Polyethylene terephthalate (PET) is a linear aromatic polyester obtained by esterification of purified terephthalic acid with ethylene glycol followed by polycondensation. Bottle-grade resins are specified by intrinsic viscosity values between 0.76 dL/g and 0.84 dL/g when measured according to ISO 1628-2 or ASTM D4603. The polymer exhibits a glass transition near 78 °C, a crystalline melting peak between 245 °C and 255 °C for bottle formulations, and solid-state density in the range 1.34–1.40 g/cm³. Virgin PET intended for food contact is evaluated under 21 CFR 177.1630 in the United States and under Regulation (EU) 10/2011 in the European Union. The central compliance question for food-grade recycled PET approvals—FDA NOL, EFSA and decontamination—is whether a recycling process can remove or sufficiently dilute post-consumer contaminants from the polymer matrix to levels below regulatory thresholds.

How Does Decontamination Efficiency Determine Regulatory Acceptance?

Recycled PET for food contact is not approved on polymer identity alone. Both the FDA NOL pathway and the EFSA opinion process require a surrogate contaminant challenge test that measures the ability of the recycling process to reduce model contaminants deliberately introduced into washed post-consumer flakes. The contaminant set represents volatile polar, volatile non-polar, non-volatile polar, and non-volatile non-polar chemistries. The FDA threshold-of-regulation approach evaluates the resulting residual concentration so that estimated daily intake remains below 1.5 µg/person/day for non-carcinogenic contaminants, corresponding to 0.5 µg/kg food for an assumed 3 kg/day intake under 21 CFR 170.39. EFSA’s evaluation criterion for non-authorised contaminants is expressed as a maximum migration of 0.1 µg/kg food, based on conservative exposure modelling. Challenge test results are converted into a process decontamination efficiency, which must be high enough to ensure that worst-case residual contamination does not exceed the applicable threshold when the recycled material is used at the intended addition level.

A super-clean recycling line achieves decontamination through sequential unit operations: hot caustic washing, surface drying, extrusion with vacuum degassing, melt filtration, and solid-state polycondensation. A typical twin-screw extruder with L/D 40–48 and vacuum venting below 20–50 mbar removes volatile contaminants from the melt phase. Downstream solid-state polycondensation reactors operated at 210–230 °C under 0.5–2 mbar for 8–16 h further reduce residual volatile compounds and increase intrinsic viscosity by post-polycondensation. Diffusion-limited removal of high-molecular-weight non-polar contaminants is the rate-limiting step. Published data for specific diffusion coefficients of tetracosane-class surrogates in PET across solid-state polycondensation conditions are limited, but industrial challenge tests generally show that volatile surrogates are reduced to below analytical detection limits while non-volatile alkanes require the full solid-state polycondensation residence time to meet regulatory residual criteria.

A single-screw or twin-screw melt-phase decontamination stage alone is rarely sufficient for direct food-contact approval. Extrusion vacuum degassing at 20–50 mbar with residence time of 60–180 s removes most residual acetaldehyde and low-boiling contaminants, but the short melt residence time and high viscosity of PET at 280–295 °C limit mass transfer of non-volatile contaminants. Melt filtration through 20–40 µm screen packs or depth media addresses particulate contamination, not molecular diffusion. Consequently, approval dossiers for food-grade recycled PET typically rely on the solid-state polycondensation stage to deliver the majority of the demonstrated decontamination efficiency. Batch solid-state units operate with slow agitation under high vacuum, while continuous designs use countercurrent inert gas flow; both must prevent particle agglomeration and maintain consistent crystallinity above 35% before entering the reactor. Feedstock with residual moisture above 50 ppm can hydrolyse the polymer during solid-state processing, reducing intrinsic viscosity and increasing free monomer content.

Solid-State Polycondensation Kinetics Govern the Decontamination Window

In a solid-state polycondensation reactor, decontamination proceeds by Fickian diffusion of contaminant molecules from amorphous regions of the semicrystalline PET matrix to the pellet surface, followed by evaporation into the vacuum or gas stream. The diffusion coefficient of a contaminant in PET depends on molecular size and temperature, with published apparent activation energies for small-molecule diffusion commonly in the range 80–120 kJ/mol. For that reason, increasing solid-state polycondensation temperature by 10 °C within the permissible window can reduce the required residence time for volatile contaminants, but it also accelerates thermal degradation reactions such as acetaldehyde formation and vinyl end-group generation. Industrial reactors are therefore operated at 210–230 °C with vacuum below 2 mbar or inert gas flow sufficient to maintain a hydrocarbon dew point below -40 °C. The reaction mass must be pre-crystallised to at least 35% crystallinity to prevent sticking or fluidisation. The combined effect of diffusion and surface evaporation produces decontamination factors that are time-dependent and sensitive to pellet geometry; small pellets with a diameter of 2.0–3.5 mm show faster contaminant removal than thick sheet regrind.

