Black Specks and Yellowing in Recycled PET: A Diagnosis Order

In post-consumer recycled polyethylene terephthalate (rPET), black specks and yellowing are rarely independent defects. A diagnostic order that separates solid-phase contamination from melt-phase thermo-oxidation is required because the two visual failures share overlapping thermal histories but differ in corrective action. Black specks in rPET are commonly attributed to carbonized polymer residues, metal-catalysed oxidation, crosslinked gel, or foreign particulates, whereas yellowing is frequently linked to thermal-oxidative chromophore formation, residual acetaldehyde, residual PVC or polyamide contamination, and extended drying above 180 °C. The sequence below moves from low-energy flake inspection to melt-rheology and molecular-weight analysis. Published data for highly mixed municipal bale flake is limited; thresholds should be validated against the local feedstock mix and the target food-contact or fibre grade.

Feedstock Contamination Screening Precedes Melt-Phase Diagnostics

Visual sorting under D65 illumination is only first-pass screening; it does not quantify black specks smaller than 200 µm. Flake retained on a 2 mm sieve is preferred because the <2 mm fraction concentrates label adhesive, aluminium foil, and polyolefin dust. Sink-float separation in water followed by 1.5 g/cm³ brine isolates polyolefin and some aluminium; PVC and heavily filled materials require X-ray or near-infrared systems. PVC contamination should be kept below 50 ppm in bottle-grade flake because dehydrochlorination during extrusion generates hydrochloric acid, accelerates chromophore formation, and produces carbonaceous black specks. Differential scanning calorimetry per ISO 11357-3:2018 at 10 K/min under nitrogen identifies a second melting endotherm near 165 °C from polypropylene or near 220 °C from polyamide 6. Ash content per ISO 3451-1:2019 above 0.05 wt% should trigger inductively coupled plasma optical emission spectroscopy per ISO 11885 for copper, iron, and aluminium, because transition metals lower the onset of thermo-oxidative yellowing. For flake intended for direct food-contact use under EU Regulation (EU) No 10/2011 or US FDA 21 CFR 177.1630, the input stream should be traceable to bottle-grade post-consumer feedstock and free of flake from non-food containers.

When colorimetric data are collected on flake rather than pellets, particle size and surface roughness mask the true yellowness index. Cryomilled flake should be dried under vacuum at 80 °C for 4 h to avoid moisture scattering, then measured as a 5 mm path-length quartz cell against a black or white backing per ASTM E313-20 and ASTM D6290-19 under illuminant D65/10°. A b* value above 4.0 in rPET bottle flake is generally a rejection flag for clear packaging, while yellowness index above 8.0 under D65/10° indicates that melt-phase stabilisation cannot restore optical neutrality. Pellet colour should be measured according to ASTM D6290-19 on injection-moulded plaques of 3 mm thickness after drying to <50 ppm moisture, because residual water in the polymer phase increases haze and depresses b* in a way that is not representative of extruded material. If the b* value decreases by more than 1.0 unit after vacuum degassing, the yellowness is partly due to volatile aldehydes rather than conjugated chromophores.

What Instrumental Thresholds Separate Hydrolytic Yellowing from Thermo-Oxidative Discoloration?

Hydrolytic degradation in rPET is identified by intrinsic viscosity loss without a proportional increase in yellowness index. Intrinsic viscosity is measured according to ISO 1628-5:1998 or ASTM D4603-18 using a 60:40 phenol/1,2-dichlorobenzene solvent at 25 °C; bottle-grade rPET typically falls between 0.70 and 0.80 dL/g after solid-state polymerisation, while injection moulding grades may be accepted at 0.60 to 0.68 dL/g. A carboxyl end-group concentration above 35 mmol/kg measured by titration per ASTM D7409-15 indicates hydrolytic chain scission, even when IV remains in specification. Thermal-oxidative yellowing, in contrast, raises b* and produces unsaturated carbonyl species, often accompanied by acetaldehyde concentrations above 5 µg/g measured by static headspace gas chromatography per ASTM F2013-10. When CEG is below 20 mmol/kg but b* exceeds 6.0, the dominant mechanism is usually oxidative chromophore formation rather than hydrolysis. These two pathways require different corrective actions: hydrolysis demands lower flake moisture and reduced barrel temperature, whereas oxidative yellowing demands tighter vacuum venting, lower melt temperature, and, in severe cases, phosphite-based stabilisation. Stabiliser selection is constrained by food-contact approvals; only substances listed in EU Regulation (EU) No 10/2011 or US FDA 21 CFR 177.1630 should be used for food-grade rPET.

