Polyethylene terephthalate is manufactured through the direct esterification of purified terephthalic acid with monoethylene glycol at 240–260 °C, followed by vacuum polycondensation under an absolute pressure below 2 mbar. The melt-phase output is an amorphous pellet with intrinsic viscosity limited by melt viscosity; for bottle applications, this pellet is crystallized and solid-state polymerized to a higher molecular weight resin. In procurement documentation, melt-phase material is frequently listed as PET polymer, while the solid-state advanced product is specified as PET resin. The distinction is not semantic because the two stages differ in molecular weight, pellet morphology, carboxylic end group concentration, and acetaldehyde content. A converter buying the melt-phase stage accepts further polycondensation duty in a solid-state reactor; a converter buying the resin stage pays for that duty already performed.

Intrinsic viscosity is measured per ASTM D4603 by dissolving dried PET in a 60/40 phenol/1,1,2,2-tetrachloroethane solvent mixture at 30 °C and using a glass capillary viscometer. For international comparison, ISO 1628-5 provides a dilute-solution viscosity number. The measured intrinsic viscosity is not a linear function of molecular weight; small differences above 0.80 dL/g correspond to larger differences in melt relaxation time and preform injection pressure. The value therefore functions as a proxy for number-average molecular weight and must be read alongside moisture, acetaldehyde, and carboxylic end group data.

PET Polymer or PET Resin? Buying the Right Stage of the Chain

At the monomer stage, purified terephthalic acid and monoethylene glycol are purchased by integrated melt-phase polymer lines. At the polymer stage, amorphous PET is purchased by solid-state upgrading operators or fiber spinners. At the resin stage, solid-state-polymerized bottle-grade pellets are purchased by preform injection molders and stretch blow molders. Purchasing at the wrong stage transfers capital burden and quality risk. Melt-phase PET with an intrinsic viscosity of 0.62 dL/g cannot be injection stretch blow molded into a carbonated soft drink bottle without an SSP step because the preform would have insufficient tensile strength and would display excessive creep under carbon dioxide pressure. Conversely, an SSP bottle resin with an intrinsic viscosity of 0.84 dL/g is poor economics for fiber melt spinning, where excessive viscosity increases spinneret pressure and limits drawability.

The selection variable is not intrinsic viscosity alone. Residual acetaldehyde in bottle-grade resin is specified because acetaldehyde migrates into filled beverages and imparts a sweet, degraded taste. Melt-phase base polymer typically carries higher free acetaldehyde because the high-temperature polycondensation step generates vinyl ester and acetaldehyde byproducts. Solid-state polymerization at 200–220 °C under nitrogen strip removes acetaldehyde by evaporation and does not require the same melt-phase temperature that regenerates it. Carboxylic end group concentration also falls during SSP through esterification with residual hydroxyl end groups. For this reason, standards such as ASTM F2013 and ASTM D4603 appear together on bottle resin certificates of analysis.

Melt-Phase Polycondensation Ends Where Solid-State Diffusion Begins

Melt-phase polycondensation is terminated when the melt viscosity becomes too high for pumping, pelletizing, and removal of ethylene glycol. For bottle base resin, the melt-phase target is normally 0.60–0.65 dL/g. Above this range, melt index declines and reactor agitator torque rises. The polycondensation equilibrium is limited by ethylene glycol removal; as molecular weight increases, diffusion of the glycol through the increasingly viscous melt becomes rate-controlling. Shifting the reaction to the solid phase reduces thermal degradation but imposes a pellet-scale diffusion limitation. The reaction then takes place in amorphous regions between crystalline lamellae, and the byproducts ethylene glycol and water must diffuse from the pellet core to the pellet surface. Pellet radius and crystallinity therefore become processing variables that are not present in melt-phase design.

Before an amorphous melt-phase pellet can enter an SSP reactor, it must be crystallized to prevent surface sticking. Crystallization is typically carried out at 160–180 °C in a fluidized bed or agitated crystallizer. The crystallinity then rises to 30–40%, which raises the sticking temperature. Without crystallization, amorphous pellets heated to SSP temperature soften and agglomerate. SSP reactors operate at 200–225 °C under vacuum or nitrogen sweep. The polymerization rate is controlled by the rate of removal of ethylene glycol and water; deeper vacuum or higher nitrogen flow increases the rate. At higher temperature the rate constant increases, but the margin to pellet sticking narrows. Published industrial data for continuous SSP reactors indicate residence times in the order of 10–20 h to raise intrinsic viscosity from 0.62 dL/g to 0.82 dL/g, with the exact time dependent on pellet diameter, catalyst residue, and gas purity.

