Menu
Continuous polycondensation of fibre-grade PET from purified terephthalic acid and ethylene glycol is operated as a multi-stage melt process in which the carboxy end group concentration of the final polymer is not a single-stage output but an accumulated consequence of esterification equilibrium, residence-time distribution, thermal exposure, hydrolytic side reactions, and stabilizer/catalyst stoichiometry. In direct melt spinning plants, polymer with intrinsic viscosity of 0.62–0.68 dL/g and carboxy end group content below 30 meq/kg is typical for apparel and technical yarns; higher carboxy end group values are associated with increased die drool at spinnerets, altered disperse dye uptake, and reduced hydrolytic stability during downstream wet processing. The concentration of carboxy end groups is routinely measured by titration according to ASTM D7409-15, which uses alcoholic alkali in a solvent system and reports milliequivalents per kilogram. Process engineers set the final carboxy end group target through a combination of esterification conversion, polycondensation temperature profile, vacuum staging, catalyst addition, and phosphorus-based thermal stabilizer additions, each of which has an operating window that becomes narrower when the product is bright, cationic-dyeable, or TiO₂-delustered.
PET produced by direct esterification retains terminally unreacted acid functionality from incomplete conversion of purified terephthalic acid. During the first esterification stage, the apparent esterification conversion is controlled by the molar EG/PTA feed ratio, typically 1.10–1.30 for continuous paste feed reactors, by the pressure profile between 1.2 bar and 2.8 bar, and by the temperature from 260 °C to 280 °C. Unreacted purified terephthalic acid has extremely low solubility in the oligomer melt, so the end group count entering the prepolycondensation stage reflects the efficiency of water removal and the dissolution of residual purified terephthalic acid crystals rather than a true homogeneous equilibrium. If the esterification conversion is allowed to fall below 95 %, the resulting oligomer may carry total carboxy ends above 350 meq/kg, which cannot be practically corrected later because polycondensation relies on ester interchange and glycol end capping, not direct esterification of free acid at the same rate. Conversely, an excessively high EG/PTA ratio above 1.6 increases diethylene glycol formation and shifts the final prepolymer carboxyl pool through ether-containing diester sequences; these sequences affect the activity of the antimony catalyst and alter the carboxy end group response to stabilizer addition. The esterification stage therefore limits the minimum achievable carboxy end group concentration in continuous lines more than the finisher vacuum can compensate for, because the acid ends that survive the first reactors are diluted but not selectively removed during chain extension.
In a modern four- to six-reactor continuous train, the residence time in the esterification zone is usually distributed over two or three agitated vessels with internal coils and recycle loops. The oligomer from the final esterifier can be held at 275–285 °C for 90–180 min, and the carboxyl end concentration is only minimally reduced after the final esterifier; the subsequent prepolycondensation stage is therefore decisive because it must simultaneously lower free ethylene glycol, remove water generated by polycondensation, and build molecular weight under a pressure cascade from 100 mbar to 10 mbar. If this cascade is initiated too late or the prepolycondensation reactor is operated with poor plug-flow characteristics, low-molecular-weight fractions containing higher carboxy end concentrations bypass the finisher and appear in the spun yarn as periodic intrinsic viscosity defects and carboxy end group excursions. These defects are observed on production-scale spin beams as variations in pack pressure rise and threadline breaks at draw ratios above 3.0, and their source is frequently traced to agitator bypass in the prepolymeriser rather than to the finisher itself. Continuous lines that use a horizontal prepolycondensation vessel with perforated baffles and a centrally mounted agitator are less prone to such bypass than unbaffled vertical units, but only when the liquid level is maintained within the designed range of 50–70 % of vessel volume.
