Colour Mismatch Between Polyethylene Batches: Root Causes and Remedies

Batch-to-batch colour variation in natural and pigmented polyethylene (PE) is rarely attributable to a single variable. The melt-state optical response of high-density polyethylene (HDPE), low-density polyethylene (LDPE), linear low-density polyethylene (LLDPE), and medium-density polyethylene (MDPE) is modulated by density, crystalline superstructure, melt flow rate, additive package, and thermal-mechanical history. When a 2 mm compression-moulded plaque is read in a spectrophotometer with d/8° geometry according to ASTM E313, yellowness index differences of 2.0 units are visually perceptible in neutral-coloured compounds. Pigmented compounds commonly require CIELAB ΔE*ab below 1.0 to 2.0 units against a customer-approved standard, depending on end-use.

A typical production specification for natural PE pellets requires yellowness index below 5.0 per ASTM D6290; however, this limit does not prevent visible batch shift in unpigmented film because haze and surface reflectance are not captured by YI alone. Resin source changes, even within the same commercial grade family, can introduce measurable shifts because catalyst residues, comonomer distribution, and thermal stabiliser levels are not held constant across production sites. Metallocene-catalysed LLDPE may contain fewer transition-metal residues than chromium-catalysed HDPE, but its narrower molecular weight distribution can alter shear heating and melt temperature by 3 to 6 °C in a high-speed extruder, which influences oxidation rate and final colour.

Catalyst residues provide a less obvious but documented root cause. Chromium-catalysed HDPE often contains trace chromium oxide and polyunsaturated sites along the backbone; these can form coloured complexes during oxidation even before visible gel formation. Ziegler-Natta grades may retain titanium and aluminium residues that combine with phenolic antioxidants to form faint yellow or pink complexes. Metallocene and supported non-metallocene single-site catalysts generally leave lower residues, but their narrower melt flow behaviour makes them more sensitive to shear-induced temperature rise in twin-screw compounding. A resin lot with ash content above 0.05 wt% per ISO 3451-1:2019 should be quarantined for colour testing even if melt flow rate and density remain in specification.

Base resin density differences of 0.002 g/cm³ measured by ISO 1183-1:2019 are enough to change the ratio of crystalline to amorphous light scattering in translucent articles. For opaque white or coloured compounds, density-driven changes in scattering are masked by pigment, but pigmented thin-wall parts may still show a shift in appearance when the same masterbatch is used across resin lots with different molecular architecture.

What Limits Dispersion when the Masterbatch Carrier Resin Differs from the Matrix?

In production-scale compounding, colour mismatch often emerges from differences in masterbatch letdown and dispersion rather than from pigment loading alone. A concentrate based on a low-viscosity LDPE carrier let down at 2 wt% into HDPE with a melt flow rate of 0.3 g/10 min per ISO 1133-1:2022 can produce visible streaks because the low-molecular-weight carrier melts and migrates before the high-viscosity HDPE matrix reaches full shear thinning. On a 75 mm co-rotating twin-screw extruder with a 38:1 L/D ratio and side feeding at zone 7, the measured pressure before a 100/60/20 mesh screen pack may rise from 8 to 12 MPa when pigment agglomerates remain intact. Maintaining melt temperature within ±5 °C of the target across the die is a common control boundary; above the upper boundary, oxidation accelerates, while below the lower boundary, viscosity rise can reduce shear stress below the threshold for agglomerate rupture. Specific energy input is a more useful control metric than screw speed alone. A shift in specific energy from 0.18 to 0.22 kWh/kg can indicate dispersion changes that precede visible colour shift.

The geometry and metallurgy of the melting zone affect dispersion before optical instruments detect a problem. Nitrided barrels and bimetallic screws with reduced clearances of 0.10 to 0.15 mm increase shear stress and improve pigment microdispersion, but they also raise energy input. A worn screw with clearance exceeding 0.40 mm may fail to break agglomerates, causing colour drift that cannot be corrected by masterbatch concentration alone. Comparative data from production lines show that screw replacement restores colour consistency faster than reformulation when the same masterbatch lot is used across two machines with different wear states.

