Introduction to Migration Modelling - European Commission · JRC Guideline on migration modelling...
Transcript of Introduction to Migration Modelling - European Commission · JRC Guideline on migration modelling...
Introduction to
Migration Modelling
► Historical Context
► Scope and condition of use
► Migration model for monolayer
materials
► Migration from multilayer packaging
► Regulatory context
► Modelling for compliance purpose
► JRC Guideline on migration modelling
Content
► Historical Context
Short History
Adolf Fick: - Physiologist employed at the
University of Zurich
– published in 1855 in "Poggendorf's
Annalen der Physik“ an article
titled: „About Diffusion".
- investigated diffusion of water
through membranes.
- theoretical / phenomenological
approach – today we would call it
“linear-response theory”.
Short History
Albert Einstein: - Investigations on Theory of
Brownian Movement (1926).
- Thermal induced random walk of
molecules.
- Driving force is a difference in
the chemical potential between
two adjacent media.
► scientific work world wide in the 80th and 90th
- FDA: Limm and Holifield
- EU: Zweifel, Piringer
- others
► various EU projects
- Migration Modelling
- Recyclability
- Certified Reference Materials
- Foodmigrosure
- Migresives
► various national projects
- Multilayer Modelling (DE)
- others
Short History
► Scope and conditions of
use
► to predict migration processes
- conservative, upper bound migration values for
compliance purposes
- realistic migration values for packaging development
and exposure estimation
► for monolyer plastics
- defined boundary conditions
- limited to monolyer in contact with a well mixed liquid
- "simple" mathematics
► for multilayer plastics
- defined boundary conditions
- complex packaging structures
- complex numerical mathematics
Scope and conditions of use
► Migration model for
- monolayer and
- multilayer materials
Diffusion
Convection
Evaporation
Reaction
Partitioning
Migration - mass transfer
>>> change of concentration
with time
t
c
Diffusion
Convection
Reaction
Evaporation
Partitioning
. . .
Migration ~ Diffusion
Mass transfer
simplifying assumption
Migration ~ Diffusion
Migration process (Mass transfer)
● Diffusion process is the rate
determining step
● Diffusion process is determined
by:
- mobility of the polymer
(AP, polymer specific constant)
- size of the migrant
(Mr, molecular weight)
- temperature
(T, temperatur)
2
2
x
cD
t
c
Fick's 2nd law of diffusion
(one dimensional):
c - concentration
t - time
x - distance
D - diffusion coefficient
D
P - polymeric material
M - contacting medium
P M
Mass transfer ~ diffusion
Diffusion models
» monolayer materials
(monolayer)
D/K
» multilayer materials
(multilayer)
(D/K)n
D,K - mass transfer constants
P M
migrant
DP
KP,M
general diffusion model:
D/K/.../DD/K - polymer/liquid
D/K/D - polymer/solid
D/K/D/K - polymer/coating/liquid
...
(D/K)n/D - general
Diffusion models
polymeric materials in contact with ...
D - diffusion coefficient [cm²/s]
D0 - pre-exponential factor
EA - activation energy [J]
R - gas constant [8,314 J/mol K]
T - temperature [K]
RT
EA
eDD
0D RT
EA
eDD
0D
Diffusion coefficient (how fast is the migration)
P M
migrant
DP
KP,M
FüllgutPackstoff
cP,0
t = 0
migration
t =
cP,0 cF,0 = 0 cP, cF,
,
,
,
F
P
FPc
cK
Partitioning
Partition coefficient, KP,F
K - partition coefficient
c - concentration
P - polymeric material
M - contacting medium
- at equilibrium
Partition coefficient (how far goes the migration)
,
,
,
M
P
MPc
cK
P M
migrant
DP
KP,M
Solution of the diffusion equation
» » analytical solution
- only monolayer
- only mean concentration in the
contacting medium and the
polymeric material
(no concentration profile available)
- no exchange cycles can be
simulated
→ see J. Crank
("Mathematics of Diffusion")
D/K
H.S.Carslaw &
J.C.Jaeger: Conduction of heat in solid
J. Crank: The Mathematics of Diffusion
Solution of the diffusion equation
12
2
220,
,exp
1
121
1 n P
nP
n
PPP
tU
d
qtD
qdc
A
m
UP
PU
K
VV
,
/
Analytical solution of the diffusion equation
mU,t/A in [µg/cm2] - Migration
t in [s] - contact time
A in [cm2] - contact area
cP,0 in [mg/kg = ppm] - initial concentration of the migrant in the plastic
P and F in [g/cm3] - density of plastic and food or simulant
dP in [cm] - thickness of plastic
DP in [cm2/s] - diffusion coefficient of the migrant in the plastic
VP and VF in [cm3] - volume of plastic and food or simulant
KP,F = cP,P / cF,F - partition coefficient (condentration relation of migrant
(w/v) in plastic and food at equilibrium)
tan qn = - qn, - qn positive roots of the trigonometric equation
(D/K)n
» » numerical solution
- for multilayer materials
- concentration profile available
- exchange cycles can be
simulated
→ see standard textbooks for
Numerical Mathematics
e.g. Finite Elements und Finite
Differences algorithms
Solution of the diffusion equation
► must solve the diffusion
equation numerically (partial
differential equation, PDE)
→ the analytical solution of the
diffusion equations serves as
reference for validation
→ validation required, i.e.
