Practical guidelines on the application of

migration modelling for the estimation of

specific migration
In support of Regulation (EU)
No 10/2011 on plastic food
contact materials

Eddo J. Hoekstra (Ed.), Rainer Brandsch, Claude

Dequatre, Peter Mercea, Maria-Rosaria Milana,
Angela Störmer, Xenia Trier, Olivier Vitrac
Annette Schäfer and Catherine Simoneau

2015

Cin +1 − Cin Cin+1 − 2Cin + Cin−1

=D
∆t h2

EUR 27529 EN
Practical guidelines on the application of
migration modelling for the estimation of
specific migration
This publication is a Technical report by the Joint Research Centre, the European Commission’s in-house science service. It aims to
provide evidence-based scientific support to the European policy-making process. The scientific output expressed does not imply a
policy position of the European Commission. Neither the European Commission nor any person acting on behalf of the Commission is
responsible for the use which might be made of this publication.

JRC Science Hub

https://ec.europa.eu/jrc

Contact information
E-mail: [email protected]

JRC98028

EUR 27529 EN

ISBN 978-92-79-52790-6 (PDF)

ISSN 1831-9424 (online)

doi:10.2788/04517 (online)

© European Union, 2015

Reproduction is authorised provided the source is acknowledged.

Printed in Italy

All images © European Union 2015

How to cite: E.J. Hoekstra, R. Brandsch, C. Dequatre, P. Mercea, M.R. Milana, A. Störmer, X. Trier, O. Vitrac A. Schäfer and C.
Simoneau; Practical guidelines on the application of migration modelling for the estimation of specific migration; EUR 27529 EN;
doi:10.2788/04517
 Table of contents
1 Introduction.................................................................................................................................................. 5
2 Concepts of migration modelling......................................................................................................... 7
3 Estimation of diffusion coefficient and polymer specific parameters.................................. 9
3.1 Estimation of diffusion coefficients .......................................................................................... 9
3.2 Estimation of polymer specific parameters ........................................................................ 11
3.2.1 Polyolefines................................................................................................................................. 11
3.2.2 Polystyrenes ............................................................................................................................... 12
3.2.3 Polyesters .................................................................................................................................... 13
3.2.4 Polyamides .................................................................................................................................. 13
3.2.5 Polyvinylchloride...................................................................................................................... 14
3.2.6 Other polymers, migrants and other parameter range ............................................ 15
4 Estimation of partition coefficient .................................................................................................... 16
4.1 Monolayers ....................................................................................................................................... 16
4.2 Multi-layers ...................................................................................................................................... 16
5 Other diffusion model input parameters........................................................................................ 18
5.1 Initial concentration ..................................................................................................................... 18
5.2 Food and food simulant............................................................................................................... 18
5.3 Contact time-temperature conditions................................................................................... 19
5.4 Contact surface, food volume and effective thickness of food .................................... 19
5.5 Remaining input parameters .................................................................................................... 19
5.6 Plastic multi-layer materials ..................................................................................................... 19
6 Procedures, practical applications, examples .............................................................................. 20
6.1 Compliance testing of substances with specific migration limits (SML) ................ 20
6.2 Optimising compliance control ................................................................................................ 21
7 REFERENCES.............................................................................................................................................. 25
8 Annex – List of symbols and abbreviations ................................................................................... 28
9 Annex – Transport equations .............................................................................................................. 30
9.1 Analytical solution to the diffusion equation ..................................................................... 30
9.2 Numerical approaches ................................................................................................................. 30
10 Annex – List of substances from the positive list of amended Regulation (EU) No
10/2011, indicating their eligibility for migration modelling ......................................................... 32
11 Annex PM/REF-numbers, chemical names, molecular mass, SML-values and highest
concentrations of some additives used in polymers ............................................................................ 33
Executive summary

The aim of this document is to assist users of diffusion models to predict

conservative, upper bound specific migration values from plastic food contact
materials for compliance purposes by providing guidance on the relevant input
parameters. This document replaces the previous version "Estimation of specific
migration by generally recognised diffusion models in support of EU Directive
2002/72/EC" and updates to the current legal basis (Regulation (EU) No 10/2011)
and enables the use of migration models for the estimation of specific migration
from plastic multi-layers.
This document represents the current validity of the model input parameters
based on constant periodical evaluations of new experimental migration data
performed by the Task Force on Migration Modelling chaired by the Joint Research
Centre of the European Commission on behalf of Directorate General Health and
Food Safety. The members of the Task Force are E.J. Hoekstra, R. Brandsch, C.
Dequatre, P. Mercea, M.R. Milana, A. Störmer, X. Trier, O. Vitrac A. Schäfer and C.
Simoneau and they are the authors of this practical guidance document.
1 Introduction
The European legislation requires verification of compliance for migration of
substances from plastic food contact materials with existing specific and overall
migration limits. To do so, there are migration tests to carry out using food/food
simulants under the test conditions specified in Regulation (EU) No 10/2011.
Numerous scientific investigations during the last two decades have demonstrated
that migration from food contact materials into food and food simulants follow
predictable physical processes. In the absence of specific interactions with food,
migration is a mass transfer of substances from a plastic material into foodstuffs
and obeys in most cases to Fick´s laws of diffusion (Crank, 1979; Vieth, 1991).
Hence, in addition to experimental methods an alternative tool based on
theoretical migration estimations can be applicable for verifying the compliance of
a migrant. The European Union introduced this option to use generally recognised
diffusion models as a novel compliance and quality assurance tool with Directive
2001/62/EC as an amendment of Directive 90/128/EEC. Directive 90/128/EEC
was repealed by Directive 2002/72/EC which was repealed by Regulation (EU) No
10/2011. Chapter 2.2.3. of Annex V of this Regulation states that migration
modelling can be applied as screening tool for specific migration as long as the
method is considered more severe1 than the verification method:
“To screen for specific migration the migration potential can be calculated
based on the residual content of the substance in the material or article
applying generally recognised diffusion models based on scientific evidence
that are constructed such as to overestimate1 real migration”.
Article 18.3 regulates the compliance check using migration models:
“For materials and articles not yet in contact with food screening of
compliance with the specific migration limit can be performed applying
screening approaches in accordance with the rules set out in Chapter 2,
Section 2.2 of Annex V. If a material or article fails to comply with the
migration limits in the screening approach a conclusion of non-compliance
has to be confirmed by verification of compliance …”.
A generally recognised model must be based on scientific evidence. The realisation
of this requirement was achieved within the European project SMT-CT98-7513
“Evaluation of Migration Models in Support of Directive 90/128/EEC” under the 5th
Framework Programme for monolayers in contact with well mixed liquids
(Hinrichs and Piringer, 2002).
The major objectives of this project were
• To demonstrate that a correspondence between the specific migration limit
(SML) and a permitted maximum initial concentration (MIC) of a substance in
the finished product can be established within some boundary limits and

