The coating is called ceramic. It is not ceramic.
A ceramic is a material formed by heating clay or other minerals to temperatures exceeding 1,000°C. The resulting product — a tile, a plate, a kiln-fired pot — is chemically inert. It does not react with food. It does not degrade with use. It has been in human kitchens for eight thousand years.
The coating on a "ceramic" nonstick pan is a sol-gel — a synthetic hybrid of silicone polymer, silica nanoparticles, and organosilane binders, applied at 200-430°C.1 The American Ceramic Society classifies these coatings as "quasi-ceramic" and notes that proprietary formulations may contain "organic polymers with varying toxicity."2 The functional nonstick component is PDMS — polydimethylsiloxane — a silicone. Its surface energy is 19-21 Dyne/cm.1 PTFE's surface energy is 18 Dyne/cm. The "ceramic" alternative to Teflon works by nearly identical physics. It is silicone marketed under a word that means clay.
No fired clay is involved. No consumer is told this.
And in at least one commercial formulation — described in patent US7879449B2, granted 2011 — the top coat of the sol-gel system, the layer that touches your food, contains fluoroalkoxysilane at 0.3-12% by weight.3 Fluoroalkoxysilane is a fluorinated compound. A consumer who switched from PTFE to "ceramic" to escape fluorinated chemistry may have purchased fluorinated chemistry under a name that evokes the oldest inert material in the kitchen.
This report is not about one coating. It is about what your kitchen transfers into your food — across every surface, every meal, every day — and why nobody has ever measured the total.
What Comes Off
In 2018, researchers at the US National Institute of Standards and Technology purchased commercially available "ceramic" nonstick pans and subjected them to three tiers of simulated use: washing with scrubbing pads, scouring with steel wool, and abrasion with a tungsten carbide burr — a rotary grinding tool that represents the most aggressive mechanical condition.4
Under that most aggressive condition, they measured what came off.
One hundred million titanium dioxide nanoparticles per square decimetre. Median diameter: 250 nanometres.4 Under the less aggressive conditions — washing and scouring with conventional tools — nanoparticle release at this concentration was not confirmed.4 The 108 figure represents a laboratory upper bound, not routine kitchen use. What it establishes is the nanoparticle reservoir within the coating: the particles are there, in quantity, and mechanical stress liberates them. The question is how much stress, and how often.
At 250 nanometres, these particles can cross cellular membranes that larger particles cannot — a property documented in animal studies and central to EFSA's 2021 conclusion that titanium dioxide could no longer be considered safe as an oral food additive [animal, in vitro].5 The study also detected ten million silicon dioxide nanoparticles per square decimetre, median diameter 460 nm.4
An independent study by Golja, Dražić, and colleagues at the Jožef Stefan Institute, published in Food Additives & Contaminants, found titanium migration of up to 861 micrograms per litre from "ceramic" coatings into 3% acetic acid — a standard food simulant — confirming that titanium enters food from these coatings in measurable quantities under standard test conditions.6
The EU banned titanium dioxide (E171) from food, effective August 2022.7 The ban applies to titanium dioxide intentionally added to food — as a whitening agent in sweets, sauces, supplements. It does not apply to titanium dioxide that migrates into food from a cookware coating. The particle sizes overlap. The entry route is the same: ingestion. The regulatory treatment is different because one is classified as a food additive and the other falls under food contact material regulation — which, for coatings, has no specific EU measures.8
The same substance. Comparable particle sizes. Banned when added deliberately. Unregulated when released incidentally.
What It Replaced
"Ceramic" nonstick was marketed as the exit from PTFE. The clean alternative. The pan you buy when you learn that Teflon raises questions.
What PTFE raises:
In 2022, researchers at Flinders University and the University of Newcastle used Raman imaging to count the particles released from PTFE-coated nonstick cookware. A single surface crack released approximately 9,100 microplastic particles. A visibly damaged coating section released 2.3 million.9 Particle sizes ranged from 13 to 318 micrometres.
