Per- and polyfluoroalkyl substances (PFAS) are a large and structurally diverse class of synthetic chemicals containing fully fluorinated carbon groups. They include small, mobile compounds, surface-active chemicals, gases, and high-molecular-weight polymers. Their resistance to heat and chemical attack, together with their unusual behavior at interfaces, has supported applications ranging from firefighting foams and protective clothing to semiconductor manufacture and industrial seals. Some PFAS have been manufactured since the mid-twentieth century.
PFAS are also important environmental contaminants. Many resist ordinary biological and chemical degradation, while others transform into highly persistent fluorinated products. Releases during manufacture, use, and disposal have contaminated drinking-water sources, soils, wildlife, and human populations. The expression “forever chemicals” refers to this persistence but is not a precise chemical classification: PFAS differ substantially in mobility, accumulation, toxicity, and susceptibility to treatment.
Understanding PFAS requires distinguishing the properties of individual substances from those of the class. Evidence is particularly extensive for perfluorooctanoic acid and perfluorooctanesulfonic acid, whereas many other PFAS remain less thoroughly characterized. Research and regulation therefore address both well-studied individual compounds and the broader problems of persistent emissions, complex mixtures, incomplete measurement, and management across product life cycles.
Definition, nomenclature, and principal families
PFAS are defined by molecular structure rather than by a single use or toxicological property. A widely used definition published by the Organisation for Economic Co-operation and Development in 2021 includes fluorinated substances containing at least one fully fluorinated methyl or methylene carbon atom, without hydrogen, chlorine, bromine, or iodine attached to that carbon. This encompasses substances with a or group. Definitions used in particular laws, monitoring programs, and scientific studies can be narrower, so reported counts of PFAS depend on the classification and chemical inventory.
In a perfluoroalkyl group, fluorine replaces all hydrogen atoms on the corresponding alkyl group. Polyfluoroalkyl substances contain fully fluorinated carbon groups alongside carbon atoms that are not fully fluorinated. A simplified perfluoroalkyl group can be written . The remainder of the molecule determines whether the substance behaves as an acid, a neutral volatile compound, a surface-active agent, or part of a polymer. “PFAS” is consequently not synonymous with “fluorinated chemical”: many fluorinated substances lack the structural groups required by a given PFAS definition.
Important nonpolymeric families include perfluoroalkyl carboxylic acids, perfluoroalkane sulfonic acids, fluorotelomer alcohols, sulfonamides, and fluorinated ether acids. Perfluorooctanoic acid (PFOA), with formula , has eight carbon atoms when its carboxyl carbon is included. Perfluorooctanesulfonic acid (PFOS), conventionally written , has an eight-carbon perfluoroalkyl chain. In most environmental waters, these acids occur predominantly as negatively charged ions, although their acid names are commonly retained when discussing measurements.
Polymeric PFAS include fluoropolymers, such as polytetrafluoroethylene (PTFE), and side-chain-fluorinated polymers, which carry fluorinated groups attached to a different polymer backbone. These architectures matter: a fluorinated side chain may be released through degradation even when the backbone remains intact, whereas an intact, high-molecular-weight fluoropolymer has different transport and biological-accessibility properties from a small fluorinated acid.
The terms “long-chain” and “short-chain” have specific conventions in much PFAS literature. Long-chain perfluoroalkyl carboxylic acids generally have eight or more total carbons, while long-chain perfluoroalkane sulfonic acids have six or more perfluorinated carbons. These categories are useful for describing trends, not for establishing universal boundaries between hazardous and harmless substances. Branching, ether oxygen atoms, functional groups, and molecular charge can be as important as chain length.
Molecular properties and environmental persistence
The carbon–fluorine bond is central to PFAS chemistry. Fluorine is highly electronegative, and carbon–fluorine bonds in many fully fluorinated structures are difficult to break under ordinary environmental conditions. The fluorinated portion of a molecule also changes the accessibility and reactivity of its carbon skeleton. Nevertheless, persistence cannot be inferred from bond strength alone: the complete molecular structure and the availability of viable reaction mechanisms determine whether degradation occurs.
