There is a plastic coating inside every food can, and it is the most heavily regulated few microns in your kitchen

Open a tin of tomatoes and look at the inside of the empty can. It is not bare metal. There is a smooth, usually pale gold or off-white film covering every square millimetre of it, a few microns thick, and your food has spent its entire shelf life in contact with that film and never once touched the steel underneath.

That coating is doing a job you would notice immediately if it failed — acidic tomato in contact with metal corrodes the can and tastes of it — and it has been reformulated across the industry twice in living memory, not because it stopped working but because of what it was made from. It is a good way into a subject most people never think about: everything that touches food transfers something into it, and the entire regulatory apparatus around packaging exists to argue about how much.

Nothing in contact with food is inert

This is the idea that reorganises the whole topic, and it is the opposite of the intuition most people carry.

There is no such thing as a packaging material that transfers nothing. Polymers contain residual monomer, oligomers, catalyst residues, stabilisers, plasticisers and processing aids. Coatings contain crosslinkers and whatever did not fully react during cure. Metals give up ions. Paper gives up fibres and whatever the fibres carried. Glass is the closest thing to inert available and it still exchanges a little at the surface. The transfer of substances from a material into the food it touches has a name — migration — and it is a physical certainty, not a defect.

So food-contact regulation does not ask whether anything migrates. It asks how much, of what, into which food, under what conditions, and compares that against a limit. Two kinds of limit do the work:

• The overall migration limit caps the total mass of everything that transfers, regardless of what it is. In the EU framework it is 60 mg per kg of food, or 10 mg per square decimetre of contact area.
• Specific migration limits apply to individual substances that have their own toxicological assessment, and they are set far lower — often in the microgram range — derived from the dose considered tolerable.

How you measure something you cannot put food in a lab about

Testing with actual food would be unworkable — there are too many foods, and recovering a trace substance from a real food matrix is analytically brutal. So testing uses food simulants: standardised liquids that behave towards the packaging the way a category of food does.

The EU set, which most of the world's testing follows in structure, runs roughly like this. Dilute ethanol stands in for aqueous foods. Three per cent acetic acid stands in for acidic foods. Higher-strength ethanol covers alcoholic and some fatty products. Vegetable oil is the simulant for fatty foods, and a synthetic adsorbent powder covers dry foods.

Two things follow, and they are worth knowing as a consumer rather than just as a specifier.

Fat is the aggressive one. Most of the substances that migrate out of a polymer are more soluble in oil than in water, so a fatty food pulls far more out of a container than the same volume of water would. The same box that is entirely fine for biscuits may not be fine for ghee.

Heat is the multiplier. Migration is diffusion, and diffusion rates climb steeply with temperature. This is why a test is defined as a time and a temperature representing the worst foreseeable use, and why "microwave safe" is a migration statement rather than a claim about melting. A container that passes at room temperature for two hours has been told nothing about ten minutes at 100 °C.

Why a can is lined at all

Three reasons, and the first one connects directly to what a coating does about rust.

Acidic food and metal make an electrochemical cell. Without a barrier the can corrodes from inside, the metal ends up in the food, and eventually the container fails. Second, sulphur-containing foods — peas, sweetcorn, fish — react with exposed iron or tin to form black sulphides, which are harmless and look alarming enough to ruin a product. Third, flavour: metal ions in food taste of metal, and the threshold is low.

The lining is therefore a barrier coating, and the defect logic from any barrier coating applies unchanged: a hole is a hole, and a small exposed anode surrounded by a large cathode corrodes fast. The industry's response is a test with a satisfying directness to it. An enamel rater fills the can with an electrolyte, applies a small voltage between the solution and the can body, and measures the current. A perfect lining conducts essentially nothing. The measured current is a direct read of how much bare metal is exposed through pinholes, cut edges and seam damage — a defect count expressed in milliamps.

The BPA story, told properly

For decades the standard can lining was an epoxy-phenolic coating whose resin is built from bisphenol A and epichlorohydrin. It is genuinely excellent at the job: it adheres to metal, tolerates the severe deformation of can forming, survives sterilisation, and resists acid for years.

Bisphenol A is also endocrine-active, and the assessment of what dose is tolerable has moved a very long way. The European Food Safety Authority's 2023 re-evaluation cut its tolerable daily intake by orders of magnitude relative to the earlier figure, and in December 2024 the EU adopted a regulation banning BPA in food-contact materials, with transition periods by application. In India, the FSSAI packaging regulations prohibit BPA in infant feeding bottles; the packaging rules are amended periodically, so the current consolidated text is the one to check rather than any summary of it.

The replacements are polyester, acrylic and oleoresin linings, and non-BPA epoxies. Several perform well. But there is a caveat here that matters more than the headline, and it is the most useful thing in this article for an ordinary shopper.

"BPA-free" is a statement about one molecule. It says the product does not contain that specific compound. It does not say the substitute has been studied as extensively — and in some applications the early substitutes were close structural analogues such as BPS and BPF, which are themselves under active toxicological scrutiny for the same class of effects. Toxicologists have a name for this pattern, regrettable substitution: replacing a scrutinised chemical with an unscrutinised relative and gaining reassurance rather than safety. A label claiming the absence of one substance is not a safety assessment of what took its place.

A food-contact regulation never asks "does anything transfer". It asks how much, into which food, at what temperature, and whether that number sits under a limit. A supplier who answers the first question has answered the wrong one.

The routes that surprise people

Printing ink on the outside. In 2005 a photoinitiator called ITX, used in ultraviolet-cured carton printing, was found in baby milk in Europe. It had not been added to anything near the food. Printed material is stacked or wound into reels while curing, and ink transfers from the printed face onto the food-contact face of the layer pressed against it — an effect known as set-off — from where it migrates into the contents. The outside of a package is a food-contact issue.

