PET Migration Testing: Overall and Specific Migration Limits

Polyethylene terephthalate (PET) is a thermoplastic polyester produced by esterification and polycondensation of ethylene glycol with purified terephthalic acid, followed in most bottle-grade routes by solid-state polymerisation to raise intrinsic viscosity to 0.80–0.85 dL/g. In monolayer food-contact applications such as carbonated soft-drink bottles, still water bottles, trays, and hot-fill containers, regulatory assessment is not limited to mechanical or barrier performance. The finished article must be evaluated for migration. Migration testing separates into overall migration and specific migration. Overall migration quantifies total non-volatile residues; specific migration quantifies identifiable substances that carry specific migration limits. The dominant European Union instrument is Regulation (EU) No 10/2011, as amended, while the United States maintains food-contact status through 21 CFR 177.1630. Compliance therefore requires testing the finished article or representative specimens under standardised time-temperature conditions with assigned food simulants.

The repeating unit of PET contains an aromatic ring and a short aliphatic diol. The polymer has a glass transition temperature near 70–80°C and a melting temperature near 240–260°C. These properties matter in migration testing because test temperature and polymer crystallinity control permeant diffusivity. Conversion from resin pellet to finished article also changes residual monomer concentration, molecular orientation, and thermal history; for this reason migration testing is performed on the finished article or on coupons cut from it, not on unprocessed granulate.

How Are Overall Migration Results Expressed and Enforced?

Article 12 of Regulation (EU) No 10/2011 fixes the overall migration limit at 10 mg/dm² of food-contact surface area. An equivalent limit of 60 mg/kg is permitted for food-contact articles intended for infants and young children, and for containers with a capacity between 500 mL and 10 L, where the surface-to-volume relationship cannot readily be applied. The conventional conversion assumes 6 dm² of contact surface per 1 kg of food. The area-based result is therefore multiplied by 6 dm²/kg to obtain the mass-based equivalent for the appropriate article categories.

Overall migration is not a sum of individually identified substances. It is a gravimetric determination of the non-volatile residue left after evaporation of the simulant. The EN 1186 series establishes the test programme. EN 1186-1:2002 guides selection of exposure conditions and simulants; EN 1186-3:2002 specifies total immersion in aqueous food simulants; EN 1186-2:2002 covers total immersion in olive oil. For fatty-food simulations where oil sorption into PET can distort measurements, isooctane or 95% ethanol substitutes are used under published substitute protocols. The residue is dried at 105°C to constant mass and weighed on an analytical balance with a limit of determination below 1 mg/dm².

Overall migration data for PET are typically low because the polymer is semicrystalline and biaxially oriented. These structural features restrict diffusion of higher-molecular-weight species. The residues that do migrate are often cyclic and linear oligomers, particularly the cyclic trimer of ethylene terephthalate. The cyclic trimer itself has no separate harmonised specific migration limit in Regulation (EU) No 10/2011; its contribution is assessed through overall migration. In routine third-party laboratory reports, PET bottles exposed to 10% ethanol or 3% acetic acid at 40°C for 10 days commonly produce overall migration values below 1 mg/dm² when resin processing is stable and surface contamination is absent. Overall migration failures are more often associated with masterbatch residues, printing-ink transfer, excessive slip additives, or contamination from recycled material than with degradation of the base PET polymer itself.

Specific migration limits apply to monomers, degradation products and catalyst residues

Specific migration testing under EN 13130-1:2004 quantifies individual substances in food simulants after the same exposure. For PET, the substances of primary regulatory interest are residual monomers and process-derived by-products. Ethylene glycol is controlled at 30 mg/kg; terephthalic acid at 7.5 mg/kg; isophthalic acid, when used as a comonomer, at 5 mg/kg; acetaldehyde, a thermal decomposition product, at 6 mg/kg; and antimony, a residual from antimony trioxide catalyst, at 0.04 mg/kg. These values appear in Annex I Table 1 of Regulation (EU) No 10/2011, and they apply as migration into food or food simulant, not as residual content in the polymer.

