The intrinsic viscosity of fibre-grade poly(ethylene terephthalate) (PET) is deliberately held at 0.62–0.68 dL/g after continuous melt-phase polycondensation in a finisher operating at 270–280 °C and 1–3 mbar absolute pressure. This window is selected because textile melt-spinning processes demand a lower melt viscosity than bottle-grade preform injection, and the fibre-grade resin therefore tolerates less molecular-weight loss before filament breakage, draw resonance, and insufficient orientational strength appear. Measurement per ASTM D4603-18 is made in a 60/40 w/w phenol/1,1,2,2-tetrachloroethane solvent at 30 °C; a decline from 0.68 dL/g to 0.58 dL/g is associated with a measurable reduction in spinline elongational viscosity and in the molecular orientation achievable at 4,000–6,000 m/min take-up. In integrated fibre plants, chip drying to 30 ppm residual moisture is required before single-screw extrusion because ester hydrolysis consumes water at the ester linkage and is autocatalyzed by the carboxylic acid end groups produced.

What Intrinsic Viscosity Measurements Are Required Before Spinning at 0.62–0.68 dL/g?

Incoming fibre-grade chip should be characterized for intrinsic viscosity before transfer to the melt-spinning hopper. ASTM D4603-18 specifies flow-time measurements on a 0.5 g/dL solution in 60/40 w/w phenol/1,1,2,2-tetrachloroethane at 30 °C using an Ubbelohde glass capillary viscometer. The calculated intrinsic viscosity is used as a process-release criterion, not as a direct molecular-weight value; typical fibre-grade deliveries are accepted only within 0.62–0.68 dL/g, with a batch-to-batch variation of no more than ±0.01 dL/g on the same spinning line. At the lower bound of 0.62 dL/g, melt-processing treatments such as drying, extrusion, and transport through 30:1 L/D single-screw extruders leave very little chain-scission margin before spinline tension becomes unstable. This is the first reason textile grade tolerates less than higher-IV PET: the starting molecular weight is already near the minimum required for adequate melt strength.

Pre-drying of fibre-grade PET is ordinarily performed in a desiccant-wheel hopper dryer with air dew point of -40 °C or lower and an inlet temperature of 160–170 °C for 4–6 h. Residual moisture is verified by ASTM D6869-17 before extrusion, with a control limit of 30 ppm. Above 80 ppm moisture at a melt temperature of 285 °C, hydrolysis during a mean residence time of 8–12 min can reduce intrinsic viscosity by more than 0.05 dL/g. In production-scale single-screw extruders with a 30:1 L/D screw and six heated zones, the result is a lower pressure drop across the spin pack and a visible increase in filament breaks at the quench chamber. This operational boundary explains why textile grade tolerates less moisture exposure than bottle-grade chip stored under the same silo conditions.

When Hydrolytic Degradation Drops IV Below 0.58 dL/g During Melt Holding

Once fibre-grade intrinsic viscosity falls below 0.58 dL/g, the melt-spinning operation enters a fault region even when extruder temperatures are unchanged. At a spinneret throughput of 0.8–1.2 g/min/hole and a take-up speed of 4,500 m/min, the spinline stress may no longer be sufficient to achieve the required molecular orientation. Filament tenacity measured by ISO 2062:2023 can decline below 30 cN/tex, and the secondary draw ratio must often be reduced from 1.6–1.8 to avoid breakage, lowering final modulus and dimensional stability. In addition, the auto-catalytic nature of the hydrolysis reaction means a small initial moisture excursion can propagate through the melt stream; carboxylic acid end groups generated during one pass further accelerate chain scission in any downstream regrind. Published kinetic data for melt-phase PET hydrolysis generally report activation energies near 75–85 kJ/mol, which implies that a 10 °C increase in melt temperature may roughly double the rate of IV loss. The lower starting IV of fibre-grade material therefore leaves a narrower margin before the critical 0.56 dL/g threshold for commercially acceptable drawn yarn is reached.

