The Safe Substitute

Three nanograms per kilogram of body weight per day.

That is the chronic oral reference dose the US Environmental Protection Agency established for GenX -- formally known as hexafluoropropylene oxide dimer acid, or HFPO-DA -- in its final Human Health Toxicity Assessment, published October 2021 and updated March 2023.¹ The reference dose represents the amount EPA considers safe for daily consumption over a lifetime. For GenX, that amount is 3 ng/kg/day.

The reference dose EPA established for PFOA -- the compound GenX replaced -- is 20 ng/kg/day.²

A lower reference dose means the compound requires a stricter safety standard. It means the agency determined that harm occurs at a lower threshold. GenX, the compound introduced as PFOA's safer replacement, has a safety threshold 6.7 times more restrictive than the compound it was designed to replace.

The label on the pan says "PFOA-free."

It does not say what replaced it.

The Claim

"PFOA-free" appears on the packaging of nearly every non-stick pan sold in major retail markets. It is the only PFAS-related information available to the consumer at the point of purchase.

The claim has no federal legal definition. No regulation in the United States, the European Union, or the United Kingdom defines what "PFOA-free" means, what it requires, or what it permits.³ A manufacturer can apply the term to any product that does not intentionally contain PFOA -- regardless of what other per- and polyfluoroalkyl substances the product does contain. California's Assembly Bill 1200, effective January 2024, is the sole exception: it prohibits manufacturers from claiming a product is "free of" a specific chemical when the product contains other chemicals from the same hazardous family.⁴ No other jurisdiction imposes this constraint.

The consumer reads "PFOA-free" and reasonably concludes: the dangerous chemical was removed. The problem was solved.

The consumer cannot determine -- from the label, the product page, the manufacturer's website, or any regulatory database -- which specific PFAS compounds are present in the coating. Whether GenX, PFBS, or another short-chain processing aid was used in manufacture. Whether migration testing has been performed on the finished product. What the EPA's safety assessment of the replacement compounds says.

The open question -- "Is there a dangerous chemical in my pan?" -- has been answered. "PFOA-free." The inquiry closes. Kruglanski and Webster's research on need for cognitive closure describes this mechanism: when the brain receives a definitive resolution to an open question, it seizes on the answer and freezes further information-seeking.⁵ The follow-up question -- "What PFAS are in my pan?" -- never forms, because the prior question was resolved.

What Replaced PFOA

PFOA was phased out of US manufacturing under the EPA's PFOA Stewardship Program, completed in 2015.⁶ The primary replacement in fluoropolymer manufacture is GenX (HFPO-DA), a compound developed by DuPont beginning in 2009 and commercially introduced in 2010 -- before PFOA's formal phase-out was complete.⁷ When Chemours was spun off from DuPont in 2015, GenX's intellectual property transferred to the new entity, according to public corporate filings and patent records.⁸

GenX, PFBS, and other short-chain PFAS serve as processing aids in the manufacture of polytetrafluoroethylene -- PTFE, the polymer that forms the non-stick coating. These processing aids remain as residuals in the finished coating.

All of them are PFAS. All share the carbon-fluorine bond -- the strongest single bond in organic chemistry, the bond that makes the coating non-stick, and the bond that does not break down in the environment or the body.

The EPA's own toxicity assessment of GenX found the following adverse effects in animal studies: liver toxicity as the primary target -- hepatocellular hypertrophy, single cell necrosis, hepatic fat infiltration; kidney toxicity; immune suppression; developmental toxicity including neonatal mortality and low birthweight in rats; liver and pancreatic tumours; and endocrine disruption.¹ [animal, in vitro]

In April 2024, the EPA set the maximum contaminant level for GenX in drinking water at 10 parts per trillion -- evidence the agency considers GenX a public health concern at concentrations measured in the trillionths.⁹ (Note: in May 2025, the EPA announced intent to reconsider this standard; the MCL's future is uncertain as of March 2026.)

A 2025 study on HFPO-DA tissue distribution in mice found the compound rapidly absorbed into the bloodstream and distributed across all tested tissues, including penetration of the blood-brain barrier. The highest concentrations accumulated in the liver, with significant reprogramming of hepatic lipid metabolism -- disruption of the liver's normal capacity to process fats.¹⁰ [animal]

This is the compound behind "PFOA-free." The label communicates an absence. The chemistry delivers a replacement that the regulator subsequently determined requires a stricter safety standard than the compound it replaced.

