The Pan and the Two Words
A nonstick frying pan sits on a hob: twenty-eight centimetres across, four hundred grams of pressed aluminium, with the charcoal-grey semi-matte finish that is the modern look of polytetrafluoroethylene — PTFE, the polymer the world still calls Teflon. The box on the shelf carries one printed phrase that, more than the brand or the price, is doing the work of converting a stranger into a buyer.
PFOA-free.
I bought one of the pans the week I started this report. I stood in the cookware aisle at a Tesco superstore in west London on a Tuesday afternoon and read the back of seven boxes in a row. Six of the seven carried PFOA-free on the front of the package. Three of those six also carried PTFE-free somewhere on the back; three did not. The seventh was a cast-iron skillet and said nothing about either, because it did not have to. The phrase is the modern shorthand for a chemistry story the consumer learned to fear two decades ago: perfluorooctanoic acid, the C8 surfactant linked in the DuPont Cape Fear and Parkersburg litigation to cancer, thyroid disease, and developmental harm, the molecule the EPA phased out under the 2006–2015 Stewardship Program, the molecule that put Dark Waters into cinemas in 2019.1 PFOA-free is the cookware aisle's response to that history. The consumer reads it and breathes out.
The consumer is doing the right thing. The shelf check Sarah does on a Saturday afternoon when she is choosing a replacement for the scratched pan in her cupboard is exactly the check the consumer-protection system asks her to do. She has read the news cycle. She has picked the pan that names the absence of the molecule she was taught to avoid. There is no other check available to her at the shelf. The label is the signal she has.
This report is about what that signal is — and is not — measuring.
The spine object of the report is the instrument behind the label: the regulatory migration test that fluoropolymer cookware is cleared under. In the United States this is the food-contact substance framework codified at 21 CFR §175.300 and §177.1550.2 In the European Union it is Regulation (EU) 10/2011 on plastic materials and articles intended to come into contact with food, with the framework Regulation (EC) 1935/2004 behind it.3 Both regimes work the same way. A piece of the cured coating is cut out, weighed, and immersed for a defined number of hours at a defined temperature in a food simulant — typically distilled water, dilute aqueous ethanol at one or two concentrations, dilute acetic acid, and an oily simulant (vegetable oil, isooctane, or Miglyol depending on the protocol). The simulant is then chemically analysed for extractables: the residual monomers, oligomers, and named processing aids that the regulator has decided should be quantified for that polymer class. The pan passes if the measured extractables sit below the migration limits. The label that follows from passing this test is the PFOA-free line on the box.
The assay measures the dissolved-phase migration of named molecules into liquid simulants under controlled time-and-temperature conditions. Three other things now known to leave the pan in your hand — and to enter your food, your kitchen air, and, over decades, your blood — sit outside what the assay was built to detect.
What the Test Was Built To See
The fluoropolymer food-contact framework in the United States was codified in the 1960s and elaborated through the 1970s. 21 CFR §177.1550 — Perfluorocarbon resins — was first promulgated under the Food Additives Amendment of 1958, and the migration-testing apparatus of 21 CFR §175.300 was developed in the same period under FDA's Bureau of Foods.2 The framework was a product of its analytical era. Migration, as the regulators of the 1960s and 1970s understood it, meant a molecule dissolving out of a plastic into a food. The instruments available were gas chromatography, paper and thin-layer chromatography, ultraviolet–visible spectrophotometry, and the first generations of high-performance liquid chromatography. The conceptual model was dissolution: a polymer in contact with a liquid food simulant releases molecules into the liquid, and one then measures what is in the liquid.
This was the right model for the questions the regulator was asking. The fluoropolymer of concern in 1965 was PTFE itself; the suspected leachate was unreacted tetrafluoroethylene monomer or low-molecular-weight PTFE oligomer. The food simulants were chosen to span the chemistry of foods that an American family might cook in 1965 — water for boiled vegetables, dilute ethanol for wines and broths, dilute acetic acid for vinegars and acidic sauces, vegetable oil for fats. The temperatures and contact times were the temperatures and times of long-contact food storage and conventional stovetop cooking.
