Polyethylene terephthalate is bought and sold on certificate-of-analysis values that map directly to downstream processing confidence. The phrase PET Resin Specification: The Four Numbers That Decide Price identifies intrinsic viscosity, diethylene glycol content, carboxylic end-group concentration, and CIE b* color. These four values are not academic descriptors; each one corresponds to a measurable failure mode in injection molding, stretch blow molding, fiber spinning, or film extrusion. A specification sheet can list viscosity, color, and comonomer values, but the price negotiation is commonly anchored to how far each value sits inside the buyer’s operating window. A resin lot with an intrinsic viscosity of 0.80 dL/g, a diethylene glycol value of 1.2 wt%, a carboxylic end-group value of 20 meq/kg, and a b* value of 1.0 will be priced differently from a lot at 0.78 dL/g, 1.7 wt%, 45 meq/kg, and 2.8, even if both are classified as bottle grade. The reason is not grading label inconsistency; it is the process capability of the customer’s equipment. A high-cavitation preform injection line running polyethylene terephthalate closures may tolerate a wider range, whereas a hot-fill container line with a tight acetaldehyde specification may not. In import and export transactions, the certificate of analysis is the basis for price deductions or premiums, and the four numbers are reviewed before sampling begins.

What Intrinsic Viscosity Reveals About Molecular Weight

When the certifier reports intrinsic viscosity, the value indicates the hydrodynamic volume of the polymer chain in solution. For PET, ASTM D4603-18 specifies a 0.50% solution in 60/40 phenol/1,1,2,2-tetrachloroethane at 30°C. The unit is dL/g, and the reported single-point value is converted from relative viscosity through the standard’s calculation sequence. Fiber-grade resin typically occupies the 0.58–0.68 dL/g range; film grades may be sold at 0.62–0.70 dL/g; bottle grades commonly fall between 0.78 dL/g and 0.86 dL/g. The price increment for higher intrinsic viscosity is not arbitrary. Raising molecular weight above the melt-phase ceiling requires solid-state polycondensation, which adds batch residence time, nitrogen flow, and energy. A resin lot at 0.84 dL/g may carry a price premium over a lot at 0.80 dL/g because the solid-state reactor’s throughput is lower and the downstream molder can thin-wall the preform without sacrificing drop impact.

The melt-phase polymerization equilibrium is limited by ethylene glycol removal. Horizontal disc-ring finishers operating at 270–285°C with vacuum below 2 mbar can deliver a melt intrinsic viscosity of roughly 0.60–0.68 dL/g; higher values require a solid-state polycondensation step. In continuous solid-state units, pellets are crystallized and then held at 200–220°C under nitrogen or vacuum to raise intrinsic viscosity and reduce carboxylic end groups. The tradeoff is residence time: a bottle-grade campaign targeting 0.84 dL/g may require several hours longer than a 0.80 dL/g campaign, which reduces output and raises cost. Published data for specific energy consumption per 0.01 dL/g increase is limited. Intrinsic viscosity is an average quantity; it does not report polydispersity, branching, or melt strength anomalies. Two lots with the same intrinsic viscosity can behave differently in a hot runner if one has been thermally damaged and shows a broader molecular weight distribution.

A moisture content above 0.4% during dissolution can invalidate intrinsic-viscosity determination because hydrolysis may occur, so sample preparation enters the measurement. For intrinsic-viscosity comparison between suppliers, the certificate should state the dissolution temperature and solvent ratio; otherwise apparent differences below 0.02 dL/g may be artifacts rather than molecular weight changes. Contract negotiations therefore specify ASTM D4603-18 rather than “inherent viscosity” without method qualification. In a twin-screw extrusion pelletizing line with an L/D 32:1 screw, a resin lot at the high end of the intrinsic-viscosity band can raise specific mechanical energy, and the cutter must be tuned to avoid fines; fines contribute to subsequent color measurement error. This is why price deductions may be applied for intrinsic viscosity outside the contracted band even if the mean value is acceptable.

On a production unit, the same premium is earned only if the high-viscosity lot does not simultaneously increase barrel pressure and cycle time. Equipment technical bulletins for high-cavitation preform molding list melt residence management as a greater constraint than clamp force; the melt pressure required to fill a 24-cavity tool increases with intrinsic viscosity, and the molder must retune screw recovery, back pressure, and hot-runner temperature. Published data for this specific configuration is limited, but the directional effects are consistent: intrinsic viscosity changes of ±0.02 dL/g alter preform weight control, especially when the machine’s barrel L/D ratio is 24:1 or higher and residence time is above 6 min.

