The Acetaldehyde Number in PET: Why Water Tastes of Plastic Above 5 ppm

Polyethylene terephthalate (PET) for still-water bottles is controlled not only by intrinsic viscosity and crystallinity but also by an acetaldehyde number, expressed as micrograms of acetaldehyde per gram of polymer (ppm). The phrase “The Acetaldehyde Number in PET: Why Water Tastes of Plastic Above 5 ppm” identifies a sensory boundary located in the polymer phase. At a wall concentration of 5 ppm, a 500 mL bottle with 20 g of PET carries 100 µg of acetaldehyde. If complete equilibration into the contained water occurred, the resulting water concentration would be 200 µg/L. Published sensory thresholds for acetaldehyde in water range from 20 µg/L to 40 µg/L, so migration from a 5 ppm polymer reservoir can exceed the taste threshold while remaining far below the EU specific migration limit for acetaldehyde of 6 mg/kg under Regulation (EU) No 10/2011.

Commercial water-grade PET is produced by continuous esterification and polycondensation followed by solid-state polymerisation (SSP). After SSP, pellet acetaldehyde is typically below 1 ppm when measured according to ASTM F2013-10(2023). That pellet value does not predict the acetaldehyde number in a moulded preform because melt processing regenerates acetaldehyde from chain-end chemistry. Preform specifications therefore control still-water applications, commonly at ≤5 ppm, with sensitive low-mineral waters often targeted at ≤3 ppm.

What Manufacturing Steps Regenerate Acetaldehyde in Bottle-Grade PET?

Acetaldehyde in PET arises from intramolecular back-biting at 2-hydroxyethyl ester end groups, decomposition of vinyl ester end groups, and oxidative cleavage of ethylene glycol units. The principal thermal route is the cyclic elimination of acetaldehyde from the hydroxyethyl ester end group, leaving a carboxyl chain end. During polycondensation at 270–285 °C under vacuum, acetaldehyde is continuously removed with ethylene glycol and water. In SSP, pellet acetaldehyde is stripped at 200–220 °C in an inert-gas or vacuum environment. This produces a resin with low pellet AA but does not eliminate the chain-end precursors that regenerate acetaldehyde during subsequent melting.

Injection moulding of preforms re-exposes the polymer to temperatures of 270–295 °C, depending on barrel settings and hot-runner design. The regenerated acetaldehyde concentration is governed by melt temperature, residence time, shear heating, oxygen ingress, and residual moisture. Moisture above 50 ppm before extrusion hydrolyses ester linkages, increasing hydroxyl and carboxyl chain ends and lowering intrinsic viscosity; the newly created hydroxyethyl ester ends can then back-bite to acetaldehyde. Oxygen in the melt accelerates oxidative degradation, particularly at surfaces and at hot-runner gates. Desiccant drying with a dew point of −40 °C and a residence time of 4–6 h is a practical safeguard, but drying cannot correct acetaldehyde generated by excessive melt temperature or extended hold time.

Bottle-grade PET often contains isophthalic acid at 1.5–2.5 wt% to suppress crystallisation, and diethylene glycol at 0.8–1.4 wt% as a byproduct. These comonomers alter melt stability; diethylene glycol units can contribute to acetaldehyde release under high-temperature processing. Antimony trioxide catalyst residues of 150–250 ppm Sb are common, although titanium-based catalysts are used in some grades. Higher catalyst residues can accelerate thermal degradation, but acetaldehyde generation is more strongly influenced by temperature and residence time.

Quantification of the acetaldehyde number in pellet and preform samples is usually performed by cryogrinding under liquid nitrogen followed by headspace gas chromatography with flame ionisation detection, as described in ASTM F2013-10(2023). The method reports residual acetaldehyde in µg/g (ppm). It does not measure the acetaldehyde-generation potential of the resin under melt processing; resin suppliers commonly apply a separate thermal challenge test at a defined melt temperature and residence time, although published data for this specific configuration is limited because test protocols are proprietary. Water-phase acetaldehyde after bottle contact is measured by headspace or purge-and-trap gas chromatography before sensory panel evaluation. Regulatory compliance with EU No 10/2011 does not guarantee sensory acceptability because the SML for acetaldehyde is 6 mg/kg while sensory thresholds are 20–40 µg/L.