The targeted analyte list was the second decision. The regulator named the molecules that the test would quantify in the simulant. For fluoropolymer cookware, this list has historically been tetrafluoroethylene monomer, hexafluoropropylene where copolymerised, low-molecular-weight PTFE telomers, and the named polymerisation processing aid of the era — through 2015, for the major Western manufacturers, PFOA.4 A molecule not on the list was not measured. A coating that passed the test was a coating whose listed analytes had migrated below their migration limits into the simulants tested. The clearance under §177.1550 is a clearance against the listed analytes, by extractables migration, into the listed simulants, at the listed temperatures, for the listed times.
The EU framework, finalised as Regulation (EU) 10/2011, did not break from this architecture. Annex I lists the authorised substances and assigns each a Specific Migration Limit (SML) or, where no SML is set, an overall migration limit of 10 mg of substance per square decimetre of contact surface.3 The food simulants are similar — water, ethanol at three concentrations, acetic acid, vegetable oil — chosen to cover aqueous, alcoholic, acidic, and fatty foods. The test apparatus is a piece of cured coating immersed in a simulant for a specified period at a specified temperature; the simulant is then analysed for the named substances. EFSA, the European Food Safety Authority, can issue scientific opinions to update the SMLs and the authorised list, but the underlying assay architecture — dissolved-phase migration of named analytes into liquid simulants — has not been replaced.
The test measures what it was designed to measure. The clearance it produces is true within the four corners of the assay; PFOA-free on a pan that passed it is technically accurate because PFOA is not in the targeted analyte list as a quantified residue above the migration limit. The architecture does not overstate itself. The gap is in everything the architecture was not asked to look at.
The Three Pathways the Assay Cannot See
Three exposure pathways from a domestic nonstick pan are characterised in the peer-reviewed analytical-chemistry literature, the federal occupational-health literature, and the federal toxicological-profile literature. Each has been measured. None of them is what the migration assay tests for.
Particle shedding
In 2022, Yunlong Luo and colleagues at Flinders University in South Australia and Newcastle University published a paper in Science of the Total Environment titled "Raman imaging for the identification of Teflon microplastics and nanoplastics released from non-stick cookware."5 The group used Raman microspectroscopy — a vibrational-spectroscopy technique that identifies a polymer particle by its molecular fingerprint, not by chemical extraction — to quantify the release of PTFE microplastic and nanoplastic particles from coated cooking surfaces under simulated cooking. The paper reports, in its central finding, that a single surface crack on a Teflon-coated frying pan could release approximately 9,100 plastic particles in a single cooking event, and a broken coating could release approximately 2.3 million microplastics and nanoplastics over the cooking lifetime measured. The instrument used was a Raman microscope with deep-learning-assisted particle identification; the particle counts are PTFE, identified by its spectral fingerprint at characteristic Raman shifts.
The Flinders work sits within a broader analytical-chemistry literature on micro- and nanoplastic release from food-contact polymers, including earlier studies on plastic-fragment release from coated paper cups, plastic kettles, and infant-feeding bottles. Mohammad Sajid and Muhammad Ilyas's 2017 perspective in Environmental Science and Pollution Research established the analytical framework — Raman, pyrolysis–gas chromatography–mass spectrometry (pyr-GC-MS), and single-particle inductively coupled plasma mass spectrometry (sp-ICP-MS) — that allows polymer particles below the size cutoff of the older optical methods to be identified and counted.6 The detection envelope of these instruments extends well below 1 micrometre, into the nanoparticle range.