The EPA's draft reference dose for GenX, published in 2018, was 80 ng/kg/day -- higher than PFOA's 20. By the time the final assessment was published in 2021, the value had dropped to 3 ng/kg/day: a 26.7-fold reduction as additional evidence accumulated.¹ The direction of revision has been consistently downward. As more data arrives, GenX looks worse, not better.

A note on what the reference dose comparison represents. GenX's lower RfD reflects both emerging evidence of harm and greater scientific uncertainty -- the EPA applied a 3,000-fold total uncertainty factor, including a 10-fold factor for database deficiency, meaning the agency itself considers the evidence base incomplete.¹ The two RfDs are based on different critical effects (liver lesions for GenX, developmental toxicity for PFOA) and different levels of scientific confidence. A lower RfD can mean "we know less and are being more cautious" as well as "this compound appears more potent." Both interpretations lead the same direction: the safety margin for the replacement is narrower than for the original. For a consumer whose cookware may contain GenX without disclosure, the practical consequence is the same regardless of which interpretation prevails.

The Release Surface

The engineering term for a non-stick coating is a "release surface." It is designed to release food.

It also releases itself.

Non-stick cookware delivers fluorinated contamination into food through two independent pathways, and no consumer-facing source examines the combined exposure.

Pathway One: Chemical Migration.

PFAS processing-aid residuals migrate from the coating into food during normal cooking. Not at the 260C degradation threshold that safety advice warns against. At 70C -- the temperature of a simmering sauce -- in acidic food simulant, researchers detected PFOA in 6 of 35 tested cookware samples at concentrations of 2.12 to 8.86 nanograms per gram.¹¹ [consumer-use migration]

In the only published study to measure PFAS migration from non-stick cookware over repeated use, AbulFadl et al. (2019) found that migration increases with every cooking cycle. In acidic food with salt: PFOS at 18.30 micrograms per kilogram on the first use. 39.55 after five uses. 60.33 after ten -- approximately a threefold increase.¹² [consumer-use migration]

A 2024 analysis using thermal desorption gas chromatography-mass spectrometry detected seven distinct PFAS compounds in migration fluid from frying pans, including PFNA at 1.76 to 2.12 micrograms per litre.¹³ [consumer-use migration]

The study measured to ten uses. A consumer who cooks five meals a week for three years uses the same pan approximately 780 times. Research covers less than 1.3 percent of the pan's use life. Whether the migration curve continues upward -- as processing-aid residuals deplete or as coating degradation exposes new surfaces -- or plateaus, no one has studied. The gap between the last data point and the last use is the gap this report orbits.

Pathway Two: Physical Release.

The coating sheds.

Cole et al. (2024), published in Science of the Total Environment, found that even intact, new non-stick cookware releases an estimated 2,409 to 4,964 PTFE microplastic particles per year into homecooked food. Non-plastic cookware tested as a control introduced none.¹⁴ [consumer-use migration]

A surface crack -- the kind of scratch from a metal utensil -- releases approximately 9,100 particles per cooking event.¹⁵ A broken or severely damaged coating may release up to 2.3 million micro- and nanoplastics, according to Luo et al. (2022), who used Raman imaging to identify PTFE particles from non-stick cookware for the first time.¹⁵ [consumer-use migration]

These are not inert foreign bodies. They are PTFE -- a fluorinated polymer. Fluorinated fragments entering the body as physical contaminants alongside the chemical PFAS migration from the same surface. Two pathways. One pan. One meal.

The non-stick coating is not a sealed surface. It is a slow-release membrane. What enters the food is not a single substance through a single route. It is a dual delivery: dissolved chemistry migrating through the liquid phase, and solid fluorinated fragments breaking from the polymer surface. Two streams from the same source, into the same meal, into the same body.

This is the Release Surface: a product surface engineered for one form of release -- food from coating -- that simultaneously performs a second, undisclosed form of release: chemical and physical contaminants from coating into food. The consumer experiences the intended release. The unintended release is invisible, unmeasured, and undisclosed.

