and membrane processes such as nanofiltration and reverse osmosis can remove PFAS from water, but they achieve no chemical destruction. They simply transfer the pollutants into concentrated waste streams that require high-temperature incineration, often exceeding 1100 degrees Celsius, or indefinite landfill storage. The review, led by Muhammad Haris Khan of Wuhan Textile University, emphasizes that the decisive measure of success is not parent-compound removal but defluorination and mineralization, the conversion of organic fluorine into inorganic fluoride ions, carbon dioxide and sulfate. On this metric, commercial adsorbents score approximately zero, because they cleave no carbon–fluorine bonds at all.

MOFs are crystalline hybrid materials built from metal nodes connected by organic linkers through strong coordination bonds, yielding periodic three-dimensional networks with exceptional porosity. More than 20,000 distinct MOFs have been reported since the field’s foundational milestone in 1999, when Omar Yaghi and colleagues reported MOF-5. Their modular architecture allows precise control over pore size, surface functionality and electronic structure. In the context of PFAS, the metal nodes serve a dual role: they act as Lewis acid sites that coordinate directly to the carboxylate or sulfonate head groups of PFAS molecules, and they function as catalytic centers for subsequent carbon–fluorine bond activation. The hydrophobic aromatic linkers, meanwhile, create pore environments that attract the perfluorinated carbon tails through van der Waals forces. MIL-101(Cr), with a surface area up to 5900 square metres per gram, and UiO-66 have demonstrated PFAS uptake capacities exceeding those of conventional activated carbons and ion exchange resins.

The mechanistic core of MOF-based PFAS destruction lies in coordinative activation. Density functional theory calculations have shown that the terminal carboxylate group of PFOA coordinates in a bidentate bridging mode at Cr(III) Lewis acid sites in MIL-101(Cr), withdrawing electron density from the alpha-carbon and weakening the adjacent carbon–fluorine bond. Such Lewis acid-mediated activation reduces the bond dissociation energy by 15 to 25 kJ per mole in computationally modeled systems, enough to facilitate catalytic defluorination at ambient temperatures. Pore confinement amplifies this advantage: PFAS molecules held inside UiO-66 mesopores exhibited degradation rate constants approximately 3.8-fold higher than surface-adsorbed molecules on amorphous silica supports, because confinement increases the reactive collision frequency between pollutant and catalyst.

Photocatalytic degradation adds a second dimension. In photoactive MOFs, the organic linkers serve as light-harvesting antennae while the metal nodes provide electron-accepting conduction band states. Pristine UiO-66(Zr) possesses a wide band gap of approximately 4.0 electronvolts, restricting its activity to ultraviolet light, but amino-functionalization of the linker narrows the gap to roughly 2.75 electronvolts, enabling visible-light absorption up to 450 nanometres. Upon photoexcitation, electrons react with dissolved oxygen to generate superoxide radical anions, while valence band holes produce hydroxyl radicals with an oxidation potential of +2.80 volts. Hydroxyl radicals attack the hydrophilic head group, triggering oxidative decarboxylation, whereas superoxide mediates reductive defluorination by injecting electrons directly into the carbon–fluorine antibonding orbital. Lin and colleagues demonstrated that superoxide was the predominant reactive species in a MIL-101/rGO photocatalytic system, achieving 72.6 percent defluorination after 150 minutes under visible light.

Iron-containing MOFs bring Fenton chemistry into the picture. The Fe(II)/Fe(III) redox couple activates hydrogen peroxide to generate hydroxyl radicals at pH 3 to 6, and MIL-53(Fe) achieved 79.4 percent PFOA removal within 90 minutes, with a pseudo-first-order rate constant approximately 4.2-fold higher than free Fe(III) at equivalent concentrations. Extended systems employing persulfate or peroxymonosulfate generate sulfate radicals, which possess longer half-lives of roughly 30 to 40 microseconds and selectivity toward electron-rich functional groups. Fe-BTC activated peroxymonosulfate at pH 5.0 achieved 83.5 percent PFOS removal and 48.2 percent defluorination within 60 minutes. The most striking demonstration of coupled mechanisms came from a solar photo-electro-Fenton platform in which Fe/Co bimetallic MOFs grown on carbon nanofiber membranes achieved 99 percent PFOA removal within 120 minutes, later integrated with a glucose fuel cell to create a biomass-powered treatment system.