The lower limit of carboxy end groups in a continuous line is set by the kinetic balance between antimony-catalyzed ester interchange and the thermal and hydrolytic generation of new acid ends. Antimony trioxide is typically fed as a glycolate solution at 150–300 mg/kg antimony in the prepolymeriser, giving a final antimony content between 180 ppm and 250 ppm. At these levels, the apparent polycondensation rate constant is sufficient to reach intrinsic viscosity 0.62–0.64 dL/g within finisher residence times of 60–120 min, but the catalyst also accelerates side reactions that generate carboxy end groups if local melt temperatures exceed 292 °C. Phosphorus-based stabilizers such as phosphoric acid, triethyl phosphate, or triethylene glycol phosphate are added after esterification at phosphorus-to-antimony molar ratios from 0.3 to 1.0. Too little phosphorus leaves the antimony catalytically active enough to promote oxidative carboxy end group formation in dead zones; too much phosphorus above a phosphorus-to-antimony ratio of 1.5 reduces polycondensation rate and forces either higher finisher temperature or longer residence time, which can paradoxically raise carboxy end group content. The accepted operational window is therefore a phosphorus-to-antimony ratio of 0.4–1.2, with the lower end used for titanium-catalyzed melts and the upper end used where spinneret pack temperatures are high. Published data for titanium phosphate equilibria in continuous melt lines are limited; however, batch autoclave studies consistently show a nonlinear carboxy end group response to added phosphorus that depends on the water content of the ethylene glycol feed and on the order of addition relative to the antimony glycolate solution.
When the melt enters the finisher at an intrinsic viscosity of 0.30–0.45 dL/g, the concentration of dissolved water and ethylene glycol in the melt phase controls the rate of carboxy end group formation because hydrolysis of PET in a melt with residual water above 100 ppm is autocatalytic at temperatures above 270 °C. Continuous finishers designed with cage or disk-ring agitators generate thin-film surface renewal and remove ethylene glycol through a vapour line connected to a vacuum system, typically achieving absolute pressures below 1.0 mbar. In such equipment, the mass-transfer coefficient for ethylene glycol removal depends on the square root of the agitator tip speed and the melt viscosity at the operating temperature; if the viscosity is too high because of excessive TiO₂ addition or low temperature, the effective interfacial area decreases and carboxy end group content rises due to hydrolytic chain scission even when the vacuum gauge reads 0.8 mbar. Production experience indicates that finisher wall deposits containing crosslinked gel also create stagnant zones where carboxy end group content can exceed 50 meq/kg; these gels are periodically released and manifest as yarn breaks in the draw zone, a failure mode that can be reduced by maintaining finisher temperature below 295 °C and phosphorus levels within the stabilizer window.
After the finisher, fibre-grade melt is transferred through jacketed lines, booster gears, and static mixers to the spin beam, and this liquid-phase residence at high temperature is a dominant source of late-stage carboxy end group formation in direct spinning. Thermal degradation of PET proceeds through random chain scission at ester linkages, producing a vinyl ester end and a carboxyl end; the reaction rate increases exponentially with temperature and follows first-order kinetics in the polymer melt. In a transfer line operated at 285–290 °C, the carboxy end group generation rate has been reported in plant audits to be broad and dependent on residual moisture, antimony concentration, oxygen content of the blanketing gas, and the surface-to-volume ratio of the transfer line; published data for a specific continuous configuration are limited, but the qualitative effect is reflected in the common specification that total transfer residence from finisher outlet to spinneret be kept below 30 min. Static mixers and spin pumps contribute relatively little holdup volume but can create local shear heating of 2–5 °C, which accelerates degradation at the wall. To limit carboxy end group formation, melt lines are designed for plug flow with minimum hold-up volume, polished internal surfaces, and nitrogen blanketing in feed zones. Any increase in carboxy end group content during transfer is amplified in fibre properties; a final carboxy end group increase of 5 meq/kg can raise pressure drop through a 40 µm filtration pack by 10–15 % and shift disperse dye exhaustion by 2–4 % at 130 °C dyeing, according to comparative dye uptake trials using ISO 105-C06 wash-fastness evaluation as the confirming test.