1 Note: a future revision of Regulation (EU) No 10/2011 may change "more severe than" into "at
least as severe as"
 • To establish documentation that demonstrates the validity of underlying
migration models for compliance purposes. Consequently, parameters used
in the migration model were selected in a way that an “upper-bound”
estimate of migration rate was generated.
The final report of this project was published in a peer-reviewed scientific journal
(Begley et al., 2005). This research project validated the diffusion model, the
mathematical equations to be applied, the estimation procedure for the mass
transfer coefficients and the conditions for their appropriate application with
special focus on monolayer polyolefin plastics.
The established diffusion models have the ability to estimate upper bound
migration values, provided the boundary and simplifying assumptions are fulfilled.
The model is subjected to a continuous update and refinement process and should
be practiced only by users with appropriate skills and training.
According to the current state-of-the-art, the scope and applicability of migration
prediction comprises the mass transfer of the migrants listed in Annex 10 from
plastics when in contact with food simulants according to Regulation (EU) No
10/2011, provided that all the settled conditions are fulfilled.
For other polymers and situations not listed in section 3.2 of this document,
diffusion modelling can be used provided that the procedures described in section
3.2.6 have been followed and properly documented to demonstrate that the use of
the model in that particular case leads to an overestimation1 of the specific
migration compared to experimental testing.
This document represents a continuous update and validation of the fields of
application regarding the estimation procedure for diffusion and partition
coefficients for a given number of polymers and migrants.
2 Concepts of migration modelling
Migration is a global term to describe a net mass transfer of chemical substances
from a packaging material into the food (Crompton, 2007). In the case of plastic
food contact materials (FCM) covered by Regulation (EU) No 10/2011, migration
includes several macroscopic mass transfer mechanisms including: i) mass
diffusion in and through the different plastic materials as well as the liquid or gas
phases separating the primary source from the food, and ii) desorption/sorption at
the interface between each crossed medium. When it involves fluid phases,
migration may also cover an additional transport or mixing effect by advection. In
the frame of the FCM legislation, migration modelling is an abstract process aiming
at calculating with various simplifications and assumptions the maximum amount
of substances which might be transferred to the food in contact. It is important to
note that the migration models to be used for the purpose of this document do not
seek to reproduce all the details of the real mechanisms (e.g. exact concentration
profile in the food or in a particular element of the packaging assembly), but a
sufficient realistic and upper value of the concentration in food or in food simulant
to reach a decision (Brandsch et al., 2002; Vergnaud and Rosca, 2006; Vitrac and
Hayert, 2007a; Gillet et al., 2009b).
To predict the specific migration from plastic food contact materials into a contact
medium for a given substance, one needs a scientific model as physical description
of the mass transport (see Annex 9), a value or a generally recognized estimation
procedure for the diffusion coefficient in plastics (see section 3), and assumptions
on partitioning conditions (see section 4) at the plastic-food and the plastic-plastic
(multilayer structures) interface.
Migration modelling itself covers several aspects: setting simplifications and
coding them along with the geometry and contact conditions in mass balance,
initial and boundary conditions; doing calculations; clarifying uncertain
parameters, refinement and hinting conclusions.
Examples of constitutive equations for migration problems with constant and
uniform diffusion coefficient in each constitutive packaging layer can be found in
various text books (Crank, 1975; Piringer and Baner, 2000, 2008; Vergnaud, 1991;
Vitrac and Hayert, 2007) and in peer-reviewed journals with sometimes different
conventions, as reviewed by Helmroth et al. (2002) and Pocas et al. (2008). For
monolayer materials, the most general formulation and analytical solutions with a
general boundary condition enabling to simulate solid or liquid food is given by
Vitrac and Hayert (2006). They extend earlier solutions proposed by Sagiv (2001,
2002). For multilayer materials, no general analytical solution exists and numerical
modelling must be used. Among proposed numerical schemes (Roduit et al., 2005;
Tosa et al., 2008; Torres et al., 2012), a general formulation including a very
efficient scheme combining semi-analytical solutions and an efficient free-volume
formulation is reported by Nguyen et al. (2013).
The margin of overestimation of a specific migration calculated by a migration
model can vary with the combination of polymer and migrant, and the contact
conditions; it is generally higher in cases of lower diffusion (e.g. high molecular
mass migrants, non-polyolefines) but might not overestimate1 in all cases.
Consequently the model shall be used in the range of migrant molecular masses,
temperatures and contact conditions that were experimentally tested and used for
the derivation of the “upper-bound” diffusion coefficients. Useful explanations to
provide guidance to the user of the model are reported in the following
paragraphs.
3 Estimation of diffusion coefficient and polymer specific
parameters
As mentioned above one key mass transfer parameter necessary for migration
modelling in plastics is the diffusion coefficient of the migrant in the plastic while
the other is the partition coefficient of the migrant between the plastic and the
food (simulant). Both parameters play a crucial role in determining the level of
migration in a real food packaging application (Reynier et al., 1999; Roduit and
Dudler, 2005, Piringer and Baner, 2008). Due to a lack of knowledge of exact values
of diffusion coefficients in any specific case, it is of primary importance from a
regulatory standpoint to establish these values in such manner that reliable “worst
case” scenarios with respect to migration are estimated. To meet this requirement
this section details a simple approach to overestimate1 diffusion coefficients.

3.1 Estimation of diffusion coefficients

The literature reports a series of models for the theoretical estimation of diffusion
coefficients in polymers (Mercea, 2008) but these models are, at least today, too
complicated for practical applications. Therefore, a simpler approach was
developed in the 90's. In a first approximation to estimate diffusion coefficients
(DP) they were correlated with the relative molecular mass (Mr) of the migrant, by
introduction of a temperature dependent polymer specific constant (AP) and the
absolute temperature (T) based on empirical relationships and experimental data
see equation 1). Similar approaches had been used before (Piringer, 1994;
Hamdani et al., 1997; Limm and Hollifield, 1996). To pursue the goal of obtaining a
simple formula for the estimation of DP, a refined equation for polyolefines and
some other plastic materials was developed (Brandsch et al., 2002). With this
equation a polymer specific upper-bound diffusion coefficient, DP*, of a migrant in
the polymer matrix can be estimated and used instead of the actual diffusion
coefficient DP. It must be emphasized that DP ≤ DP*. Therefore, using such a DP* for
migration estimations leads to overestimated migration values. From
phenomenological derivations and a statistical evaluation of experimental
diffusion and migration data (Mercea and Piringer, 1998) DP* can be estimated by
the following Arrhenius type equation (Brandsch et al., 2002):
 10454R 
1) DP* = exp AP* − 0.1351M r2 / 3 + 0.003M r −  (m2/s)
 RT 
and
τ
AP* = AP'* −
2) T
where Mr is the relative molecular mass of migrant (Da), τ a polymer specific
“activation temperature” increment (K), R the gas constant (8.3145 J mol−1 K−1)
and (10454 R) a reference activation energy (J mol−1). The asterisk * indicates an
upper bound value and the apostrophe ' indicates the parameter is temperature
independent.
From equation 1 it can be recognised that there are key variables, which determine
the diffusion in a polymer. Two of them are not linked to the polymer that are the
relative molecular mass of the migrant and the absolute temperature.
The parameter AP is linked to the polymer and describes the basic diffusion
behaviour of the polymer matrix in relation to the migrants. In soft/flexible
polymers, such as low density polyethylene (LDPE), AP values are high reflecting
high diffusion behaviour (DP) and hence high migration through the polymer, while
stiff chain polymers such as polyesters have lower AP values due to the lower
diffusion behaviour, and hence lower migration of the same migrant. Where AP can
vary with temperature, AP' is a temperature independent term. AP* (and hence
AP'*) are upper-bound values, which have been derived in a first step statistically
from individual values generated through equation 1 from diffusion coefficients
determined experimentally, so that in the second step for compliance assessment
equation 1 generates upper-bound diffusion coefficient DP* satisfying the condition
DP ≤ DP*. Using these DP*, the migration will be overestimated and consequently
overestimated migration rates will be calculated by the migration model based on
the diffusion law within certain temperature ranges.
The parameter τ, together with the “reference activation energy” of 10454∙R in
equation 1, both contribute to the diffusion activation energy, EA = (10454+τ)R.
For simplification reasons it is assumed in this model that the activation energy is
the same for all molecules in the applicable molecular mass range.
Upon analysing EA data from literature for a large series of migrants in many
polymer matrices, it was concluded that τ = 0 for many polymers. Thus, setting τ =
0 as a first approximation for LDPE gives EA = 86.92 kJ mol−1, which is in good
agreement with the mean value of EA = 87 kJ mol−1 found from literature data
(Mercea, 2000).
For other important groups of plastics relevant to food packaging, e.g. high-density
polyethylene (HDPE) and polyethylene terephthalate (PET), a higher activation
energy is generally observed. A good mean value of EA for these matrices is 100 kJ
mol−1 corresponding to a τ of 1577 K.
It is known that in a given polymer and temperature range each migrant has
different diffusion activation energies (Mercea 2000). For feasibility reasons the
polymer related mean values are used for migration modelling. Deviations of the
migrant specific activation energies from the mean value are accounted for during
the translation of diffusion data in polymer specific constants followed by the
statistical evaluation.
The same approach and estimation procedure as described for monolayer plastics
applies (see above) for each layer in a multilayer plastic.
3.2 Estimation of polymer specific parameters
The experts participating to the European project SMT-CT98-7513 ‘Evaluation of
Migration Models in Support of Directive 90/128/EEC’ agreed to consider the
prediction tool fully validated for polyolefines (PO), on the basis of the large
number of consistent results. During this project, experts agreed that although
much less data were available for the non-PO such as polystyrene (PS), high impact
polystyrene (HIPS), polyethylene terephthalate (PET), polyethylene naphthalate
(PEN), polyamide (PA), compared to PO, the basis was considered sufficiently solid
due to the fact that well-defined migration experiments were selected and
performed by internationally recognised laboratories. For the non-PO samples
covering the market situation, it could be shown that these equations overestimate
the experimental migration values. For polycarbonate (PC) and plasticised
polyvinylcholoride (PVC), experts at the time of the SMT project considered the
data insufficient to create a reliable set of parameters for migration modelling.
This current update is the result of a dedicated Task Force co-ordinated by the JRC
which considered experimental migration data produced since the project SMT
(i.e. period 2000-2004) and it takes into the outcome of the evaluation of new
(until 2008) available data, which is in particular the case for PET polymer. Newer
data have not been assessed yet and will be part of a future revision.
Where appropriate, previous values of the key parameters have been refined. In
addition, the migration behaviour of new polymers in terms of their polymer
specific constant (AP'*-value) was evaluated based on experimental data and their
relevant parameters introduced in this update.
Consequently, the updated requirements for polymer specific migration modelling
are described in the following paragraphs.