Composite and polymetallic architectures have pushed defluorination efficiencies from the 30 to 40 percent range achievable by pristine single-metal MOFs to greater than 70 percent. MOF@metal oxide heterojunctions such as MIL-53(Fe)/ZnO achieved 94.1 percent PFOA removal and 65.8 percent defluorination, substantially outperforming either component alone, because the staggered band alignment spatially separates oxidizing and reducing equivalents and suppresses charge recombination by roughly 40 percent. Z-scheme MOF/graphitic carbon nitride configurations preserve strongly oxidizing holes and strongly reducing electrons at spatially separated sites, allowing oxidative decarboxylation and reductive defluorination to proceed in parallel; a MIL-88B(Fe)/g-C3N4 Z-scheme composite reached 89.7 percent PFOA mineralization under visible light. Trimetallic frameworks push the concept further: PCN-250(Fe-Ni-Cu) couples Fe(III)/Fe(II) Fenton activity, Ni(II)-mediated visible-light sensitization and Cu(II)/Cu(I) reductive defluorination within the same catalytic pocket, while maintaining surface areas of 1650 to 1870 square metres per gram.

The persistent bottleneck is short-chain PFAS. Regulatory restrictions on long-chain compounds have shifted global production toward short-chain variants, which combine weak hydrophobic adsorption with high thermodynamic stability. Reported degradation efficiencies increase from roughly 45 percent for PFBA (C4) to 91 percent for PFOA (C8), and the C4-to-C3 conversion step remains rate-limiting in most photocatalytic systems. Reductive defluorination offers the most credible route forward because it attacks internal carbon–fluorine bonds rather than relying on terminal decarboxylation; a copper-doped Zr-MOF, Cu@UiO-66, achieved 71.3 percent PFOA defluorination via Cu(I)/Cu(II) redox cycling under simulated solar light. Yet no current system achieves complete short-chain mineralization, and the review is explicit that these strategies represent leading research directions rather than a solved problem.

Significant obstacles remain before laboratory performance translates into deployable technology. Only about 35 percent of surveyed MOFs maintained stable crystal structures after 24 hours of water immersion, and Fe leaching rates of 0.5 to 3.2 milligrams per litre per treatment cycle under acidic Fenton conditions exceed the World Health Organization drinking water guideline for iron. Pilot-scale systems typically report 30 to 60 percent lower efficiency than batch benchmarks, and laboratory-grade MOFs costing 20 to 200 dollars per kilogram must fall by an order of magnitude to compete with granular activated carbon at roughly 0.10 to 0.50 dollars per cubic metre of treated water. The review identifies artificial intelligence-guided framework design, continuous-flow reactor engineering and electrocatalytic coupling as transformative priorities. Graph neural networks and generative models trained on the Cambridge Structural Database’s more than 12,000 experimentally characterized frameworks could compress discovery timelines from years to months. With more than 12,000 contaminated sites in the United States alone and drinking water limits tightening to sub-nanogram levels, the societal mandate for complete, permanent PFAS destruction has rarely been clearer.

Subject of Research: MOF-based catalytic systems for PFAS degradation

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

Article References: Khan, M. H., Ejaz, U., Sharafat, S., & Khan, M. H. (2026). MOF-based catalytic systems for PFAS degradation: mechanisms, composite design and future challenges. Frontiers in Environmental Engineering, 5, Article 1907166. https://doi.org/10.3389/fenve.2026.1907166

Image Credits: AI Generated

DOI: 10.3389/fenve.2026.1907166

Keywords: PFAS, metal-organic frameworks, defluorination, photocatalysis, Fenton-like reactions, water treatment, environmental remediation, trimetallic MOFs, short-chain PFAS, mineralization, electrocatalysis, AI-guided design