At the spinneret pack, the polymer is filtered through layers of sintered metal or sand media before passing through precision capillaries; this zone has a short residence time, but the high surface area of the filter medium can adsorb stabilizers and change the local phosphorus-to-antimony balance. The pressure drop across a 20–40 µm filtration pack varies with melt viscosity and gel content, and when carboxy end group content rises above 35 meq/kg, the increased carboxyl group density promotes adhesion of degraded oligomers to the metal surfaces, reducing pack life. Mechanical cleaning of spin packs by pyrolysis at 450–500 °C removes polymer deposits but can leave ash residues that catalyze carboxy end group formation in subsequent campaigns; therefore, pack cleaning procedures normally specify an acid wash after pyrolysis, such as 5–10 % aqueous acetic acid at 60–80 °C, to remove catalytic metal oxides. The pressure tap at the spin pump discharge provides an early indication of carboxy end group-related melt instability: a monotonic pressure rise over 8–12 h in a direct-spinning line is frequently associated with increasing carboxy end group content and gel accumulation, and the pack is replaced when pressure exceeds the initial value by 1.5–2.0 times.
High-vacuum polycondensation is kinetically coupled to carboxy end group concentration through the reversible reaction of terminal carboxyl and hydroxyl groups to form ester linkages and water; removal of water drives chain extension, while accumulation of water reverses the reaction and increases both hydroxyl and carboxyl end groups. In a continuous finisher, the apparent equilibrium is not a single thermodynamic value but a distribution governed by the rate of interfacial transport relative to the rate of chemical generation. The mass-transfer coefficient for ethylene glycol and water removal from a PET melt at 280 °C and 1 mbar is on the order of 0.01–0.05 kg m⁻² h⁻¹ in mechanically agitated thin-film equipment, depending on surface renewal frequency. Cage-type agitators with tip speeds of 0.5–1.5 m/s generate a falling film over the vessel wall that increases the interfacial area, but if the melt viscosity at 0.30 dL/g is above 200 Pa·s, the film becomes unstable and the vacuum only removes volatiles from a limited zone. Residence time distribution in the finisher is determined by the ratio of vessel working volume to throughput and by the internal baffle arrangement; a finisher with a Peclet number below 10 has substantial axial mixing, allowing a fraction of melt to leave with insufficient chain extension and elevated carboxy end group content. For this reason, fibre-grade lines often specify a final finisher with a working volume that gives a mean residence time of 60–120 min and design the agitator to achieve a Bodenstein number above 20, ensuring plug-flow character. The vacuum system itself is important: the pressure at the finisher vapour outlet should be 0.5–1.5 mbar, and the ethylene glycol spray condenser should maintain a temperature of 10–20 °C to prevent backflow of condensate into the melt. If the vacuum line is undersized or condensate drains poorly, the effective water partial pressure at the melt surface can exceed 0.2 mbar even when the vacuum pump pressure is low, and carboxy end group values remain above target.
| Process variable | Operating range / equipment type | Observed carboxy end group response | Mechanistic basis and operational limitation |
|---|---|---|---|
| EG/PTA molar feed ratio in paste esterifier | 1.10–1.30 in agitated vertical reactors at 260–280 °C | Ratios below 1.05 increase carboxy end groups; ratios above 1.6 reduce carboxy ends but raise DEG | Unreacted purified terephthalic acid persists as suspended acid if diol is insufficient; excess ethylene glycol caps acid ends but increases ether formation |
| Prepolymerisation pressure cascade | 100 mbar to 10 mbar over two stages | Delayed or shallow cascade raises carboxy end group content in final melt | Residual water and ethylene glycol remain in melt and hydrolyze ester linkages |
| Final finisher pressure | 0.5–1.5 mbar absolute at vapour outlet | Pressure above 2.0 mbar increases carboxy end group content | Higher water partial pressure shifts polycondensation equilibrium backward |
| Finisher temperature | 275–295 °C | Sustained operation above 292 °C raises thermal carboxy end group generation | Random chain scission and oxidative degradation accelerate exponentially with temperature |
| Antimony catalyst concentration | 150–300 mg/kg antimony | Low catalyst can leave carboxy ends uncapped; high catalyst accelerates side reactions | Catalyst increases both polycondensation rate and degradation rate in dead zones |