The extractables assay at 21 CFR §175.300 and at EU 10/2011 measures dissolved-phase chemistry in a liquid simulant. A PTFE microparticle suspended in a simulant is not in the dissolved phase; it is solid polymer. The targeted-analyte chromatography that quantifies leachable molecules does not register a polymer particle as a positive on a named analyte. Ask the regulator's test did any registered analyte migrate above its specific migration limit, and it answers no. Ask the Raman microscope did intact PTFE particles leave the surface and enter the food, and it answers in the thousands per cooking event from a scratched pan, in the millions from a degraded one. Two correct answers to two different questions.
The cookware in the Flinders study was regulator-compliant. The legal status of the pans tested is not at issue; the instrument is.
Thermal decomposition
PTFE is thermally stable at room temperature and through the temperature window most stovetop cooking occupies. Above approximately 260°C the polymer begins to decompose. The decomposition rate accelerates as temperature rises and becomes rapid above 350°C. The products of decomposition include tetrafluoroethylene monomer (TFE), hexafluoropropylene (HFP), perfluoroisobutylene (PFIB), carbonyl fluoride (COF₂), hydrogen fluoride (HF), and ultrafine particulate matter in the respirable size range.
The thermal-decomposition profile of PTFE is among the most thoroughly characterised in the polymer literature. The Agency for Toxic Substances and Disease Registry's Toxicological Profile for Perfluoroalkyls — the federal public-health reference for PFAS toxicology — documents the decomposition temperature window and the principal decomposition products.7 The National Institute for Occupational Safety and Health has maintained an occupational-health page on PTFE since the 1970s; the NIOSH literature documents polymer fume fever — a recognised flu-like illness — in workers exposed to PTFE pyrolysis products at industrial temperatures, and notes that pet birds (canaries, parakeets, parrots) die from exposure to PTFE thermal decomposition products at concentrations and durations consistent with overheated domestic cookware.8
The 260°C threshold sits below the surface temperature of an empty pan preheated on a standard gas burner. The 350°C threshold sits at the temperature of high-heat searing, oven roasting, and any preheat-and-forget scenario where the pan is brought up to temperature without food. The marketing imagery on a nonstick-pan box typically shows a steak hitting a hot surface; the small-print care instructions, where they exist, advise against empty preheating and against use above 230°C. The gap between the picture on the box and the line in the care instructions is the temperature gap the polymer cares about.
The migration assay is conducted at the simulant temperatures the protocol specifies — typically 40°C, 70°C, or 100°C, the temperatures of long-contact food storage and conventional aqueous cooking. The protocol does not require sampling at the surface temperatures where decomposition begins. The analyte panel is dissolved-phase residues in liquid simulants. Gas-phase decomposition products entering kitchen air at the moment a pan reaches 280°C are not in the simulant; they have left the system the assay is measuring. Detecting them requires headspace gas chromatography–mass spectrometry (headspace GC-MS) or Fourier-transform infrared spectroscopy (FTIR) sampling at the decomposition temperatures — methods the extractables assay does not employ.
The occupational-health regime, run by NIOSH and OSHA, detects these species and has detected them for fifty years. The food-contact regime does not. The kitchen sits between the two regimes: a domestic space whose pan reaches surface temperatures that an occupational regulator would treat as hazardous if a worker were standing over them, but the framework that cleared the pan for sale into that kitchen does not test for what the occupational regulator measures.
Replacement processing aids
The third pathway is the chemistry that replaced PFOA.
When PFOA was phased out by the major Western fluoropolymer manufacturers under the 2006–2015 EPA Stewardship Program, the polymerisation surfactants used to manufacture PTFE were substituted. The replacement chemistries, depending on the manufacturer, include hexafluoropropylene oxide dimer acid — GenX, marketed by Chemours and produced primarily at the Fayetteville Works facility in North Carolina — together with ADONA (a Solvay-developed processing aid), EEA-NH₄, and various perfluorobutanesulfonate (PFBS) and short-chain PFAS surrogates used by Daikin, AGC, and other producers.9 The polymer applied to the pan after the substitution carries the same PTFE backbone — the same carbon–fluorine bond architecture, the same thermal-decomposition curve, the same mechanical-shedding behaviour. What changed sits in the manufacturing pathway, not in the finished coating's bulk chemistry.