When Acetaldehyde Control Shifts from Resin Drying to Injection Molding Parameters

Bottle-grade PET resin must be dried to below 50 ppm moisture before melt processing because hydrolytic chain scission during extrusion or injection molding reduces molecular weight and generates carboxylic end groups. Drying is typically performed at 160–170 °C for 4–6 h in desiccant dryers with a dew point of −40 °C or lower. Even with dried resin, acetaldehyde is regenerated during injection molding by thermal degradation of vinyl ester chain ends. The melt temperature therefore becomes a critical setpoint. Preform injection molding is typically conducted at 270–290 °C. Above 295 °C, acetaldehyde formation accelerates sharply; published thermal degradation studies place the apparent activation energy for acetaldehyde generation in the 150–180 kJ/mol range. To avoid exceeding the threshold, hot-runner manifold temperatures are limited, and barrel residence time is minimized. Shear heating in the screw and hot runner can raise local melt temperature above the barrel setpoint, so injection speed and back pressure must be adjusted on multi-cavity preform molds rather than relying solely on barrel temperature.

Carbonated soft drink preforms require an intrinsic viscosity of 0.82–0.86 dL/g to withstand carbon dioxide pressure and drop impact after orientation. Water bottles may be produced with intrinsic viscosity of 0.74–0.80 dL/g because the mechanical load is lower. However, water containers have more stringent acetaldehyde sensory requirements than carbonated soft drink containers, so resin selection is not a simple downgrade by intrinsic viscosity. Some water grades are solid-state polymerized to lower intrinsic viscosity but with extended stripping to achieve residual acetaldehyde below 2 ppm by ASTM F2013. Published data for specific commercial resins is limited because acetaldehyde specifications are often supplier-customer agreements rather than public standards.

Substitution of fiber-grade PET resin into a preform injection operation illustrates the stage boundary. Fiber-grade polymer with an intrinsic viscosity of 0.62–0.68 dL/g is formulated for high draw ratio melt spinning and often contains 0.3–2.0 wt% titanium dioxide as a delustrant. The pellet is not solid-state polymerized to bottle-grade intrinsic viscosity, and its residual acetaldehyde is not controlled to beverage specifications. If such material is injection stretch blow molded, the low molecular weight reduces tensile yield and burst pressure, while the titanium dioxide raises haze in a transparent bottle wall. Under ASTM D638-14, tensile yield stress of amorphous PET sheet is commonly reported in the 50–60 MPa range, but low-intrinsic-viscosity fiber resin without SSP lies at the lower boundary and does not compensate for the loss of molecular orientation in an under-stretched preform. The reverse substitution—bottle-grade resin into fiber spinning—increases melt viscosity and spinneret pressure without a corresponding benefit in tenacity. Stage-specific chemistry is therefore inseparable from downstream equipment specification.

Solid-State Reactor Temperature Limits and Pellet Sticking Thresholds

The operating window of a continuous SSP reactor is bounded by two failure modes: agglomeration at high temperature and uneconomically slow kinetics at low temperature. Crystallized PET pellets begin to soften and adhere when bed temperature approaches 220–225 °C, depending on copolymer content and crystallinity. Below 190 °C, the rate of transesterification and esterification falls, and intrinsic viscosity lift may require residence times beyond the vessel design. The practical setpoint range is therefore narrow. In a nitrogen-purge SSP vessel, nitrogen serves as heat carrier and stripping gas; its dew point and oxygen content must be controlled. Oxygen levels above 100 ppmv promote oxidative discoloration and acetaldehyde formation. Water and ethylene glycol in the recycle gas must be removed in a purge-drying loop so that the reverse reaction is suppressed. Pellet temperature, gas inlet temperature, bed height, and pellet velocity are the control variables that determine intrinsic viscosity gain. Published data for production-scale SSP configurations is limited; the boundaries stated here derive from equipment manufacturer bulletins and peer-reviewed studies of polycondensation kinetics.

Post-consumer recycled PET flakes are an additional procurement stage. They enter the chain after solid-state resin, but their intrinsic viscosity distribution is heterogeneous, and the flakes contain polyolefin cap residues, labels, and paper fibers. Solid-state reprocessing of recycled PET at 200–210 °C raises intrinsic viscosity but does not remove non-PET contaminants. Purchasing recycled PET as a drop-in replacement for virgin bottle resin requires proportional blending and preform testing because the melt filtration and SSP history of recycled flake differ from virgin polymer.

Food-contact compliance for PET resin is evaluated under FDA 21 CFR 177.1630 in the United States and EU Regulation 10/2011 in the European Union. Under EU 10/2011, the specific migration limit for antimony from PET is 0.04 mg/kg food simulant, which constrains antimony catalyst residue. Suppliers therefore limit antimony trioxide catalyst concentration and may switch to titanium-based catalysis where national regulations or customer specifications require lower antimony. Migration testing is conducted in food simulants assigned in EU 10/2011, and overall migration must not exceed 10 mg/dm² for plastic materials in contact with food. The certificate of analysis for bottle-grade PET therefore includes intrinsic viscosity by ASTM D4603, acetaldehyde by ASTM F2013, moisture, and often residual catalyst metals. A resin supply chain that skips SSP cannot meet these specifications merely by blending pellets because acetaldehyde and carboxylic end groups are distributed throughout the pellet and are not corrected by dry mixing.