Fluorinated chains have low polarizability and interact differently with water, hydrocarbons, and other materials than ordinary hydrocarbon chains do. Some PFAS combine a fluorinated, water-repelling segment with an ionic or otherwise water-compatible head group. Such molecules are surfactants: they concentrate at interfaces and reduce surface or interfacial tension. This behavior explains their effectiveness in wetting, coating, emulsification, and foam formation. It also affects environmental transport, because an interface between air and water can retain substantial quantities of surface-active PFAS.
Many perfluoroalkyl acids are terminal environmental transformation products. Their structures lack readily accessible sites for the enzymatic reactions that normally initiate breakdown of organic contaminants. Conventional hydrolysis, exposure to sunlight, and ordinary microbial metabolism often provide no efficient route to mineralization. In contrast, some polyfluorinated precursors contain nonfluorinated portions that can be attacked by enzymes or atmospheric oxidants. Those reactions may shorten or modify the original molecule while leaving a persistent perfluorinated acid.
Persistence, mobility, and bioaccumulation are separate properties. A short-chain acid may remain largely dissolved and travel readily with water, while a longer-chain substance may associate more strongly with sediment or biological proteins. Many ionic PFAS do not follow the simple lipid-partitioning behavior associated with conventional hydrophobic pollutants. Their distribution can involve protein binding, electrostatic interactions, and interfacial accumulation. Neutral precursors, meanwhile, can exhibit appreciable volatility and reach distant locations through atmospheric transport.
Environmental persistence also differs from biological elimination time. A compound may remain chemically intact in water for a very long period but leave a particular organism relatively quickly. Conversely, selective reabsorption and protein binding can retain a compound in the human body for years without implying that every PFAS behaves similarly. These distinctions explain why the class cannot be represented accurately by one degradation rate, one partition coefficient, or one biological half-life. They also explain why replacing a strongly bioaccumulative substance with a more mobile one can exchange one environmental management problem for another.
Industrial development and manufacturing routes
Industrial fluorochemistry expanded during the twentieth century as manufacturers sought materials resistant to corrosive chemicals, high temperatures, and demanding mechanical conditions. In 1938, Roy Plunkett discovered PTFE while working at DuPont. Its combination of chemical resistance, electrical insulation, and low friction subsequently supported a wide range of industrial and consumer applications. Large-scale development of fluorinated surfactants followed during the 1940s and 1950s, creating additional families of nonpolymeric PFAS and fluorinated surface treatments.
Two historically important manufacturing routes are electrochemical fluorination and fluorotelomerization. In electrochemical fluorination, an organic feedstock is fluorinated electrochemically in anhydrous hydrogen fluoride. The process can replace hydrogen with fluorine and generate mixtures of linear and branched products, as well as compounds of different chain lengths. The product mixture depends on the starting material and operating conditions. Substances produced by this route include precursors used in PFOS-related chemistry.
Fluorotelomerization commonly uses tetrafluoroethylene to build fluorinated chains, followed by reactions that introduce a functional group. Many products have an even-numbered fluorinated segment and a short nonfluorinated segment. A designation such as “6:2” indicates six fluorinated carbons and two nonfluorinated carbons in the relevant chain. These labels describe structural segments rather than toxicity categories. Fluorotelomer alcohols can be used to manufacture surface-treatment chemicals, and some can transform into perfluoroalkyl carboxylic acids.
Fluoropolymer manufacture involves polymerizing fluorinated monomers. Historically, PFOA or related salts served as processing aids in some aqueous polymerization processes. A processing aid is not the same substance as the finished polymer: residual low-molecular-weight chemicals, manufacturing emissions, and end-product properties must be distinguished. Some replacement processing aids are themselves PFAS, including hexafluoropropylene oxide dimer acid and its salts, associated with the trade name GenX.
Production patterns changed as environmental evidence accumulated. In 2000, 3M announced a phaseout of its principal PFOS-related chemistry. In 2006, the United States Environmental Protection Agency launched a voluntary PFOA stewardship program under which participating companies committed to reducing and ultimately eliminating specified PFOA-related emissions and product content by 2015. Such programs did not amount to a worldwide end to production or use. Manufacturing shifted geographically, replacement chemistries entered commerce, and older substances remained in products, contaminated sites, and waste streams. Historical production routes therefore continue to influence present-day contaminant mixtures and the interpretation of environmental measurements.
