PFAS in Food Packaging and How Testing Caught Up

For most of the period in which PFAS were used in food packaging, the analytical question was framed as a list. Which of these specific compounds is present, and at what concentration? Laboratories built methods for a few dozen named substances, regulators wrote limits for a handful, and testing programmes reported results against that list.

The list was never the problem it was meant to solve. The family of per- and polyfluoroalkyl substances contains thousands of distinct chemicals, most of which have never had an analytical standard made for them. A method that looks for forty compounds is blind to the rest by construction, and manufacturers responding to restrictions on named substances could substitute unnamed ones without ever appearing on a test report.

The shift that changed this was conceptual rather than technological. Instead of asking which fluorinated compounds are present, laboratories began asking how much fluorine of any organic kind is present at all. That single change turned an unwinnable identification problem into a tractable screening one, and it is the reason PFAS testing of packaging has become useful in the past several years.

What PFAS Compounds Are Used For

The defining feature of this chemical family is a carbon chain in which hydrogen atoms have been replaced by fluorine. The carbon to fluorine bond is among the strongest in organic chemistry, and a chain of them is chemically and thermally very stable, which is what makes these substances persist in the environment and gives the family its popular nickname.

That same structure produces an unusual surface property. A fluorinated chain has very low surface energy, meaning that both water and oil struggle to wet it. Most water-repellent treatments are not oil-repellent, and most oil-repellent treatments are not water-repellent. Fluorinated chemistry does both at once, cheaply, in a thin coating that does not change how the underlying material feels.

For food packaging this was close to ideal. Moulded fibre bowls, greaseproof paper wrappers, microwave popcorn bags, pizza box liners, pastry bags and fast food containers all need to hold hot, wet, greasy food without going soggy or leaking. Paper and moulded fibre are otherwise poor at this. A fluorinated treatment gave a compostable-looking fibre product the performance of a plastic-lined one.

The applications extend beyond packaging into non-stick cookware coatings, textile treatments, firefighting foams and industrial processing aids, and the environmental burden comes from all of these together. Packaging matters disproportionately for dietary exposure, though, because it is in direct contact with food, often hot food, which is the condition under which transfer is most likely.

Why Individual Compound Testing Fell Short

Targeted analysis by liquid chromatography with tandem mass spectrometry is a mature, sensitive and specific technique. It identifies a compound by its chromatographic retention and its characteristic fragmentation, quantifies it against a calibration standard, and can reach very low detection limits. Where the target is known, it is the right tool.

The constraint is structural. A targeted method can only measure compounds for which a pure analytical standard exists, because that standard defines the retention time, the fragmentation pattern and the calibration. Standards exist for the substances that have been studied, which means the ones that were commercially dominant and became regulatory priorities. For the great majority of the family, no standard has ever been synthesised.

Approach	What it measures	Strengths	Blind spots
Targeted LC-MS	Named compounds against standards	Specific, sensitive, quantitative, regulator-friendly	Anything without a standard, including deliberate substitutes
Total organic fluorine	Fluorine in organic form, as an element	Catches unknown compounds, fast screening	No identity, includes non-PFAS organofluorine
Extractable organic fluorine	Organic fluorine that solvent removes	Closer to what could migrate	Depends on extraction conditions chosen
Migration testing	What transfers into food or a simulant	Directly relevant to exposure	Simulant may not match real food behaviour
Screening for total fluorine by other means	All fluorine including inorganic	Very fast, inexpensive	Inorganic fluoride confounds the result badly

Two failure modes followed from relying on the targeted approach alone. The first was regrettable substitution: as specific long-chain compounds were restricted, manufacturers moved to related substances outside the analytical panel, and test reports came back clean while the treated packaging performed exactly as before. The second was the precursor problem. Many treatments are applied as larger polymeric or side-chain fluorinated materials that are not themselves on any target list but which degrade over time in the environment or in a landfill into the small compounds that are.