3.2.1 Polyolefines
The most important polyolefines (PO) used for food packaging are low density
polyethylenes (LDPE and LLDPE), high-density polyethylene (HDPE) and various
types of polypropylenes (PP-random, PP-homo and PP-blockcopolymer)2. These
materials have specific temperature ranges for which the integrity of the food
package is maintained. Using product knowledge of these packages, the
temperature range to use PO is generally limited to less than 100°C (or 121 °C in
case of homo- and random PP) which is also valid for the applicability of migration
modelling (Table 1). Under these temperature conditions and with an initial
migrant concentration, cP,0, not higher than about 1%, the migration process in
PO’s follows the general physical law of diffusion.

2 The copolymers with non-olefinic monomers, e.g. acrylics, vinylics, etc., are not yet evaluated.
Table 1 Parameter ranges for the applicability of the migration model for
selected PO.
Polymer T (°C) Mr (g mol−1) AP'* τ (K)
LDPE < 80 30 - 2000 11.5 0
LLDPE <100 30 - 2000 11.5 0
HDPE < 90 30 - 2000 14.5 1577
PP (homo + random) < 120 30 - 2000 13.1 1577
PP (blockcopolymer) < 100 30 - 2000 11.5 0

The actual “upper-bond” values of AP’* and respectively τ from equation (1) for
PO's were first determined empirically using a database with diffusion coefficients
reported in the literature over the last four decades (Mercea, 2000). In addition,
the results from recent migration measurements were used to confirm and
validate these AP’* and τ values. Further details on the methodology of the
evaluation of the model were published by Begley et al. (2005). The Ap’* value is
based on the 95% percentile of calculated Ap’ values3.

3.2.2 Polystyrenes
Polystyrenes used for food packaging applications can be roughly subdivided into
three general categories: general-purpose polystyrene (PS) high impact
polystyrene (HIPS) and styrene-butadiene-styrene block-copolymer (SBS). Using
an evaluation of experimental migration data of these polymers, the temperature
range for the applicability of migration modelling, as required by the general
requirements given before, is listed in Table 2. In these cases and an initial migrant
concentration, cP,0, not higher than about 1%, the migration process in PS’s follows
the generally accepted physical law of diffusion.

Table 2 Ranges of parameters for the applicability of the migration model for
PS , HIPS and SBS.
Polymer T (°C) Mr (g mol−1) AP'* τ (K)
PS < 70 104 - 647 -1 0
HIPS < 70 104 - 430 1.0 0
SBS < 70 84 - 689 10.5 0
PS/SBS blend # < 70 84 - 689 AP'*(blend) = −1 + 0.115∙%SBS 0
# blending PS with SBS results in a linear relationship of AP’* value as a function of

the percentage of SBS (% SBS) added to PS. The upper bound AP’* value for a blend
of PS with SBS can be described by the given relationship

For PS, HIPS and SBS the actual values of AP'* and τ could be determined
empirically from the database of diffusion coefficients and verified by well-defined
migration experiments reported in recent years (Hinrichs and Piringer, 2002;
Begley et al., 2005). Applying these values of AP'* and τ (see Table 2) in equations 1

3 The current AP’* were generated according to Begley et al., 2005.

and 2 results in “upper bound” diffusion coefficients, DP*. These DP* lead to
overestimation of the experimental migration data available

3.2.3 Polyesters
The polyesters mainly used for food packaging applications are polyethylene
terephthalate (PET), and polyethylene naphthalate (PEN). Using an evaluation of
experimental migration data of these polymers, the temperature range for the
applicability of migration modelling, as required by the general requirements
given before, is listed in Table 3. In these cases the migration process in these
polyesters follows the generally accepted physical law of diffusion.
For PET and PEN the actual values of AP'* and τ from equation 2 were determined
empirically using migration data from well-defined migration experiments
reported in recent years (Hinrichs and Piringer, 2002; Begley et al., 2005). Using
these values of AP'* and τ in equations 1 and 2 results in “upper bound” diffusion
coefficients (DP*) which lead to overestimation of experimental migration data.
This was confirmed more recently by systematic studies on migration from PET
bottles into soft drinks (Franz and Welle, 2008). The results from this study
support the assignment of AP'* values for PET having a temperature below the
glass temperature (Tg) of 70°C (Error! Reference source not found.). The split
between two ranges of temperature was derived from a statistical evaluation of
new data (Feigenbaum et al., 2005; Pennarun et al., 2005).
The temperature cut-off was selected with a safety margin to reflect the lowest Tg
values for PET on the market (typical Tg range for PET is 70-80°C). In the case of
small molecules generically referring to below 50 g mol−1, the value of AP’* should
be taken as 6.4 in both cases. In these cases and an initial migrant concentration,
cP,0, not higher than about 1%, the migration process follows the generally
accepted physical law of diffusion.
Table 3 Ranges of parameters for the applicability of the migration model for
PET and PEN
T (°C) Mr (g
Polymer AP'* τ (K)
mol−1)
PET 70<T<175 >32 6.4 1577
PET ≤70 <50 6.4 1577
PET ≤70 ≥50 3.1 1577
PEN <175 >32 5.0 1577

3.2.4 Polyamides
For polyamides (PA) only few data are available both for diffusion coefficients and
migration data. Furthermore, the water content of the food or food simulant can
strongly influence the mechanism of the transfer.
As current data had only addressed PA in contact with olive oil and isooctane, but
not PA in the swollen state, this Task Force considers that the current data does
not represent all the relevant foodstuffs. The model is therefore not considered
fully validated for PA. The following tables relating to food simulant D2 and
isooctane are provided for informative purposes only.
Table 4 Ranges of parameters for the applicability of the migration model for
polyamides#
Polymer T (°C) Mr (g mol−1) AP'* τ (K)
PA6 <100 113 @ 0 0
PA6,6 <100 32-587 2.0 0
PA12 <100 197 $ 2.6 0
# use of parameters limited to simulant D2 and isooctane. PA was previously

equilibrated with air at room temperature, within an approximate range of 40-

65% relative humidity. @ caprolactam. $ laurolactam
For PA6, PA6,6 and PA12 the actual values of AP'* and τ from equation 2 were
determined empirically using migration data from well-defined migration
experiments reported in recent years (Hinrichs and Piringer, 2002; Begley et al.,
2005; Stoffers et al., 2003, 2005) Using these values of AP'* and τ in equations 1 and
2 results in “upper bound” diffusion coefficients, DP*, which lead to overestimation
of the experimental migration data. The initial migrant concentration, cP,0, shall not
be higher than 1%.

3.2.5 Polyvinylchloride
3.2.5.1 Rigid PVC
For rigid PVC migration data and diffusion coefficients are available (unpublished
data of FABES) for the range of temperatures 20-70°C. Using an evaluation of
experimental migration data of these polymers, the temperature range for the
applicability of migration modelling, as required by the general requirements
given before, is listed in Table 5. In these cases and under the condition that the
initial migrant concentration, cP,0, shall not be higher than 1%, the migration
process in PVC follows the generally accepted physical law of diffusion.

Table 5 Ranges of parameters for the applicability of the migration model for
rigid PVC
Polymer T (°C) Mr (g mol−1) AP'* τ (K)
PVC (rigid) <70 >225 -1.0 0

3.2.5.2 Plasticised PVC

For plasticised PVC only few data are available both for diffusion coefficients and
migration data. Current data has only addressed migration of individual
plasticisers. Therefore this Task Force considers that the current data do not
present relevance to the presence of mixtures of additives and the model is
therefore not considered fully validated for plasticized PVC. The following tables
and formula are provided for informative purposes only.
Since plasticisers are used at high concentrations, the diffusion coefficient is
expected to vary with the total concentration of plasticiser or mixture of
plasticisers in PVC. The diffusion coefficient based on the initial total concentration
of plasticisers is overestimated. It is assumed that the migration follows the
general law of diffusion and that the proposed diffusion coefficient is constant and
always overestimates the real diffusion coefficient.
A linear relationship of AP’* value as a function of % plasticiser added to PVC exists
(Table 6). It should be noted that the parameters in this equation will also vary
with the plasticiser molecular mass. For 4 plasticisers with a content of up to 30%
(w/w) and a Mr of 370-419 g mol−1, a calculated AP’* for plasticised PVC of 14.6 is
considered reasonable. The formula can be used for other plasticisers only with
molecular masses above the range tested and an initial migrant concentration, cP,0,
of maximum 30%.