experimental examples must be
reproduced correctly
Software Tools
Estimation of mass transfer constants
» Diffusion coefficients (DP)
Arrhenius DP=f(D0,EA,T)
Piringer DP=f(AP',tau,Mr,T)
Brandsch DP=f(Tg,M,T) - new
» Partition coefficients (KP,M)
worst case K=1, (VP<<VM)
Piringer K=f(p, Ma, Wa, GF)
Brandsch K=f(SW)
K=f(PO/W) - new
DP - diffusion coefficient
D0 - pre-exponential factor
(Arrhenius)
EA - activation energy (Arrhenius)
T - temperature [K]
AP' - polymer specific constant
(Piringer)
tau - polymer specific temperature
constant (Piringer)
M - molecular weight [g/mol]
Mr - relative molecular weight
Tg, - glass temperature of polymer
KP,M - partition coefficient
VP - volume of polymer
VW - volume of medium
p - vapour pressure of migrant
SW - water solubility of migrant
PO/W - octanol/water- partition
coefficient of migrant
► Estimation of diffusion
coefficients
Migration ~ Diffusion
Migration process (Mass transfer)
● Diffusion process is the rate
determining step
● Diffusion process is determined
by:
- mobility of the polymer
(AP, polymer specific constant)
- size of the migrant
(Mr, molecular weight)
- temperature
(T, temperatur)
D - diffusion coefficient [cm²/s]
D0 - pre-exponential factor
EA - activation energy [J]
R - gas constant [8,314 J/mol K]
T - temperature [K]
RT
EA
eDD
0D RT
EA
eDD
0D
Diffusion coefficient
P M
migrant
DP
KP,M
DP - Diffusion coefficient (D0 = 104 cm²/s)
AP= AP„-/T - material specific constant
( - material specific temperature constant)
Mr - relative molar mass of migrant in Dalton
T - temperature in K
EA - reference activation energy
(= R·10454 = 86,9 kJ, R = 8,314 J/K·mol)
TR
RMMADD rrPP
10454003.01351.0exp
3/2
0
Estimation of diffusion coefficients (Piringer)
Polymer AP´ T [°C] cP,0 [%]
LDPE 11.5 0 < 80 < 1
LLDPE 11.5 0 < 100 < 1
HDPE 14.5 1577 < 90 < 1
PP(homo) 13.1 1557 < 120 < 1
PP(random) 13.1 1557 < 120 < 1
PP(rubber) 11.5 0 < 100 < 1
FOOD CONTACT MATERIALS
PRACTICAL GUIDE
“A PRACTICAL GUIDE FOR USERS OF EUROPEAN DIRECTIVES ”
Upper limit AP*-values (polyolefines)
Polymer AP´ T [°C] cP,0 [%]
PS 0 0 < 70 < 1
HIPS 1 0 < 70 < 1
PET 6 1577 < 175 < 1
PEN 5 1557 < 175 < 1
PA 6,6 2 0 < 100 < 1
FOOD CONTACT MATERIALS
PRACTICAL GUIDE
“A PRACTICAL GUIDE FOR USERS OF EUROPEAN DIRECTIVES ”
Upper limit AP*-values (non-polyolefines)
Table 6 Statistical validation of AP’-values for migration modelling under ’worst case’
conditions
Polymer
AP’
s
AP’(max)
AP’(min)
N
t
AP’*
LDPE 10.0 1.0 11 7.0 27 1.7 11.7 0
HDPE 10.0 1.9 12.6 5.0 49 1.68 13.2 1577
PP 9.4 1.8 12.9 6.2 53 1.68 12.4 1577
PET 2.2 2.5 7.2 -4.3 58 1.67 6.35 1577
PEN -0.34 2.4 3.8 -5.5 38 1.7 3.7 1577
PS -2.8 1.25 0.0 -6.5 32 1.7 -0.7 0
HIPS -2.7 1.67 0 -6.2 33 1.7 0.1 0
PA (6,6) -1.54 2.0 2.3 -7.7 31 1.7 1.9 0
Other Polymers (EU-Project Migration Modeling)
AP= AP'-/T Food Additives and Contaminants, 2005; 22(1): 73–90
Diffusion properties of different
polymers can be compared
based on their AP-value,
i.e. mobility of the polymer
► high AP-values account for high
mobility of the polymer (flexible
polymers) and high diffusion
coefficients respectively
► low AP-values account for low
mobility of the polymer (rigid
polymers) and low diffusion
coefficients respectively
Migration process (Mass transfer)
DP AP
[cm²/s]
gases ~ 10-1
liquids ~ 10-5 20
viscous liquids ~ 10-6 18
soft PVC ~ 10-7 16
Polymere T > Tg
LDPE ~ 10-9 11
HDPE ~ 10-10 9
PP ~ 10-11 7
Polymere T < Tg
PA ~ 10-13 2
PS ~ 10-14 0
PET ~ 10-15 -2
rigid PVC ~ 10-16 -4(Tg - glas temperature)
Diffusion coefficients (at T=20°C, Mr=300 g/mol)
► Regulatory context
EU legislation (FCM 2002/72/EC, Article 8)
real diffusions coefficient: DP <--> AP
„upper limit“ diffusion coefficient : DP* <--> AP*
- an „upper limit“ diffusions coefficient DP* gives a „worst
case“ migration estimation
Legal requirements for AP-values
Functional barrier concept
Article 7a
» Only glass and some metals may
ensure complete blockage of
migration. (absolute barrier)
» Plastics may be partial functional
barriers with properties and
effectiveness to be assessed and
may help reducing the migration of
a substance below a SML or a limit
of detection.
