Forever Chemicals Decoded: How PFAS Conquer Water, Soil and Air

Few classes of industrial chemicals have earned a nickname as ominous as “forever chemicals,” yet the label is scientifically deserved. Per- and polyfluoroalkyl substances, or PFAS, are a vast family of synthetic compounds built around carbon-fluorine bonds so strong that they resist oxidation, reduction, hydrolysis and virtually every natural degradation pathway. A comprehensive new review published in Results in Chemistry by Balasubaramaniyanyan Seethalakshmi, Ganeshkumar Govindasamy, Sunantha Ganesan and Krishnasamy Sivagami assembles the molecular evidence for why these compounds persist in every environmental compartment, how they travel from factory effluent to Arctic ice, and why the short-chain replacements now flooding the market may be even harder to contain than the legacy chemicals they replaced.

The scale of the problem begins with the chemistry. Since the 1950s, PFAS have been manufactured on an industrial scale through two routes: electrochemical fluorination, which yields mixtures of branched and linear isomers, and telomerization, which produces shorter, linear chains with even numbers of carbon atoms. The OECD expanded the formal definition in 2021 to include any fluorinated substance containing at least one fully fluorinated methyl or methylene carbon atom, and a 2018 OECD/UNEP global inventory identified 4,730 PFAS-related CAS numbers that may have been on the world market. The structural secret of their staying power lies in the carbon-fluorine bond itself. With a bond dissociation energy of roughly 485 to 544 kilojoules per mole, a bond length of just 1.35 angstroms, and fluorine’s unmatched electronegativity of 4.0, the bond is among the strongest polar covalent bonds in organic chemistry. Although each individual bond is highly polarized, the symmetric tetrahedral arrangement of multiple carbon-fluorine bonds cancels the dipoles, leaving the perfluoroalkyl tail nonpolar, hydrophobic, lipophobic and nearly immune to attack by external electric fields.

That tail is only half of the story. Every PFAS molecule pairs its water-repellent fluorinated backbone with a polar, hydrophilic head group, typically a carboxylate or sulfonate, giving the compounds an amphiphilic, surfactant-like character that lowers the surface tension of water. This dual nature is precisely what makes PFAS so mobile and so troublesome: the head group drags the molecule into aqueous phases, while the tail refuses to let go of interfaces. In living systems, the consequences are severe. PFAS accumulate in serum proteins and hepatic tissue, disrupt the hypothalamic-pituitary-thyroid axis, interfere with androgen and estrogen signaling, activate the PPAR-alpha nuclear receptor and perturb lipid metabolism, and generate reactive oxygen species that oxidize lipids and damage DNA. Human half-lives stretch over years, and the compounds have been detected in blood, lung and kidney tissue. The United States Environmental Protection Agency responded on April 10, 2024 by finalizing drinking water standards for six PFAS compounds, while long-chain variants were phased out under REACH between 2011 and 2016.

The phase-out, however, created an unexpected paradox. The short-chain PFAS introduced as replacements bioaccumulate less in humans and animals, but they are dramatically more mobile in water, harder to remove in treatment plants, prone to uptake by plants and capable of climbing the food chain. They have already been flagged as substances of very high concern under REACH. The review emphasizes that this structural transition has been addressed only superficially in the literature, even though chain length and functional group chemistry fundamentally control how PFAS partition between air, water, soil and living tissue. Understanding those partitioning behaviors is now the central challenge of PFAS science.

At the air-water interface, the governing parameter is the dimensionless Henry’s law constant, the ratio of a compound’s gas-phase concentration to its aqueous concentration. Neutral, volatile precursors such as fluorotelomer alcohols exhibit constants between 0.31 and 2.82 and can travel thousands of kilometers through the atmosphere before degrading into persistent perfluorinated carboxylic acids. Ionic terminal compounds like PFBS show negligible volatility and remain locked in water. One of the most striking findings compiled in the review concerns sea spray aerosols. As waves burst bubbles at the ocean surface, perfluoroalkyl acids become enormously enriched in the ejected droplets relative to bulk seawater, and field studies show a significant positive correlation between PFAA concentrations and sodium ions, with PFOA correlating at r = 0.77. Aerosol particles a few micrometers in radius can carry PFAS more than 300 kilometers, with atmospheric residence times ranging from days to weeks, effectively allowing the ocean to re-emit these chemicals into the air alongside sea salt.

Beneath the surface, the vadose zone, the unsaturated soil layer between ground level and the water table, acts as a delayed-release reservoir. Two retention mechanisms operate simultaneously: sorption to soil solids and accumulation at air-water interfaces within soil pores. Because adsorption at these interfaces is nonlinear, lower concentrations paradoxically produce higher retardation factors; when PFOA concentrations dropped from 1.0 to 0.01 milligrams per liter, the retardation factor rose from 2.0 to 2.8, and PFOS retardation climbed from 1.35 to 3.69 under comparable dilution. In saturated soil, the soil-water distribution coefficient Kd governs mobility, and it is exquisitely sensitive to pH. At high pH, negatively charged organic matter and minerals electrostatically repel anionic PFAS head groups, weakening sorption; at low pH, protonated clay and organic carbon bind the compounds more tightly. Long-chain compounds such as PFOS show Kd values falling from 42 to 17 liters per kilogram across the pH 3 to 9 range, while short-chain PFBA barely responds, remaining stubbornly mobile throughout.