A 2024 study by Plymouth Marine Laboratory, using a jelly-based food simulant to capture particles during cooking, estimated that daily use of PTFE cookware releases 2,409 to 4,964 microplastic particles into food per year.10 Cookware made of non-plastic materials — cast iron, stainless steel, carbon steel — released zero microplastics.10
PTFE begins decomposing at 260°C.11 An empty pan on a high burner exceeds this temperature within minutes. Stir-frying routinely operates above 260°C. The decomposition products include perfluorooctanoic acid, tetrafluoroethylene, trifluoroacetic acid, and perfluoroalkyl carboxylic acids spanning C3 to C14 — compounds within the PFAS family that persist in the human body with serum half-lives of two to eight years [human biomonitoring].12 Chronic low-level PFAS exposure is associated with thyroid disruption, immunosuppression, and developmental effects [epidemiological].12
Pet birds die from PTFE fume exposure at concentrations that produce no symptoms in humans.11 They are the kitchen's unintentional sentinel species.
The consumer who switched from PTFE to "ceramic" made a rational decision based on available information. They eliminated one particle stream. They introduced another. They traded PTFE microplastics, 13-318 micrometres in diameter, for titanium dioxide nanoparticles, 250 nanometres in diameter — particles two to three orders of magnitude smaller, less studied, and with different biological behaviour: PTFE microplastics largely transit the gastrointestinal tract, while TiO₂ nanoparticles at 250 nm can cross cellular membranes [animal].5
The exposure changed character. It did not stop.
What Was Always There
Stainless steel is the cookware that nobody questions.
In 2013, Kamerud, Hobbie, and Anderson at Oregon State University published the first comprehensive measurement of metal leaching from stainless steel cookware into food. They cooked tomato sauce — a standard acidic food — in a new Grade 316 stainless steel saucepan and measured what entered the food.13
From a new pan: 483 micrograms of nickel and 67.5 micrograms of chromium per 126-gram serving, after twenty hours of cooking.13
Twenty hours is an extreme laboratory protocol — far longer than any home cooking session. But the finding is not limited to extreme durations. After ten cooking cycles of standard length, when the pan's surface had been conditioned by use, nickel leaching was still 88 micrograms per serving under acidic conditions — fifteen to twenty-six times above baseline levels measured in food cooked in glass.13
The twenty-hour figure quantifies the upper bound. The eighty-eight-microgram figure describes what happens in practice.
To contextualise: the European Food Safety Authority sets a tolerable daily intake for nickel at 13 micrograms per kilogram of body weight per day. For a 70-kilogram adult, that is 910 micrograms per day.14 Kamerud's 88-microgram figure, from a conditioned pan, represents roughly 10% of this limit from a single serving of one food from one pan.
That margin exists. It is real. It is also calculated for nickel alone, from one material, in one meal.
Grade 316 stainless steel contains 10-14% nickel — more than the Grade 304 (8-10% nickel) used in most domestic cookware. The leaching levels from a typical home kitchen pan may be lower than Kamerud's figures. But Grade 304 was also tested in the study using NIST reference materials, and it too leached nickel at levels well above baseline.13 The phenomenon is not grade-specific. It is a property of chromium-nickel stainless steel in contact with acidic food.
Ten to twenty per cent of the population is sensitised to nickel [human biomonitoring].15 For these individuals, there is no established safe threshold — sensitisation is a binary immune response, not a dose-response curve. No cookware carries a nickel warning.
The chromium in Kamerud's data raises a harder question. Chromium exists in two forms: trivalent chromium (Cr III), a nutritional trace element, and hexavalent chromium (Cr VI), classified by the International Agency for Research on Cancer as a Group 1 human carcinogen.16 Kamerud's study measured total chromium. It did not distinguish between the two forms.13
In materials science, stainless steel corrosion in mildly acidic conditions predominantly releases trivalent chromium. The Council of Europe's 2024 Technical Guide on metals in food contact now specifies its chromium limit as Cr(III) only — an implicit expert assessment that cookware chromium is predominantly trivalent.17 This is likely correct. But "likely" is not "measured." No published study has speciated chromium from cookware leaching under cooking conditions. The gap has persisted for more than a decade since Kamerud's publication.