Table 6 Ranges of parameters for the applicability of the migration model for
plasticised PVC. plast, plasticiser
Polymer T (°C) Mr (g mol−1) AP'* τ (K)
PVC (plasticised) <70 370-419 AP'*(PVC) = −1 + 0.52∙%plast 0

3.2.6 Other polymers, migrants and other parameter range

For other polymers and parameter ranges not listed in this chapter, migration
modelling can be used for compliance purpose provided that it is demonstrated
that the mass transfer (migration process) from the plastic into the food or food
simulant follows the law of diffusion and that the parameters (AP'* and τ) were
determined according to recognised and validated procedures. Such procedures
may be described in future guidelines. These guidelines should demonstrate how
DP values are generated and how they are translated into AP'*-values based on the
range of applicability in terms of contact conditions and molecular mass range.
Furthermore the documentation should bring conclusive proof that the use of the
model in that particular case leads necessarily to an overestimation in the 95th
percentile of all applicable cases.
4 Estimation of partition coefficient

4.1 Monolayers
In absence of specific data, in order to model worst case scenarios, the partition
coefficient of the migrant between polymer P and food F should be taken as KP,F =
1, which means that the substance is well soluble in F. This option leads to the
highest migration values, i.e. complete transfer of the migrant from the food
contact material to food at equilibrium. The question whether this equilibrium
partitioning will be reached in a practical application, depends on the polymer type
and more specifically on the diffusion behaviour of the polymer under the practical
contact conditions. For all other cases, that is for which the migrant is only
sparingly soluble in F the partition constant should be set at KP,F = 1000, e.g. for
lipophilic substances when the polymeric material is in contact with aqueous
food/simulant.

4.2 Multi-layers
In absence of specific data, the worst case scenario needs to be modelled. The
partition coefficients4 of all polymer layers between the polymer layer containing
the migrant (Pi) and the food (F) shall be taken as KPi,Pi-1 = 1 (i = i→1). The
partition coefficients of all polymer layers between the polymer layer containing
the migrant and the outer polymer layers (see Figure 1) or functional barrier shall
be taken as KPi+1,Pi = 0.001 (i = i→n-1).

n i+8 i+7 i+6 i+5 i+4 i+3 i+2 i+1 i 2 1 0

Pi+1 Pi P1 F

KPn,Pi+8 KPi+1,Pi KPi,P2 KP2,P1 KP1,F

Figure 1 Forced specific migration of migrant in one layer (i) to food. K in black
is 0.001 and K in red is 1

If the migrant is present in more layers, several worst case scenarios may exist and
need to be modelled in order check which is the worst case. The partition
coefficients of all polymer layers between the innermost polymer layer containing
the migrant (Pi) and the food (F) shall be taken as KPi,Pi-1 = 1 (i = i→1), and all
polymer layers between the outmost polymer layer containing the migrant (Pj) and

4 Note: and , where C is the concentration of the migrant in the respective

layer or food at equilibrium.
the outermost polymer layer (Pn) or functional barrier shall be taken as KPj+1,Pj =
0.001 (j = j→n-1). For each scenario the partition coefficient between one
particular polymer layer containing the migrant (Px) and its outer layer is set to K
Px+1,Px = 0.001 (x = x→j-1), whereas all the remaining partition coefficients are set
at KPx,Px-1 = 1 (x = i+1→x).

Scenario 1: only migration from Pi to food

n i+8 i+7 j i+5 i+4 x i+2 i+1 i 2 1 0

Pi+1 Pi P1 F

KPn,Pi+8 KPj+1,Pj KPj,Pj-1 KPx+1,Px KPx,Px-1 KPi+1,Pi KPi,P2 KP2,P1 KP1,F

Scenario 2: migration from Px and Pi to food

n i+8 i+7 j i+5 i+4 x i+2 i+1 i 2 1 0

Pi+1 Pi P1 F

KPn,Pi+8 KPj+1,Pj KPx+1,Px KPx,Px-1 ← → KPi+1,Pi KPi,P2 KP2,P1 KP1,F

Scenario 3: migration from Pj, Px and Pi to food

n i+8 i+7 j i+5 i+4 x i+2 i+1 i 2 1 0

Pi+1 Pi P1 F

KPn,Pi+8 KPj+1,Pj KPj,Pj-1← → KPx+1,Px KPx,Px-1 ←→ KPi+1,Pi KPi,P2 KP2,P1 KP1,F

Figure 2 Example of scenario’s of forced specific migration of one migrant that

is present in three layers (i, j, x), to food. In this example three
scenario's are necessary to be calculated. K in black is 0.001 and K in
red is 1
5 Other diffusion model input parameters

5.1 Initial concentration

The initial concentration (cP,0) of all substances listed in Annex 10 shall not be
higher than 1%. A level above 1% may cause a plasticising effect in given polymers
and diffusion modelling cannot be used. If a plasticising effect is observed, the use
of the migration model is possible only with an AP-value accounting for the
plasticising effect and determined by experiments.
As an exception the concentration of plasticisers in plasticised PVC may be up to
30% under the condition that the relative molecular mass is 225 g mol−1 or higher
(see section 3.2.5.2).
When a substance is blooming out of the polymer, the substance is not
homogenously distributed in the polymer anymore and modelling is not
applicable. The problem of blooming must be considered case by case with respect
to the polymer/migrant combination investigated. It is well known that antistatic
and antifogging agents typically incorporated into polyolefines deliberately
migrate at the surface of the polymeric materials. The same substance does not
bloom out from a more polar polymeric material like polyester or polyamide. The
user of the migration model is strongly advised to carefully consider the possibility
of blooming, to avoid application of the migration model for special cases where
blooming influences the level of migration.

5.2 Food and food simulant

Food and food simulant are not as such selected in the diffusion model. They are
represented by the density and the diffusion coefficient in food. Regulation (EU) No
10/2011 sets by convention the density of food at 1 kg dm−3. For foods with a
higher density the specific migration will be underestimated when the
conventional value is used. Therefore it is recommended to use a density of food of
1 kg dm−3 for polymers in contact with unknown foods, for foods with a lower
density then the conventional one or for liquid food simulants. If the density of the
food is known and higher than the conventional density then the use of real
density of the worst case food is recommended.
To fulfil the condition of “well-mixing” the diffusion coefficient in the food or liquid
food simulant is set as default at 10-6 m2 s−1.
Migration models cannot be used when swelling of the polymer occurs. Examples
of swelling are polyolefines in contact with iso-octane or other combinations of
polymers with test media having a high swelling power, because according to the
initial and boundary conditions, it is assumed that diffusion in the polymer is
uniform and does not vary with time.
5.3 Contact time-temperature conditions
The contact time and temperature shall be chosen such that they represent the
worst foreseeable conditions, following relevant sections of Chapter 2 of Annex V
of Regulation (EU) No 10/2011. The highest modelled specific migration shall be
compared with the SML.

5.4 Contact surface, food volume and effective thickness of food

The contact surface and volume of food shall be chosen such that they represent
the worst case foreseeable surface-to-volume ratio or 6 dm2/kg for articles
specified in Article 17.2 of Regulation (EU) No 10/2011. For caps, gaskets, stoppers
and other sealing Article 17.3 additional rules are applicable (see also Technical
guidelines for compliance testing in the framework of Regulation (EU) No 10/2011
on plastic food contact materials, 2016)
The effective thickness of food is the ratio of the volume of food and the contact
surface. For a surface to volume ratio of 6 dm²/kg this corresponds to a food
thickness of 16666,7 µm in the model.

5.5 Remaining input parameters

For the remaining input parameters, i.e. thickness and density of polymer or
polymer layers, the real values have to be used.

5.6 Plastic multi-layer materials

During manufacturing and storage of plastic multi-layer materials diffusion and
partition processes between the layers of the material already occur.
If the plastic multi-layer material is stored by piling sheets or articles on each
other, diffusion between the different sheets occurs. There are three different
possibilities for piling: 1) all food contact layers come into contact with the non-
food contact layer of the next sheet (set-off), 2) the food contact layers come into
contact with the food contact layer of the next sheet and 3) the food contact layers
come randomly into contact with the food contact layer or the non-food contact
layer of the next sheet.
If the plastic multi-layer material is put and stored on a roll the food contact layer
comes into contact with the non-food contact layer (set-off).
For migration modelling of plastic multi-layer materials these processes and the
resulting phenomena before contact with food need to be taken into account.
Otherwise the migration model result may underestimate the real specific
migration.
6 Procedures, practical applications, examples
The model allows the following estimations:
(i) Based on knowledge of the existing initial concentration of a migrant of
known molecular mass in a polymer, its specific time and temperature
dependent migration into a given food simulant or food can be
calculated from equation 3 (Annex 9.1) or numerical solutions.
(ii) Reversely, based on a given migration limit or SML value from
legislation, the maximum initial concentration (MIC) of a migrant of
known molecular mass in a polymer that can be used in a food contact
can be estimated from equation 7 (Annex 9.1) or numerical solutions.
As a general rule: in cases where the migration estimation scheme outlined above
leads to results which are above the specific migration limit (SML), an
experimental test of compliance is compulsory.
If the upper-bound polymer specific AP’* value is not known or applicable from the
tables given in Chapter 3.2, migration model cannot be used and the procedures
given in section 3.2.6 shall be carried out in order to extent the applicability of the
migration models.