Representative bands are compiled from commercial certificate-of-analysis data and do not replace contract limits; exact values vary by reactor train, catalyst package, and regional supply.

Vacuum Depth and End-Group Control

Melt-phase polycondensation fails to deliver bottle-grade molecular weight unless vacuum depth and temperature staging are controlled simultaneously. Diethylene glycol is not an inert diluent in the PET chain. It forms during esterification when two ethylene glycol molecules dehydrate to produce a diol that is incorporated into the polyester backbone. The feed point of diethylene glycol to the polymerization reactor is controlled at the monomer recovery stage, and the value reported on a certificate of analysis is measured by ASTM D4607-14 after depolymerization and gas chromatography. A diethylene glycol value near 1.2 wt% in bottle-grade PET is associated with a crystallization half-time that is slower than that of a homopolymer, which improves preform clarity but can extend cycle time in injection molding. At 1.8 wt% or above, the melting point depression can require lower barrel temperatures; published trade literature indicates that areal draw ratio limits in stretch blow molding may need to be reduced because the thermal crystallization window shifts. For textile fiber, the same diethylene glycol content opens dye uptake sites; for film, excessive diethylene glycol is undesirable because the lower melting point reduces dimensional stability in hot-fill applications. Published data for this specific configuration is limited, and no single diethylene glycol value can be applied without matching the copolyester composition and process layering.

The formation of diethylene glycol is favored by excess ethylene glycol in the esterification vessel and by high hydroxyl-to-carboxyl ratios. Reactor temperature alone is not a single predictor; the ratio of free ethylene glycol to water in the recycle column determines how much diethylene glycol is generated. A rise of 0.2 wt% diethylene glycol can move the melting point of the copolymer lower by several degrees Celsius, but the exact suppression is composition-dependent. Because the certificate value is a bulk average, a segregated preform in a hot runner can experience a slightly different crystallization rate if wall slip and shear heating drive local temperature above the crystallization onset. Published data for this specific configuration is limited. These are the tradeoffs that sit behind the second number on the specification sheet.

Carboxylic end-group concentration is the third number in the price calculation. Carboxylic end groups act as hydrolytic chain-scission sites. The value is measured by titration after dissolution in 50/50 phenol–chloroform; supplier methods use potassium hydroxide at 0.1 N. A carboxylic end-group value of 20 meq/kg is equivalent to an acid number of 1.12 mg KOH/g using the molecular weight of potassium hydroxide. This conversion is used when comparing recyclate lots across different certificate formats. Bottle resin certificates commonly show ≤ 35 meq/kg, while specialty film and high-modulus fiber specifications may require ≤ 20 meq/kg. The price impact is not linear: a lot at 15 meq/kg may be accepted for premium applications, whereas a lot at 40 meq/kg may be downgraded to strapping or staple fiber depending on the buyer’s melt filtration unit.

Process conflicts arise when carboxylic end groups are pushed lower in the melt phase: increasing residence time and vacuum to reduce end groups can simultaneously raise diethylene glycol if the ethylene glycol recycle loops are not purged adequately, and can increase color. Solid-state polymerization is therefore used to reduce carboxylic end groups because the lower temperature restricts side reactions. Continuous solid-state units with nitrogen flow and temperature staging at 200–220°C are the standard post-reactor route for bottle-grade intrinsic viscosity and end-group improvement. Resin with carboxylic end groups above 35 meq/kg should be dried immediately before processing; exposure to ambient relative humidity above 60% can increase surface moisture and accelerate hydrolysis during plastication. Desiccant dryers with air dew point below −40°C and pellet residence times of 4–6 h at 160–180°C are the standard barrier against processing hydrolysis. The operational boundary is strict: high-carboxylic-end-group resin should not contact nylon or polycarbonate purge materials unless the drying system can maintain 50 ppm moisture or lower.

Recyclate streams complicate the carboxylic end-group number. Bottle flake that has undergone repeated melt recycling can show higher carboxylic end-group values because hydrolysis occurs in wash and extrusion. A blend of virgin and recyclate may meet the average intrinsic viscosity specification but exceed the customer’s carboxylic end-group limit. When the 30 meq/kg limit is exceeded, the price drops for applications such as strapping or fiber, where hydrolytic stability requirements are lower. Some converters install melt filtration and vacuum venting to handle high-carboxylic-end-group recyclate; twin-screw extruders with vacuum vents can strip some volatiles, but they do not reduce carboxylic end groups below the value already formed. This is an operational boundary sometimes missed in purchase contracts.

