Polyethylene Terephthalate (PET) Glass Transition Near 78 °C: Why Sterilisation Sets the Limit

Polyethylene terephthalate (PET) is a semi-crystalline polyester obtained from the polycondensation of terephthalic acid or dimethyl terephthalate with ethylene glycol. Bottle-grade resins are commonly characterised by intrinsic viscosity values between 0.80 dL/g and 0.84 dL/g as determined by ASTM D4603, while fibre and film grades may show values between 0.55 dL/g and 0.65 dL/g. Differential scanning calorimetry per ASTM D3418 on amorphous specimens typically locates the glass transition near 78 °C, with published onset values from 67 °C to 81 °C depending on residual moisture, comonomer content, cooling rate, and degree of crystallinity. The melting endotherm is observed between 240 °C and 255 °C, far above the thermal conditions used in most terminal sterilisation processes. Nevertheless, the operational ceiling for transparent amorphous PET is not the melt temperature; it is the glass transition and the associated cold-crystallisation and stress-relaxation phenomena. Steam autoclave cycles at 121 °C or 134 °C exceed the glass transition by 43 °C to 56 °C, which places the polymer in a rubbery state where dimensional stability, optical clarity, and molecular weight retention are compromised. Sterilisation therefore sets the limit for PET in medical and pharmaceutical packaging applications.

Which Molecular and Mechanical Transitions Occur Above 78 °C?

Dynamic mechanical analysis per ASTM E1640 at 1 Hz generally resolves the loss modulus peak of bottle-grade PET between 80 °C and 85 °C, close to the calorimetric glass transition. At 121 °C, the amorphous phase has sufficient free volume for large-scale segmental motion, stress relaxation, and orientation losses. The heat distortion temperature of unannealed amorphous PET under a flexural stress of 0.455 MPa is typically reported between 70 °C and 75 °C per ASTM D648; under 1.82 MPa the value falls to 60 °C to 65 °C. The practical implication is that an amorphous PET tray or bottle sidewall under light mechanical load can deform before the chamber reaches terminal sterilisation conditions. Differential scanning calorimetry on fully amorphous PET at a heating rate of 10 °C/min shows a cold-crystallisation exotherm with onset near 120 °C and a peak between 135 °C and 150 °C. Steam sterilisation at 121 °C lies just at the leading edge of this crystallisation window. Long hold times, thick walls, or pre-existing nucleation sites can promote additional crystallinity during the cycle, producing haze, shrinkage, and embrittlement.

In steam autoclave operations validated according to ISO 17665, typical cycles expose the load to 121 °C for 15 min to 20 min or 134 °C for 3 min to 5 min. The combination of heat and saturated moisture is more damaging than dry heat at the same nominal temperature because the ester linkages in PET can hydrolyse. Multi-cycle testing of injection-moulded PET bars has shown measurable intrinsic viscosity loss after repeated saturated steam exposure, although published data for this specific configuration is limited. The common production consequence is a drop in tensile elongation and impact resistance when measured according to ASTM D638 and ASTM D256. Dimensional failure is even more immediate for oriented PET. When an injection stretch blow-moulded container is exposed to 121 °C, residual orientation stresses relax and the article can shrink toward its preform geometry. Heat-setting during blowing at 130 °C to 150 °C reduces this tendency, but heat-set hot-fill containers qualified at 85 °C to 95 °C are not automatically validated for autoclave exposure. Hospital central sterile supply departments frequently reject amorphous or lightly heat-set PET trays after visual and dimensional inspection because the chamber temperature exceeds the heat distortion threshold by more than 46 °C. The glass transition is therefore the boundary condition for terminal steam sterilisation of amorphous PET.

Thermoforming lines running crystallised PET (CPET) deliberately cross the cold-crystallisation window. Amorphous sheet is heated to 130 °C to 150 °C and the formed part is held in a 150 °C to 180 °C mould to build a crystallinity level between 30 % and 40 % when measured by density per ASTM D1505. The resulting structure tolerates reheating to 200 °C and exhibits far less dimensional change than amorphous PET. The trade-off is opacity, reduced notched impact performance when tested per ASTM D256, and a higher modulus. The medical packaging implication is direct: if a PET component must survive autoclave sterilisation while retaining transparency, crystallisation is not an acceptable stabilisation route, and a low-temperature sterilant must be selected.