Applications and performance requirements
PFAS have been used where ordinary materials cannot easily provide the same combination of oil repellency, water repellency, chemical resistance, low surface energy, and thermal stability. These properties serve different functions in different products. A fluorinated surfactant used at low concentration in a manufacturing bath is not functionally equivalent to a bulk fluoropolymer component, even though both can fall within a broad PFAS definition. Evaluating an application therefore requires identifying the actual substance, its role, and the conditions under which releases occur.
A prominent use has been aqueous film-forming firefighting foam, or AFFF, developed for fires involving flammable liquids. Fluorinated surfactants help an aqueous film spread over certain hydrocarbon fuels, suppress fuel-vapor release, and complement the foam blanket. Historically, repeated training and emergency use at airports, military facilities, refineries, and fire-training grounds released PFAS directly to soil and water. Older PFOS-based formulations and later fluorotelomer-based formulations have different chemical profiles, and foam concentrates can contain complex precursor mixtures rather than only a few routinely measured acids.
In textiles, carpets, leather, and paper treatments, fluorinated chemistry has provided resistance to both water and oil. That combination is technically distinct from water repellency alone, which can often be obtained using other materials. Side-chain-fluorinated polymers have been used to modify surfaces while preserving the mechanical properties of a substrate. Food-contact paper and packaging historically represented another application, although restrictions and voluntary changes have reduced certain uses in several jurisdictions.
Fluoropolymers have applications in gaskets, seals, valves, chemical-processing equipment, wire insulation, membranes, and nonstick coatings. Their resistance to aggressive chemicals can be important in industrial systems. In semiconductor manufacturing, PFAS have been used in specialized formulations, equipment components, and processes requiring precise control of contamination and surface behavior. Other applications have included metal plating, photographic materials, lubricants, and selected medical and laboratory devices.
Alternatives are evaluated against the particular performance requirement rather than against PFAS as an undifferentiated category. Fluorine-free firefighting foams, nonfluorinated surface treatments, and alternative process designs can replace some uses, but their performance depends on the fuel, substrate, equipment, and operating conditions. Substitution can also change energy consumption, service life, worker exposure, or waste generation. These comparisons are life-cycle questions: low emissions during a product’s normal operation do not by themselves account for releases during raw-material production, manufacture, maintenance, or disposal.
Environmental sources, transport, and transformation
PFAS enter the environment through manufacturing emissions, industrial wastewater, firefighting-foam use, product wear, waste disposal, and atmospheric releases. Wastewater-treatment plants can redistribute them among effluent, sludge, and air without destroying the fluorinated structures. Land application of contaminated sludge can transfer PFAS to agricultural soils. Landfills receive PFAS-containing products and industrial wastes, and their leachate can return contaminants to wastewater-treatment systems, producing a continuing circulation among waste-management facilities.
Movement through soil depends on molecular structure and site conditions. Negatively charged PFAS interact with mineral surfaces, organic matter, dissolved ions, and air–water interfaces. Chain length and functional group often influence retention, but soil pH, salinity, water saturation, and the presence of other chemicals also matter. In unsaturated soil, accumulation at air–water interfaces can delay downward movement. Changes in moisture conditions can then mobilize part of the retained inventory, sustaining releases long after the original source has ceased.
Once PFAS reach groundwater, persistent compounds can form extensive contaminant plumes. Short-chain acids often move relatively readily, whereas some longer-chain compounds and precursors are retained more strongly. Transport through an aquifer is therefore not adequately described by water velocity alone. Desorption, diffusion into less permeable zones, and slow precursor transformation can prolong contamination. A site may show continuing concentrations even after source removal because contaminated soils and low-permeability materials remain secondary reservoirs.