6.1 Compliance testing of substances with specific migration limits

(SML)
One major objective of this document is to give guidance for compliance testing.
Consequently, one major field of application concerns the control for compliance of
substances listed in the Regulation (EU) No 10/2011 and its amendments with
respect to their SML’s.
It must be emphasised that at the present stage of knowledge the migration model
is only suitable for the polymers and under the conditions described in chapter 3.2.
In addition, Annex 10 lists all substances in Regulation (EU) No 10/2011 and
indicates their theoretical eligibility for modelling based on the criteria listed
below. However, before applying the model one always has to ensure that the
model assumptions (Chapter 2) are fulfilled for the substance(s) considered.
The following criteria for eligibility for migration modelling are applied:
All organic, non-gaseous substances with a well-defined molecular mass, soluble in
the polymeric matrix, are eligible for migration modelling.
All polymeric additives with a well-defined molecular mass distribution, i.e.
accessible by gel permeation chromatography, mass spectrometry or any other
analytical technique capable to deliver the molecular mass distribution are eligible
for migration modelling. Migration modelling is applicable provided that the actual
molecular mass distribution of the polymeric additive under investigation is
available by application of a suitable analytical technique. Polymeric additives are
marked with "pa" in Annex 10. An overestimated modelled migration result is
obtained when the lowest molecular mass of the mass distribution is used for the
estimation of the diffusion coefficients.
All organic substances known to deliberately bloom out from some polymeric
materials, e.g. antistatic or antifogging agents incorporated in polyolefines, were
included in the list under reserve. Migration modelling is applicable only in those
cases (polymer/migrant combinations), in which blooming does not occur.
Note: Blooming out of a given component from the plastic occurs if the difference in polarity
between the component and the polymeric matrix is high resulting in low solubility of the
component in the polymer. Furthermore blooming often occurs when substances have a surfactant
structure (polar head, non-polar tail). Due to blooming the component is migrating in short time at
the polymer/air interface resulting in high migration values for short contact times. From migration
modelling point of view the assumption that a migrant is distributed homogeneously in P, is not
fulfilled anymore and accordingly if blooming occurs migration modelling cannot be applied.
All organic mixtures with undefined molecular mass, typically derived from
natural sources like fats and oils, rosins, waxes, starch, proteins, cellulose,
cotton are not eligible for migration modelling. However a specific substance
with well-defined molecular mass below 2000 g mol−1 that is a component of a
mixture can be addressed by migration modelling estimations.
All inorganic substances, metals, metal oxides, metal salts, etc. are generally not
eligible for migration modelling unless differently indicated in Annex 10.
This list in Annex 10 is a reference for users, who perform estimations. Once
selected from this list, migration modelling can be applied for the substance,
provided the substance is contained in one of the polymers specified in chapter 3.2
and also the proper value for the partition coefficient , KP,F is applied.

6.2 Optimising compliance control

When the necessary inputs to the model are available, migration models can be
used to optimise compliance control strategies.
Tables of FCM and PM/REF-numbers, chemical names, Mr-, SML-values and highest
concentrations (cP,0) of some additives usually used in polyolefines and non-
polyolefines can be found in Annex 11.
It is the responsibility for every company involved in the production, conversion,
import and retail sale of FCM to demonstrate compliance with existing EU
legislation. National authorities in turn are responsible to enforce that the
legislation is followed. To fulfil their obligations, both companies (or the
contracted control laboratories) and enforcement employ compliance testing of
FCMs, e.g. by experimental chemical testing of migration and/or by migration
modelling.
The migration results obtained by modelling are however only as good as the data
put into the model, and only valid if the assumptions of the model are fulfilled. In
order to use migration models successfully, it is essential to have a well-described
FCM. This typically requires good traceability and information flow through the
production and supply chain, from raw material producers, to the food industry
and to the “seller” of the finished FCM article. It should also be emphasized, that
migration models can determine only the migration of known substances with
known initial concentrations.
When testing compliance of a plastic FCM, the first question to ask is which type of
polymer is at hand – if in doubt infrared spectroscopy (FTIR) can possibly be of
some help. If the polymer is listed in tables of this guide, migration modelling can
be applied. If not, the polymer specific constants (AP'* and τ) must be determined
as described in section 3.2.6 and then the migration model can be applied.
Otherwise experimental migration-chemical testing needs to be performed.
Apart from the polymer identity, it is crucial to know
1. Which releasable substances are present in the objects to test (e.g. additives,
residual amounts of monomers)?
2. What are their initial concentrations, e.g. amount of additive as seen from
formulation, or determined in an experimental test (see guidance in Feigenbaum
et al. (2002))?
3. What is the “worst-case” foreseeable use (type of food, maximum temperature
and maximum packaging time) in practical life, based on the function of FCM and
any given advisory to the user?
4. What is the intended shape of the final article, specifically what surface area will
contact what portion of food, i.e. the surface-to-volume ratio?
5. Which of these substances can be used in practice in the polymer and fulfil the
limitations of substances that can be modelled, e.g. blooming agents are not
homogeneously distributed in the polymer and hence cannot be modelled?

In practice it is a challenge to obtain all the necessary information (Petersen et al.,

2005), of which 1, 3 and 4 are inherent requirements to any control laboratories
that do experimental testing, but 2 and 5 are specific to migration modelling.
Typically two situations exist (Figure 3):
• All necessary information is available. An example is when a raw material
producer tests the compliance of polymer samples representative for its final
intended use, which is known to them, e.g. coffee cups.
• Necessary information is missing. An example is when somewhere in the
marketing stage of the product, an intermediate product manufacturer has
not documented the contacts of the producer of their starting material for
retrieval of the supporting documents for the purpose of the enforcement
authorities. This situation is against Art. 5 of Regulation (EC) No 2023/2006
and Art. 17 of the Framework Regulation (EC) No 1935/2004.
The quality of information received should always be judged critically, as is the
case with experimental results. When the data are found to be trustworthy, the
application of a migration model is straightforward and extremely time saving. In
particular if the migration model predicts migration below the SML, no further
action needs to be taken. If the migration model predicts migration above the SML,
experimental migration testing must be conducted. An additional useful option
offered by the migration model to producers is to use the migration model for the
calculation of the maximum allowed initial concentration (MIC). This information
allows for the reformulation of the FCM or for the restriction of use of the product
(e.g. food types, lower contact temperatures, times and/or surface-to-volume
ratios) to keep migration below the SML.
As a first step in enforcement the authorities have to make an effort to retrieve the
information from the production chain. If the information is available migration
modelling can follow as described above. If the substances that migrate, are
known, but their cP,0 are unknown, then there are three possibilities. The producer
needs to retrieve from the company the usual max. concentrations (cP,0) of
additives that can be used in various polymers (listed in tables 4.2.1 (PO) and 4.2.2
(non-PO)). These substances are extracted from the table in Annex 10, based on
the usual compositions for a given polymer. The tables were prepared by
consulting the most important producers of plastic materials and the secondary
literature referring additives for plastic materials (Zweifel, 2001). In addition to
the identification numbers the upper limits of initial concentrations reported to be
used in plastic materials for food contact are shown. Nevertheless, as fully
specified by Milana and Piringer (2002), “this table should be considered only as an
example to offer a first guide in selecting specific additives if no other information is
available about the composition of a sample to be tested”. Whereas enforcement has
the legal right to obtain all information asked for, private companies may
encounter difficulties in practice due to confidentiality issues in relation to the
composition.
As shown in Figure 3 the endpoint to decide on compliance of a material is the
experimental migration testing. This means, that in practice there should be an
(accredited for enforcement) experimental migration test available to decide if the
FCM is legal or not.
In conclusion, migration modelling can be time and cost saving, in particular to
those who have readily access to all necessary information about the FCM.
 Compliance material

NO
Plastic monolayer? Plastic multi-layer?

YES YES NO
Polymer(s) known? Out of scope
YES NO

Is (are) the polymer(s) listed in Experimental determination of

Chapter 3.2 of this guide? polymer identity
NO
YES

Is all the necessary YES Establish polymer NO

information available? specific constants

YES NO

Can substance be Assume 100%

modelled? YES migration
NO of Cp,0?
NO
YES NO
Traced missing
information in the
supply chain?

YES NO

Information Cannot get missing

received information

NO Experimental
SM<SML? migration testing

YES

YES SM<SML?