Preforms used in injection stretch blow moulding are reheated in quartz infrared ovens to 100 °C to 115 °C and stretched at area ratios of 6:1 to 10:1. Heat-setting is performed by holding the blown container in a mould heated to 130 °C to 150 °C for 2 s to 6 s. This step increases crystallinity of the oriented sidewall and reduces thermal shrinkage during subsequent hot filling. Even with heat-setting, a 134 °C autoclave cycle approaches the cold-crystallisation peak and can exceed the stabilisation capacity of the oriented structure. Published data for specific container geometries under saturated steam at 134 °C is limited; qualification therefore requires dimensional comparison before and after the full sterilisation cycle using coordinate measuring or optical comparator systems.

Dry-Heat and Steam-Induced Hydrolytic Degradation Pathways

Dry-heat sterilisation according to ISO 20857 commonly employs 160 °C for 120 min or 170 °C for 60 min. These temperatures lie 82 °C to 92 °C above the nominal glass transition and coincide with the cold-crystallisation exotherm maximum. Amorphous PET develops haze and shrinks under these conditions before useful sterilisation can be completed. Thermogravimetric analysis per ASTM E1131 shows that the main degradation mass loss in nitrogen occurs well above 300 °C; however, moisture-induced chain scission is thermally activated at much lower temperatures. PET moulding resins must be dried before processing in desiccant dryers operating at 160 °C for 4 h to 6 h with a dew point of -40 °C or lower, corresponding to residual moisture below 50 ppm as measured by ISO 15512. During dry-heat sterilisation, insufficiently dried parts or trapped moisture in packaging can hydrolyse the ester chain, reducing molecular weight and surface craze resistance. Repeated dry-heat cycles can also produce visible yellowing from thermo-oxidative degradation, even though the cycle temperature is below the melting point. Dry-heat sterilisation is therefore restricted to crystallised, heat-stabilised PET structures and is generally unsuitable for transparent amorphous PET medical devices.

Ethylene oxide sterilisation is performed at chamber temperatures from 37 °C to 63 °C per ISO 11135, with relative humidity between 40 % and 80 %. The thermal load remains below the PET glass transition for most combinations, avoiding the dimensional relaxation and cold-crystallisation failures observed in steam or dry-heat cycles. However, ethylene oxide absorption into amorphous PET can occur, and the process must be followed by aeration to meet residue limits under ISO 10993-7. Hydrogen peroxide vapour and gas plasma systems operate at 40 °C to 60 °C and are compatible with many amorphous PET components, although penetration into narrow lumens and closed containers remains an operational constraint. These low-temperature routes decouple sterilisation from the 78 °C glass transition. Their primary limitation is not thermal deformation but sterilant distribution, compatibility with the packaged load, and cycle validation.

When Radiation Sterilisation Remains Below the Thermal Limit but Introduces Radical Chemistry

Gamma sterilisation per ISO 11137 commonly delivers doses from 25 kGy to 50 kGy; the process is essentially athermic at the product level, with local temperature rises generally below 40 °C in industrial irradiators. Electron beam irradiation at 10 MeV or lower produces similar dose levels with shorter exposure times. Because the cycle remains below the PET glass transition, amorphous transparent parts do not undergo the same thermal dimension change seen in autoclave processing. The chemical and radiation physics of PET are, however, affected. Ionising radiation generates free radicals along the chain, some of which recombine, while others remain trapped and contribute to dose-dependent yellowing. PET is regarded as one of the more radiation-tolerant packaging polymers, but formulation additives, colour bodies, and co-packaged adhesives can respond differently. Tensile and impact testing per ASTM D638 and ASTM D256 should be performed on the final part after the selected dose because the response is grade-specific. Radiation sterilisation is therefore a lower-temperature alternative for PET when the medical device or packaging cannot tolerate 121 °C steam, but it transfers the risk from the glass transition to radical-induced degradation and extractables.

Production-scale injection moulding of amorphous PET medical consumables typically uses hot-runner valve-gated tools on machines with clamp force between 1200 kN and 2500 kN. Mould temperatures are held at 10 °C to 20 °C to quench the melt and retain clarity, but this leaves residual thermal and flow-related stress in the component. Steam autoclave exposure at 121 °C raises the part above the glass transition and mobilises these frozen-in stresses, producing anisotropic warpage and dimensional variation. Dimensional audits on autoclave-qualified PET parts are commonly performed with coordinate measuring machines or vision systems before and after cycling; acceptance criteria for critical envelope dimensions are often 1.0 % to 1.5 %. Amorphous PET parts without heat-setting or a crystallised shell routinely exceed this limit when autoclaved. The material-level boundary remains the glass transition near 78 °C; the sterilisation method selected for a PET device determines whether the component remains within its usable dimensional and molecular-weight envelope.