EPA's Integrated Risk Information System (IRIS) finalised a chronic oral reference dose for GenX (HFPO-DA) in 2021 at 3 × 10⁻6 milligrams per kilogram per day, based on hepatotoxicity (liver) endpoints in animal studies.10 The reference dose is lower than PFOA's. EPA's 2024 final National Primary Drinking Water Regulation set a maximum contaminant level for GenX (HFPO-DA) in drinking water of 10 nanograms per litre — 10 parts per trillion — and a Hazard Index limit for mixtures of GenX, PFHxS, PFNA, and PFBS.11 The European Chemicals Agency added HFPO-DA and its ammonium salt to the REACH Candidate List of Substances of Very High Concern in 2019, on grounds of vPvB (very persistent, very bioaccumulative) and equivalent-level-of-concern toxicity.12
GenX is, by the regulatory record of the past five years, a substance of substantial toxicological concern. It is also not on the targeted-analyte list of the standard fluoropolymer-cookware extractables migration assay under 21 CFR §175.300 or EU 10/2011 Annex I. The assay quantifies named analytes; if GenX is not a named analyte for cookware clearance, the assay does not measure it. A coating manufactured with GenX as a polymerisation aid can pass the assay and be marketed as PFOA-free without any test having looked for GenX. The targeted-analyte architecture is functioning as built: the test answers the question did the named analytes migrate above their limits, and the named analyte list reflects the regulatory state of the chemistry at the time the list was compiled. The chemistry has changed. The list updates more slowly.
The same pattern is structurally available for the next substitution. If GenX is replaced under regulatory pressure by a new short-chain PFAS surrogate, the targeted-analyte list will not include that new molecule on the day the substitution happens. The label will not change. The polymer will not change. The test will continue to pass.
What the Label Encodes
Three exposure pathways, none of them in the assay; the assay is the instrument the label rests on. What does PFOA-free on the box actually certify, then, on the box's own terms?
It certifies that the cookware producer has substituted away from PFOA as a polymerisation aid. The substitution is verified through manufacturer attestation, supply-chain documentation from the fluoropolymer producer (Chemours, Daikin, AGC, or 3M), and continued compliance with the extractables migration assay under §175.300 / §177.1550 in the United States and the Specific Migration Limits in Annex I of EU 10/2011 in the European Union. The Federal Trade Commission's Guides for the Use of Environmental Marketing Claims (the Green Guides) require that a free-from claim be substantiated by evidence that the named substance is absent, and the standard substantiation is supplier attestation backed by the migration-assay clearance.13 EU Regulation 1935/2004 Article 3 requires food-contact materials not to transfer constituents to food in quantities that could endanger human health, with the migration-assay framework of EU 10/2011 as the operative test.3
The label is regulatorily compliant and technically accurate at the molecular level it names. It is read by the consumer as a class-level safety claim, and it refers to a molecule retired more than a decade ago.
The wider clean-cosmetic and clean-cookware vocabulary that has grown up around this kind of labelling produces a structurally similar architecture in other categories. BPA-free plastics, the wave of 2008–2012, were marketed on the absence of bisphenol A from polycarbonate baby bottles and reusable water containers. Bisphenol A was replaced by bisphenol S, bisphenol F, and other bisphenol analogues. A 2015 review in Environmental Health Perspectives documented that the replacement bisphenols exhibit comparable or, on some endpoints, greater endocrine-disrupting activity than the substance they replaced.14 The structural parallel between BPA-free → BPS and PFOA-free → GenX is a labelling-and-substitution pattern, offered here as analogue rather than as proof. Whether consumers reading PFOA-free in 2026 are interpreting it the way consumers reading BPA-free in 2010 interpreted that label is a question primary consumer-research evidence would resolve; the most rigorous reading of the literature is that this question has not been directly tested for the PFOA-free claim, and the BPA-to-BPS pattern is the available analogical reference.
