Additive Migration in Plastics: How It Works, How to Model It and 7 Ways to Reduce It

Additive migration is the movement of an additive through a polymer by diffusion, to the surface of the part or into whatever the part touches, and it is the reason a film gets its slip, a gasket loses its plasticizer and a food-contact compound has to prove it stays under a legal limit. Every additive migrates to some degree, so the formulator's question is never whether it moves, but how fast and how far.

Migration has two faces. Slip agents, migrating antistatic agents and antifog additives are designed to migrate: they do nothing until they reach the surface. Plasticizer exudation, stabilizer chalking, taste and odour transfer and a failed specific migration limit are the same transport process working against the formulation instead of for it.

This page defines the vocabulary that separates migration from blooming, exudation, volatilisation, extraction and leaching, explains why additives migrate through diffusion and solubility in the amorphous phase, names the six variables that set the rate, and shows how the Piringer model in EUR 27529 turns the polymer, the temperature and the molecular mass of a migrant into an upper-bound diffusion coefficient, with a worked value for seven common additives in LDPE. It then covers which additive families migrate by design and which migrate as a defect, the seven levers that reduce migration, how migration is measured, and which limits under Regulation (EU) No 10/2011 an article has to meet.

The table below sets out the 5 physical routes an additive can leave a plastic part, plus the one non-physical route.

A sixth route is not physical at all: an additive can simply be consumed by the reaction it was added to stop. That is chemical depletion, covered on polymer degradation.

What Is Additive Migration?#

Additive migration is the diffusion-driven movement of a substance that was intentionally added to a plastic out of the polymer matrix, either to the surface of the article or into a phase it is in contact with. The definition follows Regulation (EU) No 10/2011 Article 3(7): an additive is a substance intentionally added to achieve a physical or chemical effect during processing or in the final material or article, and intended to be present in it. Migration is not the same thing as the presence of a non-intentionally added substance (NIAS); a NIAS is an impurity, a breakdown product or a reaction by-product rather than the additive itself, and it follows the same diffusion physics without being part of the original formulation.

Migration is a property of the additive-polymer pair, not of the polymer alone, so it has to be read family by family across all 43 families of plastic additives.

Migration, Blooming, Exudation, Extraction and Leaching: What Is the Difference?#

The five terms describe the same transport step with five different destinations: migration ends in a contacting phase, blooming and exudation end on the surface, volatilisation ends in the air, extraction and leaching end in a liquid, and plate-out ends on the tool. Table T1 above sets out the driving force and a typical example for each route, and that vocabulary is used consistently for the rest of this page.

Blooming happens when an additive's concentration exceeds its solubility in the polymer at the use temperature, at a rate set by its diffusion coefficient, according to Nouman, Saunier, Jubeli and Yagoubi in Polymer Degradation and Stability (2017). Exudation is the same mechanism for an additive that behaves as a liquid at the surface rather than a solid. Bleeding describes the transfer of a migrated additive onto a contacting material, such as a printed film picking up slip agent from the web stacked underneath it. Leaching is the release of an additive into a liquid, whether a food simulant, groundwater or the wider environment; the consumer-facing definition and mechanism of leaching are covered as a separate term.

Which Additives Migrate and Which Stay in the Polymer?#

Low-molecular-mass, weakly compatible additives migrate readily, while high-molecular-mass, oligomeric and polymer-bound additives stay in the matrix: molecular mass is the single strongest predictor a formulator controls. Primary antioxidants are generally 300-1,000 g/mol, and within that range the effect is direct: low-molecular-weight grades such as BHT, at 220.35 g/mol, lose more mass to volatilisation and migration than heavier grades in the same family.

Additives that migrate readily include:

• Plasticizers, which are not chemically bound to the polymer
• Fatty-acid amide slip agents, such as erucamide
• Migrating antistatic agents, such as glycerol monostearate
• Antifog surfactants
• Low-molecular-weight phenolic antioxidants, such as BHT
• Monomeric HALS, such as Tinuvin 770
• Low-molecular-weight UV absorbers

Most polyolefin antioxidant packages have moved to a heavier replacement for exactly this reason: Irganox 1010, at 1,177.6 g/mol, has replaced BHT in most polyolefins because the larger molecule diffuses and volatilises far more slowly.