The selection of food simulants for specific migration is tied to the intended food type. Aqueous acidic foods use 3% acetic acid; beverages with alcohol up to 10% use 10% ethanol; fatty foods use vegetable oil or a permitted substitute. For a PET water bottle tested for long-term storage at room temperature, the accelerated exposure of 40°C for 10 days in 10% ethanol and 3% acetic acid is used. For hot-fill tea or sports drinks, the test temperature is raised to 70°C for 2 h. The analytical finish for terephthalic acid and isophthalic acid is reverse-phase high-performance liquid chromatography with UV detection; ethylene glycol is typically determined by gas chromatography with flame ionisation detection after derivatisation, or by liquid chromatography with refractive index detection. Antimony requires inductively coupled plasma mass spectrometry because the specific migration limit is close to instrumental background in some simulants.

Migration kinetics in polymer matrices for PET follow Fickian diffusion when the migrant is soluble in the simulant and the polymer is below its glass transition temperature. The diffusion coefficient for ethylene glycol in oriented PET is low; the test is therefore designed to provide a conservative overestimate by using accelerated time-temperature conditions. Because PET is semicrystalline, crystallinity and orientation further retard migration. However, residual monomers located in amorphous regions remain mobile. Solid-state polymerisation and high stretch ratios reduce but do not eliminate their migration.

When recycled PET is introduced as a monolayer or multilayer constituent, migration testing must also include validation of the recycling decontamination process. In the European Union, recycled plastics must meet the same overall and specific migration requirements. Contaminants from the feedstock can survive melt-phase recycling if the process lacks adequate decontamination parameters; migration testing of recycled PET articles therefore often includes gas chromatography–mass spectrometry screening for non-intentionally added substances in addition to the standard overall and specific migration suite. This supplementary analysis is not a replacement for positive-list verification.

When hot-fill and retort conditions expose PET to thermal stress

Hot-fill and retort applications create a simultaneous thermal and diffusion challenge. A hot-fill PET bottle must be heat-set during stretch blow moulding to maintain dimensional stability at fill temperatures between 82°C and 95°C. The heat-set step increases crystallinity, reduces free volume, and slows migration. In production-scale reheat stretch blow moulding, preforms are heated by infrared ovens to surface temperatures around 105–115°C, blown in a mould at 120–150°C, and held under pressure to induce spherulitic crystallinity. A preform injection process uses reciprocating screw machines with an L/D ratio near 24:1, barrel temperatures between 270°C and 285°C, and hot-runner temperatures up to 295°C. At these temperatures hydrolytic and thermal chain scission compete with crystallisation: moisture above 50 ppm in the dried granulate hydrolyses ester linkages, increasing free ethylene glycol and lowering intrinsic viscosity. This is a compliance-relevant processing boundary, not merely a mechanical defect, because the released ethylene glycol can appear in specific migration tests.

Acetaldehyde formed above 280°C is volatile and can diffuse to the preform surface. Acetaldehyde migration is a known organoleptic failure mode in PET water bottles; sensory detection in still water is low and is often below the regulatory specific migration limit, but it is measured alongside ethylene glycol. Antimony migration is influenced by acidic simulants: 3% acetic acid typically extracts more antimony than 10% ethanol. Production-scale experience shows that antimony migration from antimony-catalysed PET remains below 0.04 mg/kg under normal bottle processing, but the margin can narrow when recycled PET with multiple heat histories is used or when hot-fill exposure is combined with long acid-contact shelf life. Published data for specific antimony migration from organometallic alternative catalysts in niche retortable monofilms is limited.

Resin drying and process-gas monitoring are therefore deterministic for migration compliance. Dehumidifying dryers for PET must deliver air at a dew point below −40°C, and the dried granulate must feed the injection machine with residual moisture below 50 ppm. Incompatibility with amine-based processing aids and polyamide-based barrier layers should be treated as formulation-specific migration risk: reactive interactions can alter acetaldehyde scavenging or delaminate barrier layers, creating locally elevated migration. The use of ethylene glycol or terephthalic acid as migration markers in quality control is valid only when the polymer is fully dried; wet granulate can generate both compounds during melt processing, creating false positives relative to the finished article. Overall and specific migration results are therefore interpreted not as intrinsic resin constants but as process-dependent values for the finished food-contact article.