The shelf communicates the absence of one named molecule. The polymer identity of the coating is not on it, nor the identity of the processing aid that replaced PFOA, nor the temperature at which the coating begins to decompose, nor the methodology under which migration was assessed, nor any quantification of particle shedding under standardised wear. None of these is required disclosure. Under FTC's Green Guides, EU 1935/2004, and the FDA's Food Contact Substance framework, the label is doing what it is required to do.
What the Industry Position Looks Like
The fluoropolymer-cookware industry has a structured response to the analytical-chemistry literature this report draws on. It runs on its own evidence and deserves engagement on its terms.
The position has three load-bearing components. The first is the polymer-of-low-concern framework, articulated most rigorously by Barbara Henry, Theresa Carlin, Jessica Hammerschmidt, Robert Buck, Lance Buxton, Heidi Fiedler, Stephen Seed, and Oscar Hernandez in a 2018 paper in Integrated Environmental Assessment and Management titled "A critical review of the application of polymer of low concern and regulatory criteria to fluoropolymers."15 Applying the OECD's polymer-of-low-concern (PLC) criteria to PTFE, the Henry et al. paper argues that PTFE meets the OECD criteria — high molecular weight, low residual monomer, insolubility, biological inertness — and should be regulated as a PLC rather than grouped with low-molecular-weight PFAS for restriction purposes. The FluoroCouncil, an industry trade body operating within the American Chemistry Council, cites Henry et al. and the OECD framework in its public position on fluoropolymer regulation.16
The second component is the gut-passage argument. PTFE particles, the industry position holds, are biologically inert. Ingested PTFE passes through the gastrointestinal tract without absorption or systemic uptake. The Henry et al. review summarises animal feeding studies in which PTFE administered orally produced no measurable systemic effects. EFSA's 2020 scientific opinion on PFAS in food likewise excludes fluoropolymers from the scope of its dietary-exposure assessment on the basis that the polymer is not bioavailable through ingestion.17
The third component is the temperature-of-use argument. PTFE decomposition is real, the industry position grants, but begins at surface temperatures (260°C) that the producer's instructions for use exclude. A consumer who follows the manufacturer's instructions — preheat with oil, do not exceed recommended temperatures, do not preheat empty pans on high heat — will not produce the decomposition products NIOSH documents.
On the polymer-of-low-concern framework: the OECD PLC criteria address the polymer itself, not the particulate state. A polymer meeting the PLC criteria in its high-molecular-weight bulk form is not, by that fact alone, established as low-concern in its sub-micrometre particulate state. The toxicology of microplastics and nanoplastics generally — across polymer classes — is an active research area. The 2022 EFSA Scientific Opinion on the risks to public health related to the presence of microplastics and nanoplastics in food described the available evidence on absorption, translocation, and toxicity as "limited" and called for additional research.18 The polymer-of-low-concern designation, on its 2018 articulation, did not address particle-state behaviour because the particle-shedding evidence base on cookware was not yet published. Luo et al. is from 2022. Henry et al. is from 2018.
On the gut-passage argument: this report's exposure architecture is not solely oral. The thermal-decomposition pathway is inhalational; the gas-phase products NIOSH documents enter kitchen air and the respiratory tract. The gut-passage defence, by construction, addresses one of the three pathways and does not address the other two. The particle-shedding pathway is dominantly oral but raises particle-state questions the bulk-polymer studies do not.
On the temperature-of-use argument: the producer's small-print instructions do exclude empty preheating and high-heat use. The producer's marketing imagery — the steak hitting the hot surface, the stir-fry over the flame, the oven roast — does not. Both sit on the same package. The consumer replicating what the package showed them doing is not violating instructions in a culpable sense.
The industry position is structurally true at the gut-passage level. It is a defence of one of the three exposure routes. It does not address what enters the kitchen air, and it does not address what the assay does not measure.
What This Looks Like Together
A nonstick pan in domestic use sits inside three test regimes simultaneously.