Regulatory submissions for food-grade recycled PET must be supported by migration testing or modelling conducted under appropriate food simulant conditions. Under Regulation (EU) 10/2011, aqueous and low-alcohol simulants are assigned for PET bottles: 10% ethanol for aqueous foods, 20% ethanol for foods containing up to 20% alcohol, and 50% ethanol for higher-alcohol products. Testing at 40 °C for 10 days is commonly used as the worst case for long-term ambient storage, while 70 °C for 2 h represents hot-fill and pasteurisation conditions. Migration data are compared with specific migration limits for authorised substances and the 0.1 µg/kg criterion for non-authorised contaminants. If challenge test modelling already demonstrates residual contamination below the threshold under a 100% migration assumption, additional migration testing may be reduced, but the dossier must still present a complete mass-transfer calculation.

FDA NOL and EFSA Evaluation Pathways for Post-Consumer rPET

The FDA no objection letter submission is a premarket process in which the recycling company files chemistry data, process description, surrogate challenge test results, and migration calculations. If the agency issues a no objection letter, the letter identifies the specific process and may specify conditions such as source control, maximum recycled content, and intended food-contact applications. The EFSA pathway under Regulation (EC) No 282/2008 and Regulation (EU) 2022/1616 follows a scientific opinion format: the European Commission requests EFSA to evaluate the safety of a defined recycling process, EFSA reviews the challenge test and modelling, and the Commission may authorise the process for the EU market. The two frameworks differ in exposure assumptions and procedural status but share the same underlying decontamination assessment logic.

Feedstock source control is the first quantifiable decontamination barrier. A deposit-refund collection stream that segregates PET bottles from co-mingled municipal waste reduces the burden of migrating contaminants relative to mixed plastic bales. Near-infrared sorting systems with air-jet ejection remove PVC, polyolefins, and aluminium to a residual PVC concentration typically specified below 50 ppm in food-grade recycled PET feedstock. Residual PVC above 100 ppm can generate hydrochloric acid during extrusion, causing acid-catalysed chain scission and off-spec intrinsic viscosity. Multi-layer bottles, barrier-coated containers, and opaque pigmented packaging present additional sorting and melt-processing constraints. Black PET trays containing carbon black cannot be reliably identified by conventional near-infrared sorting and are typically excluded from food-grade recycled PET feedstock. These limitations are operational boundaries rather than regulatory exemptions.

When Solid-State Polycondensation Alone Fails to Meet Challenge Test Limits

If the challenge test shows that a non-volatile non-polar surrogate exceeds the EFSA 0.1 µg/kg migration criterion at the requested use level, the process must either extend solid-state polycondensation residence time, reduce feedstock contamination, or lower recycled content. Extended solid-state polycondensation beyond 16 h at 230 °C may increase thermal degradation by-products and acetaldehyde reformation, creating a process window bounded by both decontamination kinetics and polymer quality retention. Vacuum rotary dryers with jacket temperatures of 190–210 °C and pressures below 5 mbar can serve as pre-drying and surface devolatilisation steps, but they do not provide the same solid-state molecular weight build-up as solid-state polycondensation. Melt-phase vacuum extrusion with two-stage venting and high-surface-area devolatilisation screws lowers residual benzene, toluene, and chloroform-class contaminants before solid-state processing.

Batch-to-batch variance in feedstock contamination is a recurring production-scale issue. A twin-screw extrusion line with L/D 48 and multiple vent ports may show stable decontamination under a uniform surrogate challenge test, while real post-consumer flakes vary in contaminant concentration by a factor of up to 10 between collection regions. For that reason, approval dossiers include a safety margin between measured decontamination efficiency and the regulatory threshold. The FDA NOL submission typically reports challenge test data using intentionally contaminated flakes at concentrations far above expected post-consumer levels, often 200–1,000 mg/kg for each surrogate, to demonstrate robust removal capacity. EFSA opinions similarly require challenge test concentrations that reflect a worst-case misuse scenario. The resulting data are reviewed with the conservative assumption that all residual contaminant migrates into the food simulant; therefore, the demonstrated decontamination factor must be high enough to compensate for this conservative exposure model.

Food-grade recycled PET should not be blended with amorphous glycol-modified PET or polyolefin cap fractions unless the downstream process is specifically designed for such blends. Glycol-modified PET contamination reduces crystallinity and can cause agglomeration during solid-state polycondensation, while polyolefin fractions degrade and form black specks at PET processing temperatures. Chlorinated paraffins, silicone-based release agents, and amine-based additives in non-food PET articles can survive washing and generate odour-active by-products during extrusion. Production lines processing recycled PET at high recycled content require strict pre-drying at 160–180 °C for 4–6 h to maintain moisture below 50 ppm; exceeding 190 °C during drying risks thermal oxidation and yellowing. These operational boundaries are evaluated case by case in food-contact compliance submissions because they directly affect decontamination performance and final article migration.