In polyethylene blown film intended for direct food contact, the overall migration limit of 10 mg/dm² of food-contact surface area, as specified in Article 12 of Commission Regulation (EU) No 10/2011, is the controlling gravimetric endpoint for non-volatile substances that transfer to food simulants. The measurement, performed by total immersion according to EN 1186-2:2002 for fatty simulants and EN 1186-3:2002 for aqueous simulants, does not differentiate between polymer-derived oligomers, processing aids, slip agents, or degradation products. Molecular weight parameters—M_n, M_w, M_z, and the low-molecular-weight tail—control both the intrinsic concentration of mobile polymer-derived species and the rate of diffusion through the semi-crystalline matrix. In polyolefins, lower molecular weight increases the number of chain ends available to create local free volume discontinuities, shortens the average chain length that must be disentangled for segmental motion, and can decrease the average diffusion path tortuosity if crystallinity is reduced. Conversely, higher molecular weight generally reduces migrant diffusivity in the amorphous phase but can increase melt elasticity and processing severity, which may introduce new degradation products if barrel residence time or melt temperature is not controlled.

The operational significance of molecular weight is most visible when migrating species have molecular weights between 300 g/mol and 1,200 g/mol, a range that includes paraffinic oligomers, low-molecular-weight waxes, fatty amides, and phthalate-free plasticisers. In high-pressure low-density polyethylene, chain branching and broad molecular weight distribution produce a measurable oligomeric fraction; in linear low-density polyethylene made by Ziegler-Natta or metallocene catalysis, the low-molecular-weight tail is generally narrower but can still contain oligomers formed by chain transfer. During overall migration testing in 95% ethanol or isooctane substitutes at 40 °C to 60 °C, these low-molecular-weight species are extracted from the amorphous regions and can approach the 10 mg/dm² limit if the total additive package and polymer oligomer content are not optimised. A film produced from a grade with M_w below 80,000 g/mol may show higher gravimetric residue than a film of identical comonomer content and thickness produced from a grade with M_w above 180,000 g/mol, although exact results depend on additive solubility, surface area-to-volume ratio, and simulant type. Intentionally added substances such as erucamide, oleamide, silica antiblock, and fluoroelastomer processing aids migrate through the amorphous phase at rates governed by their molecular size, shape, and compatibility. The polymer molecular weight alters the free volume but does not uniformly affect all migrants; polar migrants in non-polar polyethylene may be excluded from crystalline regions and migrate through amorphous channels, whereas non-polar oligomers may co-crystallise or reside in interlamellar zones. In a blown film with a high freezeline quench rate, crystallinity is often lower and crystal size is smaller, increasing the amorphous phase fraction and therefore the available volume for migration. This effect can partially counteract the benefit of higher molecular weight because the amorphous phase remains continuous and more permeable. The overall migration limit of 10 mg/dm² therefore cannot be predicted from polymer molecular weight alone; it requires actual extraction testing under the intended food simulant because the method captures the combined effects of polymer morphology, additive exudation, and simulant swelling.

What mechanistic role does the number-average molecular weight play in diffusion-controlled migration through semi-crystalline blown film?

Diffusion of low-molecular-weight migrants in polyethylene is often described by the free volume model D = D₀ exp(−B/f_v), in which f_v is the fractional free volume of the amorphous phase. Chain ends contribute free volume because they terminate the covalent sequence and allow local segmental rearrangement without the constraints of mid-chain entanglements. As M_n decreases, the concentration of chain ends per unit volume increases; for a hydrocarbon polymer with M_n of 20,000 g/mol compared with 40,000 g/mol, the molar chain-end concentration is approximately doubled, although the effect on f_v is partially offset by crystallite interfaces and chain loop conformations in the interlamellar region. In semi-crystalline blown film, the diffusion pathway for a migrant is restricted to the amorphous phase, and the crystallites act as impermeable obstacles that increase tortuosity. Higher molecular weight tends to increase tie-molecule density between lamellae and may decrease the long-spacing amorphous thickness in rapidly quenched blown film, further reducing the effective diffusivity. Published diffusion data for alkane and oligomer probes in polyethylene indicate that an increase in M_w by a factor of two can reduce the diffusion coefficient by roughly 0.5 log units when crystallinity and orientation are held constant; however, published data for exact diffusion coefficients in a given commercial blown film configuration is limited because crystallinity, comonomer distribution, and additive partitioning all shift under industrial quenching rates.