PFAS also move through rivers, coastal waters, and the atmosphere. Neutral volatile precursors can undergo long-range atmospheric transport and later oxidize into persistent acids. Ionic PFAS can be transported in aerosols and water droplets, including particles generated from contaminated sea spray. Ocean circulation provides another long-distance pathway. Detection in remote regions does not necessarily indicate nearby manufacture; it can reflect transport over years through interconnected atmospheric and aquatic systems.
Environmental transformation complicates source attribution. A measured increase in a terminal acid may result from precursor conversion rather than a new release of that acid. Conversely, a decrease in one monitored compound may represent conversion to an unmeasured PFAS rather than destruction. Researchers therefore distinguish disappearance of the parent molecule from complete mineralization, in which organic carbon is converted to inorganic products and fluorine is recovered as fluoride. The persistence of transformation products makes chemical mass balances essential when interpreting field observations and treatment experiments.
Human exposure and biological disposition
Human exposure occurs through drinking water, food, household dust, consumer products, and occupational environments. Their relative contributions vary with geography, age, diet, occupation, and proximity to contaminated sources. Drinking water can dominate exposure in communities affected by industrial releases or firefighting sites. Food pathways include contaminated fish, crops, livestock products, and transfer from some food-contact materials. Occupational exposure can be substantially higher in fluorochemical manufacturing and certain activities involving PFAS-containing materials.
Exposure assessment distinguishes external intake from the amount reaching internal tissues. Many well-studied perfluoroalkyl acids are readily absorbed after ingestion. Their distribution differs from that of pollutants that accumulate mainly in fat: binding to blood and tissue proteins, particularly serum albumin, is important. Some PFAS concentrate in blood, liver, and kidneys. Molecular size, charge, protein affinity, and transport across biological membranes influence which tissues receive the greatest exposure.
Elimination involves renal and biliary processes, including active transport and reabsorption. A substance filtered or secreted toward urine may be transported back into the body, prolonging retention. Human elimination half-lives for PFOA, PFOS, and perfluorohexane sulfonic acid are generally measured in years, although estimates differ among populations and study methods. Other PFAS can be eliminated much faster. Half-lives measured in laboratory animals may differ markedly from those in humans and can vary by sex within an animal species.
Some PFAS cross the placenta and are present in breast milk, establishing exposure pathways during early development. Transfer depends on the compound and maternal concentration. Blood concentrations also change with growth, pregnancy-related physiological changes, and elimination through menstruation or lactation. These processes complicate comparisons across age groups and populations. A lower measured concentration does not necessarily mean that earlier exposure was low, particularly where production changes or migration have altered exposure over time.
Biomonitoring measures chemicals in biological samples, most commonly serum or plasma for established PFAS. Population surveys have documented widespread exposure to several legacy compounds, and concentrations of some have declined in countries following production restrictions and phaseouts. However, routine panels cover only a fraction of the class. A serum result represents selected internal exposures at the sampling time; it does not inventory every PFAS in the body, identify all sources, or by itself determine whether a particular disease was caused by exposure.
Toxicological and epidemiological evidence
Health evidence is strongest for a limited group of well-studied PFAS, especially PFOA and PFOS. Investigations combine laboratory toxicology, mechanistic experiments, occupational studies, and community-based epidemiology. These evidence streams answer different questions. Controlled animal experiments can establish effects under specified conditions, while human observational studies examine associations under real-world exposures but must account for correlated chemicals, other risk factors, and uncertainty in past exposure.
Reported human associations for selected PFAS include higher serum cholesterol, reduced antibody responses to some vaccines, altered liver-enzyme measures, and certain pregnancy-related outcomes. The strength and consistency of evidence vary by compound and endpoint. Immune effects have been particularly important in several risk assessments. A measured reduction in vaccine antibody response is a population-level biological finding; its interpretation is not identical to diagnosing immune deficiency in an individual.
Experimental studies have identified effects on lipid metabolism, liver function, development, and immune signaling. Some PFAS interact with nuclear receptors, including peroxisome proliferator-activated receptors, which regulate metabolic pathways. Other mechanisms involve cellular signaling, transport proteins, and endocrine-related processes. No single mechanism accounts for all observed effects across the class. Species differences in receptor activity, internal dose, and elimination are major considerations when extrapolating laboratory findings to humans.