Additives that stay in the polymer include:

• Oligomeric HALS, such as Chimassorb 944 and Tinuvin 622
• High-molecular-weight hindered phenols, such as Irganox 1010
• Polymeric plasticizers
• Reactive and polymer-bound additives
• Mineral fillers and reinforcements
• Pigments, which are particles rather than dissolved solutes

Why Do Additives Migrate? Diffusion, Solubility and the Amorphous Phase#

An additive migrates for two reasons at once: thermodynamics says it would rather not be where it is, and kinetics says the amorphous parts of the polymer let it move. Solubility sets whether the additive wants to leave the polymer at all; diffusion, governed by Fick's laws, sets how fast it can. A third factor, crystallinity, decides how much of the polymer volume is even available for that movement, because additives dissolve and travel only in the amorphous regions of a semi-crystalline polymer. Together the three explain why the same migrant can sit stable in one polymer and be gone within days from another.

Fick's Laws: Why Early Migration Follows the Square Root of Time#

Adolf Fick set out the two laws of diffusion in 1855, and the second one gives the result every formulator needs: early migration grows with the square root of time, not linearly with it. Fick's first law states that the diffusive flux is proportional to the concentration gradient. For a semi-infinite medium, the concentration profile as a function of depth x and time t is:

c(x,t) = c0 x erfc(x / (2 x sqrt(D x t)))

where D is the diffusion coefficient in square metres per second and erfc is the complementary error function. The diffusion length, 2 sqrt(D t), sets how far the migration front has travelled at time t, and it is this square-root term that governs early-stage migration from a part thick enough to behave as a semi-infinite source.

The practical consequence follows directly from the square root: doubling the contact time does not double migration, it multiplies it by about 1.41, and quadrupling the time doubles it. That single relationship is the most useful sentence on this page for a shelf-life argument, because it means a compound that passes a 10-day migration test is not automatically safe at 40 days without repeating the calculation.

Solubility: Why the Additive Leaves the Polymer at All#

Solubility decides whether the additive wants to leave: an additive dissolved above its solubility limit at the service temperature is supersaturated, and the excess moves out. Solubility of an additive in a polymer is higher in the melt than in the solid state, so a part compounded at processing temperature and then cooled can end up holding more additive than the solid polymer can actually dissolve. Nouman, Saunier, Jubeli and Yagoubi, writing in Polymer Degradation and Stability (2017), describe this supersaturation as the mechanism behind blooming: the excess additive diffuses toward the surface until the bulk concentration falls back to the solubility limit.

Solubility in a fat-like medium also predicts extraction behaviour. Songwon publishes SONGNOX 1010 at below 0.05 g per 100 g of squalane and below 0.1 g per 100 g of n-hexane and ethanol at 25 °C, while SONGNOX 1680 reaches 10.0 g per 100 g of n-hexane at the same temperature, a difference specific to those two named grades rather than to the antioxidant family as a whole.

Why Crystallinity and Free Volume Set the Speed#

Additives dissolve and travel only in the amorphous regions of a semi-crystalline polymer, which is why the same migrant moves at very different speeds through LDPE, HDPE and polypropylene. A more crystalline matrix concentrates a given additive load into a smaller amorphous volume and changes how that additive partitions and moves through the part. The Piringer model encodes this difference directly in its polymer constants: LDPE carries A_P'* of 11.5 with a temperature-activation term tau of 0 K, HDPE carries 14.5 with tau of 1,577 K, and PP homopolymer and random copolymer carry 13.1 with tau of 1,577 K. Those constants translate into an activation energy of about 86.92 kJ/mol for LDPE, close to the literature mean of 87 kJ/mol, against about 100 kJ/mol for HDPE and PET. The constants differ by polymer rather than by a single crystallinity rule; Table T3 below gives the full set.