The framework is not unique to cookware. It applies wherever a functional coating is also an exposure pathway:

• PFAS-treated food packaging that resists grease but migrates processing chemicals into food
• "Water-resistant" outdoor gear whose durable water repellent coatings shed PFAS into sweat and soil
• "Stain-resistant" textiles and carpets where PFAS treatment migrates through skin contact

The Release Surface names what the surface does, not what the chemistry is. It identifies the structural condition in which a product's defining functional feature is also its contamination mechanism.

Pan to Liver

The body pathway traces a direct line from the coating surface to the organs.

Entry. Ingestion. The consumer swallows both dissolved PFAS and solid PTFE particles with their meal. Every meal cooked on a non-stick surface delivers both pathways simultaneously.

Travel. PFAS enter the bloodstream through gastrointestinal absorption. Once in the blood, they bind to human serum albumin -- the transport protein that carries them through the circulatory system.¹⁶

Accumulation. Blood is the transport medium, not the primary storage site. PFAS preferentially partition out of blood into protein-rich tissues. The primary accumulation site is the liver, where PFAS bind strongly to liver fatty acid binding protein, known as L-FABP.¹⁶ Secondary sites: kidney, lung, and brain. Organ-level concentrations exceed blood concentrations.¹⁷ [animal, human biomonitoring]

An autopsy study of 99 tissue samples from Tarragona, Spain -- the first to measure PFAS accumulation across multiple human organs from the same subjects -- found PFAS in liver, kidney, lung, brain, and bone. Lung tissue showed the highest concentrations.¹⁷ [human biomonitoring]

For GenX specifically: the 2025 mouse tissue distribution study found HFPO-DA concentrations highest in liver, followed by kidney, lung, heart, and spleen -- with blood-brain barrier penetration confirmed.¹⁰ The liver showed the most significant functional impact: reprogramming of hepatic lipid metabolism. The liver's ability to process fats -- triglycerides, cholesterol -- was disrupted. [animal]

Harm. "Chemicals migrate into food" is mechanism. What follows is harm.

GenX's primary target is the liver: hepatocellular hypertrophy, single cell necrosis, apoptosis, fat infiltration.¹ [animal] The European Food Safety Authority's 2020 scientific opinion shifted the critical endpoint for PFAS from cholesterol effects to a more sensitive measure: reduced response of the immune system to vaccination in children. EFSA set a group tolerable weekly intake of 4.4 nanograms per kilogram of body weight per week for the sum of four PFAS -- PFOA, PFOS, PFNA, and PFHxS.¹⁸ [epidemiological, human biomonitoring] PFOA exposure is associated with kidney cancer.¹ GenX developmental toxicity in animal studies includes neonatal mortality and low birthweight.¹ [animal]

To be precise about the evidence: these health outcomes are established in animal studies and occupational/environmental epidemiology for PFOA and PFOS. For GenX, the evidence base is thinner -- animal studies are the primary source, and the EPA's 3-fold database deficiency factor reflects this. No human epidemiological study has linked non-stick cookware use specifically to any health outcome. The harm associations come from the compound class, not from the cookware-specific exposure route.

TIMELINE OF PERSISTENCE

• Time in your pan (use life): 3-5 years
• Time in your meal (per serving): minutes (dissolved); ingested with food
• Time in blood (PFBS, single dose): approximately 44 days
• Time in liver (GenX): emerging data; organ concentrations exceed blood
• Time in environment (C-F bond): 1,000+ years
• Time on label: "PFOA-free" (no expiry)

The pan wears out. The C-F bond does not.

The Half-Life Illusion

The strongest defence of short-chain PFAS as safer replacements rests on pharmacokinetics: PFBS has a serum half-life of approximately 44 days, compared to PFOA's 2.7 to 3.8 years.¹⁹ The body clears the replacement faster. This is accurate. For a single dose, the advantage is genuine.

But the consumer does not receive a single dose. The consumer cooks on this surface daily. Under continuous exposure, even a compound with a 44-day half-life reaches steady-state blood concentration. It takes approximately seven half-lives -- about ten months -- to eliminate 99 percent of a single dose of PFBS. A new dose arrives with tomorrow's dinner. The half-life advantage matters only if the exposure stops. The exposure does not stop.