Both problems share a shape. The analysis was measuring a proxy, the presence of specific well-studied molecules, for the thing anyone actually cared about, which was whether fluorinated chemistry had been applied to this packaging at all.

Total Organic Fluorine as a Screening Proxy

The alternative measures the element. Fluorine in an organic compound is fluorine regardless of what molecule it sits in, so a method that liberates and quantifies it detects any fluorinated substance, named or not.

The dominant technique combines combustion with ion chromatography. A sample is burned completely at high temperature in an oxygen-rich atmosphere, which breaks every carbon to fluorine bond and converts the fluorine to hydrogen fluoride. The combustion gases are trapped in an absorbing solution, and that solution is injected onto an ion chromatograph, where fluoride is separated from other anions and quantified by conductivity detection.

The result is a total fluorine figure. To make it meaningful, the inorganic contribution has to be handled, because naturally occurring inorganic fluoride is present in many materials and would otherwise dominate. The usual approach is to remove or measure inorganic fluoride separately, often by extracting the sample or by washing it, so that the reported number refers to organic fluorine.

A common variant reports extractable organic fluorine instead, in which the sample is extracted with a solvent and only the extract is combusted. This is more directly relevant to whether anything could leave the packaging, and less relevant to whether the packaging was treated at all. The two figures answer different questions and both have their place.

The trade the method makes is explicit and worth stating plainly. Total organic fluorine gives no identity. A high result says fluorinated chemistry is present in quantity; it does not say which compounds, how hazardous they are, or whether they would migrate. It is a screening tool, and the sensible workflow uses it to decide which samples deserve the expense of targeted analysis. Screening broadly and confirming selectively is a far better use of laboratory capacity than running an incomplete targeted panel on everything.

There is a specificity caveat too. Non-PFAS organofluorine compounds exist, including some pharmaceuticals and agrochemicals, and while they are rarely relevant to packaging they are not impossible, so an unexpected result warrants confirmation rather than assumption.

Migration Testing From Packaging to Food

Presence in packaging and transfer into food are different questions, and the second is what determines exposure.

Migration testing places the packaging material in contact with either the actual food or a food simulant under defined conditions of temperature and time, then measures what has moved into the food side. Simulants are standardised liquids chosen to represent categories of food behaviour: an aqueous simulant for watery foods, an acidic one for acidic foods, an alcohol solution for fatty foods, and a dry substitute material for dry foods.

Conditions are chosen to represent the worst reasonably foreseeable use. For a container intended to hold hot food, the test runs hot. For something intended for a microwave, the test includes microwave heating. The idea is not to replicate an average meal but to bound what could happen, so that a pass provides meaningful assurance.

Several factors drive how much moves across. Temperature is the largest, since heat increases the mobility of small molecules and can degrade the coating. Contact time matters, particularly for foods that sit in packaging for extended periods. Fat content matters a great deal, because the substances of interest partition into fat more readily than into water, so a greasy hot food is a far more aggressive test than a cold aqueous one. Physical damage to the coating, including creasing and abrasion during manufacture and transport, opens paths that an intact surface does not have.

The awkward part is that migration testing with simulants can understate what happens with real food, because real foods contain surfactants, emulsifiers and fats in combinations that no simulant reproduces exactly. Where a result sits close to a limit, testing against the actual food is more informative than testing against a simulant that was chosen for standardisation rather than realism.

Detection Limits and Background Contamination

Fluorinated materials are so widespread in laboratory equipment that measuring low levels of fluorine requires eliminating the laboratory itself as a source, and this is a genuine and persistent difficulty.

The offenders are familiar. Fluoropolymer tubing, seals, O-rings, thread tape, container liners and coated stir bars are all standard laboratory items, and all are made from the material being measured. Solvents can carry trace contamination from their own containers or from manufacturing. Even the water supply and the air handling can contribute in a laboratory that also runs other fluorine work.