What Controls the Rate? 6 Migration Drivers#

Six variables control how much additive migrates: temperature, contact time, the surface-to-volume ratio, the polymer, the molecular properties of the additive, and what the plastic is touching. These are the same six factors the EU food-contact literature uses to frame a migration assessment, and the table below sets out the direction of each effect and what, if anything, a formulator can do about it.

Only two of the six drivers sit inside the formulator's control at the formulation stage: the polymer and the additive. Temperature, contact time, surface-to-volume ratio and the contacting medium are set by the application, not by the recipe. That asymmetry is exactly why every lever in the 7-levers section below is a molecular choice rather than a processing instruction.

How Is Additive Migration Calculated? The Piringer Model#

Food-contact migration is calculated in the EU with the Piringer model, published by the European Commission's Joint Research Centre as EUR 27529, which estimates an upper-bound diffusion coefficient from nothing more than the polymer, the temperature and the molecular mass of the migrant. Brandsch, Dequatre, Mercea, Milana, Stoermer, Trier, Vitrac, Schaefer, Simoneau and Hoekstra published the guideline, Practical Guidelines on the Application of Migration Modelling for the Estimation of Specific Migration, in 2015, building on the diffusion model that Mercea and Piringer set out in 1998 and that Brandsch and colleagues extended in 2002. The model never measures a real migration value; it produces a deliberately conservative, upper-bound estimate that regulators accept as a screening tool.

The Upper-Bound Diffusion Coefficient D_P*#

The model estimates an upper-bound diffusion coefficient D_P in square metres per second from three inputs: two polymer constants, the temperature in kelvin and the relative molecular mass of the migrant.* The equation is:

D_P* = exp(A_P* - 0.1351 x M_r^(2/3) + 0.003 x M_r - 10454 / T)
A_P*  = A_P'* - tau / T

A_P'* and tau are the polymer's own constants, read from Table T3 below. M_r is the relative molecular mass of the migrant in g/mol. T is the absolute temperature in kelvin. D_P* comes out in square metres per second, and it is always an upper bound, never a measured value.

Take LDPE at 40 °C (313.15 K), where A_P'* is 11.5 and tau is 0. For BHT, with a relative molecular mass of 220.35 g/mol, the equation gives an upper-bound D of about 4.4 x 10⁻¹² m²/s. For Irganox 1010, at 1,177.6 g/mol, it gives about 3.1 x 10⁻¹⁵ m²/s: roughly 1,400 times slower, from one substitution. Put into the diffusion length 2 sqrt(D t), that is a migration front about 3.9 mm deep after 10 days for BHT against about 0.10 mm for Irganox 1010.

The values below were calculated by PlasticAdditives.net from the JRC equation and constants. They are upper-bound screening estimates, not measurements.

Calculated by PlasticAdditives.net with the JRC EUR 27529 equation and the LDPE constants A_P' = 11.5 and tau = 0 K, valid below 80 °C for migrants of 30-2,000 g/mol. Relative molecular masses from our substance database. These are upper-bound screening values; real diffusion coefficients are lower.*

Polymer Constants: A_P'*, tau and the Validity Limits#

Each polymer carries two constants in the model: A_P'*, which sets the size of the diffusion coefficient, and tau, an activation-temperature increment in kelvin that makes the coefficient more temperature-sensitive. The table below gives the full set published in JRC EUR 27529.

PET below 70 °C uses A_P' 6.4 for migrants under about 50 g/mol and 3.1 above it. Read the JRC tables before applying a value.*

Three limits apply before any row above is used. Every row holds only inside its stated temperature and molecular-mass window. The plasticised-PVC and polyamide rows are informative only in the JRC guideline, not a compliance screening basis. PET below 70 °C carries two different A_P'* values depending on whether the migrant sits above or below about 50 g/mol, so the JRC tables have to be read before a PET value is applied to a real substance.

The Partition Coefficient K_P,F#

The diffusion coefficient says how fast the additive travels inside the plastic; the partition coefficient says how willingly it crosses into the food. K_P,F = 1 is the worst-case value, used when the migrant is well soluble in the food. K_P,F = 1000 is used when the migrant is only sparingly soluble in the food, for example a lipophilic additive migrating into an aqueous food. Both are defined cases inside the JRC guideline, not free parameters a formulator can tune to improve a result.