Environmental persistence is identical across chain lengths. The C-F bond -- the structural feature that makes the coating non-stick -- does not break down regardless of whether the PFAS molecule is long-chain or short-chain. Every PFAS molecule released from a pan into food, into a body, into a landfill, into groundwater, persists for an estimated 1,000 years or more.²⁰ "Shorter body half-life" describes how fast the compound moves through a human body. It does not describe how long it lasts in the world.

A 2024 paper by Lawrence Wackett in Microbial Biotechnology reframes PFAS persistence as an evolutionary problem, not merely a chemical one.²¹ The C-F bond can be cleaved by microbial enzymes -- the bond strength is not the primary barrier. The barriers are evolutionary: C-F bond cleavage releases fluoride ions, which are toxic to all cells, creating negative selective pressure against organisms that develop the ability to degrade PFAS. The metabolic energy gained from breaking down PFAS is negligible.

And the degradation pathway requires multiple simultaneous enzymatic innovations -- transport mechanism, cleavage enzyme, fluoride sensor, fluoride export -- all of which must evolve together. We created a class of compounds that evolution cannot learn to decompose because the learning process itself kills the learner.

And the half-life hierarchy inverts across the placenta.

A 2024 birth cohort study from Zhoushan, China, measured 24 PFAS in paired maternal and cord blood. Short-chain PFAS showed higher placental transfer efficiencies than intermediate-chain compounds: PFBA at 146 percent, PFPeA at 118 percent, PFHxA at 110 percent, PFBS at 97 percent.²² [human biomonitoring]

A transfer efficiency above 100 percent means the cord blood concentration exceeds the mother's blood concentration. The compounds marketed as "safer" because they clear the adult body faster reach the developing foetus at higher relative concentrations than the compounds they replaced.

The shorter half-life that makes the replacement appear safer in an adult is the property associated with higher placental transfer efficiency to the foetus. The compound clears the mother's blood faster. It crosses into the foetal blood more efficiently. The safety feature is the exposure feature, reversed across a generation.

What the Test Missed

The regulatory tests that allow this product to exist -- and why they miss what matters.

The FDA Food Contact Notification programme evaluates individual PFAS compounds for food contact use. It does not evaluate finished products. It does not test what migrates from the pan into the meal. The programme approved the compound. Nobody is required to test the pan.³

There is no mandatory PFAS migration testing for finished cookware under real-world cooking conditions in any major market. The studies documenting migration at normal cooking temperatures -- AbulFadl et al., the Taylor & Francis TD-GC-MS analysis, the Toptanci detection at 70C -- are academic research, not regulatory requirements.¹²¹³¹¹ They exist because scientists chose to investigate, not because any authority mandated measurement. There is no EU-harmonised specific regulation requiring PFAS migration testing of PTFE-coated metal cookware.²³

There is no regulatory standard for PTFE microplastic release from cookware. The physical pathway -- thousands of fluorinated polymer particles entering food per year -- has no testing requirement, no disclosure obligation, and no limit. One of the two contamination pathways has no regulatory framework at all.

GenX is measured in NHANES -- the US population biomonitoring programme added HFPO-DA in the 2017-2018 cycle.²⁴ But GenX was detected in fewer than 0.4 percent of blood specimens -- fewer than 1 in 250 people -- despite the majority of American households using non-stick cookware.²⁵ By comparison, PFOS is detected in 98 to 100 percent of specimens across all cycles.²⁶

This creates a paradox. If cookware delivers GenX to consumers, why does the biomonitoring data not show it? Three partial explanations:

• Not all "PFOA-free" cookware uses GenX specifically -- many manufacturers may use other processing aids (PFBS, ADONA, proprietary alternatives), so the NHANES GenX measurement tests a narrower question than "does cookware expose you to PFAS"
• The NHANES limit of detection may be too high to capture low-level chronic exposure from cookware
• GenX may be present as a processing-aid residual in amounts too small to create detectable blood levels -- which would mean the cookware contribution to GenX exposure is genuinely small

None of these explanations resolves the paradox completely. The report does not claim cookware is the dominant GenX exposure route. It claims the consumer cannot determine what PFAS are in their pan, and the regulatory system does not require anyone to tell them.

The Strongest Defence

The opposition deserves its best case. Here it is.

"The dose makes the poison."