NO
Migration substance Migration substance
compliant non-compliant

Figure 3 Decision scheme of how to handle a FCM when doing compliance

testing.
7 REFERENCES
Baner, A. L. (2000). The Estimation of Partition Coefficients, Solubility Coefficients,
and Permeability Coefficients for Organic Molecules in Polymers. Food
Packaging, American Chemical Society. 753: 37-55.
Baner, A. L. and O. G. Piringer (2008). Partition Coefficients. Plastic Packaging:
Interactions with Food and Pharmaceuticals. O. G. Piringer and A. L. Baner,
Wiley: 89-121.
Begley, T., L. Castle, A. Feigenbaum, R. Franz, K. Hinrichs, T. Lickly, P. Mercea, M.
Milana, A. O'Brien, S. Rebre, R. Rijk and O. Piringer (2005). "Evaluation of
migration models that might be used in support of regulations for food-
contact plastics." Food Additives and Contaminants 22(1): 73-90.
Brandsch, J., P. Mercea, M. Ruter, V. Tosa and O. Piringer (2002). "Migration
modelling as a tool for quality assurance of food packaging." Food Additives
and Contaminants 19: 29-41.
Crank, J. (1979). The Mathematics of Diffusion, Clarendon Press.
Crompton, T. R. (2007). Additive Migration from Plastics Into Foods: A Guide for
Analytical Chemists, Smithers Rapra Technology.
Feigenbaum A., Dole P., Aucejo S., Dainelli D., De La Cruz Garcia C., Hankemeier T.,
N’Gono Y., Papaspyrides C. D., Paseiro P., Pastorelli S., Pavlidou S., Pennarun
P. Y., Saillard P., Vidal L., Vitrac O., Voulzatis Y. (2005) Functional barriers:
Properties and evaluation. Food Additives and Contaminants 22(10), 956–
967
Feigenbaum A., Scholler D., Bouquant J., Brigot G., Ferrier D., Franz R., Lillemark L.,
Riquet A.M., Petersen J.H., Van Lierop B., Yagoubi N. (2002) Safety and
quality of food contact materials 1: Evaluation of analytical strategies to
introduce migration testing into good manufacturing practice. Food
Additives and Contaminants, 19(2), 184-201
Franz R., Welle F. (2008) Migration measurement and modelling from
poly(ethylene terephthalate) (PET) into softdrinks and fruit juices in
comparison with food simulants. Food Additives and Contaminants 25(8),
1033 – 1046.
Gillet, G., O. Vitrac and S. p. Desobry (2009a). "Prediction of Solute Partition
Coefficients between Polyolefins and Alcohols Using a Generalized
Flory−Huggins Approach." Industrial & Engineering Chemistry Research
48(11): 5285-5301.
Gillet, G., O. Vitrac, D. Tissier, P. Saillard and S. Desobry (2009b). "Development of
decision tools to assess migration from plastic materials in contact with
food." Food Additives and Contaminants Part a-Chemistry Analysis Control
Exposure & Risk Assessment 26(12): 1556-1573.
Hamdani M., Feigenbaum A., Vergnaud J.M. (1997) Prediction of worst-case
migration from packaging to food using mathematical models. Food
Additives and Contaminants, 14(5), 499-506
Helmroth, E., R. Rijk, M. Dekker and W. Jongen (2002). "Predictive modelling of
migration from packaging materials into food products for regulatory
purposes." Trends in Food Science & Technology 13(3): 102-109.
Hinrichs K., Piringer O. (editors) (2002) Evaluation of migration models to used
under Directive 90/128/EEC. Final report contract SMT4-CT98-7513,
 European Commission, Directorate General for Research, Brussels. EUR
20604 EN
Hoekstra E.J. (2016) Technical guidelines for compliance testing in the framework
of Regulation (EU) No 10/2011 on plastic food contact materials. in draft
Limm W., Hollifield H.C. (1996) Modelling of additive diffusion in polyolefines.
Food Additives and Contaminants 13(…), 949-967
Mercea P., Piringer O. (1998) Data bank to validate a mathematical model to
estimate the migration of additives and monomers from polyolefines into
foodstuffs and food simulants. EU-contract No. ETD/97/501370 (DG3).
Mercea P. (2008) Models for diffusion in polymers. In: Plastic Packaging -
interactions with food and pharmaceuticals (Piringer O.-G., Baner A.L.
editors). 2nd edition. WILEY-VCH, Weinheim, New York
Mercea P. (2000) Appendix I – Diffusion data for LDPE, HDPE and PP. In: Plastic
Packaging Materials for Food - Barrier Function, Mass Transport, Quality
Assurance and Legislation ( Piringer O.-G., Baner A.L. editors) (Weinheim:
Wiley-VCH)
Milana M.R., Piringer O. (2002) Guidance Document for Enforcement Laboratories:
How to Use Migration Models. In: Evaluation of migration models to used
under Directive 90/128/EEC (Hinrichs K., Piringer O. (editors)). Office for
Official Publications of the European Communities, Luxembourg. EUR
20604 EN, 222-249
Nguyen, P.-M., A. Goujon, P. Sauvegrain and O. Vitrac (2013). "A computer-aided
methodology to design safe food contact materials and related materials."
AIChE Journal: accepted.
Og zal, T. A., C. Peter, B. Hess and N. F. A. van der Vegt (2008). "Modeling Solubilities
of Additives in Polymer Microstructures: Single-Step Perturbation Method
Based on a Soft-Cavity Reference State." Macromolecules 41(13): 5055-
5061.
Pennarun P.Y. Saillard P., Feigenbaum A., Dole P. (2005) Experimental direct
evaluation of functional barriers in PET recycled bottles: Comparison of
migration behaviour of mono- And multilayers. Packaging Technology and
Science 18(3), 107-123
Petersen J.H., Trier X.T., Fabech B. (2005) Mathematical modelling of migration: a
suitable tool for the enforcement authorities? Food Additives and
Contaminants 22(10), 938 – 944Piringer O.G., Baner A.L. (2008) Plastic
Packaging - interactions with food and pharmaceuticals. 2nd edition.
WILEY-VCH, Weinheim, New York.
Piringer O. (1994) Evaluation of plastics for food packaging. Food Additives and
Contaminants 11(2), 221-230
Piringer, O. G. and A. L. Baner (2008). Plastic Packaging: Interactions with Food and
Pharmaceuticals, Wiley.
Reynier A., Dole P., Feigenbaum A. (2002) Integrated approach of migration
prediction using numerical modelling associated to experimental
determination of key parameters. Food Additives and Contaminants
19(Supplement), 42 – 55
Reynier A., Dole P., Feigenbaum A. (1999) Prediction of worst case migration:
presentation of a rigorous methodology. Food Additives and Contaminants
16(4), 137-152.
Roduit, B., C. H. Borgeat, S. Cavin, C. Fragniere and V. Dudler (2005). "Application of
Finite Element Analysis (FEA) for the simulation of release of additives from
multilayer polymeric packaging structures." Food Additives and
Contaminants 22(10): 945-955.
Sagiv, A. (2001). "Exact solution of mass diffusion into a finite volume." Journal of
Membrane Science 186(2): 231-237.
Sagiv, A. (2002). "Theoretical formulation of the diffusion through a slab-theory
validation." Journal of Membrane Science 199(1-2): 125-134.
Simoneau C. (2010) Estimation of specific migration by generally recognised
diffusion models in support of EU Directive 2002/72/EC.
Stoffers N.H., Dekker M., Linssen J.P.H., Störmer A., Franz R. (2003) Alternative
fatty food simulants and diffusion kinetics of nylon 12 food packaging. Food
Additives and Contaminants 20(10), 949 – 959.
Stoffers N.H., Brandsch R., Bradley E., Cooper J., Dekker M., Störmer A., Franz R.
(2005) Feasibility study for the development of certified reference
materials for specific migration testing. Part II: Estimation of diffusion
parameters and comparison of experimental and predicted data. Food
Additives and Contaminants 22(2), 173-184
Torres, A., A. Guarda, N. Moraga, J. Romero and M. J. Galotto (2012). "Experimental
and theoretical study of thermodynamics and transport properties of
multilayer polymeric food packaging." European Food Research and
Technology 234: 713-722.
Tosa, V., K. Kovacs, P. Mercea and O. Piringer (2008). "A Finite Difference Method
for Modeling Migration of Impurities in Multilayer Systems." Numerical
Analysis and Applied Mathematics 1048: 802-805.
Vergnaud, J. M. (1991). Liquid transport processes in polymeric materials:
modeling and industrial applications, Prentice Hall.
Vergnaud, J. M. and I. D. Rosca (2006). Assessing Food Safety of Polymer Packaging,
Rapra Technology.
Vieth, W. R. (1991). Diffusion in and Through Polymers: Principles and
Applications, Hanser.
Vitrac, O. and M. Hayert (2006). "Identification of Diffusion Transport Properties
from Desorption/Sorption Kinetics: An Analysis Based on a New
Approximation of Fick Equation during Solid−Liquid Contact." Industrial &
Engineering Chemistry Research 45(23): 7941-7956.
Vitrac, O. and M. Hayert (2007). Design of safe food packaging under uncertainty.
New trends in chemical engineering research. P. Berton. New York, Nova
Science Publishers: 251-292.
Zweifel H. (2001) Plastic Additives Handbook, 5th Edition. HANSER Publishers,
Munich
8 Annex – List of symbols and abbreviations

α dimensionless parameter α = VF / VP
K P,F

A area of P in contact with F (dm2)

Ap temperature dependent polymer specific constant (non-dimensional,
parameter describes the basic diffusion behaviour of the polymer
matrix) AP = AP' − τ
T