This is honest uncertainty. It belongs in the report because it belongs in the evidence.
What the Name Obscures
"Stainless steel" is not a single material.
A tri-ply pan — the construction marketed as "professional grade" — is a sandwich: two layers of Grade 304 stainless steel (18% chromium, 8-10% nickel) permanently bonded around an aluminium core (typically 99.7% purity or 3003 alloy). The bonding process — high-pressure cold rolling at 18,000 to 22,000 psi — creates a metallurgical bond exceeding 300 newtons per square millimetre, a permanent fusion described in ASM International's handbook on clad metals.18
"18/10 stainless steel" sounds like a material. It is a ratio — describing the chromium and nickel content of one layer of a multi-material composite. The marketing presents a fraction as an identity.
This has a direct end-of-life consequence. Tri-ply cookware cannot be recycled as stainless steel because it is not stainless steel. The aluminium core cannot be separated from the steel. Sent to a steel recycler, the aluminium contaminates the melt. Sent to an aluminium recycler, the steel contaminates it. Few curbside recycling programmes accept cookware of any kind.19
The pan marketed as "recyclable stainless steel" is metallurgically neither recyclable nor stainless steel. It is a bonded composite with no commercial separation pathway.
"PFOA-free" is the third name that obscures. PFOA — perfluorooctanoic acid — is one compound in a class of thousands of per- and polyfluoroalkyl substances. The Ecology Center's 2020 "What's Cooking?" study, using X-ray fluorescence and infrared spectroscopy, found that 79% of nonstick cookware tested contained PTFE.20 PTFE is itself a PFAS compound. A pan labelled "PFOA-free" may contain PTFE and remain fully compliant with the claim.
Three names. Three gaps between what the word suggests and what the material is.
The Temperature Nobody Tests At
For metals and alloys in contact with food, there is no harmonised EU migration test. No standardised temperature. No required protocol. No specific limit.
EU Regulation 1935/2004 covers 17 categories of food contact material.8 Specific measures have been adopted for 4: plastics, recycled plastics, ceramics (lead and cadmium only), and regenerated cellulose. The remaining 13 categories — including metals, coatings, silicones, wood, and rubber — have only the general framework and, for metals, a non-binding Council of Europe guidance document that member states may adopt, adapt, or ignore.17
For the four materials that do have tests, the gap is not temperature per se — EU Regulation 10/2011 specifies test conditions up to 225°C for plastics in high-temperature applications, using vegetable oil as the food simulant above 100°C.21 But aqueous food simulants — the acetic acid and ethanol solutions that replicate the acidic foods driving metal leaching — are limited to 100°C or reflux conditions.21 Tomato sauce at a simmer is 85°C. Tomato sauce in a pan over medium-high heat can exceed 100°C at the pan surface. The test captures the simmer. It does not capture the sear.
Kamerud's nickel data was generated at sustained cooking temperature.13 The NIST nanoparticle study simulated mechanical use, not thermal exposure.4
| Method | Temperature |
|---|---|
| Simmering | 85-95°C |
| Frying | 175-190°C |
| Searing | 200-260°C |
| Stir-frying | 250-300°C+ |
| Empty pan on high heat | Exceeds 300°C in minutes |
Chemical reaction rates increase with temperature — the Arrhenius relationship, a foundational principle of thermodynamics. As a general rule, reaction rates roughly double for every 10°C increase, though the precise acceleration depends on the specific reaction, the food matrix, and the surface chemistry involved. Even conservatively, the gap between 85°C and 200°C implies a substantial acceleration in migration rates that no study has quantified under real cooking conditions.
The temperature gap for aqueous food tests is the difference between a 100°C reflux and a 200°C pan surface. For metals, there is not even a test to be inadequate.