The first is the food-contact regime: 21 CFR §175.300, §177.1550 in the United States; Regulation (EU) 10/2011 and 1935/2004 in the European Union. This regime tests dissolved-phase migration of targeted analytes into liquid simulants. It does not test particles, does not test gas-phase decomposition, and does not include the replacement processing aids on its analyte list. The pan passes.
The second is the analytical-chemistry literature: Raman microspectroscopy, pyrolysis–gas chromatography–mass spectrometry, single-particle ICP-MS, headspace GC-MS. This literature measures particle release directly from cookware surfaces and identifies polymer fragments at sub-micrometre resolution. The instruments exist, the methods are validated, the results are published, and the cookware tested includes regulator-compliant pans. The instrumentation is not what the regulator runs.
The third is the occupational-health regime: NIOSH's PTFE pages, OSHA's polymer fume fever literature, the ATSDR Toxicological Profile for Perfluoroalkyls. This regime characterises gas-phase decomposition at the surface temperatures domestic stovetops reach, documents biological effects in exposed workers and in pet birds (the canary species par excellence), and does not extend its detection apparatus into the food-contact framework.
The kitchen sits at the geometric centre of the three regimes. Two of the three see the pathways; the one that produces the label does not.
The food-contact regime was built by competent regulators with the instruments available in the 1960s and 1970s, asking the questions those instruments could answer. The analytical-chemistry literature arrived later, with instruments those regulators did not have. The occupational-health regime was built around different exposed populations — industrial workers — and uses different sampling apparatus. The three regimes have not been integrated because no single regulatory agency owns the integration mandate. FDA owns food-contact migration; NIOSH owns occupational exposure; EPA owns environmental and drinking-water exposure. A pan in a domestic kitchen falls into the FDA's jurisdiction only. The other two regimes' findings sit outside what FDA is asked to consider when clearing a coating.
This investigation continues below.
Want the next one in your inbox?
The label PFOA-free sits at the output of the first regime. It is what that regime is structurally capable of producing as a consumer signal, and it cannot produce a signal about the other two regimes — the test it rests on does not see what they measure.
What Would Change This Analysis
The methodology of this report is capable of producing the answer no meaningful gap. Six things, any of which would update or substantially weaken the analysis as it stands, are named here.
First — an FDA rulemaking, formal guidance, or scientific consultation that updates 21 CFR §175.300 or §177.1550 to include particle-phase or gas-phase analytes in the cookware migration test. If the targeted analyte list is extended to include Raman-quantifiable PTFE microparticles per cooking event, headspace-measurable decomposition products at relevant temperatures, and the current generation of replacement processing aids (HFPO-DA, ADONA, EEA-NH₄), the definitional gap closes. The instrument would then see what currently sits outside it. As of the publication date of this report, no such rulemaking has been issued.
Second — a peer-reviewed apportionment study quantifying the cookware-derived contribution to total US or UK PFAS body burden, separated from drinking-water exposure, food-packaging exposure, textile exposure, and firefighting-foam exposure. NHANES finds PFAS in the serum of nearly all Americans tested, per the CDC National Biomonitoring Program.19 If a published apportionment establishes that cookware contributes a trivially small fraction of that body burden compared with drinking water and food packaging, the cookware-specific framing is disproportionate to the exposure share, and the analysis should weight accordingly. To this report's research review, no peer-reviewed apportionment study at this resolution is currently published.