Migration levels from cookware under normal conditions are in the nanogram-per-gram range -- far below acute toxicity thresholds. No one gets sick from a single meal cooked on a non-stick pan. For single-exposure acute toxicity, this is correct. What it does not address: chronic low-dose exposure over years of daily use. Steady-state accumulation under continuous exposure. Combined dual-pathway loading from both chemical migration and physical particle release. The EPA's own GenX reference dose of 3 ng/kg/day was set precisely because chronic low-dose effects -- liver lesion constellations in mice over 53 to 64 days of exposure -- matter at concentrations the "dose makes the poison" framework was not designed to evaluate.¹

"PTFE is FDA-approved for food contact use."

It is. The approval reflects a genuine assessment process. What the approval covers: the polymer PTFE, assessed as an individual food contact substance. What it does not cover: migration testing of the finished pan under real-world cooking conditions with acidic food, salt, and repeated use. The combined exposure from processing-aid residuals and coating degradation. The dual pathway. The approval is real. The gap between what was approved and what the consumer is exposed to is also real.

"Consumer cookware exposure is small relative to drinking water and food packaging."

This may be true. PFAS-contaminated drinking water affects an estimated 200 million or more Americans, according to Environmental Working Group extrapolation from USGS sampling data.²⁷ Food packaging can deliver higher PFAS concentrations per serving than cookware migration. The relative contribution of cookware to total PFAS body burden is genuinely unknown -- no published study has isolated it. This is an honest data gap, not a straw man. The report acknowledges the uncertainty. But uncertainty about relative contribution does not establish that the cookware contribution is negligible -- it establishes that the cookware contribution has not been measured. The absence of evidence is not evidence of absence. The dual pathway -- chemical PFAS migration plus PTFE microplastic release -- is specific to cookware. Drinking water does not deliver fluorinated microplastics. Multiple smaller sources compound; the existence of larger sources does not make a smaller source acceptable.

"Short-chain PFAS clear the body faster than long-chain."

True for single-dose elimination. The pharmacokinetic improvement is genuine. What it does not cover: continuous daily exposure from cookware use, which produces steady-state concentrations regardless of half-life. Environmental persistence that is identical across chain lengths. And the placental transfer data: the compounds that clear the adult body faster cross into the foetal blood at higher relative concentrations.²²

"PTFE microplastics are biologically inert."

PTFE is among the most chemically inert materials known. It is used in surgical implants, heart valves, and vascular grafts because of its biocompatibility. Ingesting PTFE particles that pass through the gastrointestinal tract may be less concerning than having PTFE permanently implanted in the cardiovascular system. This defence has genuine force. However: nano-scale PTFE particles may behave differently from bulk PTFE, as surface-area-to-volume ratio changes chemical behaviour. PTFE cookware particles may carry adsorbed PFAS processing-aid residuals on their surface, creating a particle-plus-chemical combined exposure. And the absence of evidence of harm from ingested PTFE microplastics reflects the absence of studies, not the absence of risk. The report distinguishes clearly: the chemical pathway carries the toxicological weight. The physical pathway quantifies an additional, uncharacterised exposure.

"Sixty years of mass use and no cookware-specific epidemiological signal."

PTFE-coated cookware has been in consumer kitchens since the early 1960s. Billions of meals. No published study has linked non-stick cookware use to any specific health outcome in a human population. This is an honest limitation. But PFAS exposure comes from so many sources simultaneously that isolating cookware's contribution epidemiologically is nearly impossible -- the confounding is inherent. The substitution timeline complicates it further: GenX-era cookware has been in wide use for less than a decade, insufficient time for chronic-exposure epidemiology to detect effects. And the declining PFAS blood levels in NHANES reflect removal of PFOA and PFOS, not safety of their replacements.

What Would Change This Analysis

Three conditions, any of which would require substantial revision.

First: A controlled dietary intervention study that isolates the PFAS contribution from non-stick cookware use -- measuring blood PFAS levels in participants who switch from non-stick to PFAS-free cookware for six to twelve months, compared to controls who continue using non-stick. If cookware's contribution to total PFAS body burden is less than five percent of total PFAS intake from all sources, the material concern in this report changes magnitude significantly. The dual pathway would still exist, but its public health significance would be substantially reduced.