The comparative data below systemise common root-cause variables and corrective actions for colour mismatch during compounding.

Melt flow rate and molecular weight distribution influence colour beyond simple mixing. A lot with a lower melt flow index than the standard sheet dissipates more viscous energy during extrusion, leading to higher melt temperature and accelerated degradation of both polymer and pigment. This is observed on single-screw injection moulding machines with barrel diameters from 25 to 80 mm, where a drop in melt flow index from 8 to 5 g/10 min can raise the measured melt temperature by 2 to 4 °C at constant screw speed. The resulting yellowness increase may not appear immediately in pellets but becomes visible after a second heat history in thermoforming or blow moulding. For white compounds containing rutile titanium dioxide, the photocatalytic surface of the pigment can accelerate oxidation if the antioxidant package is consumed; this interaction is not visible as a colour shift until the part enters extended indoor lighting or outdoor exposure. Therefore, a colour approval should not rely solely on pellet or plaque readings after the first processing pass; a second-pass evaluation according to ASTM D6290 or CIELAB measurement is required when the part undergoes regrinding or multiple thermal cycles.

When Thermal Oxidation During Multiple Heat Histories Alters Yellowness Index

The yellowing mechanism in unstabilised or partially stabilised PE is dominated by thermo-oxidative chain scission and the formation of conjugated carbonyl species. In processing lines that operate at die temperatures above 220 °C, residence times longer than 5 min at high shear can consume phenolic antioxidants and generate quinone-type chromophores from oxidised hindered phenols. Transmission FTIR analysis of compression-moulded films, following ASTM D5576-00(2021)e1, may show an increase in the carbonyl absorption band at 1710–1740 cm⁻¹ of 0.02 absorbance units per mm thickness; this shift frequently accompanies a measurable yellowness index increase. The relationship is not strictly linear because light scattering from spherulites and surface defects also contributes to perceived colour.

Remediation requires a reduction in peak melt temperature, purging of degraded polymer after shutdown, and verification that the antioxidant masterbatch is fully dispersed. In blown film extrusion, where bubble stability limits the ability to reduce melt temperature, a hindered phenolic and phosphite stabiliser package is commonly adjusted within the formulation rather than relying on process changes alone. Pre-drying of hygroscopic filler-containing masterbatch is recommended when ambient relative humidity exceeds 60%; moisture entering the extruder can hydrolyse phosphite stabilisers and reduce their colour-protection efficiency.

Process-control remedies include nitrogen blanketing of the feed throat, reduction of screw speed, and installation of melt coolers for high-output lines. However, reducing screw speed without increasing feed rate may reduce specific energy input but extend residence time, creating a trade-off that must be monitored with in-line yellowness index measurement. In multi-machine comparisons, an in-line spectrophotometer installed after the die can detect YI drift earlier than pellet sampling; when the in-line signal shifts by 0.5 YI units, corrective action is typically initiated before the pellet specification is violated.

In cast and blown film, measured colour is not solely a material property; surface topography and thickness distribution alter spectral reflectance and scattering. A textured chill roll at 18 °C to 24 °C may produce lower gloss and less apparent colour difference than a polished roll at the same temperature, even when the identical compound is processed. Gloss measurements per ISO 2813:2014 and haze per ASTM D1003-21 should be recorded alongside spectral data, because changes in surface roughness can shift the CIELAB values without any change in pigment dispersion. In multi-layer coextrusion, layer thickness variation of ±10% in a white skin layer can create visible colour mismatch across a film roll; closed-loop thickness control or target layer ratio adjustment is therefore part of the colour correction workflow.

Thickness and surface effects are especially relevant in white PE film for labels and hygiene applications. A calcium carbonate or titanium dioxide filler increases opacity but also changes the reflectance curve. When filler content is reduced to lower density, the hiding power drops and the base resin yellow cast may become visible even though the pigment weight percentage remains unchanged. For consistent colour, the masterbatch formulation should be adjusted for the final thickness; a concentrate approved at 50 µm thickness may not provide sufficient hiding power at 20 µm and will expose yellowness from the core layer.