Melt Filtration, Acetaldehyde, and Metal Residue Signatures

Black-spec severity is best quantified on a continuous screen changer rather than by visual pellet inspection. A 40 µm nominal retention melt filter placed after the extruder gear pump retains carbonized polymers, metal flakes, and crosslinked gel. Differential pressure across the filter is a more sensitive indicator than pellet count because small black specks can be dispersed below visual detection. An increase of more than 0.4 bar/min at constant throughput or a total differential pressure above 40 bar is treated as an industrial screening alarm, although alarm limits must be reduced for high-IV resin at lower melt temperature. Published data for this specific alarm limit is limited; the value is a screening default rather than a standard requirement. Filter cake analysis by Fourier transform infrared spectroscopy separates oxidised PET from polyamide or polyolefin contaminants. Simultaneous acetaldehyde analysis on vacuum-degassed strands or pellets by ASTM F2013-10 quantifies residual aldehyde, and values above 3 µg/g in food-contact preforms may exceed sensory thresholds for neutral water. Metal residues determined by ISO 11885 on ash from ISO 3451-1:2019 should be reported not only as total ash but as individual copper, iron, zinc, and aluminium concentrations because copper is a stronger oxidation catalyst than iron on a molar basis. For antimony trioxide-catalysed bottle-grade PET, total antimony levels are commonly below 300 mg/kg and are not sufficient to explain yellowing unless reduced sulfur species are present from label adhesives or degradation of polyamide barrier layers.

Differential pressure monitoring alone does not distinguish black specks generated in the extruder from black specks entering with flake. When pressure rise begins before barrel zone 5 reaches 260 °C, the source is usually solid-phase contamination; when pressure rise begins only after melt temperature exceeds 280 °C, the source is usually thermal degradation in the screw or die. This distinction can be checked by sampling the melt before and after a static mixer or by stopping the screw and inspecting the flight clearances for carbonised deposit. Screw design matters: an L/D ratio of 32:1 with a single vacuum vent in the metering section is common for rPET, but high-shear kneading blocks in the melting zone can raise local melt temperature by 10–15 °C and trigger black speck formation even when barrel setpoints remain below 270 °C. Replacement of kneading blocks with gear mixing elements and reduction of screw speed from 600 min⁻¹ to 400 min⁻¹ often reduces black-spec counts without changing throughput. No single published standard defines black-spec counts in rPET pellets; internal optical scanner limits based on particle size larger than 150 µm are common and should be linked to downstream preform or sheet rejection rates.

If Black Specks Appear Only After Solid-State Polymerization, Target Diffusion-Limited Degradation

Solid-state polymerisation can raise intrinsic viscosity but also converts pre-existing black specks into larger carbonaceous centres because high temperatures under low residual oxygen are not neutral. SSP reactors operating at 210 °C under vacuum below 1 mbar increase IV through polycondensation, but if flake retains ethylene glycol, low-molecular-weight oligomer, or trace PVC, the same conditions can generate brown and black side products. A diagnostic method is to compare black-spec counts per 100 kg on flake before SSP and on pellets after SSP using an optical scanner set to detect particles larger than 150 µm. If the count increases by more than 20% through SSP, flake washing and sorting should be improved rather than increasing vacuum level. Yellowing after SSP is often diffusion-limited; surface b* may be 2.0 to 3.0 units higher than core b* when the chip is cryogenically fractured and measured under D65/10°. A gradient of more than 1.5 units between surface and core indicates that surface-bound contaminants or oxidative attack are present. In such cases, the process should be diagnosed with time-resolved melt shear rheometry under nitrogen, not only with IV, because crosslinked gel and black specks can coexist with acceptable intrinsic viscosity.

Capillary rheometry at 280 °C and apparent shear rates from 100 s⁻¹ to 1000 s⁻¹ detects crosslinked gel not always visible in filtration data. A melt viscosity ratio between 100 s⁻¹ and 1000 s⁻¹ above 3.0 in high-IV rPET often indicates long-chain branching or gel; when combined with black specks larger than 150 µm, the diagnosis shifts from simple foreign contamination to thermo-oxidative gel formation. Oscillatory time sweeps at 270 °C and 1 rad/s under nitrogen monitor gel formation during extended residence; an increase in storage modulus by more than 20% over 20 min is a practical upper boundary for extrusion residence time. This test requires a controlled-stress rotary rheometer with parallel-plate geometry and a gap of 1 mm, and should be run on material dried to below 50 ppm moisture. When gel is absent, black specks may be traced to dead zones in the die adapter or screen changer. Inspection of the die adapter after 24 h of continuous operation often reveals carbonised deposits at points where flow velocity drops below 0.1 m/s, especially at melt temperature above 285 °C. Reducing die volume and eliminating non-streamlined transitions lowers black-spec formation more effectively than raising melt filtration area.

Interpreting Intrinsic Viscosity Loss During Vacuum Drying

Drying of rPET flake at 170 °C for 6 h in a desiccant dryer with dew point below -40 °C is required to bring moisture below 50 ppm before extrusion. However, high residual water in ambient flake can hydrolyse the polymer before drying is complete, so IV measured on incoming flake at 0.72 dL/g may fall to 0.63 dL/g during the first hour at 160 °C before moisture is removed. This early IV loss is a moisture problem, not a thermal-oxidative problem. If IV continues to fall after the flake reaches below 50 ppm moisture, the cause is usually oxygen ingress through leaking seals, insufficient nitrogen purge, or excessive drying temperature above 190 °C. In a closed-loop desiccant dryer, the dew point should be checked at the hopper inlet and outlet; an outlet dew point above -20 °C indicates moisture breakthrough. Yellowing during drying is rarely caused by hydrolysis, because hydrolysis primarily produces chain ends rather than chromophores. Therefore, yellowed flake with IV above 0.65 dL/g and CEG above 35 mmol/kg should not be diagnosed as purely hydrolytic; the yellowing is more likely due to residual tallow-derived or adhesive contaminants.