When a Model Replaces a Test Under EU 10/2011#

Regulation (EU) No 10/2011 lets a model stand in for a migration test in one direction only: a calculated result may be used to show compliance, because the model over-estimates, but a calculated failure has to be confirmed by a real test under Article 18(3). Annex V chapter 2.2.3 sets that condition explicitly: modelling is accepted as a screening tool provided it over-estimates real migration. Modelling first entered EU law through Directive 2001/62/EC, which amended Directive 90/128/EEC, and the approach was validated in the EU project SMT-CT98-7513 and checked against experimental migration data at 95 % confidence by Begley and colleagues in Food Additives and Contaminants (2005).

Screen a formulation with the additive migration estimator before committing a compound to a full compliance-test programme; the tool applies the same equation and constants described above to a specific part geometry, temperature and contact time.

Which Additives Migrate? Migration Behaviour by Family#

Additive families fall into three migration classes: additives that only work once they have migrated, additives whose migration is a slow loss of performance, and additives that are physically incapable of migrating. The table below sets out where each of the nine families with meaningful migration behaviour falls, the typical timescale involved and what that migration causes.

Each family's own migration behaviour is covered in depth on its hub. Timescales are typical values for the polymer and dosage named in the source, not specifications.

The split matters for formulation strategy: a family migrating by design is tuned for speed and dose, a family migrating as a side effect is tuned to slow the migration down, and a family that cannot migrate is never the cause of a migration complaint.

Migration by Design: Slip, Antistatic and Antifog Additives#

Three additive families do nothing at all until they have migrated: slip agents, migrating antistatic agents and antifog additives all have to reach the surface before they work. Slip agents such as erucamide are dosed at 0.05-0.12 wt% in LDPE and LLDPE film; most of the coefficient-of-friction drop happens within 24-48 hours, and the value settles to its final level after 7-10 days. A shorter fatty-acid amide chain blooms faster but is less thermally stable, and the loading has to be reduced as film gauge increases to reach the same target, typically a coefficient of friction of about 0.2 on polyolefin film. Erucamide forms a surface bilayer about 4 nm thick on a polypropylene fibre, and reaching the torque-reduction target on HDPE closures needs about 15.7 µg/cm² of erucamide against about 1.7 µg/cm² of behenamide, roughly a ninefold difference between two closely related fatty-acid amides.

Migrating antistatic agents need about 2 days to build a monolayer in LDPE and LLDPE and reach an antistatic surface resistivity of 10⁹ to 10¹³ ohm. Synthetic silica antiblock particles in the same formulation adsorb erucamide and antistats and retard their migration, which is why an antiblock package and a slip package have to be dosed together rather than independently.

Migration by design carries five practical consequences:

• Conditioning time is a specification, not a delay
• Coefficient of friction is measured after a stated time, not immediately after extrusion
• Slip loading has to be re-set whenever the film gauge changes
• An antiblock or filler in the same formulation competes for the migrating additive
• A migrating antistat has a finite service life, because the reservoir in the bulk eventually depletes

When Migration Is a Defect: What It Costs a Formulation#

When migration is not part of the design, it shows up as five different failures: lost performance, a spoiled surface, a bond that does not hold, a compliance breach and a problem handed to the next processor. Plasticizer loss is the clearest performance failure: because plasticizers are not chemically bound to the polymer, their migration out of a flexible PVC part causes loss of flexibility, embrittlement and cracking. Antioxidant and HALS blooming leaves a chalky surface, and flame-retardant systems can whiten in the same way; an AlPi/MPP flame-retardant system in PA66 showed a white frost after ageing at 85 °C and 85 % relative humidity. Solvent dyes can migrate in polyolefins above the polymer's glass transition temperature, which is the reason solvent dyes are not used to colour PE and PP.

