Content uploaded by Vincent Dudler
Author content
All content in this area was uploaded by Vincent Dudler on Jan 14, 2016
Content may be subject to copyright.
Mitt. Lebensm. Hyg. 96, 281– 297 (2005) 281
Heat stability and migration from
silicone baking moulds
Roger Meuwly, Kurt Brunner, Céline Fragnière, Fritz Sager and Vincent Dudler
Swiss Federal Office of Public Health, P.O. Box, 3003 Bern
Received 27 May 2005, accepted 23 August 2005
Introduction
Flexible silicone baking moulds, available in different forms and colours, have
been marketed for some years as alternatives to traditional metal bakeware. Accord-
ing to the manufacturers, silicone baking moulds offer numerous practical advan-
tages: they are non-stick, can be used in the microwave, are dishwasher-safe and
resistant to a large temperature range.
Silicones constitute a group of polymeric chemical substances containing poly-
siloxanes characterized by Si-O-Si and Si-C bonds. Silicones include a range of
products with a variety of properties and applications: silicone liquids (release
agents, impregnating agents for textiles, additives, etc.), silicone pastes (lubricants,
etc.), silicone resins (heat-resistant and release coatings, etc.) and silicone elastomers
(sealants, baking moulds, etc.).
Silicones, for food contact applications, are not regulated at the EU level but at
the national level for example in Germany (1), France (2), etc. Annex 1 of the Regu-
lation (EC) 1935/2004 includes Silicones within a list of materials and articles which
may be covered in the future by specific measures (3). In Switzerland, there are no
specific regulations for articles in silicones but they must fulfill the general require-
ments of the Swiss Ordinance concerning “Articles of Daily Use” (4) which states,
as in the article 3 of the Regulation (EC) 1935/2004, that articles should not release
to or form in foodstuffs any substance in a quantity that poses a risk to human
health or that adversely affects the organoleptic properties of the food. The Council
of Europe, in which Switzerland actively participates, issued “Resolution AP (2004)
5 on silicones used for food contact applications” which contains some specific
requirements and an inventory list of substances used in the manufacture of the
products (5). One requirement states: “The total of all substances migrating into
food from silicone materials or articles should not exceed 10 mg/dm2of the surface
Original papers
Meuwly 19.10.2005 13:12 Uhr Seite 281
area of the final material or article or 60 mg/kg of food, this being considered as the
overall migration limit”. The same limit is given in the French legislation (2).
Even though silicone elastomers demonstrate a high degree of thermal stability and
excellent resistance to aging, high temperatures lead to depolymerization of the elas-
tomer, with subsequent volatilization and migration of certain substances. The few pub-
lications concerning the suitability of silicones as food contact materials have indeed
shown that a certain quantity of substances migrates from silicone-based articles (6, 7).
The objective of this work was to study the release of chemical substances from
the silicone baking moulds at various temperature, especially around 200–220°C,
temperatures attained in an oven when cakes and pastries are baked. Some tests at
higher temperatures up to 280°C were also performed to check the statement of the
manufacturers concerning the heat stability of the moulds. The tests currently spec-
ified for migration for food contact applications involve temperatures not exceeding
175°C. It is thus very important to develop migration procedures that make possi-
ble the investigation in a real temperature range.
The stability of the silicone moulds at high temperature was checked by two
tests: measurement of the loss of volatile components during heating as described
in the German recommendations (1) and the French legislation (2) and overall
migration (OM) studies with modified polyphenylene oxide as a food simulant.
Some indications about the nature of the migrants were also collected during this
study.
Experimental
Samples
Silicone baking moulds of various shapes were received or bought from differ-
ent retail stores (Table 1). These kitchenware are used in food preparation, especially
for baking cakes, pastries, etc., in a microwave or in a conventional oven and for
storage of food in the freezer. According to the manufacturers, the moulds can with-
stand extreme temperatures such as the ones reported in Table 1.
282 Mitt. Lebensm. Hyg. 96 (2005)
Meuwly 19.10.2005 13:12 Uhr Seite 282
Table 1
Characteristics of silicone baking moulds samples
Sample Nr Type Temperature Thickness Colour
mould size (cm) min/max (°C)1(mm)2
A4 Baking forms –25 to 220 1.97 Red
16153
B12 Mini-Gugelhofs –25 to 250 1.17 Red
3022.53.5
C4 Mini-puddings –60 to 260 1.39 Blue
17175.7
D6 Muffins –40 to 280 0.95 Orange
3017.57
E24 Minicakes –40 to 280 1.02 Black
60403
F9 Minicakes –40 to 280 0.97 Red
60403
G9 Minicakes –60 to 280 1.20 Red-brown
3017.53.5
1according to the manufacturers
2average value, measured at different positions
Chemicals and apparatus
Modified polyphenylene oxide (MPPO, Tenax, temperature stability: Tmax=
350°C) 60 to 80 mesh (Supelco): before the first use MPPO was extracted 6 h with
diethylether, transferred in a Petri dish which was placed in a fume hood to allow the
diethylether to evaporate. The complete evaporation of the solvent was achieved by
placing the Petri dish into an oven at 160°C for 6 h. Dry MPPO was stored into a
closed flask. Diethylether p.a. (Merck) was distilled over 1 g sodium hydride per liter
diethylether. The samples were aged in an oven (Heraeus T 5042 EK) without air cir-
culation. The oven was regulated within the permitted tolerance (±5°C) on the tem-
perature indication given by an additional thermocouple (type K) positioned in the
middle of the oven at the sample location. The temperature profile of each experiment
was recorded by a datenlogger (Ecolog TN4 with an accuracy of ±0.6°C at 200°C).