Pigment Degradation, Masterbatch Dilution, and Additive Interference Boundaries

Not all pigments survive PE processing under the same thermal and shear conditions. Phthalocyanine blue and green pigments generally withstand extrusion above 250 °C for short residence times, but certain monoazo yellows and reds begin to decompose above 220 °C. Pigment suppliers usually report heat stability limits, but these are often determined in a different carrier or with shorter residence time than production lines. When a masterbatch is let down at 1 wt% instead of 3 wt%, the effective wall-slip and local shear conditions can change because the concentrate acts as a low-viscosity melt stream; the pigment may experience higher local temperature in the kneading blocks. In a 40:1 L/D corotating twin-screw extruder, increasing the number of kneading elements after a side feeder usually improves colour strength but can raise melt temperature by 3 to 5 °C, offsetting the dispersion gain.

The compatibility of carrier resin with pigment and base resin is another controlled variable. An ethylene-vinyl acetate or acid-modified carrier can improve wetting of difficult pigments, but residual acid groups may react with certain chromophores or stabilisers and shift hue. The use of oxidised PE waxes as dispersants is recognised to improve carbon black dispersion, yet overuse can introduce yellow-brown components. For critical colour applications, the masterbatch carrier and dispersant package should be fixed and not substituted without a full CIELAB comparison on plaques and finished parts.

Some additive systems create colour drift independent of pigment. Overloading with zinc stearate or calcium stearate above 0.2 wt% can leave visible plate-out on die lips and may shift opacity in white compounds. In PE compounds intended for gamma sterilisation, high doses of certain phenolic antioxidants can form yellow quinoidal species; irradiation at 25 kGy has been observed to increase yellowness index in some medical-grade PE formulations. Where gamma compatibility is required, the complete additive package must be tested on a 2 mm compression-moulded plaque according to ASTM E313 after irradiation, because unirradiated YI values do not predict post-sterilisation colour.

In outdoor PE applications, ultraviolet stabilisers influence initial colour and long-term colour drift. Hindered amine light stabilisers are generally non-colouring at 0.1 wt% to 0.3 wt%, but their interaction with phenolic antioxidants under high processing temperatures can produce a slight yellow tint in natural formulations. Carbon black-loaded PE shows less yellowing because carbon black screens ultraviolet radiation, but the particle size and dispersion of carbon black affect jetness and surface appearance. In masterbatches with carbon black, poor dispersion appears as brownish or greyish streaks rather than achromatic black. Dispersion can be assessed by pressure-rise measurement and by scanning electron microscopy on cryomicrotomed specimens.

Post-consumer recycled PE introduces batch-to-batch optical variability that cannot be corrected by pigment concentration alone. Contaminants such as polypropylene, ethylene-vinyl acetate, and residual adhesives shift melt rheology and create microscopic phase domains that scatter light. A target addition rate of 15 wt% to 25 wt% recycled content may require a redesign of the colour concentrate because the recycled fraction carries its own pigmentation and stabiliser residues. In injection moulding, a melt-temperature rise of 5 °C may be required to handle a recycled lot with lower melt flow index, and that rise can alter colour of temperature-sensitive pigments. Compliance with food-contact requirements under FDA 21 CFR 177.1520 or European Commission Regulation (EU) No 10/2011 adds further constraints on the use of recycled streams and masterbatch ingredients.

Recycled content may also carry residual printing inks, paper labels, and surface coatings that generate grey or yellow tones after remelting. Melt filtration using a screen changer with 80 to 120 mesh screens can remove larger contaminants, but dissolved inks and nano-sized residues will pass through and affect colour. Deodorisation and stabiliser replenishment do not automatically neutralise colourants from recycled packaging. A reliable strategy includes sourcing post-industrial scrap with known resin grades and pre-colour sorting by near-infrared detection. When this is not possible, the final compound must be tested on production-scale equipment because laboratory blending alone does not reproduce the colour change caused by contaminant degradation in the melt phase. Colour drift can remain below 0.8 ΔE*ab when post-industrial streams are tightly source-controlled; without source control, published data for this specific configuration is limited and pilot-scale trials are required.