Mathematical Diffusion Modelling of Non Intentionally Added Substances through Polyolefin Barrier Layers in Flexographic Food Packaging
Mathematical diffusion models quantify NIAS transfer through polyolefins, enabling rapid, low-cost compliance validation for flexographic food packaging.
Origin
Flexographic packaging lines running UV-curable and solvent-based ink sets generate chemical degradation products during high-speed drying and crosslinking. Retained ethyl acetate, isopropyl alcohol, and ethoxypropanol solvents trapped within ink layers act as aggressive plasticizers inside adjacent polyolefin films. Energy-curable flexographic inks rely on photoinitiators like polymeric hydroxyketones, aminoketones, and phosphine oxides.
Unreacted photoinitiators undergo side-chain cleavage under ambient light or residual UV exposure. This photo-oxidation yields low-molecular-weight breakdown products that lack intentional functional purpose in the finished structure.
Polyurethane laminating adhesives joining polyolefin films to outer substrates introduce secondary non-intentionally added substances. Residual aromatic monomeric isocyanates undergo hydrolysis when exposed to atmospheric humidity during room-temperature curing. The resulting aromatic amines migrate rapidly through low-density polyolefin networks.
Thermal degradation of ink resins, wax additives, and slip agents during hot-nip lamination further populates the migrant spectrum with cyclic aliphatic oligomers, oxidized fatty acid amides, and substituted alkylphenols.
Retained solvents lower polymer glass transition temperatures. Film swelling increases native free volume.
- Photoinitiator Cleavage Products generate benzaldehyde, substituted benzoic acids, and alkylaminobenzenes through Norrish Type I and Type II photochemical reactions when curing lamps deliver uneven spectral irradiance across wide-web presses.
- Adhesive Cyclic Oligomers form via unreacted side-ring condensations during two-component polyurethane curing, producing lipophilic species between 300 Daltons and 700 Daltons that bypass classical primary barrier layers.
- Polyolefin Degradation Homologues arise from thermal-oxidative scission during extrusion and corona surface pretreatment, yielding linear aldehydes, ketones, and carboxylic acids within the first 15 micrometers of the inner sealant surface.
- Solvent Breakdown Derivates occur when ester solvents undergo acid-catalyzed hydrolysis in the fountain solution, leaving free acetic acid and trace aliphatic alcohols inside the printed ink matrix.
Standard testing under European Standard EN 13130 shows residual isopropyl acetate concentrations above 5 milligrams per square meter increase migrant diffusion velocity across low-density polyethylene by 40 percent at 40 degrees Celsius.
Incomplete solvent removal during flexographic drying ovens creates localized solvent reservoirs within multi-layer structures. These retained volatiles migrate through polyolefin layers into dry fatty food simulants. Converting lines operating at line speeds above 300 meters per minute face elevated risks of solvent retention when drying capacity fails to keep pace with ink laydown volumes.
Failure to quantify baseline contaminant concentrations prior to lamination guarantees incorrect risk assumptions in downstream migration assessments. Undetected primary aromatic amines or photoinitiator fragments breach absolute migration limits upon reaching the packaging surface. Compliance failures triggered by unmodeled breakdown products force expensive product recalls, immediate stock quarantines, and complete loss of brand equity across retail supply networks.
Flux
Mass transfer through polyolefin barrier layers operates under Fickian diffusion rules when the permeant does not alter the physical matrix structure. Mass transport inside solid polyolefin films follows Fick’s second law for one-dimensional non-steady-state conditions. Differential change in migrant concentration over time links directly to the spatial concentration gradient across thickness x.
The mathematical form of Fickian transport inside a planar film appears as:
∂C/∂t = D · (∂²C/∂x²)
Concentration C depends on spatial depth x and elapsed time t. Variable D represents the diffusion coefficient of the migrant in square centimeters per second. Boundary conditions at polyolefin surfaces dictate real-world transfer rates.
The outer boundary at x = 0 connects to the printed ink or adhesive layer containing initial migrant concentration C_0. The inner boundary at x = L meets the food phase or food simulant.
Estimating diffusion coefficient D requires empirical correlation models calibrated against experimental migration data. The modified Piringer model calculates diffusion coefficients based on migrant molecular weight M_r and absolute temperature T in Kelvin. Matrix-dependent parameter A’_p describes the inherent polymer chain mobility.