GC-MS analyses were performed using a Hewlett-Packard GC 5890 equipped
with an HP 5972 MSD detector. The capillary column was a DB-1ht (J & W Scien-
tific) with the dimension 30 m0.25 mm i.d. and a coating film thickness of 0.1 µm.
The GC initial settings were: injector temperature 270°C and column temperature
70°C followed by a temperature program of 20°C/min up to 360°C. The mass spec-
trometer was operated at a transfer line temperature of 280°C with EI at 70eV at a
scan mass-range 50–400 amu. For spectra interpretation a Wiley 7 library was used.
MALDI-TOF/MS spectra were obtained from Solvias AG, 4002 Basel, using an
LDI-1700 spectrometer (Linear Scientific, Inc. Reno, USA) equipped with a laser
operating at 337 nm. The matrix used was 2,5-dihydroxybenzoic acid. The mass-
range was 10000 Da (4 calibrating substances from 1300 Da to 7000 Da) with a mass
accuracy of ±0.1%. Single charged ions ([M+Na]+) were detected and analyzed.
Mitt. Lebensm. Hyg. 96 (2005) 283
Meuwly 19.10.2005 13:12 Uhr Seite 283
Overall migration at high temperatures
According to the Resolution of the Council of Europe (5), migration tests
should be conducted according to the EC Plastic Directives. The overall migration
tests of the silicone baking moulds were performed according to the Commission
Directive 97/48/EC laying down the basic rules necessary for testing migration of
the constituents of plastic materials and articles intended to come into contact with
foodstuffs (8). The experimental methods are described in the European Standard
EN 1186-13: “Test methods for the determination of the overall migration at high
temperatures” (9). The determination of the overall migration into fatty food simu-
lants from plastic material articles is performed by total immersion of test specimens
in a fatty food simulant, normally olive oil, at temperatures up to 175°C for selected
times. At high temperature the migration tests with olive oil introduce a number of
analytical difficulties (oxidation of oil, ...). Replacement of olive oil by an appropri-
ate absorbent material is an alternative accepted by the legislation: MPPO is listed
like isooctane and ethanol as a test medium in substitute fat test (8). In this substi-
tute procedure, the mass of the components absorbed on modified polyphenylene
oxide (MPPO) is taken as the measure for the assessment of the overall migration
into a fat simulant. This procedure with MPPO has allowed to test silicone baking
moulds up to 280°C, maximum temperature according to the manufacturers of the
heat stability of some articles.
When an article is intended to come into repeated contact with foodstuffs like
these silicone baking moulds, tests shall be carried out three times (M1, M2 and M3)
on a single sample using a fresh sample of simulant on each test as described in the
European Standard EN 1186-1: “Guide to the selection of conditions and test meth-
ods for overall migration” (10). Its compliance shall be checked on the basis of the
level of the migration found in the third test (M3).
The residue of the migration was analyzed by GC-MS and MALDI-TOF to
investigate the nature of the migrants.
Principle
The surface of the sample to be tested was covered with modified polyphenylene
oxide (MPPO) and kept at the selected time/temperature test conditions in an oven.
The MPPO were then extracted with diethylether. The extract was evaporated to
dryness using a nitrogen stream and the residue mass was determined gravimetri-
cally. The residual mass does not contain volatiles.
Sample preparation
The determination of the overall migration of silicone baking moulds was
performed on the flat bottom part of the article in contact with the food. The dust
on the samples was removed with compressed air and the samples were treated with
antistatic cloth.
284 Mitt. Lebensm. Hyg. 96 (2005)
Meuwly 19.10.2005 13:12 Uhr Seite 284
Procedure
The surface of the sample to be tested was covered with MPPO taking 4 g
MPPO per dm2of surface area. For the blank determination, an empty Petri dish
was taken and the same mass of MPPO was placed on it. The examination of a blank
was carried out in parallel. The test sample was covered with a glass plate and put in
the preheated oven at the required temperature. The sample was left in the oven for
the selected period of time (M1, 1 h in general) as soon as the temperature in the
oven had once again reached a temperature within the permitted tolerance (±5°C)
for the test temperature. The average time for the oven to reach the test temperature
again was 13 min at 175°C, 16 min at 220°C, 23 min at 260°C and 27 min at 280°C.