Melt Temperature, Residence Time, and the 5 ppm Preform Limit

Production-scale preform injection uses reciprocating screws with L/D ratios of 20:1–24:1 and barrel temperature profiles between 270 °C and 285 °C for bottle-grade PET of 0.80 dL/g intrinsic viscosity. Hot-runner temperatures are often maintained at 275–285 °C; higher settings can generate acetaldehyde at the gate region before the preform is injected. Melt residence times from plastication to injection are ordinarily controlled between 60 s and 120 s. When valve-gate hot runners or unbalanced cavity filling extend residence time beyond 120 s, preform acetaldehyde rises.

On a 72-cavity preform tool with a clamp force of 450 t, infrared gate temperature checks frequently show a 5–10 °C rise above hot-runner set point at the valve gate during high-speed injection. In production audits, cavity-to-cavity acetaldehyde variation above 1 ppm has been observed on multi-cavity tools when hot-runner manifold temperature offsets exceed ±5 °C. The same audits show that reducing hot-runner set point from 285 °C to 275 °C often lowers preform AA by 1–2 ppm, but the magnitude depends on gate design, comonomer content, and catalyst residues. If resin is exposed to plant air above 60% RH for more than 30 min, pre-drying must be extended to avoid moisture-driven IV loss and subsequent acetaldehyde regeneration.

Acetaldehyde migration into water is controlled by polymer-phase concentration, diffusion coefficient, wall thickness, contact area, fill volume, storage temperature, and headspace. For a 500 mL bottle with a 20 g PET wall at 5 ppm AA, the total polymer reservoir is 100 µg. Complete partitioning into water would produce 200 µg/L, a concentration five to ten times higher than sensory thresholds of 20–40 µg/L. In real packages, only a fraction migrates over commercial shelf life. However, acetaldehyde located within the first 50–100 µm of the product-contact surface can diffuse to the water interface during the first days of storage. This is why preform AA measured on the whole wall is not sufficient; surface-near distribution and gate-region hot spots matter. Storage at 40 °C raises the diffusion rate and can produce sensory failure even for preforms that pass at ambient temperature.

When the 5 ppm Barrier Is Breached in the Bottle Wall

When a preform exceeds 5 ppm acetaldehyde, the contained water may taste of plastic even though acetaldehyde is a naturally occurring fruit aroma at lower concentrations. The sensory response in water is not governed solely by acetaldehyde concentration; other aldehydes and degradation products contribute. At a wall concentration above 5 ppm, the probable headspace concentration inside a closed bottle after 24 h at 25 °C can exceed 20 µg/L, depending on bottle volume and headspace ratio. The exact amount varies with thickness and crystallinity; highly oriented sidewalls have lower diffusivity than amorphous gate or shoulder regions, so migration from the contact surface is not uniform.

The 5 ppm boundary should not be used as a food-contact regulatory limit. A package can comply with EU No 10/2011 SML of 6 mg/kg and still exceed sensory threshold by two orders of magnitude. For this reason, sensory evaluation remains a separate specification for water brands, particularly for low-mineral and sparkling waters where acetaldehyde notes are less masked by sweetness or flavour.

The 5 ppm Value Is a Polymer Specification, Not a Water Concentration

The phrase “above 5 ppm” describes the acetaldehyde number in the PET bottle wall, not the concentration in the water. A water concentration of 5 ppm acetaldehyde would be 125–250 times above published taste thresholds and is not expected from ordinary PET migration under ambient storage. The polymer-phase specification exists because it can be measured directly on the preform and because it correlates with the migration reservoir. The correlation is not linear; preform geometry, state of orientation, blow-moulding heat history, and distribution of AA across the wall influence the water-phase concentration. Consequently, a preform at 4.5 ppm can still generate a perceptible off-taste if the gate area contains high local AA, while a well-balanced 5.5 ppm preform may pass in a flavoured beverage but not in low-mineral still water.

Resin producers, preform moulders, and bottlers therefore use layered specifications: pellet AA after SSP below 1 ppm, preform AA after moulding at or below 5 ppm for sensitive still water, and migration or sensory testing for final package validation. The acetaldehyde number in PET is a process-control number rather than a direct regulatory limit. It protects the sensory quality of water because acetaldehyde has one of the lowest odour thresholds among PET degradation products. Published data for the exact residual-to-migration partition coefficient across commercial bottle geometries is limited, so conservative melt-processing windows and preform testing remain the operational basis for avoiding plastic off-taste above 5 ppm.