► FB consists of one or several layers,
► FB assures that the migration of positively listed
substances does mot exceed the specific migration limit
► FB assures that the migration of substances not listed
does not exceed the limit of 0,01 mg/kg food (including
set-off)
► not allowed are substances classified as proved or
suspect "carcinogenic", “mutagenic” or “toxic to
reproduction”, substances in Annex I to Council Directive
67/548/EEC
► FB prevents the migration of "not intentionally added
substances" (NIAS = impurities, decomposition products,
etc.), i.e. keeps their migration not detectable (detection
limit 0,01 mg/kg food).
Functional barrier concept
contact
medium
Direct and indirect contact
t
A
m tF ,
2d 10d
concave
I > 0
Food contact
layer
two data points:
2d@60°C
10d@60°C
Functional barrier concept
t
A
m tF ,
2d 10d
convex
concave
I > 0
I < 0
two data points:
2d@60°C
10d@60°C
printing ink
Functional barrier concept
FB
FB
D
d 2
6
1
dFB - thickness
DFB - diffusions coefficient
- "lag time"
Simulanz /Lebensmittel
Migrant
Kunststoff
DP
KFB,F
FBSimulanz /
Lebensmittel
Migrant
Kunststoff
DP
KFB,F
FBSimulanz /
Lebensmittel
Migrant
Kunststoff
DP
KFB,F
FBSimulanz /
Lebensmittel
Migrant
Kunststoff
DP
KFB,F
FBSimulanz /
Lebensmittel
Migrant
Kunststoff
DP
KFB,F
FBSimulanz /
Lebensmittel
Migrant
Kunststoff
DP
KFB,F
FB
KP,FB
DFB
"lag time"
Theoretical understanding
DEHA, Verbund 1 (NC/20µmOPP/Adh/30µmOPP)
21Tage bei 20°C, Ethanol95%
0
100
200
300
400
500
600
700
0 5 10 15 20 25 30 35 40
Zeit [Tage]
Mig
rati
on
[µ
g/d
m²]
exp.
calc. N
Functional barrier
multilayer structure:(from left to right)
ink(1µm)
OPP(20µm)
adh.(2,5µm)
OPP(30µm)
time [days]
one sided migration test
food simulant (D) substitute: 95% ethanol
temperature: 20°C
migrant: DEHA
migration kinetic
O
O
O
O
CH3
CH3
CH3 CH3
DEHA, Verbund 1 (NC/20µmOPP/Adh/30µmOPP)
10 Tage bei 40°C, Ethanol95%
0
500
1000
1500
2000
2500
0 5 10 15 20 25 30 35 40 45
Zeit [Tage]
Mig
rati
on
[µ
g/d
m²]
calc. N
exp.
time [days]
one sided migration test
food simulant (D) substitute: 95% ethanol
temperature: 40°C
migrant: DEHA
multilayer structure:(from left to right)
ink(1µm)
OPP(20µm)
adh.(2,5µm)
OPP(30µm)
Functional barrier
migration kinetic
O
O
O
O
CH3
CH3
CH3 CH3
DEHA, Verbund 1 (NC/20µmOPP/Adh/30µmOPP)
7 Tage bei 60°C, Ethanol95%
0
500
1000
1500
2000
2500
0 1 2 3 4 5 6 7 8
Zeit [Tage]
Mig
rati
on
[µ
g/d
m²]
exp.
calc. N
time [days]
one sided migration test
food simulant (D) substitute: 95% ethanol
temperature: 60°C
migrant: DEHA
multilayer structure:(from left to right)
ink(1µm)
OPP(20µm)
adh.(2,5µm)
OPP(30µm)
Functional barrier
migration kinetic
O
O
O
O
CH3
CH3
CH3 CH3
► JRC Guideline on
Migration Modelling