Controlling this means substituting materials throughout the analytical path, using polypropylene or glass in place of fluoropolymer wherever contact occurs, dedicating glassware, and running procedural blanks constantly rather than occasionally. The blank defines what the method can honestly report, and a laboratory whose blanks are variable cannot claim a low detection limit no matter what the instrument is capable of.

This has a practical consequence for interpreting results. A result reported as below the limit of detection means below what that laboratory could distinguish from its own background on that day, which may be considerably higher than what the technique can achieve in ideal conditions. Comparing results between laboratories requires comparing their blanks and detection limits, not just their numbers, and a report that does not state its detection limit is difficult to use.

Sample preparation adds its own risk. Cutting packaging with contaminated scissors, storing samples in coated containers, or homogenising with equipment that has fluoropolymer parts can all introduce the analyte after collection. In studies that survey products on the market, preparation contamination is a more likely explanation for an isolated odd result than a genuinely unusual product.

Regulatory Restrictions Across Regions

Regulatory approaches have diverged, and the divergence is itself informative about how difficult the substance class is to regulate.

Several jurisdictions have moved to restrict the whole class in food contact materials rather than named compounds, on the logic that class-wide restriction is the only approach that substitution cannot defeat. Some national and subnational governments have enacted bans on intentionally added PFAS in food packaging, typically defined by the presence of any fluorinated substance above a threshold expressed as total organic fluorine. Defining the restriction in terms of the screening measurement rather than a compound list is the regulatory expression of the analytical shift described above.

In the European Union, food contact materials are governed by a framework requiring that materials not transfer constituents to food in quantities that endanger health, alongside specific authorisation lists for plastics. Individual PFAS have been restricted through chemicals legislation, and a broader class-wide restriction proposal has been under consideration, which would represent one of the largest chemical restrictions attempted.

In the United States, the Food and Drug Administration has authority over food contact substances, and authorisations for certain grease-proofing agents have been withdrawn through voluntary market phase-out agreements with manufacturers. Several states have legislated independently and moved faster than the federal position, which is why packaging requirements differ across the country.

Elsewhere, regulatory activity has generally followed the same direction at different speeds, with environmental restrictions on manufacture and use often arriving before food contact rules specifically. The World Health Organization and national food safety agencies continue to assess dietary exposure, and the assessments consistently identify packaging as one contributor among several rather than as the dominant route for most people.

Reading Product Claims About PFAS Free

Consumer-facing claims in this area range from carefully substantiated to meaningless, and telling them apart is possible with a few questions.

Ask what was measured. A claim backed by total organic fluorine testing is substantially stronger than one backed by a targeted panel, because the targeted panel can only exclude what it looked for. A claim that specifies the analytical approach and the detection limit is a claim someone has thought about.

Ask what the claim covers. Packaging is assembled from components, and a claim about the moulded fibre body may say nothing about a lid, a lining, a window film or an adhesive. Claims scoped to a whole product are more useful than claims scoped to one material.

Ask whether the claim addresses intentional addition or presence. Intentionally added is a meaningful commitment about formulation, but recycled fibre can carry fluorinated residues from previous life cycles without anyone adding anything, and manufacturing equipment can transfer traces. A presence-based threshold expressed in total organic fluorine terms is harder to satisfy and correspondingly more informative.

For anyone weighing all this at a practical level, the reasonable summary is that packaging is one exposure route among several, that the direction of regulation is towards class-wide restriction, and that testing has improved enough that manufacturers can no longer demonstrate compliance by looking only where they know nothing will be found. Reducing time spent with hot, greasy food in coated fibre containers is a small, cheap adjustment; reorganising a diet around packaging is not proportionate to what is currently understood.

The useful shift in this whole area is a change in question. Asking which named chemicals a package contains was always going to produce reassuring answers, because the list was short and the chemistry was long. Asking how much organic fluorine is present at all is harder to evade, cheap enough to run at scale, and increasingly what regulators and buyers are writing into their requirements.