The sample was removed from the oven and allowed to cool to room temperature
without removing the glass cover. The MPPO was carefully transferred into a
200 ml Erlenmeyer flask and 20 ml of diethylether was added. The solution was
stirred for 1 minute and then allowed to stand for 1 minute without shaking. The
diethylether solution was decanted from the MPPO through a filter into a 200 ml
vial. The extraction procedure of MPPO was repeated twice with each time 30 ml
diethylether. The filter was rinsed with 10 ml diethylether and the combined
diethylether solution was concentrated with a Rotavap to approximately 5 ml. The
residue was transferred to a weighted 15 ml vial and concentrated further to dry-
ness, using a stream of nitrogen until constant weight was obtained. The mass of the
residue was determined by substracting the original mass of the vial from the mass
of vial plus residue.
The second migration test (M2) was performed with the sample used for the first
test via the same procedure as described above. The third migration test (M3) was
performed with the sample used for the first and the second tests via the same pro-
cedure. Prolonged heating time tests were performed on two samples to check
the effect of a repeated use on the overall migration. Again the overall migration
value during a repeated 1 h test at the selected temperature was determined during
100 hours. The sample was heated with MPPO for (x-1) h, MPPO was discarded, a
fresh sample of MPPO was taken and the sample was heated for a further 1 h. The
migrants in the fresh MPPO were determined to give the overall migration value for
1 h at time x.
The precision of the measurements was assessed on different samples at 220°C.
This was shown to be sample dependent. The standard deviation under repeatability
conditions oscillated between 12–28% for the first migration test (M1) and
decreased steadily for each additional heating cycle to reach 6–15% for the third
migration test (M3).
Volatiles
The total release of volatile substances was measured by simply weighing the
moulds before and after the heating during 4 hours at 200°C as described in the
German recommendations (1) and the French legislation (2). About 10 g of the sam-
Mitt. Lebensm. Hyg. 96 (2005) 285
Meuwly 19.10.2005 13:12 Uhr Seite 285
ple were cut in pieces of about 12 cm and left for 48 h above calcium chloride in a
desiccator. The pieces were weighed (precision of ±0.1 mg) and then put in the oven
for 4 h at 200°C. After cooling in a desiccator the pieces were weighed again and the
percentage of volatiles was calculated based on the ratio of the weights. The release
of volatiles should not exceed 0.5% (1, 2).
Results and discussion
Overall migration
The determination of the overall migration of the sample A was performed at
different temperatures during repeated 1 h contact cycle with MPPO (Figure 1). Up
to 100°C the sample showed low migration value, especially at the third 1 h expo-
sure. From 150°C, the values increased rapidly for the three exposure time and the
limit of 10 mg/dm2prescribed by the Resolution of the Council of Europe (5) is
reached in most cases. These results confirm the good stability of silicone elastomers
up to 150°C. All experimental points can be fitted with a sigmoidal curve of the
type:
K
MT=(1+b·e–R·T)
where T is the temperature of observation in degree Celsius, R depicts a degra-
dation rate of the silicone, K and b are two constant factors. The curve in Figure 1
shows the amount of migrants liberated by sample A at the third 1 h exposure test
(M3) used to check the compliance of articles intended to come into repeated con-
tact with foodstuffs.
286 Mitt. Lebensm. Hyg. 96 (2005)
0 50 100 150 2 00 250 30 0
Heatin
g
tempe
r
atu
r
e[°C]
0
5
10
15
20
25
30
35
40
45
50
Overall migration during one hour [mg/dm
2
]
1
st
heating cycle (M1)
2
nd
heating cycle (M2)
3
rd
heating cycle (M3)
Fit curve of M3 data
Figure 1 Overall migration of sample A during a repeated one hour test
Meuwly 19.10.2005 13:12 Uhr Seite 286
The thermal stability of the different samples was tested at temperature between
175 and 280°C (Table 2). At 175°C for the first 1 h contact all the investigated sam-
ples showed high overall migration values, between 25.5 mg/dm2and 49.2 mg/dm2.
At the third exposure the values M3 used to check the compliance of articles
intended to come into repeated contact with foodstuffs were slightly lower. Again
no significant difference was observed between the samples. The increase of the test
temperature from 175 to 220, 260 and 280°C showed in parallel a slight increase of
the migration values. The behavior of the various samples at these temperatures was
almost the same. The true values for the overall migration in fatty food simulants are
subject to uncertainty owing to the lack of precision inherent in the method. A
point to consider is that at high temperatures, the more volatile compounds of the
migrate could evaporate without being absorded on the MPPO and consequently
could not contribute to the overall migration value. The values in Table 2 possibly
underestimate the real loss of material during the heating.