Third — primary consumer-research evidence on how kitchen-shoppers actually interpret PFOA-free. The BPA-to-BPS analogue is the closest published parallel; it is not direct evidence. Survey, focus-group, or eye-tracking research that asked consumers, in the cookware aisle or in controlled experimental conditions, what PFOA-free communicated to them would either confirm the named-threat-absent reading or refute it. The behavioural-economics literature on the BPA-free claim (notably Karasik et al., 2015) supports the analogical inference; primary cookware-label data does not yet exist in the published record.14
Fourth — independent organic-fluorine analytical testing of major ceramic-coated alternative cookware (Thermolon, Greblon, brand-specific sol-gel coatings marketed as PFAS-free) using total-organic-fluorine (TOF) or extractable-organic-fluorine (EOF) methods. If the PFAS-free alternatives are found to contain organic fluorine at levels comparable to PTFE-coated cookware — from auxiliary processing aids, release agents, or curing-stage fluorinated chemistry — the path-forward layer of this report's analysis collapses; the alternatives are not materially different. Independent testing on a few ceramic products in the past five years has detected organic fluorine in some samples; a systematic, publicly reported, multi-brand study has not been issued.
Fifth — a Freedom of Information Act release of the migration-testing data underpinning current FDA fluoropolymer cookware clearances. If the FDA's record demonstrates that the cookware producers' submitted clearance dossiers do include particle-phase or gas-phase analytical data — Raman, pyrolysis-GC-MS, headspace GC-MS — beyond the published §175.300 requirements, the instrument-is-the-binding-constraint claim weakens; producers would be voluntarily generating the data the regulator does not require. This FOIA pathway has not been worked at the time of writing.
Sixth — a successful ECHA REACH derogation that permits fluoropolymer cookware to continue under the universal PFAS restriction on the grounds that the polymer is a PLC. ECHA's universal PFAS restriction proposal under REACH, submitted in 2023 by five member-state authorities, is under impact assessment with a final decision expected 2025–2026.20 A derogation grounded in the Henry et al. PLC framework would represent regulatory acceptance of the industry position on bulk-polymer concern, but would not, on its terms, address particle-state behaviour or thermal decomposition. The derogation outcome would update the regulatory landscape; the analytical-chemistry findings on the three exposure pathways would remain.
Each of these six pathways is named because, in genuine prospect, the evidence could turn through them. The answer the evidence currently supports — that the assay does not measure what the literature documents — is the answer named here. What would update it is named alongside.
The Levers
Two tiers of action follow from the chemistry: what does not require buying anything, and what to choose when something needs replacing.
Tier 1 — what does not require buying anything.
Read the small-print care instructions on every nonstick pan currently in the cupboard. The maximum-temperature line is where the polymer's thermal limit appears in writing: do not preheat empty; do not exceed 230°C (450°F) is the standard phrasing. The marketed-use imagery on the box and the website does not reflect this limit; the limit is what the polymer requires. Match cooking practice to the small print, not to the imagery: medium heat for routine use, no empty preheating, no high-heat searing. A pan that has been used above 260°C for any length of time has begun to decompose and the released chemistry has already entered the kitchen air; the immediate forward-looking action is to bring temperatures back inside the small-print envelope. Visible scratching is the second signal: per Luo et al., a single surface crack is the threshold above which particle release is measurable in the thousands per cooking event. A pan whose coating is visibly scratched is shedding measurably regardless of cooking temperature, and the small-print instructions on every nonstick package advise replacement at visible coating damage.
Ask brands which analytical method substantiates the safety claim beyond the extractables assay. The customer-service line at the brand head office is the operational pathway. The question is precise: Beyond compliance with the 21 CFR §175.300 / EU 10/2011 extractables migration assay, what analytical chemistry — Raman microspectroscopy, pyrolysis-GC-MS, headspace GC-MS, single-particle ICP-MS — substantiates the PFOA-free safety claim on this pan, and what does that chemistry measure? The brands that have the data answer within a working week; the brands that do not, do not. The asking is the diagnostic.
Tier 2 — what to choose when something needs replacing.
Sequence by what is most defensible at the chemistry layer.
The simplest path is uncoated cookware. Cast iron, carbon steel, and stainless steel are non-fluoropolymer substrates with thirty-plus-year lifespans, fully characterised release chemistry (iron release for cast iron, chromium and nickel for stainless steel, both within long-established food-contact limits), and pricing across the cookware market range from sub-£20 carbon-steel skillets to upmarket cast-iron enameled lines. The trade-off is real and worth naming: uncoated cookware requires seasoning maintenance (cast iron, carbon steel) or a different cooking technique (stainless steel demands hotter preheating and oil management before food contact). The convenience promise that the 1961 Tefal launch was built around — the wipe-clean, low-fat, low-skill pan — is what is forfeited. This is a genuine domestic trade-off, not a free substitution.