Chain entanglements also impose a threshold effect. Below the critical molecular weight for entanglements, which for polyethylene is generally cited between 3,000 g/mol and 4,000 g/mol, polymer chains do not form stable entanglements and migrate more readily as individual species. The extractable polymer fraction is therefore not limited to species below 1,000 g/mol; chains up to the entanglement threshold can be solubilised in strong simulants if the contact time is long and temperature is high. Above the entanglement threshold, self-diffusion of polymer molecules becomes constrained by reptation, and extracted free chains are rare in standard migration tests. This threshold is one reason that polymer with a low M_n near 10,000 g/mol may release more polymeric material than predicted by simple oligomer solubility arguments. For commercial blown film resins, M_n is usually well above this threshold, but the low-molecular-weight tail may still contain chains in the 3,000 g/mol to 10,000 g/mol range that are extracted under accelerated test conditions. Overall migration is not governed solely by diffusion; partition coefficients between polymer and simulant also depend on the amorphous free volume and polar comonomer content. In non-polar polyethylene, solubility of hydrophobic migrants in the polymer increases with chain length, so higher molecular weight polymer may actually retain processing aids more strongly, but the extraction time in standard tests is finite and may not reach equilibrium for large migrants. The standard overall migration method therefore reflects a kinetic snapshot, not a thermodynamic equilibrium, which enhances the practical importance of molecular weight-dependent diffusion. For a 50 µm monolayer film tested under 10 days at 40 °C in simulant D1, systems with diffusion coefficients below 10^-14 m²/s may remain below the 10 mg/dm² limit even with high total extractables, while fast-diffusing systems may exceed it early and then plateau. The diffusion front penetration depth scales with the square root of the diffusion coefficient and time; reducing diffusivity by one order of magnitude reduces penetration depth by approximately a factor of three.

Production-scale blown film evaluations on single-screw extruders with 24:1 to 30:1 L/D ratios and spiral mandrel dies of 100 mm to 400 mm diameter show that molecular weight changes are never isolable from processing variables. When a converter shifts from a melt flow rate of 4.0 g/10 min to 1.0 g/10 min at 190 °C under 2.16 kg load, the zero-shear viscosity rises approximately by a factor of 2 to 4 depending on molecular weight distribution and long-chain branching. Zero-shear viscosity scales with weight-average molecular weight to the 3.4 power, so a doubling of M_w above the critical entanglement molecular weight increases viscosity by nearly an order of magnitude. This results in higher die pressure, increased motor torque, and higher melt temperature at constant screw speed, which may force the operator to raise barrel temperatures or reduce throughput to avoid over-torque alarms. The increase in melt temperature, typically 10 °C to 20 °C at moderate screw speeds, can degrade heat-sensitive additives or promote chain scission in the low-molecular-weight tail, partially offsetting the migration benefit of the higher-molecular-weight base resin. Published data for exact die pressure and extractables responses on a given commercial line is limited because grade change trials are confidential, but directional effects are consistently reported in extrusion literature.