Recycled paperboard. Mineral oil residues from printing inks in the recovered paper stream — reported as MOSH and MOAH fractions — can migrate through board and into dry foods, which is why food-grade recycled board either uses controlled input streams or carries a functional barrier layer.

Recycled plastic. The safety question for recycled content is not the polymer, it is what the previous use put into it. This is why food-grade recycling requires a decontamination process that has been assessed for exactly that, rather than merely clean-looking flake.

Reuse. A container designed for a single fill at ambient temperature, then reused for hot oil or acidic pickle, is being used outside every condition it was tested under. The mithai box and the cold-drink bottle refilled with hot tea are the everyday versions of this.

What to ask a supplier

1. Ask for the declaration of compliance, not the phrase "food grade". "Food grade" is a marketing term with no fixed meaning. A DoC names the regulation, the substances with restrictions, the simulants used, the test conditions and the results.
2. Which simulant and which conditions? Compliance is always conditional. Ask which food type and which time-temperature the material was tested for, and check it covers what you will actually do to it.
3. Material or finished article? A resin can comply while the article made from it does not, because forming, printing, curing and adhesives all change the outcome. The compliance you need is for the object.
4. For coatings: which substrate and what cure schedule? An undercured coating has more unreacted material available to migrate. The cure profile is part of the compliance, not a production detail.
5. Was the testing done by an accredited laboratory? Ask for the report, not a summary of the report.
6. What is out of scope? Temperature ceiling, fatty foods, reuse, microwave, freezer. What a document does not cover is as important as what it does.

To make the point concrete about how a range should be organised: Smart Warrior Coatings, part of Reinste Nano Ventures — which, in the interest of disclosure, belongs to the same group as this publication — keeps food-grade preservative layers as a distinct line from its industrial corrosion coats and its surface shields, rather than presenting one coating as suitable everywhere. That separation is exactly the shape a buyer should expect, because food contact is not a performance tier above general-purpose. It is a different qualification with a documentary answer, and the technical data sheet is where that answer either exists or does not.

Why it matters for students and researchers

The open problem in this field has an unglamorous name and enormous scope: NIAS, non-intentionally added substances.

Every deliberately added substance in a food-contact material has a dossier. But a coating cures, a polymer degrades under heat, and additives react with each other and with the food — and the products of all that are substances nobody put in, nobody listed, and nobody has a specific migration limit for. Oligomers, breakdown products, reaction products, impurities in raw materials. Identifying them requires high-resolution mass spectrometry and a great deal of interpretive work, and after identification comes the harder question of what a toxicological assessment of an unknown trace substance even looks like. It is one of the more honest frontiers in applied chemistry: the field openly states that its inventory of what is in food packaging is incomplete.

Around it sit several other live areas. The toxicology of mixtures at low doses, since real exposure is to many substances at once and limits are set one substance at a time. Bio-based and compostable barrier coatings, where the barrier performance of the traditional materials has been hard to match. Active and intelligent packaging — layers that scavenge oxygen, release preservatives on a schedule, or indicate spoilage — which brings a regulatory puzzle with it, because a packaging material designed to migrate something on purpose does not fit a framework built to limit migration.

And there is a specifically Indian gap worth naming, because it is the kind of thing a research group here can actually close. Migration data is generated largely against European and North American food patterns. Indian usage is often harsher: higher fat, high acidity, long ambient storage in genuinely hot conditions, and reuse of containers as a cultural norm rather than an exception. Migration behaviour under those conditions, for the materials actually used in this market, is thinly studied — and it is measurable with equipment that many Indian institutions already own.

Frequently asked questions

Is it dangerous to microwave food in plastic?

It depends entirely on the plastic, and the label is the answer. Microwave-safe means the container has been migration-tested at the relevant temperature for the relevant food type. Containers not carrying that claim — takeaway tubs, ice-cream boxes, thin films, most single-use packaging — were tested for something much gentler, and heating them, especially with oily food, puts them well outside their assessed conditions. Glass and ceramic sidestep the question.

Is "BPA-free" safe?

It is one specific reassurance and not a general one. It tells you bisphenol A is absent; it does not tell you what replaced it or how well that has been studied. In some product categories early substitutes were close chemical relatives now under scrutiny themselves. The more useful question for a food business is which regulation the article complies with and under what conditions; for a household, the more useful habit is avoiding heat and fat in containers not intended for them, which reduces migration of everything at once.

Are steel or brass utensils safer than coated ones?

Different trade-offs rather than a clear winner. Good stainless steel is close to inert for ordinary cooking, which is a real advantage. But bare reactive metals and acidic food is the pairing that started this whole subject — this is why traditional brass and copper vessels are tinned inside, and why an untinned one holding something acidic overnight is genuinely a bad idea. The general rule holds across materials: acid, fat, heat and time are what drive transfer.

Should I be worried about canned food?

The lining exists precisely to prevent the metal contact people worry about, and canning remains one of the safest preservation methods there is — it is a sterilisation process, and it has an excellent record. The reasonable concerns are narrower: dented or rusted cans, whose lining may be breached at the deformation; storing food in an opened can rather than transferring it, since the cut edge is exposed metal; and, if it matters to you, checking which lining chemistry a brand has moved to.

Can I reuse a plastic bottle?

For cold water, refilled and cleaned, the risk is small and the more realistic problem is microbial rather than chemical. What is genuinely outside the design case is heat and fat: hot tea, hot oil, leaving it in a closed car in summer, or repeatedly washing it in very hot water, which also stresses and micro-cracks the surface. A container is qualified for a set of conditions, and reuse is fine as long as you stay inside them.