An additive that has migrated to the surface changes what can be printed, coated or welded onto it, so a migration problem in one process step becomes an adhesion failure in the next one. And an additive that migrates above its specific migration limit makes the finished article non-compliant, regardless of how well it performed mechanically.

The five failure modes in a migrating formulation are:

• Lost performance, such as plasticizer-driven embrittlement
• Spoiled appearance, such as chalking or flame-retardant frost
• Poor adhesion for print, coating or welding
• A compliance breach against the specific or overall migration limit
• A defect handed to the next processor in the supply chain

Troubleshooting a suspected migration defect starts from the additive-related defects reference, which maps each surface or performance symptom back to its formulation cause.

How Do You Reduce Additive Migration? 7 Levers#

Reduce additive migration with 7 levers: raise the molecular mass, switch to an oligomeric grade, match solubility to the medium, stop over-dosing, choose a slower host polymer, add a barrier layer, and screen with the model before you compound.

1. Raise the molecular mass. Irganox 1010 (1,177.6 g/mol) in place of BHT (220.35 g/mol) cuts the calculated upper-bound diffusion coefficient in LDPE at 40 °C by a factor of roughly 1,400.
2. Switch to an oligomeric or polymeric grade. Oligomeric hindered amine light stabilizers (HALS) such as Chimassorb 944 (Mn 2,000-3,100) and Tinuvin 622 (Mn 3,100-4,000) replace monomeric Tinuvin 770 (480.7 g/mol), which our source library records as having higher mobility and volatility than the oligomeric grades. Polymeric plasticizers, built from alternating C4-C10 dibasic acids and C2-C5 glycols and end-capped, are rated good to excellent for organic extraction and volatility resistance, and permanence rises with viscosity.
3. Match solubility to the contacting medium. Songwon publishes SONGNOX 1010 at below 0.05 g per 100 g of squalane, while SONGNOX 1680 reaches 10.0 g per 100 g of n-hexane: the same antioxidant chemistry behaves very differently against a fatty food, depending on the grade.
4. Stop over-dosing. Blooming starts when the additive exceeds its solubility at the service temperature, so a stabilizer added "just in case" is the most common self-inflicted bloom in the industry.
5. Choose the slower host. The Piringer constants put the same migrant at very different speeds in LDPE, HDPE, PP and PET, so the host polymer is itself a migration lever, not only a mechanical choice.
6. Put a barrier in the way. A coextruded skin layer or a functional barrier changes the distance the migrant has to cross before it reaches the contacting phase; the EU sets a functional-barrier limit of 0.01 mg/kg.
7. Screen before you compound. Run the Piringer screening with the real geometry, temperature and contact time, and reserve verification testing for the formulations the screening cannot clear.

For the full framework behind these choices, see how to select plastic additives.

How Is Additive Migration Measured?#

Additive migration is measured in three different ways: by what comes out of the plastic into a simulant, by what is left in the plastic after ageing, and by what has arrived on the surface. The table below names the standard behind each measurement and what it actually captures.

No single standard covers the whole picture, which is why a migration study for a food-contact compound usually runs an overall migration test alongside a specific migration test for each substance the formulation might release. The full set of simulants, exposure conditions and durations used across these standards belongs to migration testing, one level deeper than this page. The complete standards library for every additive-dependent property, not only migration, sits on testing plastic additives.

Which Migration Limits Apply?#

Migration into food is limited in four ways at once: an overall limit on everything that comes out, a specific limit for each listed substance, separate limits for migrating metals, and a limit on what may pass a functional barrier. Under EU 10/2011, the overall migration limit (OML) is 10 mg/dm², or 60 mg/kg of simulant for articles intended for infants and young children, under Article 12; the generic specific migration limit is 60 mg/kg, and a substance behind a genuine functional barrier is limited to 0.01 mg/kg.