Operational boundaries for high-molecular-weight blown film grades include reduced maximum screw speed, narrower die gap tolerance, and increased susceptibility to melt fracture if the molecular weight distribution is too narrow. For LLDPE with a narrow molecular weight distribution, increasing M_w above 200,000 g/mol without a bimodal or long-chain-branched architecture can create sharkskin surface defects at shear rates above 500 s^-1 to 1,000 s^-1 at the die lip. A bimodal high-density polyethylene film grade may process at a melt temperature of 210 °C to 240 °C with a die gap of 1.2 mm to 2.0 mm, but the low-molecular-weight mode in the bimodal distribution is intentionally added for processability and may contribute to the extractable fraction if not properly separated by comonomer placement. Converters must also avoid combining high-molecular-weight LLDPE with amine-based antifog or antistatic additives above manufacturer-specified loadings because acidic processing residues from the polymerisation catalyst can react with amines to form amides and other extractable reaction products. At relative humidity above 60%, hygroscopic additive masterbatches require pre-drying at 80 °C for 2 h to prevent moisture-induced hydrolysis during extrusion; pure polyolefin pellets themselves do not require pre-drying, but surface moisture can affect bubble cooling stability. Bubble cooling and frost line height also interact with molecular weight. High-molecular-weight resins allow the bubble to be run at higher blow-up ratios without instability, which increases transverse orientation and can reduce permeability to migrants by aligning amorphous tie molecules in the film plane. However, orientation can also create microvoids at the crystalline–amorphous interface, which may increase the accessible free volume for simulant penetration. The net effect on overall migration is not universally beneficial; published data on orientation and migration in blown polyethylene is limited because industrial film samples vary in frost line height, die diameter, and internal bubble cooling settings. The practical result is that a converter must evaluate overall migration on the same film structure and processing conditions intended for production, not on a compression-moulded plaque or a film extruded at a different draw ratio.

When the low-molecular-weight tail dominates the extractable pool in blown polyethylene

The low-molecular-weight tail of a polymer is conventionally defined by the fraction with molecular weight below 1,000 g/mol or below 10,000 g/mol, depending on the analytical purpose. In polyethylene, oligomers in the 300 g/mol to 1,000 g/mol range are sufficiently mobile to be extracted by fatty food simulants and sufficiently non-volatile to contribute to overall migration residue. Gel permeation chromatography according to ISO 16014-1:2019 can quantify the area fraction below a defined molecular weight threshold, but the detection response factors for low-molecular-weight polyolefins are strongly dependent on refractive index increments and column resolution. When the low-molecular-weight tail increases from 1.0% to 3.5% of the total polymer mass, the gravimetric extractable contribution may increase disproportionately because the shortest chains are not only thermodynamically more soluble in simulants but also diffuse several orders of magnitude faster than the bulk polymer. In blown film, the low-molecular-weight species are concentrated in the amorphous interlamellar regions during crystallisation and are rejected from growing spherulites or lamellar stacks, so their local concentration at the crystal–amorphous interface can exceed the bulk concentration measured by GPC.

Quantitative extraction of low-molecular-weight species from polyethylene can be performed by Soxhlet extraction with n-hexane or xylene under reflux, followed by gravimetric determination or gas chromatography. For polyolefin food-contact films, 21 CFR 177.1520 specifies maximum extractable fractions in n-hexane for certain olefin polymer types, but the European overall migration framework uses food simulants rather than direct solvent extraction. A film intended for fatty food contact is tested in D2 or a substitute, which swells the polymer and extracts a different spectrum of oligomers than n-hexane. Because D2 and isooctane do not fully reproduce the interaction of real foods, the 10 mg/dm² limit is an operational compliance threshold rather than a toxicological exposure estimate. Molecular weight data from GPC are therefore best used as a screening tool; if the oligomer fraction below 1,000 g/mol exceeds 2% of total polymer mass, migration testing should be performed on the actual film rather than relying on generic grade data.

These comparative classes are not compliance predictions. The overall migration result for a specific film depends on thickness, additive package, coextruded layers, and simulant. However, the table illustrates that molecular weight alone cannot be disentangled from molecular weight distribution and chain architecture. A bimodal HDPE with a high overall M_w can contain a low-molecular-weight mode that contributes to extractables, while a metallocene LLDPE with a lower M_w but a very narrow oligomer fraction may show lower extractables in practice. Published data for exact oligomer extraction from specific commercial grades is limited because additive packages and catalyst residues are proprietary.