Ap' temperature independent polymer specific constant (non-

dimensional parameter describes the basic diffusion behaviour of the
polymer matrix)
Ap* Upper bound polymer specific constant (parameter describes the
basic diffusion behaviour of the polymer matrix)
AP'* temperature independent upper bound polymer specific constant
(non-dimensional parameter describes the basic diffusion behaviour
of the polymer matrix)
CEN European Committee for Standardization
cF,∞ equilibrium concentration of migrant in F (mg/kg)
cP,∞ equilibrium concentration of migrant in P (mg/kg)
cP,0 initial concentration of migrant in P (mg/kg)
DP diffusion coefficient of migrant in P (m2/s)
Dp* Upper bound value of the diffusion coefficient
dP thickness of P (m)
ρF density of F (g/cm3)
ρP density of P (g/cm3)
EA activation energy of diffusion (kJ mol-1)
EA,ref reference activation energy (86.9 kJ mol-1)
F food simulant or food
GMP Good Manufacturing Practice
HDPE High Density Polyethylene
HIPS High Impact Polystyrene
KP,F partition constant of the migrant between P and F
KP1/P2 partition constant of the migrant between P1 (in contact with F) and
P2 (the next polymer layer)
LDPE Low Density Poly ethylene
mF,eq mass of migrant from P into F at equilibrium (mg)
mF,t mass of migrant from P into F after time t (mg)
Mr relative molecular mass (Da)
MIC maximum initial concentration in P (mg/kg)
P Polymer
PA Polyamide
PBT Polybutylene terephthalate
PEN Polyethylene naphthalate
PET Polyethylene terephthalate
PMMA Poly methyl methacrylate
PO Polyolefin
PP Polypropylene
PS Polystyrene
PVC Polyvinylchloride
qn the non-zero, positive roots of equation tan q n = −α q n

ρF Density of F
ρP Density of P
R gas constant (8.3145 J mol-1 K-1)
SME Small and Medium Enterprises
SML Specific Migration Limit (mg/kg)
τ polymer specific “activation temperature” increment (K)
t migration time (s)
T absolute temperature (K)
Tg glass transition temperature (K)
VF volume of F (cm3)
VP volume of P (cm3)
9 Annex – Transport equations

9.1 Analytical solution to the diffusion equation

The analytical solution for the specific migration of a migrant from a plastic
monolayer in contact with a well-mixed liquid of limited volume (VF) is given by
equation 4.37 in Cranck (1975):
m F ,t  α   ∞ 2α (1 + α )  D 
3) = cP ,0 ρ P d P   1 − nΣ=1 exp  − qn2 2P t  
1+ α 1 + α + α qn
2 2
A   d P 

where:
1 VF cF ,∞ ρ F VF mF ,∞
4) α= = =
K P , F VP cP ,∞ ρ P VP mP ,∞
tan q n = −α q n
5)
6) mF ,∞ + mP,∞ = mP,0

where mF,t is the mass of migrant transferred from P into F after time t (note mF,0 =
0), cP,0 the initial concentration of migrant in P, VP the volume of P, A the contact
area between P and F, P and F the densities of P and respectively F and qn the
non-zero, positive roots of equation 5.
Note 1: it is important to use the correct dimensions for the different parameters
Note 2: To obtain, for a given time t a good estimate of mF,t /A the summation in Eq. 3 should
be done until the relative difference between the last two calculations is less than 1%.
For compliance check of the initial concentration in the product, equation 3 can be
rearranged to calculate the maximum initial concentration of the migrant (MIC) in
the plastic based on the specific migration limit (SML):

−1
V ρ   α  ∞ 2α (1 + α )  D   
7) MIC = SML F F ρ P d P   1 − nΣ=1 exp − qn2 2P t   
1+ α 1 + α + α qn
2 2
A    d P   

9.2 Numerical approaches

Beside analytical solutions of the diffusion equation (see section 9.1), numerical
solutions have to be used for most cases. In general numerical solutions are used
for multilayer plastics (Roduit and Dudler, 2005, Tosa et al., 2008, Reynier et al.,
2002) and for mono-layers for more complex situations, e.g. concentration
dependent diffusion coefficients and non-linear boundary conditions (Piringer and
Baner, 2008).
How to solve partial differential equations numerically is described in standard
textbooks and goes beyond the scope of this guidance document. Numerical
algorithms need to be reliable and require validation. A reliable numerical
algorithm needs to conserve the substance mass in the system, i.e. does not build
up or consume mass of the migrant due to the applied numerical methodology.
Validation of a numerical algorithm can be achieved in two ways (1) by direct
comparison with case examples results as and/or (2) by comparison with the
analytical solution of the diffusion equation for monolayer structures or multilayer
structures which can be simplified to a monolayer.
10 Annex – List of substances from the positive list of
amended Regulation (EU) No 10/2011, indicating their
eligibility for migration modelling

This annex is published as separate document since it needs to be updated

following each revision of Table 1 of Annex I of Regulation (EU) No 10/2011
11 Annex PM/REF-numbers, chemical names, molecular mass,
SML-values and highest concentrations of some additives
used in polymers
Note: Please consider the list in Annex 10 for the eligibility of the substances to be
modelled

1.1 Polyolefines.
PM/ Chemical name Mr SML CP,0 (%)
REF (mg/kg)
38560 2,5-bis(5-tert-butyl-2-benzoxazolyl)thiophene 431 0.6
N,N’-bis(3-(3,5-di-tert-butyl-4-
38800 553 15 HDPE 0.2
hydroxyphenyl)propionyl)-hydrazide
38820 Bis(2,4-di-tert-butylphenyl)penta-erythritol diphosphite 605 0.6 PP 0.1; LDPE 0.06
38840 Bis(2,4-dicumylphenyl)pentaerythritol diphosphite 853 5 LDPE 0.06
39890 Bis(methylbenzylidene) sorbitol 386 60
46480 Dibenzylidene sorbitol 358 60
46640 2,6-Di-tert-butyl-p-cresol (BHT) 220 3 PP 0.2
48640 2,4-Dihydroxybenzophenone 214 6
48720 4,4’-Dihydroxybenzophenone 214 6
48880 2,2’-Dihydroxy-4-methoxy benzophenone 244 6
Ethylenglycol-bis(3,3-bis(3-tert-butyl-4-
53670 795 6 PP 0.2; HDPE 0.1
hydroxyphenyl)butyrate)
2,2’-Ethylidene-bis(4,6-di-tert-butyl-phenyl)-
54300 487 6 PP 0.1; LDPE 0.06
fluorophosphonite
2-(2-Hydroxy-3,5-bis(1,1-dimethylbenz-
60320 448 1.5
yl)phenyl)benzo-triazole
2-(2’-Hydroxy-3’-tert-butyl-5’-methyl-phenyl)-5-
60400 316 30 PP 0.4; HDPE 0.3
chlorobenzotriazole
2-(2’-Hydroxy-3,5’-di-tert-butylphenyl)-5-
60480 358 30 PP 0.5
chlorobenzotriazole
61600 2-Hydroxy-4-n-octylbenzophenone 326 6 PP 0.5; HDPE 0.3; LDPE 0.5
Octadecyl 3-(3,5-di-tert-butyl-4-hydroxy-
68320 531 6 PP 0.2; HDPE 0.1; LDPE 0.3
phenyl)propionate
Pentaerythritol tetrakis[3-(3,5-di-tert-butyl-4-
71680 1178 60 PP 0.2; HDPE 0.25; LDPE 0.03
hydroxyphenyl)-propionate]
Phosphorous acid, bis(2,4-di-tert-butyl-6-methylphenyl)
74010 514 5 PP 0.1; HDPE 0.05; LDPE 0.06
ethyl ester
74240 Phosphorous acid, tris(2,4-di-tert-butylphenyl)ester 647 60 PP 0.1; HDPE 0.5; LDPE 0.12
Poly(6-morpholino-1,3,5-triazine-2,4-diyl)-[(2,2,6,6-
80480 tertamethyl-4-piperidyl)imino]-hexamethylene-[(2,2,6,6- 2600 - 1.8 PP 0.5
tertamethyl-4-piperidyl)-imino]
Poly[6-[(1,1,3,3-tetramethylbutyl)-amino]-1,3,5-triazine-
2000-
81200 2,4-diyl]-[(2,2,6,6-tertamethyl-4-piperidyl)imino]- 3 PP 0.5; HDPE 0.2; LDPE 0.5
3100
hexamethylene-[(2,2,6,6-tertamethyl-4-piperidyl)imino]
Poly-[[6-[N-(2,2,6,6-tetramethyl-4-piperidinyl)-n-
81220 butylamino]-1,3,5-triazine-2,4-diyl][(2,2,6,6-tetramethyl- 2600- 5 PP 0.2; HDPE 0.1; LDPE 0.1
4-piperidinyl)imino]-1,6-hexanediyl[(2,2,6,6- 3400
tetramethyl-4-piperidinyl)imino]]-alpha-[N,N,N',N'-
 tetrabutyl-N"-(2,2,6,6-tetramethyl-4-piperidinyl)-N"-[6-
(2,2,6,6-tetramethyl-4-piperidinylamino)-hexyl]-[1,3,5-
triazine-2,4,6-triamine]-omega-N,N,N',N'-tetrabutyl-
1,3,5-triazine-2,4-diamine]
Reaction product of di-tert-butyl phosphonite with
biphenyl, obtained by condensation of 2,4 di-tert-
83595 991 18 PP 0.1; HDPE 0.05 ; LDPE 0.06
butylphenol with Friedel-Crafts reaction product of
phosphorus trichloride and biphenyl
Thiodiethanol-bis(3-(3,5-di-tert-butyl-4-
92880 643 2.4
hydroxyphenyl)propionate)
93120 Thiodipropionic acid, didodecyl ester 515 5 PP 0.1 – 0.5
93280 Thiodipropionic acid, dioctadecyl ester 683 5 PP 0.5
93520 Alpha-Tocopherol 431 60
94960 1,1,1-Trimethylol-propane 134 6
1,3,5-Trimethyl-2,4,6-tris(3,5-di-tert-butyl-4-
95200 775 60 PP 0.2; HDPE 0.1
hydroxybenzyl)benzene
2,4,6-Tris(tert-butyl)phenyl 2-butyl-2-ethyl-1,3-
95270 450 2 HDPE 0.05;LDPE 0.06
propanediol phosphite
1,3,5-Tris(3,5-di-tert-butyl-4-hydroxybenzyl)-1,3,5-
95360 triazine-2,4,6-(1H,3H,5H)-trione 784 5 PP 0.1; HDPE 0.1