Named examples show how the specific migration limits (SML) actually run: BHT is FCM 315 with an SML of 3 mg/kg, Irganox 1076 is FCM 433 with 6 mg/kg, Irganox MD 1024 is FCM 675 with 15 mg/kg, Chimassorb 944 is FCM 740 with 3 mg/kg and Tinuvin 622 is FCM 716 with 30 mg/kg, while Irganox 1010 (FCM 496) and Irgafos 168 (FCM 671) carry no SML and are controlled instead by the overall migration limit. Since Regulation (EU) 2023/1442, the phthalate SMLs are DEHP 0.6 mg/kg, DBP 0.12 mg/kg and BBP 6 mg/kg, with DINP and DIDP together at 1.8 mg/kg under group restriction 26.

Metals listed in Annex II carry their own limits: aluminium 1 mg/kg, barium 1 mg/kg, antimony 0.04 mg/kg, cobalt 0.05 mg/kg, copper 5 mg/kg and zinc 5 mg/kg of food. US food-contact plastics follow a different route entirely: FDA food contact rules set n-hexane extractables limits under 21 CFR 177.1520 rather than a per-substance migration limit.

Does Additive Migration From Plastics Affect Health and the Environment?#

Additive migration matters outside the factory in two places: what reaches food from packaging and medical devices, and what reaches the environment from plastic in use and after disposal. The specific and overall migration limits described above are the primary control mechanism for the food-contact case; for the environmental case, Hahladakis and colleagues at the University of Leeds record in the Journal of Hazardous Materials (2018) that additives can also be released during recycling and recovery and from products made from recyclate. The remainder of this page moves from formulation and physics into that outer reading, starting from the plastic formulation practice that sets the additive load in the first place.

How Much Migrates Into Food?#

Measured migration into food simulants ranges from undetectable to biologically active, depending entirely on the additive and the conditions. Zimmermann and colleagues tested 24 plastic products in water for 10 days at 40 °C and found that every migrate tested was baseline-toxic in an in vitro assay (2021). At the other end of the range, EFSA assessed an oligomeric 2-aminobenzamide acetaldehyde scavenger used in PET at 650 mg/kg and found migration into 20 % ethanol of just 0.0038 mg/kg, an example of a modern low-migration additive design working as intended.

The literature on chemicals migrating from plastic food packaging reviews these studies in depth; for the applications this page covers in a formulation context, see additives for food packaging.

Do Plasticizers Leach Out of Plastic Over Time?#

Yes: plasticizers are not chemically bound to PVC, so they move to the surface and into fatty foods, oils and solvents over time, which is why the EU limits DEHP migration into food to 0.6 mg/kg. Medical PVC carries 26.7-48.7 wt% of plasticizer depending on the formulation, according to Bernard and colleagues in PLoS One (2018), and DEHP specifically has a REACH Annex XIV latest application date of 1 January 2029 and a sunset date of 1 July 2030 under Regulation (EU) 2023/2482. The family-specific numbers behind this behaviour belong to plasticizer migration, one level deeper than this page; the device-level picture is on additives for medical plastics.

What Causes Blooming in Plastics?#

Blooming is caused by supersaturation: solubility is higher in the melt, so a cooled part holds more additive than it can dissolve, and the excess diffuses to the surface at a speed set by its diffusion coefficient. Nouman, Saunier, Jubeli and Yagoubi describe this mechanism in Polymer Degradation and Stability (2017). The levers that slow it down, molecular mass, solubility matching and dosage discipline, are the same levers described above for migration generally, because blooming and migration into a contacting phase are the same transport process ending at two different destinations.

Migration and Legacy Additives in Recycled Plastic#

Recycling carries additives forward: every stabilizer, plasticizer and compatibilizer in the input becomes a legacy additive in the output, and it migrates from the new article exactly as it did from the old one. Compatibilizers added during one recycling loop become legacy additives in the next processing stream, and Hahladakis and colleagues (2018) record that additives can be released during recycling and recovery and from products made from recyclate. A Swiss survey of PVC floorings sampled in 2021-22 found restricted ortho-phthalates, mainly DEHP, in a meaningful share of samples, alongside other ortho-phthalates in a further share; the exact percentages are under internal verification and are not stated here.

Design for recycling addresses how additive choice affects recyclability; the outer-section view of the topic continues on legacy additives in recycled plastic and on microplastics and additive leaching.