The next path is ceramic-coated (sol-gel) cookware marketed as PFAS-free. The coating is a silica–titania–zirconia composite film applied with a curing process; the chemistry is materially different from fluoropolymer chemistry at the coating layer. The caveat: published independent organic-fluorine testing on ceramic cookware has, in some cases, detected organic fluorine in finished products from auxiliary processing chemistry. This is the alternative that can be PFAS-free at the coating chemistry layer but where the consumer cannot, at the shelf, verify that the specific brand and SKU is. The brand's response to a customer-service query asking for the coating composition, the curing-process chemistry, and any independent organic-fluorine measurement is the proxy diagnostic available; brands that can answer in writing are the brands that have done the work.
The third path is the cookware that the new state-level disclosure regimes are beginning to surface. California Assembly Bill 1200, effective from 1 January 2023 for online disclosure and 1 January 2024 for on-package labelling, requires manufacturers selling cookware in California to disclose, on the product or the brand website, the presence of any chemical of concern on a designated list — including PFAS as a class.21 (The PFAS prohibition clauses within AB 1200 apply to plant-based food packaging; the cookware provisions are disclosure-and-label, not a ban.) Minnesota's Amara's Law (Statute 116.943, signed May 2023) restricts intentionally-added PFAS in eleven consumer-product categories — cookware among them — on a published timeline beginning 1 January 2025. The Cookware Sustainability Alliance filed a legal challenge in January 2025; the U.S. District Court for Minnesota denied the preliminary injunction in February 2025 and ruled against the plaintiffs' Commerce Clause challenge in August 2025.22 Cookware manufactured to the California or Minnesota standard is sold elsewhere in the United States and exported to the United Kingdom; the brand's California-compliant SKU is, by construction, the most disclosed product in that brand's line. The Saturday-afternoon action available now is to ask the brand which of their currently-sold pans meets the California AB 1200 disclosure standard and request that disclosure. The disclosure exists because the law required it; the consumer in another market can ask for the same information.
The fourth path is the disclosure rule itself. The path-forward at the structural layer — the level at which this report's Magic Wand candidate would sit — is a labelling-disclosure regime that requires, on the front of the cookware package: polymer identity (PTFE, ceramic sol-gel, anodised aluminium, uncoated metal), processing-aid identity (PFOA, HFPO-DA, ADONA, EEA-NH₄, or none), decomposition or maximum-use temperature (in the format the small-print care instructions already disclose, but on the front of the package), particle-shedding quantification under standardised wear (the analytical-chemistry methodology Luo et al. used, formalised as a regulatory test), and a migration methodology that includes particle and gas-phase analytes. The precedents exist in statute: AB 1200's disclosure architecture, Amara's Law's chemical-by-chemical restriction, the EU Cosmetics Regulation Article 19 ingredient-listing requirement that 074 and 085 in this series have walked. The instrument exists in the analytical-chemistry literature. The regulatory authority exists at the FDA under the Food Additives Amendment of 1958 and at the European Commission under EU 10/2011. What is missing is the rule that combines them.
This is what we are adding to the Magic Wand list: a cookware disclosure rule that puts polymer identity, processing-aid identity, decomposition temperature, and particle-shedding quantification on the package — together with a migration-test methodology, available now in the published literature, that the regulator can adopt to measure what the consumer's pan is actually emitting in the consumer's kitchen.
The label answers a 2005 question with a 1970s test. The pan in the hand emits things that neither the question nor the test was built to measure. The terms of the test were written before the instrument that could read the pan in your hand was invented; the new instrument exists, and the label has not yet been updated to read it.