When CIE b* Is Negotiated Before Resin Transfer

Color is negotiated before resin transfer because storage and drying do not correct an off-spec b* value. ASTM D6290-19 defines the measurement of color for plastic pellets using a spectrophotometer or tristimulus colorimeter under CIE D65/10° conditions. The fourth commercial number is CIE b*, not L*, because L* rarely falls below contract limits except in heavily contaminated recyclate streams. Water-clear bottle resin is typically sold with a b* value no greater than 1.5; b* values between 1.5 and 2.5 may be acceptable for tinted containers or opaque sheet. Higher b* values are not merely cosmetic: they correlate with thermal-oxidative chain damage in melt-phase finishing, residual catalyst metals, or acetaldehyde buildup. In a continuous solid-state polymerization unit, oxygen ingress through rotary valves raises b* even when intrinsic viscosity remains unchanged; the corrective action is not higher vacuum but improved nitrogen flow and pressure interlocks.

CIELAB differences above 0.5 units are generally detectable under controlled viewing conditions; the exact customer rejection limit is product-specific and depends on wall thickness, masterbatch dilution, and illuminant. Phosphorous-based stabilizers can reduce thermal color formation, but overdosing can inhibit catalyst activity and depress melt-phase intrinsic viscosity; the specification must balance color and molecular weight. Food-contact status for PET in the United States is covered under 21 CFR 177.1630, which addresses the base polymer rather than visual yellowness. The color value is a commercial specification, not a regulatory limit.

Pellet surface moisture also influences optical readings. A sphere-based spectrophotometer collected under ASTM D6290-19 should be used with a clean glass cell and compaction to eliminate light path scattering from pellet edges and fines. If pellets are measured in a translucent container with condensation, the reported CIE b* value can shift even when the material has not changed. The standard specifies sample preparation; raw readings without standard handling should not be compared between laboratories. Yellowness index may be reported as an auxiliary value under ASTM E313-20, but the contract number is CIE b*. This distinction matters because yellowness index is a scalar derived from tristimulus data, while CIE b* is a coordinate in the CIELAB space.

Once the dryer dew point crosses −40°C, batch-to-batch variation in the four numbers moves from laboratory paperwork to injection-line stability. A shift of ±0.02 dL/g in intrinsic viscosity requires a change in holding-pressure profile on high-cavitation preform tools; if carboxylic end groups rise at the same time, the molecular weight during plastication may fall due to hydrolysis, producing short shots and frequent preform blowouts in two-stage stretch blow molding. Residual acetaldehyde, measured by ASTM F2013-10(2018), is not one of the four price-setting numbers but is a release gate for water containers, where ≤ 3 ppm is often specified. Acetaldehyde generation is sensitive to barrel temperature and residence time; thus resin with higher intrinsic viscosity and higher b* may fail acetaldehyde limits if the injection molder compensates by raising the barrel setting rather than optimizing screw recovery. Published data for this specific configuration is limited; however, equipment manufacturers’ process notes for L/D 32:1 barrier screws recommend a melt temperature of 280–300°C for bottle resin and a back pressure of 0.5–1.0 MPa to avoid excessive shear heating.

The four numbers interact as a cascade. High diethylene glycol reduces melting point, so a molder may lower barrel temperature; lower temperature increases melt viscosity and can raise shear heating, which can increase acetaldehyde and b* in the preform. High carboxylic end groups with inadequate drying cause hydrolytic intrinsic-viscosity loss, which reduces preform burst strength. Oxygen ingress during solid-state polymerization raises b* even when intrinsic viscosity and carboxylic end groups are acceptable. The absence of a desiccant dryer with a dew point below −40°C is an operational boundary; processing with a hopper dryer at ambient humidity is not sufficient for intrinsic viscosity above 0.80 dL/g. Incompatibilities are not limited to water: acetaldehyde-producing additives, vinyl chloride contamination from PVC residues, and flame-retardant residues can interact with the polymerization history of the resin and produce plate-out on the screw root after prolonged runs. No specification number is meaningful without the corresponding sample preparation and test method attached to the certificate of analysis.