The Portfolio
Here is the kitchen, mapped.
A household that scrambles eggs in a PTFE pan at breakfast, sautés vegetables in a "ceramic" pan at lunch, simmers sauce in stainless steel at dinner, and stirs with a silicone spatula throughout is exposed to four distinct material streams in a single day: PTFE microplastics, titanium dioxide nanoparticles, nickel and chromium ions, and PDMS from the spatula.
Each of these exposures has been studied — where it has been studied at all — in isolation. Each is assessed against its own regulatory limit or, for most, against no specific limit. The body does not experience them in isolation. The body receives them as a combined dose.
This investigation continues below.
Want the next one in your inbox?
A 2023 cumulative risk assessment published in Food and Chemical Toxicology applied Maximum Cumulative Ratio methodology to 123 stainless steel food contact samples, measuring six metals simultaneously. The abstract reports a mean MCR of 1.87 — meaning the combined risk exceeded the highest single-substance risk by nearly double — and that chemical-by-chemical assessment missed the cumulative risk for 46.5% of individuals in the study population.22
That is one material. One product category. Six metals.
Extend the principle to a kitchen. Over a single year, a household using three different cookware types across three daily meals accumulates approximately 10,000 cooking-surface contacts. Each contact transfers something — a particle, an ion, a polymer fragment — at quantities assessed as safe in isolation. No framework sums the contacts. No study has measured what three surfaces, three meals a day, 365 days a year delivers to a single body. The portfolio is not theoretical. It is the description of every kitchen. It has simply never been measured as one.
This is the Migration Portfolio: the aggregate of all materials migrating from all cooking surfaces into food within a single kitchen, experienced by the body as a combined dose but assessed by regulation as isolated streams.
The Migration Portfolio is not a claim that aggregate exposure causes harm. It is a claim that the framework for detecting harm — material by material, substance by substance, temperature below cooking temperature — is structurally incapable of seeing the aggregate. The absence of evidence of portfolio-level harm is not evidence of safety. It is evidence that nobody has looked.
The Strongest Defence
The most credible counterargument: each individual migration level falls within established safety margins.
This is true. Kamerud's 88 micrograms of nickel from a conditioned pan represents approximately 10% of the EFSA tolerable daily intake.14 The FDA considers PTFE cookware safe at temperatures below 260°C. The German Federal Institute for Risk Assessment considers metal cookware safe under normal use conditions. Hard-anodised aluminium cookware — where an electrochemical process creates a 25-100 µm aluminium oxide layer resistant to acid and abrasion — leaches well below 0.01 parts per million, an order of magnitude below concern thresholds.23 The non-anodised aluminium figure of 2,144 mg/L cited in some studies applies to uncoated aluminium in acidic conditions — not representative of the anodised cookware most Western consumers use.24
These safety margins are real. They were established by competent authorities using the best available single-substance data. For any individual material, the regulatory position is defensible.
What the defence does not cover: the sum. Each material's safety margin was calculated without reference to the other materials in the same kitchen. The 10% of the nickel TDI from stainless steel does not account for the titanium dioxide nanoparticles from the "ceramic" pan used an hour earlier, or the PTFE microplastics from the nonstick pan used that morning. Each margin was set in isolation. The kitchen is not isolated.
The defence also does not address the temperature gap. The safety margins are based on studies conducted below 100°C. Whether the margins hold at actual cooking temperatures — where the Arrhenius relationship predicts substantially faster migration — has not been tested, because no testing standard requires it for metals.
The individual-material safety case is solid. The Migration Portfolio does not argue that any single material is unsafe. It argues that the assessment framework — designed for one material, one substance, one temperature — has a structural blind spot where the kitchen actually operates: multiple materials, combined exposure, cooking heat.
The Levers
The Migration Portfolio does not have a single solution. It has a set of decisions, each of which reduces a different component of the portfolio.