Comonomer drift in continuous fibre-grade PET plants introduces a second boundary. Diethylene glycol (DEG) is generated in situ or added at 1.2–1.5 wt% to suppress crystallization and improve dye uptake; the melting peak measured by ASTM D3418-21 is typically 250–255 °C at that DEG level. If DEG rises above 1.8 wt%, the melting point can fall below 247 °C, and the allowable melt temperature before thermal degradation must be reduced, while quench air must still produce a stable oriented structure at 0.4–0.6 m/s crossflow. Fibre-grade PET at 0.62–0.68 dL/g tolerates less comonomer drift than bottle-grade PET because bottle-grade preform moulding does not rely on the same spinline melt-strength window; a preform at 0.80–0.84 dL/g can absorb equivalent DEG variation with less immediate loss of pressure hold strength.

High-Speed Melt Spinning Operates Within a Narrow Rheological Corridor

Capillary rheometry on dried fibre-grade PET at 285 °C and a shear rate of 1,000 s⁻¹ places apparent melt viscosity in the range of 150–300 Pa·s, while bottle-grade resin at 0.84 dL/g commonly exceeds 350 Pa·s under identical conditions. The lower viscosity of textile grade is necessary for high spinneret hole densities and fine denier; however, it also means that small shifts in moisture, DEG, or residence time alter spinline stability more quickly. On a commercial spin beam with 4,800 holes and a 0.25 mm orifice diameter, a viscosity drop of 20 Pa·s can be sufficient to change filament diameter coefficient of variation beyond 1.5% if the quench delay and air velocity are not adjusted. At a melt temperature of 295 °C, thermal degradation begins to dominate over the viscosity benefit, leaving a practical processing window of approximately ±5 °C around the optimum. This is the central reason textile grade tolerates less: the operating corridor is narrower because the melt has been intentionally shifted to the low-viscosity end of the PET processing envelope.

Fractional gel particles, dust, and polymer fines create another tolerance boundary. A fibre-grade chip stream with 0.62–0.68 dL/g cannot be filtered as aggressively as high-IV bottle grades because the lower melt viscosity reduces filter pressure drop but also reduces the ability to encapsulate or homogenize particulates before the spinneret. Screen packs in the spin pack are typically 20–40 µm absolute, and any contamination above this size generates fibre breaks at 4,000 m/min; bottle-grade preform moulding, by contrast, uses larger gates and can tolerate occasional particle-induced defects without the same immediate downstream winding failure. This practical difference is visible in fibre plants where spin-pack service intervals are shortened when regrind content exceeds 5–10 wt% in the chip blend.

Fibre-to-Preform Transfer Crossovers in Regrind and Co-Injection

Transfer of fibre-grade PET to bottle-grade conversion is restricted by the same IV deficit. A preform resin normally requires 0.80–0.84 dL/g to pass sidewall deformation and burst resistance tests referenced in ASTM D2463-23. Adding 10 wt% of 0.65 dL/g fibre regrind to a 0.84 dL/g virgin resin may reduce the blend intrinsic viscosity to 0.80–0.82 dL/g, which can remain acceptable only if the preform is overdesigned. At 20 wt% regrind, the blend can approach 0.80 dL/g, below many commercial preform specifications. In co-injection processes, fibre-grade melt is also more sensitive to residence time in the hot runner because its lower molecular weight generates additional thermal degradation products, including acetaldehyde. Consequently, textile grade tolerates less recycled content in bottle-grade streams, and separate silos are maintained for fibre and preform chip.

In production environments where ambient relative humidity exceeds 60%, fibre-grade chip stored in unsealed silos can reach 100 ppm moisture within 24 h if dryer air is not maintained at -40 °C dew point. A spinning plant using 0.62–0.68 dL/g chip should therefore seal the chip transfer line from the day-silo to the hopper, maintain desiccant-bed regeneration at 220–250 °C, and verify chip moisture at the extruder throat with an in-line Karl Fischer analyser calibrated to ASTM D6869-17. In the quench chamber, relative humidity at 20 °C and 65% ambient may be reduced to 40% or less by chilled air handling to avoid filament hydration before drawing.