PET Polymerisation and SSP: How IV Is Actually Built

The subject PET Polymerisation and SSP: How IV Is Actually Built describes a two-stage molecular-weight architecture in which polyethylene terephthalate is first polymerised in the melt to a moderate intrinsic viscosity and then advanced in the solid state to bottle or technical-fibre grades. PET is the polycondensation product of ethylene glycol with purified terephthalic acid or dimethyl terephthalate. The glass transition is approximately 78 °C and the crystalline melting peak typically falls between 245 °C and 255 °C. Intrinsic viscosity, expressed in dL/g, is used as a molecular-weight proxy and is measured according to ASTM D4603-18 or ISO 1628-5:2015 in 60:40 phenol/1,1,2,2-tetrachloroethane at 30 °C. Melt-phase lines normally deliver 0.55–0.65 dL/g. Solid-state polycondensation is then used to reach 0.80–0.84 dL/g for carbonated-soft-drink bottles and 0.85–1.00 dL/g for tire cord and strapping.

In direct esterification, purified terephthalic acid and ethylene glycol are mixed into a paste and reacted in an esterification vessel at 240–270 °C. Water is distilled off and free ethylene glycol is recycled. In the dimethyl terephthalate route, transesterification releases methanol and requires a catalyst, but the route allows easier purification. The resulting oligomers are transferred to a pre-polycondensation vessel and then to a finisher. During polycondensation the Carothers relationship applies; the number-average degree of polymerisation is determined by the extent of reaction. Because the equilibrium constant for PET melt polycondensation is low, ethylene glycol must be removed continuously to drive chain growth. Industrial finishers use rotating disk or cage agitators, high-surface-area melt films, and vacuum systems capable of maintaining 1–3 mbar. The practical ceiling is not a single molecular-weight limit but the point at which the required residence time for ethylene glycol removal exceeds the thermal stability window of the melt.

Melt-Phase Polycondensation Cannot Exceed a Practical IV Ceiling

Melt-phase polycondensation is carried out in a high-vacuum finisher at 270–285 °C. Antimony trioxide is the most common catalyst at 150–250 ppm residual antimony; titanium alkoxide systems may operate at 5–30 ppm residual titanium and germanium dioxide at 20–50 ppm residual germanium. Catalyst selection alters the polyesterification rate, the subsequent SSP rate, and the colour shift on remelting. During the finisher stage, ethylene glycol, water, acetaldehyde, and oligomeric volatiles are removed through vacuum domes and spray condensers. The melt viscosity at the discharge is monitored as a surrogate for molecular weight. The maximum melt-phase IV is constrained by three linked factors: mass transfer of ethylene glycol from a high-viscosity melt, thermal chain scission, and acetaldehyde regeneration. At 290 °C and above, degradation accelerates; acetaldehyde is formed by β-hydrogen transfer and thermal cleavage. The result is that continuous melt lines are rarely operated above 0.65 dL/g. Beyond that point, the rate of IV increase becomes lower than the rate of thermal degradation-related IV loss.

Diethylene glycol formation during esterification is an additional control variable. The presence of 0.8–1.5 wt% diethylene glycol units in the chain lowers the crystalline melting point, reduces gas barrier, and changes SSP crystallisation kinetics. Esterification and polycondensation are also sensitive to carboxyl end-group balance. A high carboxyl end-group concentration consumes hydroxyl ends and limits the maximum attainable degree of polymerisation; industrial lines therefore monitor carboxyl end groups before the finisher and adjust the glycol feed ratio.

After melt finishing, the strand is quenched in water, cut into cylindrical pellets, and conveyed to the solid-state line. The first unit operation is crystallisation. In a fluidised-bed or paddle crystallizer, amorphous pellets are heated to 120–170 °C until the crystallinity rises to 30–50%. This prevents the pellets from sintering when the temperature is raised above their original glass transition. The crystallized pellets are then preheated to 200–225 °C under dry nitrogen and transferred to the SSP reactor. Continuous moving-bed vessels from Polymetrix, Bühler, and UOP Sinco hold the pellets for 8–20 h in a countercurrent nitrogen stream or under vacuum. The SSP reactor removes ethylene glycol, water, and acetaldehyde; this removal shifts the polycondensation equilibrium and extends existing chains. The output IV is controlled by bed height, pellet residence time, gas dew point, and reactor temperature.

What Actually Limits the Intrinsic Viscosity Build Rate in SSP?