Parameter τ represents the temperature dependence constant. Equation formulation follows:
D = D_0 · exp(A’_p – 0.1351 · M_r^(2/3) + 0.003 · M_r – 10454 / T)
Parameter A’_p values vary systematically between polyolefin classifications. High-density polyethylene features tight chain packing and high crystallinity, resulting in low A’_p values. Low-density polyethylene contains branched chains, producing high A’_p values and faster migrant transport.
| Polyolefin Barrier Grade | Density Range (g/cm³) | Crystallinity (%) | Matrix Parameter A’_p | Estimated D for 300 Da Migrant at 40°C (cm²/s) |
|---|---|---|---|---|
| Low-Density Polyethylene (LDPE) | 0.915 – 0.925 | 40 – 50 | 11.5 | 1.2 × 10⁻¹⁰ |
| Linear Low-Density Polyethylene (LLDPE) | 0.918 – 0.935 | 50 – 60 | 10.5 | 3.5 × 10⁻¹¹ |
| High-Density Polyethylene (HDPE) | 0.940 – 0.965 | 70 – 80 | 7.0 | 1.8 × 10⁻¹³ |
| Cast Polypropylene (CPP) | 0.895 – 0.905 | 50 – 55 | 8.5 | 2.1 × 10⁻¹² |
| Biaxially Oriented Polypropylene (BOPP) | 0.900 – 0.910 | 60 – 70 | 5.5 | 4.2 × 10⁻¹⁴ |
Temperature alters diffusion coefficients through Arrhenius relationships. Activation energy E_a governs rate changes during hot-fill procedures, thermal sterilization, or microwave heating. Energy values range from 30 kilojoules per mole to 100 kilojoules per mole depending on molecular volume and polymer chain rigidity.
Regulation EU 10/2011 mandates worst-case calculation assumptions where numerical modelling substitutes for analytical testing, setting migrant initial concentrations to 100 percent of raw ink mass fraction.
Diffusion inside multi-layer structures containing printed ink, polyurethane adhesive, and polyolefin barrier layers demands numerical finite difference schemes. Crank-Nicolson algorithms solve transient transport across material interfaces. Discontinuous concentration profiles occur across layer boundaries due to unequal solubility capabilities.
Concentration step changes depend directly on thermodynamic partition coefficients between adjacent materials.
Does localized temperature variation during ambient flexographic printing drying cycles introduce asymmetric diffusion profiles that invalidate isotropic Piringer parameter estimations?
Film
Do Orientation Processing Steps Alter Migrant Breakthrough Timelines?
Biaxial orientation dramatically alters polyolefin barrier performance. Stretching polypropylene or polyethylene films above their glass transition temperature aligns amorphous polymer chains parallel to the film surface. Mechanical orientation increases crystallite packing density and eliminates large amorphous pathways through which migrant molecules travel.
Biaxially oriented polypropylene cuts migrant diffusion rates by two orders of magnitude compared to unoriented cast polypropylene of identical caliper.
Mechanical stretching creates tortuous diffusion paths around impenetrable crystalline domains. The obstruction factor scales directly with volumetric crystalline fraction and crystallite aspect ratio. Higher orientation ratios yield longer breakthrough lag times for non-intentionally added substances originating from outer ink layers.
Crystalline Microstructure and Obstruction Factors
Amorphous regions inside polyolefins carry all migrant flux. Solid crystalline spherulites block migrant passage completely. Volumetric crystallinity dictates the effective cross-sectional area open to mass transport.
Fractional reduction in diffusion velocity correlates with the two-phase microstructural model.
D_effective = D_amorphous · (φ_amorphous / τ_tortuosity)
Tortuosity factor τ_tortuosity accounts for path length elongation around crystalline plates. Branching in linear low-density polyethylene disrupts crystalline domain packing, lowering tortuosity and raising permeation risks for low-molecular-weight ink additives.
- Determine polyolefin film density, layer caliper, and crystalline mass fraction using differential scanning calorimetry under ASTM D3418 conditioning.
- Calculate pure amorphous diffusion coefficient for target non-intentionally added substance using molecular weight and Piringer parameter constants.
- Apply tortuosity correction factor derived from crystallite aspect ratio measurements to determine effective film diffusion coefficient.
- Compute lag time breakthrough value t_lag using planar film geometry equation t_lag = L² / (6 · D_effective).
- Evaluate cumulative migrant flux reaching inner food-contact boundary across total intended package shelf life.
Solvent sorption during flexographic printing swell polyolefin amorphous zones. Sorbed esters or ketones increase polymer chain mobility, temporarily lowering local glass transition values. Swelling destabilizes barrier performance during early storage periods before volatile solvents evaporate into ambient surroundings.
A thick unoriented sealant layer transmits volatile migrants faster than a thin biaxially oriented film layer.
Partition
Thermodynamic equilibrium between packaging layers and food phases governs absolute mass transfer capacity. Partition coefficient K_P,F measures the concentration ratio of a migrant species distributed between polymer matrix P and food simulant F at equilibrium:
K_P,F = C_P,∞ / C_F,∞
A high partition coefficient indicates strong chemical retention inside the polyolefin layer. A low partition coefficient drives rapid mass transfer into adjacent food phases. Lipophilic non-intentionally added substances, including photoinitiators like Irgacure 907 and lubricant additives like erucamide, exhibit high affinity for dry fatty foods and olive oil simulants.
These compounds display low partition coefficients relative to polyolefin films, accelerating transfer into fatty foods.