Table 2
Overall migration during a repeated one hour test at different temperatures
Sample A B C D E F
mg/dm2mg/dm2mg/dm2mg/dm2mg/dm2mg/dm2
T=175°C
M1 25.5 25.9 29.1 49.2 33.8 25.8
M2 21.1 25.7 21.6 33.7 26.1 19.7
M3 21.8 21.8 23.9 31.7 23.2 17.7
T=220°C
M1 34.2 44.7 38.2 48.1 41.5 33.7
M2 31.8 41.9 33.4 37.2 36.0 32.1
M3 28.4 35.9 28.9 32.8 41.9 29.0
T=260°C
M1 45.3 42.7 40.9 44.7 50.5 37.7
M2 40.3 45.1 32.6 40.5 48.2 33.7
M3 34.7 31.8 27.2 27.9 34.1 24.9
T=280°C
M1 44.4 40.7 41.5 45.0 44.0 37.2
M2 32.1 33.5 38.4 39.0 37.6 32.3
M3 32.8 31.6 39.6 31.3 34.9 27.3
M1: Sample X heated 1 h (X1)
M2: Sample X1 heated 1 h (X2)
M3: Sample X2 heated 1 h (X3)
The values of the overall migration during a 1 h heating cycle at 220°C based on
the heating time of samples A and D are depicted in Figure 2a. These values give an
indication of the quantity of the migrants liberated each time the mould is heated for
1 h at 220°C. The values of the overall migration decrease rapidly during the first
exposure of the moulds with MPPO. But even after 100 h of heating, a small quan-
tity of substances continue to migrate from the moulds to the MPPO. This certainly
indicates a continual degradation (depolymerization) process which always supplies
Mitt. Lebensm. Hyg. 96 (2005) 287
Meuwly 19.10.2005 13:12 Uhr Seite 287
new substances able to migrate. Mould A is two times thicker than mould D; this
certainly plays a role in the quantity of the products of decomposition and there-
after in the quantity of the migrants. The data were not corrected as regards the
thickness of the samples in order to depict effective migration values. By plotting
the measured values as the sum of the measured migration in function of the square
root of time (Figure 2b) it appears clearly that the loss is not linear. This indicates
that the kinetics of loss is not principally governed by a diffusion process. Other
processes, such as the degradation rate of the silicone backbone or the direct
volatilization of migrants at the sample surface are likely to play a role in this meas-
urement.
The results do not take into account the “simulant D reduction factor” used
to correct the results for foodstuffs containing fat as indicated in Directive
85/572/EEC (11). For example a reduction factor of 3 would apply for confec-
tionery products in paste form with fatty substances on the surface and a reduction
factor of 5 for fresh pastry, cakes and other baker’s wares with fatty substances on
the surface. At all temperatures, even at 280°C at the third 1 h exposure contact and
when a reduction factor of 5 is used, all the investigated silicone baking moulds had
an overall migration value below the recommended limit of 10 mg/dm2(5).
288 Mitt. Lebensm. Hyg. 96 (2005)
0 20406080100120
Heatin
g
time [ho u
r
]
0
10
20
30
40
50
Overall migration [mg/dm
2
]
sample A
sample D
01234
(time)1/2
0
100
200
300
Cumula tive loss [mg/dm
2
]
b
a
Figure 2 Overall migration during one hour test at 220°C after a prolonged heating time
Meuwly 19.10.2005 13:12 Uhr Seite 288
Volatiles
The total release of volatile substances after 4 h at 200°C (Table 3) was measured
by weighing the moulds before and after the heating as described in the German
recommendations (1) and the French legislation (2). During the first heating cycle of
4 h at 200°C, the weight loss of the samples was 0.11–1.78%. As shown in Table 3
the values of the weight loss decrease rapidly after the first heating cycle. Already
during the second and during the whole of the following heating cycle, all the sam-
ples showed a value lower than 0.21%, below to the limit of 0.5% given in the Ger-
man and French recommendations. The high weight loss in the first heating cycle is
most likely due to residual solvents and/or by-products formed during the manu-
facture of the articles or due to the fact that the time for the postcuring of the article
was not sufficient. This process is easy but requires much energy. To avoid addi-
tional expenses, some firms leave out this last stage and ask their customers to heat
the mould at 230°C during 2 h without food before the first use, stating that the
possible development of some smoke is not deleterious.
Table 3
Weight loss of volatiles (%) at 200°C in heating cycles of 4 h
Number of heating cycle
Sample 1235
A 0.11±0.01* 0.04± 0.01 0.02± 0.01 0.01± 0.01
B 0.58±0.01 0.12±0.01 0.08±0.01 0.04 ±0.01
C 0.93±0.01 0.15±0.00 0.07±0.01 0.04 ±0.01
D 1.68±0.02 0.20± 0.02 0.11 ± 0.03 0.06±0.04
E 0.96±0.04 0.21±0.03 0.15±0.02 0.08 ±0.01
F 1.78 ±0.02 0.07±0.03 0.14 ± 0.03 0.12±0.09
G 0.46±0.01 0.12± 0.00 0.06 ± 0.01 0.04±0.01
* average of three measurements with the standard deviation s
By varying the temperature from 200 to 220 and then to 280°C (Tables 3–5), the
increase of the weight loss is considerable. Table 5 shows that the least stable sample
D lost 4.47% of its weight during the first 4 h of heating at 280°C and even 1.44%
during the 5th heating cycle of 4 h. These results show a strong dependence of the
amount of volatiles on the heating temperature and clearly indicate that these sili-
cone moulds are not as stable at high temperature as advertised by their manufac-
turers. The sample D, supposed to withstand temperature to 280°C, lost during the
first 12 h of heating (34 h heating cycle) 1.99% of its weight at 200°C, 2.38% at
220°C and even 8.58% at 280°C. After these prolonged thermal treatments, some
samples were discolored or brittle.