Second: Addition of GenX to the standard NHANES biomonitoring panel at a lower limit of detection, with results showing population-level GenX blood concentrations well below the EPA reference dose equivalent blood level. The current detection frequency of less than 0.4 percent may reflect measurement limitations rather than absence. If improved measurement confirms GenX exposure from cookware is far below concern thresholds, the toxicity inversion -- while still factually accurate as a comparison of regulatory standards -- becomes less significant for consumer risk assessment.

Third: Comprehensive migration testing data from cookware manufacturers, conducted under real-world conditions -- acidic food, high temperature, repeated use over hundreds of cycles, damaged coating -- showing PFAS migration at levels well below EFSA's group tolerable weekly intake of 4.4 ng/kg body weight per week even under worst-case scenarios. If such data existed, the information gap at the centre of this report would narrow. The label would still be incomplete, but the material risk would be better characterised.

All three conditions are obtainable. None currently exists.

The Levers

Tier 1: No-Cost / Low-Friction.

These require no purchase. They reduce exposure using the pan you already own.

Lower the heat. The migration studies show that temperature is a primary driver of PFAS release. Use non-stick pans for low-temperature cooking only -- eggs, pancakes, crepes. High-heat searing belongs in a different pan.

Avoid acidic and salty foods in non-stick pans. Tomato sauces, vinegar-based preparations, and salt all increase PFAS migration.¹² A stainless steel or cast iron pan handles acid without releasing fluorinated compounds.

Replace damaged pans immediately. A scratch multiplies particle release by orders of magnitude -- from thousands of particles per year to thousands per cooking event.¹⁵ The moment a non-stick coating shows visible damage, its contamination output accelerates dramatically.

Ventilate. If you heat a non-stick pan above 260C, open a window. PTFE degradation products at high temperatures include perfluoroisobutylene and fluoroethylene, which cause polymer fume fever -- flu-like symptoms that resolve in hours but indicate chemical inhalation.

Tier 2: Replacement.

When replacing cookware -- not before.

Cast iron: zero PFAS concern. Iron leaching is nutritionally beneficial for most people (iron-deficiency anaemia affects approximately 1.2 billion people globally). Requires seasoning. Will last generations.

Carbon steel: zero PFAS concern. Lighter than cast iron. Requires seasoning. Preferred by professional kitchens for responsiveness to heat.

Stainless steel: zero PFAS concern. For nickel-sensitive users, 18/0 (nickel-free) options exist. Not non-stick -- requires oil and technique.

Ceramic-coated cookware (sol-gel technology): no PFAS in the coating composition. Shorter lifespan than PTFE -- one to three years versus three to five. Proprietary formulations mean the exact composition varies by manufacturer. When choosing ceramic, look for third-party testing confirming no intentionally added PFAS. The material specification matters; the brand name does not.

The cost is real. At entry-level pricing, a PFAS-free alternative costs approximately twice a comparable PTFE pan. Ceramic's shorter lifespan increases the lifecycle cost further. Over ten years, the total cost of ceramic may be three to four times the cost of PTFE. This is a structural equity issue: the lowest-income consumers are the most likely to remain in the PFAS exposure pathway. The gap is not a matter of consumer choice. It is a consequence of a market structure where the health costs of PFAS exposure are not priced into the PTFE product.

The Landing

The consumer now knows: what is in the pan. GenX, PFBS, and other PFAS that no label discloses. How it enters the body: two pathways, both operating at normal cooking temperatures -- dissolved chemistry migrating through the food, solid fluorinated particles breaking from the coating surface. Where it goes: liver, kidney, lung, brain. What the regulator's own numbers say about the replacement: a safety threshold 6.7 times more restrictive than the compound it replaced. And what to do about it -- starting with the Tier 1 behavioural levers that cost nothing, progressing to material alternatives when the pan reaches the end of its life.

The information gap remains. No mandatory migration testing. No compound disclosure. No PTFE microplastic standard. No isolation of the cookware contribution to total PFAS body burden. The label still says "PFOA-free," and the consumer still cannot determine what that pan contains.

This is Series Report 1 of 2. Report 046, The Class Exemption, maps the regulatory architecture that produces this material reality -- why compound-by-compound assessment cannot answer the consumer's question, and what class-based restriction would change.