Regulatory compliance matrices and test method selection for 10 mg/dm² overall migration

The selection of migration test conditions is not a single fixed protocol but is determined by the foreseeable food-contact application, contact time, and temperature. Commission Regulation (EU) No 10/2011 assigns overall migration testing under Article 12 and Annex III food simulants. For aqueous, acidic, and alcoholic foods, simulants A, B, C, and D1 are used; for fatty foods, simulant D2 or the substitute 95% ethanol or isooctane may be used depending on the regulatory pathway and method availability. The total immersion method of EN 1186-2:2002 uses olive oil or alternative fatty simulant and requires gravimetric determination of extracted residue after evaporation of the simulant, with correction for volatiles and fibre loss. EN 1186-3:2002 describes total immersion for aqueous simulants, while EN 1186-14:2002 is used for substitute isooctane tests. The results are expressed in milligrams per square decimetre of food-contact surface area, and the legal limit is 10 mg/dm². For food-contact plastics, the analytical repeatability and reproducibility of these methods can be affected by polymer swelling, simulant absorption, and incomplete recovery of low-volatility oligomers, which is why the method specifies blank determinations and, for fatty simulants, a preliminary extraction step.

Interpreting these methods for blown film requires attention to surface-area-to-volume ratio. A thin film with high surface area and low thickness extracts proportionally more residue per unit mass than a thicker article, but the legal limit is expressed per surface area, so thinner films are not automatically penalised. However, the overall migration result for a 25 µm film may be dominated by surface additives such as slip agents and antiblock particles, whereas the result for a 100 µm film may include more bulk material but the extraction reaches similar surface-area normalised values if the migrants are uniformly distributed. The molecular weight effect is more pronounced in thicker films because diffusion distances are longer and the extracted quantity per unit area depends on the diffusion front penetration depth. Interlaboratory reproducibility of overall migration tests is often limited to ±20% relative standard deviation when fatty simulants are used, particularly for low migration values below 5 mg/dm². At levels near the 10 mg/dm² threshold, the result must be corrected for blank residue, water content, and simulant absorption. The use of isooctane as a substitute for olive oil can overestimate migration for polyolefins because isooctane is a stronger swelling solvent; the same molecular weight effect is therefore amplified in isooctane screening tests. If a film passes the 10 mg/dm² limit in isooctane, compliance with D2 is likely, but the inverse is not always true. The selection of time and temperature conditions in Annex III of 10/2011 must reflect the worst-case foreseeable contact, which for room-temperature storage of fatty foods is often represented by 10 days at 40 °C in D2, while hot-fill or microwave packaging may require 2 h at 70 °C or higher.

Thermal degradation and recycled material incorporation change the oligomer profile independent of virgin polymer molecular weight

In blown film production, thermal and oxidative degradation during extrusion can generate low-molecular-weight fragments that were not present in the virgin resin. Polyethylene processed at melt temperatures above 230 °C can undergo chain scission at weak links, particularly when oxygen is not excluded by nitrogen blanketing or vacuum venting. The degradation products include aldehydes, ketones, carboxylic acids, and alkanes, many of which are volatile or semi-volatile; the non-volatile fraction contributes to overall migration. A high-molecular-weight base resin does not eliminate this risk because the high melt temperatures and shear heating required to process high-viscosity material can accelerate thermo-oxidative chain scission. Post-industrial regrind and post-consumer recyclate often contain a broader oligomer distribution, residual printing inks, adhesives, and degradation products from previous heat histories. When the regrind fraction exceeds 20% to 30%, overall migration in fatty simulants may increase even if the virgin resin meets the 10 mg/dm² limit at 100% virgin level. Therefore, any change in recycle content, screw design, die temperature, or regrind drying protocol requires remeasurement of overall migration according to the applicable EN 1186 method. The polymer molecular weight of the virgin component remains important, but its effect is subordinate to the degradation history of the final melt stream.