Mitt. Lebensm. Hyg. 96 (2005) 289
Meuwly 19.10.2005 13:12 Uhr Seite 289
Table 4
Weight loss of volatiles (%) at 220°C in heating cycles of 4 h
Number of heating cycle
Sample 1235
A 0.15±0.01* 0.06±0.01 0.05± 0.01 0.02±0.01
B 0.74±0.02 0.17±0.01 0.10±0.01 0.02± 0.01
C 1.21±0.03 0.14±0.02 0.10±0.03 0.08 ±0.01
D 2.11±0.06 0.19± 0.01 0.08 ± 0.02 0.08±0.03
E 1.37±0.09 0.19±0.05 0.14±0.03 0.11 ±0.01
F 1.86 ±0.08 0.17±0.03 0.14 ± 0.02 0.16±0.03
G 0.65±0.02 0.13± 0.01 0.08 ± 0.01 0.05±0.01
* average of three measurements with the standard deviation s
Table 5
Weight loss of volatiles (%) at 280°C in heating cycles of 4 h
Number of heating cycle
Sample 1235
A 0.78±0.08* 0.30±0.02 0.28± 0.04 0.25±0.02
B 3.01±0.11 2.24±0.01 2.23±0.21 1.34 ±0.22
C 4.11±0.25 2.87±0.18 1.70±0.21 0.63 ±0.01
D 4.47±0.22 2.25± 0.48 1.86 ± 0.37 1.44±0.38
F 3.27 ±0.07 1.51±0.08 1.40 ± 0.09 1.95±0.07
* average of three measurements with the standard deviation s
Characterisation of the migrants
Silicone baking moulds are silicone elastomers made from crosslinked polydi-
methylsiloxanes. Thermal degradation of silicone elastomers have been described in
many studies (12–15). The decomposition of the polymers occurs in three stages:
1) evaporation of volatile components, 2) thermal decomposition and 3) thermal
oxidation: The second and third processes often overlap. Additives and initial
removal of volatiles by postcuring increase the stability (16).
290 Mitt. Lebensm. Hyg. 96 (2005)
(
Si-O
)
n
CH3
CH3
(
Si-O
)
n-2
CH3
CH3
OSi Si
CH3
CH3
CH3
H3C
CH3
CH3
Figure 3 Structure of the migrants
Cyclic polydimethylsiloxanes (Cn) Linear polydimethysiloxanes (Ln)
n=5– 35 Si(CH3)2-groups
Meuwly 19.10.2005 13:12 Uhr Seite 290
According to the literature, the degradation products consist primarily of
cyclic polydimethylsiloxane (Cn) and linear polydimethylsiloxane (Ln) oligomers
(Figure 3). It is assumed that linear polydimethylsiloxanes are methyl terminated linear
chain without excluding the fact that some chains are hydroxyl terminated (17, 18).
The distribution of the different siloxanes in the residue of the global migration
at different temperatures were determined from the MALDI-TOF and the GC-MS.
In the MALDI-TOF, the oligomeres have a difference of 74.2 Dalton which is the
mass of the repeated unit Si(CH3)2O-group. The migrating substances mainly con-
Mitt. Lebensm. Hyg. 96 (2005) 291
8 9 10 11 12 13 14 15 16 17 18
Time [min]
0
200
400
600
800
1000
Abundance [a.u.]
C8 C9 C10
C11
C12
C13
C14 C15
C16
C17
C18
C19
C20
C21
C22C23
L8 L9 L10
L11 L12 C24
Figure 4 GC-MS of residue of sample A at the third one hour exposition at 175°C
9 101112131415161718
Time [min]
0
40
80
120
160
200
Abundance [a.u.]
L7
L8
L9
L10
L11
L12
L13
L14
L15
L16
L17
L18
L19
C8 C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22 C23
Figure 5 GC-MS of residue of sample B at the third one hour exposition at 175°C
Meuwly 19.10.2005 13:12 Uhr Seite 291
292 Mitt. Lebensm. Hyg. 96 (2005)
Table 6
Composition of residue in % from the overall migration tests
Temp. 175°C 220°C 260°C
Sample A B C D E F A B C D E F A B C D E F
Cn MW
C7 519 1.3
C8 593 2.2 1.0 1.1
C9 667 0.6 3.5 1.7 2.9
C10 742 1.1 1.0 5.5 2.6 1.2 5.6
C11 816 3.6 1.9 8.4 3.8 3.1 8.7 0.4
C12 890 6.2 3.1 11.1 4.5 7.3 11.0 1.0 1.1 1.0
C13 964 11.0 5.1 12.4 5.0 8.8 12.2 1.7 1.3 2.8 2.6 1.7 2.9
C14 1038 15.1 7.6 13.5 5.3 10.4 13.1 4.7 4.3 7.1 5.2 4.5 7.0 0.5 0.5
C15 1112 16.4 9.6 12.9 5.1 11.3 12.7 9.1 8.9 12.8 7.9 8.9 12.0 1.4 0.9 1.4
C16 1186 14.8 9.6 10.6 4.3 10.6 10.7 13.4 12.3 16.3 9.6 12.6 14.7 3.2 1.5 2.1 3.3 0.6 1.0
C17 1261 11.9 8.1 7.9 3.4 8.3 8.2 15.6 12.9 16.1 9.7 13.9 14.7 6.4 3.2 4.3 6.7 1.7 3.0