In 2023, the International Agency for Research on Cancer classified PFOA as carcinogenic to humans, Group 1, and PFOS as possibly carcinogenic to humans, Group 2B. The PFOA evaluation drew on sufficient evidence in experimental animals and strong mechanistic evidence, together with limited evidence in humans for renal cell carcinoma and testicular cancer. These classifications identify carcinogenic hazard. They do not specify the probability of cancer at a particular environmental concentration, and the classification of one PFAS cannot automatically be assigned to every member of the class.
Dose, duration, timing, and mixtures remain central uncertainties. Exposure during development may have different consequences from the same external dose in adulthood. Closely correlated PFAS concentrations can make it difficult to separate the contribution of individual substances in observational studies. Reverse causation can also arise when a health-related change in kidney function or physiology alters chemical concentrations. Risk assessments address these issues through evidence integration and uncertainty factors, but estimates differ when agencies select different endpoints, models, or assumptions. Findings for well-studied compounds provide an important evidence base without establishing that thousands of structurally distinct PFAS share identical potency.
Sampling, chemical analysis, and exposure measurement
PFAS analysis requires unusually careful contamination control because fluorinated materials are common in laboratory and field equipment. Sampling containers, tubing, waterproof clothing, instrument components, and laboratory reagents can introduce background contamination. Validated methods specify suitable materials and include field blanks, laboratory blanks, recovery checks, and other quality controls. Collection and storage procedures also matter because compounds can adsorb to container surfaces, partition into particles, or change through precursor transformation.
Targeted analysis commonly combines liquid chromatography with tandem mass spectrometry. Chromatography separates compounds, while mass spectrometry detects characteristic ions and fragmentation patterns. Isotope-dilution methods use labeled standards to compensate for extraction losses and matrix effects. This approach can measure selected PFAS at very low concentrations in drinking water, but a reported result applies to the compounds included in the method. A nondetect means that a target was not quantified above the method’s reporting threshold, not that the sample contains no PFAS.
Standardized methods have expanded beyond a few legacy acids. United States EPA Methods 537.1 and 533 address specified PFAS in drinking water. Method 1633, finalized in 2024, covers 40 PFAS in several environmental matrices, including wastewater, solids, and tissue. Different matrices require different extraction and cleanup procedures. A method suitable for relatively clean drinking water cannot simply be assumed to perform equally well in sludge, industrial effluent, or biological tissue.
Broader analytical approaches seek substances missed by targeted panels. High-resolution mass spectrometry supports suspect screening and identification of previously unrecognized compounds, although identification confidence depends on reference standards and structural evidence. The total oxidizable precursor assay subjects a sample to strong oxidation and measures changes in selected perfluoroalkyl acids. It can reveal an oxidizable precursor contribution, but it neither identifies all original precursors nor converts every PFAS quantitatively into measured products.
Aggregate fluorine measurements provide a complementary perspective. Techniques measuring adsorbable or extractable organic fluorine estimate operationally defined fractions of fluorine-containing material. They do not directly yield the number, identity, or toxicity of PFAS present, and some measured organic fluorine may come from compounds outside the PFAS definition being used. Comparing fluorine equivalents from identified substances with aggregate measurements can expose an unexplained fraction. Such comparisons require consistent units, extraction recoveries, and attention to inorganic fluoride. The resulting “fluorine mass balance” is therefore a structured accounting exercise rather than a single definitive test for all PFAS.
Water treatment, site remediation, and destruction
PFAS treatment distinguishes separation from destruction. Separation removes contaminants from water or another medium but transfers them into a smaller waste stream. Destruction breaks down the organic fluorinated structure, ideally producing fluoride and other characterized inorganic products. A treatment system can show excellent removal from its outlet while leaving the original PFAS largely intact in spent media, sludge, foam, or concentrated liquid. The management of these residuals is therefore part of the treatment process, not a separate afterthought.
Granular activated carbon is widely used to remove certain PFAS through adsorption. Performance depends on chain length, functional group, natural organic matter, contact time, and carbon properties. Short-chain compounds often pass through earlier than longer-chain substances. Ion-exchange resins can provide high capacity for selected anionic PFAS, but competing ions and the chemical composition of the water affect performance. Both technologies require monitoring for breakthrough and eventual management or regeneration of the treatment medium.