In living organisms, partitioning follows a different logic. Linear isomers accumulate more readily than branched ones, which are eliminated faster due to their lower octanol-water partition coefficients. Protein binding, quantified through the protein-water partition coefficient, dominates bioaccumulation: PFOS binds human serum albumin with a dissociation constant of just 0.62, stronger than PFOA, while short-chain PFBA interacts weakly. Fish sampled across global studies carry approximately 62 micrograms of PFAS per kilogram, exceeding regulatory advisory limits, and more than 90 percent of human exposure occurs through food and drinking water. Molecular dynamics simulations reveal why chain length matters at the membrane: short-chain PFAS are too stubby to pack against the hydrocarbon tails of lipid bilayers and interact only as a monolayer, whereas long-chain compounds form tightly cohesive van der Waals-packed layers that lodge deep in the membrane core and resist removal.

Degrading these molecules remains one of environmental engineering’s hardest problems. Microbial transformation offers little comfort: the carbon-fluorine bond’s dissociation energy of roughly 480 to 560 kilojoules per mole exceeds what most enzymes can attack, PFAS provide essentially no metabolic energy to bacteria, and degradation typically stalls at persistent shorter-chain intermediates rather than achieving mineralization. A notable 2020 study by Yu and colleagues did document reductive defluorination of two C6 compounds by an organohalide-respiring community, with fluoride release occurring only when the fluorinated compounds served as exclusive electron acceptors, and intriguingly, the activity was driven by minor phylogenetic groups rather than the dominant Dehalococcoides. On the abiotic side, photocatalysis begins with decarboxylation, peeling CF2 units from PFOA one at a time until only carbon dioxide, water and fluoride remain, and vacuum ultraviolet systems boosted PFOA degradation from 10 to 95 percent and defluorination from 17 to 94 percent. But most destructive technologies demand extreme conditions, with complete mineralization requiring temperatures near 850 degrees Celsius, making in situ application impractical.

Conventional wastewater treatment fares little better. Primary clarification merely shuttles PFAS between phases, activated sludge leaves most compounds untouched, coagulation-flocculation fails for short-chain variants, and adsorption and ion exchange simply concentrate the chemicals onto spent media requiring disposal. Reverse osmosis and nanofiltration achieve high removal but at substantial cost, energy demand and membrane fouling risk, and they too transfer rather than destroy the contamination. This is where computational modeling is transforming the field. Machine learning tools such as PFAsorptionML, trained on 1,274 measured soil-liquid distribution coefficients covering 51 PFAS, now predict sorption from molecular weight, hydrophobicity and soil organic carbon with robust validation performance, while quantitative structure-property relationships, symbolic regression and graph-based neural networks promise to extend predictions to the thousands of PFAS never experimentally characterized. The review’s authors caution that artificial intelligence should complement, not replace, mechanistic understanding, since extrapolation to poorly represented compounds and matrices remains unreliable.

The overall picture that emerges is sobering but actionable. Long-chain PFAS retard in soils and sediments, binding to iron oxides and organic matter and acting as slow secondary sources to groundwater; short-chain PFAS race through aquifers, dominate surface waters and slip past treatment barriers. Precursor transformation can even increase apparent concentrations of terminal acids over time, meaning that fate assessments must track chemistry, not just movement. The authors argue that future research must prioritize hybrid degradation systems combining chemical and biological processes, cost-effective capture technologies tailored to short-chain compounds, and AI-guided structure-property models capable of screening the more than 15,000 PFAS variants that no laboratory budget could ever test individually. Until then, the forever chemicals will keep traveling, on sea spray, through aquifers and into the food web, exactly as their molecular architecture dictates.

Subject of Research: Environmental fate, transport and degradation of per- and polyfluoroalkyl substances (PFAS)

Article Title: Fate and transport of per -and polyfluoroalkyl substances (PFAS) in environment

Article References: Seethalakshmi, B., Govindasamy, G., Ganesan, S., & Sivagami, K. (2026). Fate and transport of per-and polyfluoroalkyl substances (PFAS) in environment. Results in Chemistry, 31, Article 103846. https://doi.org/10.1016/j.rechem.2026.103846

Image Credits: AI Generated

DOI: 10.1016/j.rechem.2026.103846

Keywords: PFAS, forever chemicals, carbon-fluorine bond, environmental transport, short-chain PFAS, bioaccumulation, air-water partitioning, vadose zone, photocatalysis, machine learning, wastewater treatment, sea spray aerosols