C18 1335 7.8 5.8 5.1 2.4 5.5 5.8 14.7 11.1 13.3 8.7 13.4 13.0 10.0 6.2 8.2 10.3 4.0 6.7
C19 1409 4.9 4.0 3.1 1.6 3.5 3.8 12.5 8.6 10.3 7.2 10.8 10.5 12.7 10.2 12.4 12.7 8.0 11.7
C20 1483 3.2 2.2 1.8 1.0 2.2 2.3 10.0 6.0 7.7 5.8 8.8 8.2 14.3 14.0 15.3 13.0 12.9 15.2
C21 1557 1.8 1.2 1.0 0.5 1.1 1.3 7.2 4.1 5.5 4.6 6.4 6.2 14.2 14.6 16.1 11.6 15.4 16.1
C22 1632 0.9 0.6 0.5 0.7 5.0 2.6 3.7 3.4 4.8 4.3 12.3 11.9 14.0 9.6 16.6 14.4
C23 1706 3.3 1.9 2.3 2.6 3.2 2.8 10.0 11.6 11.3 7.4 14.3 11.6
C24 1780 1.9 1.1 1.2 1.6 1.9 1.7 7.2 7.5 7.2 5.2 11.5 8.5
C25 1854 0.9 0.7 0.9 0.9 4.4 3.7 4.3 3.2 7.2 5.4
C26 1918 2.4 2.6 2.7 1.6 4.4 3.3
C27 2003 1.2 1.9 1.2 0.9 2.3 1.9
C28 2077 1.3 1.2
Meuwly 19.10.2005 13:12 Uhr Seite 292
Mitt. Lebensm. Hyg. 96 (2005) 293
Temp. 175°C 220°C 260°C
Sample A B C D E F A B C D E F A B C D E F
Ln MW
L7 533 1.6
L8 607 0.7 4.4
L9 681 2.3 8.5 1.6
L10 756 4.2 11.0 3.3 0.7
L11 830 5.5 10.3 6.8 1.6
L12 904 6.2 8.0 5.5 1.3 3.1 0.7
L13 978 6.3 5.7 3.9 3.3 4.7 1.3
L14 1052 5.4 3.8 2.7 5.2 5.1 1.8 0.5
L15 1126 3.9 2.3 1.7 5.4 4.6 2.0 0.4 1.0
L16 1200 2.6 1.3 0.9 4.1 3.5 1.5 1.1 1.8
L17 1275 1.4 0.6 2.8 2.4 1.0 1.4 2.4
L18 1349 0.7 0.3 1.6 1.6 2.1 2.5
L19 1423 0.5 0.8 1.0 2.1 2.1
L20 1497 0.4 0.6 1.6 1.4
L21 1571 0.3 0.2 1.6 1.0
Cn= cyclic polydimethylsiloxane, Ln= linear polydimethylsiloxane
n= number of Si(CH3)2-groups
Meuwly 19.10.2005 13:12 Uhr Seite 293
sist of siloxane oligomers of molar mass between 500 and 2100 Dalton which corre-
spond to oligomers with 7 to 28 Si(CH3)2O-groups. Figures 4 and 5 represent typi-
cal GC-MS graphs of the residue of the global migration. Figure 4 represents the
GC-MS obtained with sample A, mostly cyclic polydimethylsiloxane oligomers, at
the third one hour exposition (M3) at 175°C whereas Figure 5 represents sample B,
mixture of cyclic and linear polydimethylsiloxane oligomers, at the same condition.
Samples C, F and G show the same pattern as sample A whereas sample D and E, to
some extent, are comparable to sample B.
The composition of the residue at the third one hour exposition (M3) in fonc-
tion of the temperature is given in Table 6. The GC-MS of the residue of the samples
at different temperatures were normalized by considering that the total surface of all
silicone peaks equals 100%. The sum of cyclic (Cn) and linear (Ln) polydimethyl-
siloxane oligomers for some samples differ slightly from 100% due to the fact that
the percentage of each peak is only given to the first decimal.
By increasing the temperature during the test for the global migration, the size
of the siloxane oligomers increases. From 175°C to 220°C and finally to 260°C, the
largest peaks for the sample A increase from n=14–16 to n=16–18 and n=19–21
units of Si(CH3)2O-groups. By increasing also the temperature from 175 to 220 and
finally to 260°C, the content of the linear polydimethylsiloxane oligomers in the
residue decreases, for exemple from 39.7% to 25.2 and then 10.3% for sample B.
Compounds with a molecular weight above 1000 Dalton are commonly consid-
ered to be of low toxicological relevance because they can not normally enter the
metabolism. The percentage of the polydimethylsiloxanes lower than 1000 Dalton
varies from 21.9 to 68.1% at 175°C, between 1.7 and 14.2% at 220°C, whereas at
260°C all the polydimethylsiloxanes in the residues are superior to 1000 Dalton.
Two points could play an important role in this topic; at high temperatures: a) the
thermal degradation of silicone elastomers favours the formation of larger
oligomers and/or b) the lower oligomers are not absorbed by MPPO.