What you can do without buying anything:
Reduce cooking temperature where the food allows it. Every migration process described in this report accelerates with heat. Simmering instead of frying reduces leaching rates.
Do not preheat an empty nonstick pan on high heat. An empty PTFE pan reaches 260°C — the decomposition threshold — within minutes on a high burner.11 Add oil or food before heating. Ventilate your kitchen when cooking at high heat in any pan.
Avoid cooking highly acidic foods — tomatoes, vinegar, wine, citrus — in stainless steel or bare aluminium for extended periods. Kamerud's data demonstrates that acidity is the primary driver of nickel and chromium migration.13 Glass, enamelled cast iron, or seasoned carbon steel are lower-migration options for acidic recipes.
Allow new stainless steel pans to undergo several cooking cycles with non-acidic foods before using them for acidic cooking. Kamerud's data shows nickel leaching decreases with sequential use — conditioning the surface reduces subsequent migration.13
If you are replacing cookware anyway:
Single-material cookware eliminates the coating layer entirely. No coating means no coating degradation, no nanoparticle release, no fluorinated top coat.
Carbon steel — 99% iron, 1% carbon — is a single material that develops a natural nonstick surface through seasoning: oil heated beyond its smoke point polymerises into a hydrophobic layer bonded to the iron surface.25 The nonstick property comes from food-grade oil, not synthetic chemistry. The pan weighs 45% less than an equivalent cast iron. It is infinitely recyclable through standard scrap metal streams. It costs $40-80, purchased once.
The maintenance cost: approximately 60 seconds per use. Hand-wash. Dry immediately. Apply a thin layer of oil. Over a 50-year cooking life, that is roughly 300 hours of total maintenance — six hours per year, one minute per day.
The trade-off is real: carbon steel cannot tolerate acidic foods for extended periods (acid strips the seasoning). It is not dishwasher-safe. It rusts if left wet. These are genuine functional constraints, not minor inconveniences. For acidic cooking, enamelled cast iron or well-conditioned stainless steel remain the better choices.
I could not locate a peer-reviewed study measuring iron migration specifically from carbon steel cookware. Iron migration is inferred from cast iron literature — which documents iron enrichment of food as real and, for most people, nutritionally beneficial.26 For individuals with haemochromatosis or iron-overload conditions, any iron cookware requires medical guidance. Carbon steel is not risk-free. It is a single, well-characterised element rather than a portfolio of particles, ions, and nanoparticles.
Cast iron offers similar material properties — single element, infinitely recyclable, improves with use — at greater weight and with a rougher cooking surface.
Neither carbon steel nor cast iron requires a marketing name. They are what they are: iron. The composition is the label.
TIMELINE OF PERSISTENCE
- Time on the hob: 30 minutes
- Time in the body (nickel): hours to days (renal clearance)
- Time in the body (PFAS from PTFE degradation): 2-8 years (serum half-life)
- Time in landfill (nonstick pan): indefinite (PTFE does not biodegrade)
- Time of aggregate kitchen assessment: zero (no study exists)
What Would Change This Analysis
Two specific pieces of evidence would materially alter the conclusions of this report.
First: a comprehensive dietary exposure study measuring aggregate particle and ion migration from a representative multi-material kitchen. Cooking actual meals — not food simulants — across the cookware surfaces a household actually uses, at the temperatures those surfaces actually reach. If such a study found that combined migration from a multi-material kitchen remains below concern thresholds even when summed across material categories, the Migration Portfolio would be a theoretical construct without practical significance. That study does not exist. Its absence is the central gap this report identifies.
Second: chromium speciation of the leachate measured by Kamerud et al. If the chromium migrating from stainless steel cookware is entirely trivalent — as materials science suggests and the Council of Europe's updated guidance implies — then the chromium component of the stainless steel concern substantially diminishes, narrowing the migration issue to nickel alone. If any hexavalent chromium is present, the concern escalates. This single analytical measurement would resolve a question that has been open for more than a decade.
Both studies are feasible. Neither has been conducted.