| Phosphorus-to-antimony molar ratio | 0.4–1.2 | Below 0.3 raises oxidative carboxy end group formation; above 1.5 slows polycondensation and indirectly raises carboxy ends | Phosphorus stabilizers deactivate catalyst and require balanced thermal history |
| Transfer line residence time | Below 30 min from finisher to spinneret | Extended residence time increases late-stage carboxy end group formation | Liquid-phase thermal degradation occurs at 285–290 °C in jacketed transfer lines and spin beams |
TiO₂ delustrant at 0.3–2.0 wt% is added as a slurry before polycondensation or after prepolymerisation, and its presence alters the carboxy end group formation pathway because the rutile surface can adsorb antimony and phosphorus. In dull grades, the effective antimony activity is reduced when TiO₂ is not adequately surface-treated, requiring higher catalyst feed or lower phosphorus addition to maintain the same final intrinsic viscosity. This shifts the phosphorus-to-antimony operating window downward to 0.3–0.8 and increases the risk of thermal carboxy end group formation if the spin beam temperature is not reduced. Flame-retardant fibre grades based on phosphorus monomers or phosphinate additives interact with the polycondensation catalyst more strongly; the excess phosphorus can raise the carboxy end group level above 35 meq/kg unless the melt residence time is shortened or the final finisher vacuum is improved. Cationic-dyeable PET produced with sodium dimethyl isophthalate-5-sulfonate also exhibits elevated carboxy end group content because the sulfonate comonomer increases melt acidity and the required lower polycondensation temperature reduces the esterification conversion. In such systems, carboxy end group targets are typically relaxed to 40–50 meq/kg, but the higher carboxyl content reduces the hydrolytic stability of the fibre during jet dyeing at 130 °C, so the package is often supplemented with an after-drawing heat-setting step to increase crystallinity and reduce the accessible amorphous fraction where hydrolysis occurs.
| Fibre-grade parameter | Test method designation | Typical specification for bright apparel fibre | Relevance to carboxy end group formation or downstream performance |
|---|---|---|---|
| Carboxy end group content | ASTM D7409-15 | 15–30 meq/kg | Primary control parameter; values above 30 meq/kg affect dyeability and hydrolytic stability |
| Intrinsic viscosity | ASTM D4603-18 or ISO 1628-5:2021 | 0.62–0.68 dL/g | Molecular weight indicator; low intrinsic viscosity with high carboxy end groups suggests hydrolytic or thermal chain scission |
| Melting temperature | ASTM D3418-21 | 245–255 °C | Shift below 245 °C indicates excessive diethylene glycol incorporation or comonomer-induced sequence disruption |
| Wash fastness of dyed yarn | ISO 105-C06 | Grade 4–5 for disperse dye after heat setting | Elevated carboxy end groups reduce wet fastness and require additional heat setting |
| Yarn tensile strength | ASTM D2256-21 | 4.5–5.5 cN/dtex for draw-twisted yarn | Strength loss above 10 % after wet storage indicates autocatalytic hydrolysis driven by carboxyl ends |
A production-scale audit of continuous melt spinning lines using direct polymerisation has shown that batch-to-batch carboxy end group variation is more strongly correlated with the ratio of phosphorus to antimony and the finisher vacuum stability than with the esterification feed ratio alone. The carboxy end group variance within a lot, as measured by sampling at the spin pack every 2 h, is typically 2–4 meq/kg when the phosphorus-to-antimony ratio is controlled within ±0.05; if the phosphorus metering pump loses calibration, the variance increases to 8–12 meq/kg and the yarn shows corresponding dyeability streaks. Maintenance of the finisher vapour line and spray condenser is equally important because partial clogging of the ethylene glycol condensate return line raises the effective water backpressure and produces a slow upward drift in carboxy end group content over 6–10 h. Process analytical verification usually combines off-line titration of melt or yarn samples according to ASTM D7409-15 with in-line melt viscosity and melt temperature signals; the in-line viscosity alone cannot distinguish carboxy end group development from molecular weight loss, so a calibrated numerical model that includes feed stoichiometry, finisher vapour pressure, and melt residence time is used to infer carboxy end group content. The operational boundary for bright apparel-grade fibre is generally a final carboxy end group level of 15–30 meq/kg; above 30 meq/kg, the fibre is still spinnable but downstream dyeing uniformity and hydrolytic stability may fall outside the specification limits of ISO 105-C06 or the buyer’s wet-fastness protocol.