Reverse osmosis and nanofiltration can reject a broad range of PFAS, including many relatively mobile compounds. Their effectiveness depends on membrane characteristics, operating conditions, and the substances present. They produce a concentrated reject stream and require energy and maintenance. Foam fractionation uses the tendency of surface-active compounds to accumulate at air–water interfaces, concentrating them into a foam that can be collected. It is generally more effective for strongly surface-active PFAS than for some short-chain species.
Contaminated-site remedies include source excavation, containment, groundwater extraction and treatment, and in situ sorbent amendments. Their purpose may be to reduce migration or exposure rather than eliminate the entire contaminant inventory. Long-term behavior depends on site geology, residual sources, and the reversibility of retention. Immobilizing PFAS does not destroy them, and an amendment’s short-term laboratory performance does not by itself establish decades of field stability.
Destructive approaches under investigation or use in particular settings include high-temperature thermal treatment, supercritical water oxidation, electrochemical oxidation, plasma processes, and selected photochemical or chemical reduction methods. Performance is highly dependent on the waste matrix and operating conditions. Demonstrating destruction requires more than measuring the decline of a few target compounds: relevant evidence includes fluoride release, organic-fluorine balances, gas emissions, and identification of intermediates. A process effective for a concentrated industrial solution may be impractical for a dilute groundwater plume. Much treatment development consequently combines concentration with a separate destructive step, while evaluating energy demand, equipment corrosion, secondary waste, and incomplete breakdown.
Regulation, risk assessment, and substitution
PFAS governance operates through several distinct mechanisms: restrictions on manufacture and use, drinking-water limits, food-exposure assessments, emissions controls, contaminated-site rules, and waste-management requirements. These instruments address different questions and should not be treated as interchangeable. A drinking-water limit specifies a concentration in a defined medium; a production restriction controls entry into commerce; and a tolerable intake estimates exposure associated with a particular health-protection framework. Their numerical values depend on legal as well as scientific considerations.
The Stockholm Convention on Persistent Organic Pollutants has listed several PFAS groups for international control. PFOS, its salts, and perfluorooctane sulfonyl fluoride were listed in 2009; PFOA, its salts, and PFOA-related compounds in 2019; and perfluorohexane sulfonic acid, its salts, and related compounds in 2022. The listings differ in their provisions and exemptions. They address named substances and related chemical groups rather than establishing a single worldwide rule covering every PFAS.
In the European Union, the 2020 Drinking Water Directive established parametric values of for the sum of 20 specified PFAS and for “PFAS Total,” with implementation provisions concerning which parameter is used. Also in 2020, the European Food Safety Authority established a group tolerable weekly intake of per kilogram of body weight for the combined intake of PFOA, PFOS, PFNA, and PFHxS. Reduced vaccine antibody response was the critical effect underlying that assessment. The value applies to those four compounds together, not to all PFAS.
In the United States, EPA’s April 2024 final drinking-water rule established enforceable maximum contaminant levels of each for PFOA and PFOS, alongside provisions for several other PFAS and selected mixtures. This dated rule illustrates the distinction between compound-specific limits and mixture-based approaches. In 2023, authorities from Denmark, Germany, the Netherlands, Norway, and Sweden submitted a broad PFAS restriction proposal under the EU’s REACH chemicals framework. A proposal and its scientific evaluation are distinct from an enacted restriction.
A central policy question is whether to manage PFAS one substance at a time, in structural subgroups, or as a broad class with use-specific distinctions. Individual assessment offers chemical specificity but is slow relative to the size of the class. Broader approaches address persistence and the possibility of replacing one problematic substance with another, while requiring distinctions among polymers, mobile acids, precursors, and different exposure pathways. Substitution assessments consequently examine technical function, available alternatives, emissions across the life cycle, and the possibility that an alternative introduces a different hazard. These questions connect chemical regulation with product design, industrial process control, and the long-term stewardship of persistent materials.