Hansen solubility parameters split total cohesive energy density into dispersion forces, polar interactions, and hydrogen bonding components. Matching solubility parameters between migrant molecules and polyolefin matrix structures predicts equilibrium distribution behavior. Non-polar migrants match polyolefin parameter profiles, raising retention within the packaging film.
- Equilibrium Mass Balance Constraints dictate that total migrant quantity within closed systems remains constant, forcing exact distribution balances between ink, adhesive, polyolefin layer, and food contact phase.
- Henry’s Law Solubilities govern low-concentration behavior where migrant activity coefficient remains constant inside amorphous polyolefin domains.
- Simulant Surface Affinity shifts partition boundaries when aqueous, ethanol, or vegetable oil simulants contact inner polyolefin surfaces, altering phase equilibrium constants.
- Interface Discontinuities generate distinct concentration jumps across adjacent polymer layers according to ratio K_1,2 = C_1 / C_2.
| Migrant Chemical Species | Molecular Mass (Da) | Log Kow (Octanol-Water) | Polyolefin Matrix | Simulant Partition Coefficient K_P,F |
|---|---|---|---|---|
| 2,4-Diethylthioxanthone (DETX) | 268.37 | 5.4 | LDPE | 1.2 (Food Simulant D2: Vegetable Oil) |
| Benzophenone | 182.22 | 3.1 | HDPE | 15.0 (Food Simulant A: 10% Ethanol) |
| Bisphenol A Diglycidyl Ether (BADGE) | 340.41 | 3.8 | LLDPE | 0.8 (Food Simulant D1: 50% Ethanol) |
| 2,2-Dimethoxy-2-phenylacetophenone | 256.30 | 3.2 | BOPP | 4.5 (Food Simulant E: Poly-Tenax) |
| Tripropylene Glycol Diacrylate (TPGDA) | 300.35 | 2.8 | CPP | 8.1 (Food Simulant B: 3% Acetic Acid) |
Simulant selection directly impacts numerical model validity. Isooctane and 95 percent ethanol act as aggressive swelling agents for polyolefins, artificially lowering partition coefficients during physical validation trials. Mathematical models using conservative default partition assumptions (K_P,F = 1) overpredict migration into aqueous foods, introducing unnecessary commercial friction into package design approvals.
Dynamic mechanical analysis confirms polyolefin swelling in fatty food simulants increases matrix free volume, driving partition coefficients down by up to two orders of magnitude.
Partition boundaries inside multi-layer flexographic structures demand explicit specification in contract technical agreements. Material supply specifications defining acceptable migrant limits must include partition coefficient boundary conditions matched strictly to intended food contact types. Specifying functional barrier capabilities without defining simulant-specific partition behavior renders chemical compliance guarantees unenforceable in commercial disputes.
Dossier
Creating regulatory dossiers for flexographic food packaging relies heavily on validated mathematical diffusion models. Physical migration testing via gas chromatography coupled with mass spectrometry (GC-MS) or liquid chromatography with high-resolution mass spectrometry (LC-HRMS) carries substantial laboratory costs and long turnaround times. Diffusion calculations deliver instantaneous worst-case exposure estimates, allowing packaging engineers to screen dozens of ink, adhesive, and polyolefin film combinations during early product development.
Numerical simulation tools model concentration dynamics across complex shelf-life scenarios. Storage calculations span multi-year ambient profiles, refrigerated storage, or elevated thermal cycles encountered during processing. Software platforms apply finite element analysis to compute total migrant mass transferred per square decimeter of packaging surface.
| Evaluation Metric | Mathematical Diffusion Modelling | Analytical Testing (GC-MS / LC-MS Screening) |
|---|---|---|
| Turnaround Time per Structure | 1 to 4 Hours | 15 to 30 Business Days |
| Direct Unit Cost per Docket | €150 to €300 | €2,500 to €6,000 |
| Screening Capacity | High (Iterative chemical library testing) | Low (Limited by physical test cell availability) |
| Detection Threshold Sensitivity | Theoretical (Zero lower boundary) | 0.01 mg/kg (Analytical limit of quantification) |
| Regulatory Acceptance Scope | Accepted for positive compliance proof (EU/FDA) | Mandatory for contested migration breaches |
Modelling approaches assume uniform migrant distribution inside initial ink or adhesive layers. Unrealistic safety factors in conservative mathematical algorithms sometimes generate false-positive compliance breaches. When modelled concentrations exceed specific migration limits (SML), packaging engineers must transition to physical testing using standardized food simulants under specific time-temperature exposures.
Regulatory compliance dossiers relying on diffusion calculations remain valid only when raw ink composition data accounts for 100 percent of volatile and non-volatile raw material fractions.
A common supplier defense claims that non-intentionally added substances cannot be modelled due to unknown chemical structures or missing molecular weights. Converters bypass this limitation by applying high-resolution screening to identify major peak masses, assigning conservative molecular weight cutoffs, and running diffusion calculations using worst-case Piringer A’_p parameters to establish definitive upper migration thresholds.