Conclusion
Up to 100°C, silicone elastomeres can be considered as inert as shown by low
overall migration values. Around 150°C, silicone elastomers start to degrade at such
a rate that the limit of 10 mg/dm2prescribed in the Resolution of the Council of
Europe is reached in most cases. To keep the value within the limit, it is necessary to
take into account the “simulant D reduction factor”. Due to analytical uncertainty,
there is no significant difference in the overall migration results of the various sili-
cone baking samples. A study has shown that MPPO generally overestimates the
contamination of the foods from packaging (19). The release of volatiles is high dur-
ing the first use and then decreases rapidly.
Neither the determination of the overall migration, nor that of the volatiles gives
a complete view of the behaviour at high temperature of the silicone moulds. Each
one brings some part of the response to the thermal stability of silicone elastomers.
294 Mitt. Lebensm. Hyg. 96 (2005)
Meuwly 19.10.2005 13:12 Uhr Seite 294
Some questions concerning the reality of the high temperature migration test-
ings are still open. What is the real temperature at the food-silicone baking moulds
interface especially in the case of food containing a certain quantity of water? What
is the temperature of the paste or the cake in a normal or a microwave oven? These
questions are very significant due to the fact that increasing the temperature acceler-
ates the rate of the degradation process of the moulds. The migration levels in foods
compared to MPPO is another critical point. It is therefore intended to investigate
further the migration study with real foods instead of simulants.
Acknowledgement
We would like to thank Awilo Ochieng Pernet for carefully reading the manus-
cript.
Summary
Heat stability of silicone baking moulds used in bakery has been evaluated with
different standard methods. The results of the migration tests using modified
polyphenylene oxide (MPPO, Tenax) as a food simulant indicate that these materi-
als are stable up to 150°C. Above this temperature the limit of 10 mg/dm2pre-
scribed by the Resolution of the Council of Europe is reached in most cases. The
migration residue is essentially composed of cyclic oligomeric polydimethylsilox-
anes, however, some samples contain cyclic and linear oligomeric polydimethyl-
siloxanes. The mass of these oligomeres is between 500 and 2100 Dalton whereas the
maximum varies from 1000–1500 Dalton depending on the test temperature. The
loss of volatiles at 200°C completes the migration values and shows that some
moulds lose more than 0.5% of their weight during a 4 hour heating process.
Although the loss of the mould weight decreases rapidly with the number of uses,
all the observations indicate that silicone baking moulds are not inert enough for use
in all the range of temperature indicated by the manufacturers.
Zusammenfassung
Die Temperaturbeständigkeit von Silikonbackformen wurde mit verschiedenen
genormten Methoden überprüft. Die Ergebnisse der Globalmigrationsprüfung mit
modifiziertem Polyphenylenoxid (MPPO, Tenax) als Lebensmittelsimulanz hat
ergeben, dass das Material bis 150°C stabil bleibt. Ab dieser Temperatur überschrei-
ten fast alle geprüften Muster den Globalmigrationsgrenzwert von 10 mg/dm2, der
von dem Europarat empfohlen wird. Der Hauptanteil des Migrationsrückstandes
besteht aus zyklischen oligomeren Polydimethylsiloxanen, und in einigen Mustern
sind auch lineare vorhanden. Diese Oligomere haben Molekularegewichte zwischen
500 und 2100 Dalton, wobei der Hauptanteil je nach Testtempereratur zwischen
1000–1500 Dalton variiert. Die Bestimmung der flüchtigen Anteile bei 200°C ver-
vollständigen die Globalmigrationresultate, teilweise betragen die Gewichtsverluste
während des Erhitzens nach 4 Stunden mehr als 0.5%. Obschon der Gewichts-
Mitt. Lebensm. Hyg. 96 (2005) 295
Meuwly 19.10.2005 13:12 Uhr Seite 295
verlust der Silikonbackformen mit zunehmenden Gebrauch exponentiell abnimmt,
zeigen alle Experimente, dass, entgegen den Angaben der Hersteller, die Silikon-
backformen über den ganzen Temperaturbereich nicht genügend inert sind.
Résumé
La stabilité thermique des moules en silicone utilisés en pâtisserie a été étudiée à
l’aide de diverses méthodes types. Les résultats des essais de migration utilisant de
l’oxide de polyphénylène modifié (MPPO, Tenax) comme simulant alimentaire
indiquent que ces matériaux sont stables jusqu’à 150°C. Au-dessus de cette tem-
pérature, la valeur de 10 mg/dm2recommandée par la Résolution du Conseil de
l’Europe est atteinte dans la plupart des cas. Le résidu de migration est composé
principalement d’oligomères de polydiméthylsiloxane cycliques mais également
linéaires pour certains moules. Ces oligomères ont une masse comprise entre 500 et
2100 Dalton avec un maximum qui varie de 1000–1500 Dalton avec la température
de test. La mesure de perte des substances volatiles à 200°C complète les résultats de
migration et montre que certains moules perdent plus de 0.5% de leur poids durant
un chauffage de 4 heures. Bien que la perte de masse des moules diminue rapidement
avec le nombre d’utilisation, toutes les observations indiquent que les moules de
cuissons en silicone ne sont pas assez inertes pour une utilisation dans tout le
domaine des températures annoncées par les fabricants.