Two resistances control the SSP rate: the chemical polycondensation reaction between terminal hydroxyl and carboxyl end groups, and the diffusion of condensate molecules from the chain end to the pellet surface. In the early SSP stage, the end-group concentration is high and the kinetic regime dominates. As IV increases, the concentration of reactive ends falls, the amorphous-phase viscosity rises, and the diffusion path becomes rate-determining. Published effective diffusion coefficients for ethylene glycol in semicrystalline PET at 200–225 °C are on the order of 10^-9 to 10^-8 cm²/s. Because diffusion time scales with the square of pellet radius, chip diameter is a first-order process variable. Pellets with a diameter of 2.0–2.5 mm build IV faster than 3.5 mm pellets, but smaller pellets generate more fines and increase gas-distribution pressure drop. The melting onset of the crystallized resin sets the upper temperature limit: local hot spots above 230 °C can cause sintering. The nitrogen or vacuum system must keep the condensate partial pressure low; otherwise water and ethylene glycol re-equilibrate with the chain ends and suppress molecular-weight growth.

End-group stoichiometry influences the SSP trajectory. Polycondensation consumes one hydroxyl and one carboxyl end group per chain extension event. If the melt-phase resin enters SSP with an excess of carboxyl end groups, the attainable molecular weight can become hydroxyl-limited. Transesterification reactions during SSP can redistribute chain lengths without releasing condensate, but they do not by themselves raise the number-average molecular weight. For this reason, solid-state IV build is not simply a function of residence time; it is a function of hydroxyl-to-carboxyl balance, catalyst residue, pellet geometry, and condensate removal.

Solid-State Reactor Gas Phase and Pellet Morphology Control

Inside a moving-bed SSP reactor, a plug-flow pellet residence time distribution is maintained with internal gas distributors and conical discharge sections. The nitrogen loop is filtered, heated, dried to a dew point below -40 °C, and maintained with oxygen below 10 ppmv. Bed temperature uniformity is typically controlled within ±2 °C to avoid local overcooking. Pellet morphology affects both diffusion and mechanical stability. Crystallinity above 50% reduces sticking but restricts condensate diffusion; crystallinity below 30% risks sintering at SSP temperature. The pre-crystallization step therefore sets the amorphous-phase mobility and the effective diffusion path length. Dust below 150 μm is removed before the reactor because fines can cause channelling, filter fouling, and pressure-drop excursions. Operating failures observed on continuous lines include uneven preheating, pellet breakage, and condensation in the nitrogen recycle line, all of which broaden the IV distribution and increase residual acetaldehyde.

Intrinsic viscosity itself is a rheological measurement that depends on dissolution and solvent purity. Dried resin is dissolved in 60:40 phenol/1,1,2,2-tetrachloroethane at 105–135 °C for a time sufficient to eliminate gel-like undissolved structures, then measured in an Ubbelohde capillary viscometer at 30 ± 0.01 °C. Single-point calculations are acceptable under ASTM D4603-18; ISO 1628-5:2015 specifies the reference protocol for PET. The IV result is converted to molecular weight through the Mark–Houwink relation, but the published constants vary with solvent ratio and temperature. Acetaldehyde in bottle resin is measured by headspace gas chromatography according to ASTM F2013-10; carboxyl end groups are determined by titration according to ASTM D7409-15. These measurements matter because SSP not only raises molecular weight but also strips residual acetaldehyde and free glycol. Carbonated-soft-drink preform resin is specified below 1 ppm acetaldehyde because higher levels migrate into the beverage and create off-taste.

When SSP Is Pushed Beyond 0.85 dL/g for Tire Cord and Strapping

For high-tenacity polyester used in tire cord, technical yarn, rope, and strapping, IV values of 0.85–1.00 dL/g are required. To reach this range, SSP lines operate at the upper thermal edge, commonly 220–230 °C, with extended residence times and enhanced condensate removal. The processing window is narrow because the melting onset of the annealed semicrystalline resin may be only 5–10 °C above the reactor setpoint. A temperature overshoot above 230 °C can fuse pellets and force a shutdown. At high IV, melt processing becomes more demanding: preforms and fibres must be dried to <30 ppm moisture before extrusion to avoid hydrolytic degradation, and extrusion temperatures may be raised by 5–15 °C relative to bottle grades. Titanium-catalysed PET can show faster SSP rates than antimony-catalysed resin but may also generate more acetaldehyde on remelting. Carboxyl end-group concentration is held below 20 meq/kg for hydrolysis-resistant bottle grades and below 15 meq/kg for technical fibre grades. Under these conditions, the key process conflict is that long SSP residence times increase IV and strip acetaldehyde but also risk thermal yellowing and surface oxidation if oxygen ingresses into the reactor.

Storage and conveying after SSP introduce their own process limits. Pellets discharged from the cooler at 40–60 °C are transferred by dense-phase conveying to silos purged with dry air or nitrogen. If ambient humidity exceeds 60% relative humidity, surface moisture adsorbs on the pellets and must be removed again before injection moulding. The IV build is not stable if the resin is held at elevated temperature in the presence of oxygen; residual catalyst can accelerate oxidative discolouration. Therefore the final cooling section of the SSP line is designed to bring pellets below 60 °C before atmospheric exposure.