1,1,3-Tris(2-methyl-4-hydroxy-5-tert-butylphenyl)
95600 545 5 PE 0.1
butane

PM/REF, PM/REF-numbers; Mr, molecular mass; CP,0, highest concentrations in

polymer
1.2 Non-polyolefines
PM/ Chemical name Mr SML CP,0
REF (mg/kg) (%)

PS
40020 2,4-Bis(octylthiomethyl)-6-methyl-phenol 425 6 0.2
61440 2-(2;-Hydroxy-5’-methylphenyl)benzotri-azole 225 30 0.25
61600 2-Hydroxy-4-n-octylbenzophenone 326 6 0.2
68320 Octadecyl 3-(3,5-di-tert-butyl-4-hydroxy-phenyl)propionate 531 6 0.15
74240 Phosphorous acid, tris(2,4-di-tert-butylphenyl)ester 646 60 0.2
83595 Reaction product of di-tert-butyl phosphonite with biphenyl, obtained by 595 18 0.2
condensation of 2,4 di-tert-butylphenol with Friedel-Crafts reaction product of
phosphorus trichloride and biphenyl
94400 Triethyleneglycol-bis[3-(3-tert-butyl-4-hydroxy-5-methylphenyl) propionate] 587 18 0.2
95600 1,1,3-Tris(2-methyl-4-hydroxy-5-tert-butylphenyl) butane 545 5 0.2

HIPS
31520 Acrylic acid, 2-tert-butyl-6-(3-tert-butyl-2-hydroxy-5-methylbenzyl)-4-methyl- 395 6 0.5
phenyl ester
38560 2,5-bis(5-tert-butyl-2-benzoxazolyl)thiophene 431 0.6 0.05
40000 2,4-Bis(octylmercapto)-6-(4-hydroxy-3,5-di-tert-butylanilino)-1,3,5-triazine 589 30 0.1
40020 2,4-Bis(octylthiomethyl)-6-methyl-phenol 425 6 0.2
61440 2-(2;-Hydroxy-5’-methylphenyl)benzotri-azole 225 30 0.4
68320 Octadecyl 3-(3,5-di-tert-butyl-4-hydroxy-phenyl)propionate 531 6 0.5
94400 Triethyleneglycol-bis[3-(3-tert-butyl-4-hydroxy-5-methylphenyl) propionate] 587 3 0.04

PEN
51700 2-(4,6Diphenyl-1,3,5-triazin-2-yl)-5-(hexyloxy)phenol) 425 0.05 0.5
60320 2-(2-Hydroxy-3,5-bis(1,1-dimethylbenz-yl)phenyl)benzo-triazole 448 1.5 0.5
71680 Pentaerythritol tetrakis[3-(3,5-di-tert-butyl-4-hydroxyphenyl)-propionate] 1178 60 0.1
74240 Phosphorous acid, tris(2,4-di-tert-butylphenyl)ester 647 60 0.1
94400 Triethyleneglycol-bis[3-(3-tert-butyl-4-hydroxy-5-methylphenyl) propionate] 587 3 0.1

PET
51700 2-(4,6Diphenyl-1,3,5-triazin-2-yl)-5-(hexyloxy)phenol) 425 0.05 0.2
60320 2-(2-Hydroxy-3,5-bis(1,1-dimethylbenz-yl)phenyl)benzo-triazole 448 1.5 0.2
60480 2,2’-Methylenebis(4-methyl-6-tert-butyl-phenol) 358 30 0.2
71680 Pentaerythritol tetrakis[3-(3,5-di-tert-butyl-4-hydroxyphenyl)-propionate] 1178 60 0.08

PA
38820 Bis(2,4-di-tert-butylphenyl)penta-erythritol diphosphite 605 0.6 0.125
53200 2-Ethoxy-2’-ethyloxanilide 312 30 0.5
59120 1,6-Hexamethylene-bis(3-(3,5-di-tert-butyl-4-hydroxyphenyl)propionamide 637 45 0.5
60320 2-(2-Hydroxy-3,5-bis(1,1-dimethylbenz-yl)phenyl)benzo-triazole 448 1.5 0.5
60480 2,2’-Methylenebis(4-methyl-6-tert-butyl-phenol) 358 30 0.5
68320 Octadecyl 3-(3,5-di-tert-butyl-4-hydroxy-phenyl)propionate 531 6 0.5
71680 Pentaerythritol tetrakis[3-(3,5-di-tert-butyl-4-hydroxyphenyl)-propionate] 1178 60 0.5
74240 Phosphorous acid, tris(2,4-di-tert-butylphenyl)ester 647 60 0.5
81200 Poly[6-[(1,1,3,3-tetramethylbutyl)-amino]-1,3,5-triazine-2,4-diyl]-[(2,2,6,6- 2000- 3 0.5
tertamethyl-4-piperidyl)imino]-hexamethylene-[(2,2,6,6-tertamethyl-4- 3100
piperidyl)imino]
81220 Poly-[[6-[N-(2,2,6,6-tetramethyl-4-piperidinyl)-n-butylamino]-1,3,5-triazine-2,4- 2600- 5 0.5
diyl][(2,2,6,6-tetramethyl-4-piperidinyl)imino]-1,6-hexanediyl[(2,2,6,6- 3400
tetramethyl-4-piperidinyl)imino]]-alpha-[N,N,N',N'-tetrabutyl-N"-(2,2,6,6-
tetramethyl-4-piperidinyl)-N"-[6-(2,2,6,6-tetramethyl-4-piperidinylamino)-hexyl]-
[1,3,5-triazine-2,4,6-triamine]-omega-N,N,N',N'-tetrabutyl-1,3,5-triazine-2,4-
diamine]
83595 Reaction product of di-tert-butyl phosphonite with biphenyl, obtained by 991 18 0.25
condensation of 2,4 di-tert-butylphenol with Friedel-Crafts reaction product of
phosphorus trichloride and biphenyl
92880 Thiodiethanol-bis(3-(3,5-di-tert-butyl-4-hydroxyphenyl)propionate) 643 2.4 0.5
93120 Thiodipropionic acid, didodecyl ester 515 5 0.25
94400 Triethyleneglycol-bis[3-(3-tert-butyl-4-hydroxy-5-methylphenyl) propionate] 587 3 0.5
95200 1,3,5-Trimethyl-2,4,6-tris(3,5-di-tert-butyl-4-hydroxybenzyl)benzene 775 60 0.5

PM/REF, PM/REF-numbers; Mr, molecular mass; CP,0, highest concentrations in

polymer
 Europe Direct is a service to help you find answers to your questions about the European Union
Free phone number (*): 00 800 6 7 8 9 10 11
(*) Certain mobile telephone operators do not allow access to 00 800 numbers or these calls may be billed.

A great deal of additional information on the European Union is available on the Internet.
It can be accessed through the Europa server http://europa.eu

How to obtain EU publications

Our publications are available from EU Bookshop (http://publications.europa.eu/howto/index_en.htm),

where you can place an order with the sales agent of your choice.

The Publications Office has a worldwide network of sales agents.

You can obtain their contact details by sending a fax to (352) 29 29-42758.
 LB-NA-27529-EN-N
JRC Mission

As the Commission’s
in-house science service,
the Joint Research Centre’s
mission is to provide EU
policies with independent,
evidence-based scientific
and technical support
throughout the whole
policy cycle.

Working in close
cooperation with policy
Directorates-General,
the JRC addresses key
societal challenges while
stimulating innovation
through developing
new methods, tools
and standards, and sharing
its know-how with
the Member States,
the scientific community
and international partners.

Serving society
Stimulating innovation
Supporting legislation

doi:10.2788/04517

ISBN 978-92-79-52790-6