Key words
Silicones, baking moulds, heat stability, migration, volatiles
References
1Franck R. und Wieczorek H. (Eds): Kunststoffe im Lebensmittelverkehr, Empfehlungen des
Bundesinstitutes für Risikobewertung, XV. Silicone, 53. Lieferung, Carl Heymanns Verlag,
Köln, Januar 2004. (http://bfr.zadi.de/kse/)
2 Matériaux au contact des denrées alimentaires, Produits de nettoyage de ces matériaux: Arrêté du
25 novembre 1992. Les éditions des Journaux officiels français brochure N° 1227, 15 juillet 2002
3 Regulation (EC) No 1935/2004 of the European Parliament and of the Council of 27 October
2004 on materials and articles intended to come into contact with food and repealing
Directives 80/590/EEC and 89/109/EEC. Official Journal of the European Communities
13.11.2004 L 338
4 Verordnung vom 1. März 1995 über Gebrauchsgegenstände (GebrV, SR 817.04). EDMZ,
3000 Bern. (http://www.admin.ch/ch/d/sr/c817_04.html)
5 Council of Europe, Resolution AP (2004) 5 on Silicones used for food contact applications
(replacing Resolution AP (99) 3). Council of Europe, Strasbourg
6Piringer O. and Bürcherl T.: Extraction and Migration Measurements of Silicone Articles and
Materials coming into Contact with Foodstuffs. Document RD-6/26 from Committee of Experts
on Materials coming into Contact with Food. Council of Europe, Strasbourg, November 1994
7Lund K.H. and Petersen J.H.: Safety of food contact silicone rubber: Liberation of volatile
compounds from soothers and teats. Eur Food Res Technol 214, 429–434 (2002)
296 Mitt. Lebensm. Hyg. 96 (2005)
Meuwly 19.10.2005 13:12 Uhr Seite 296
8 Commission Directive 97/48/EC of 29 July 1997 amending for the second time Council
Directive 82/711/EEC laying down the basic rules necessary for testing migration of the
constituents of plastic materials and articles intended to come into contact with foodstuffs.
Official Journal of the European Communities 12.08.1997 L 222
9 EN 1186-13: European Standard, Materials and articles in contact with foodstuffs – Plastics –
Part 13: Test methods for overall migration at high temperatures, September 2002
10 EN 1186-1: European Standard, Materials and articles in contact with foodstuffs – Plastics –
Part 1: Guide to the selection of conditions and test methods for overall migration, April 2002
11 Commission Directive 85/572/EEC of 19 December 1985 laying down the list of simulants to
be used for testing migration of constituents of plastic materials and articles intended to come
into contact with foodstuffs. Official Journal of the European Communities 31.12.1987 L 372
12 Hilborg H., Karlsson S. and Gedde U.W.: Characterisation of low molar mass siloxanes
extracted from crosslinked polydimethylsiloxanes exposed to corona discharges. Polymer 42,
8883–8889 (2001)
13 Homma H., Kuroyagi T., Izumi K., Mirley C.L., Ronzella J. and Boggs S.A.: Evaluation of
surface degradation of silicone rubber using gas chromatography/mass spectroscopy: IEEE
Transactions on power delivery 15 (2), 796–803 (2000)
14 Radhakrishnan T.S.: New method for evaluation of kinetic parameters and mechanism of
degradation from pyrolysis-GC studies: thermal degradation of polydimethylsiloxanes.
Journal of Applied Polymer Science 73, 441–450 (1999)
15 Patel M., Soames M., Skinner A.R. and Stephens T.S.: Stress relaxation and thermogravimetric
studies on room temperature vulcanised polysiloxane rubbers. Polymer Degradation and
Stability 83, 111–116 (2004)
16 Fateh-Alavi K., Gällstedt M. and Gedde U.W.: The effect of antioxidants on the surface oxi-
dation and surface craking of crosslinked polydimethylsiloxane. Polymer Degradation and
Stability 74, 49–57 (2001)
17 Hunt S.M. and George G.A.: Characterization of siloxane residues from polydimethylsilox-
ane elastomers by MALDI-TOF-MS. Polym. Int. 49, 633–635 (2000)
18 Krivda A., Hunt S.M., Cash G.A. and George G.A.: MALDI-TOF/MS Characterization of
LMW PDMS in high voltage HTV silicone rubber insulators. IEEE CEIDP, Victoria BC
Canada 703–708 (2000)
19 Mountfort K., Kelly J., Jickels S.M. and Castle L.: A critical comparison of four test methods
for determining overall and specific migration from Microwave susceptor packaging. J. Food
Prot. 59 (5), 534–540 (1996)
Corresponding author: Dr. Roger Meuwly, Swiss Federal Office of Public Health,
Food Science Division, Food Chemistry Section, P.O. Box, 3003 Bern, Switzerland,
e-mail: roger.meuwly@bag.admin.ch
Mitt. Lebensm. Hyg. 96 (2005) 297
Meuwly 19.10.2005 13:12 Uhr Seite 297
