Czech J. Food Sci.
Vol. 24, No. 3: 143–148
Effect of Plastic Packages on Benzo[a]pyrene Concentration
in Sunflower Oil
PETER ŠIMKO1, BOŽENA SKLÁRŠOVÁ1, PETER ŠIMON2 and ELENA BELAJOVÁ1
1Food Research Institute, Bratislava, Slovak Republic; 2Department of Physical
Chemistry, Faculty of Chemical and Food Technology, Slovak University of Technology,
Bratislava, Slovak Republic
ŠIMKO P., SKLÁRŠOVÁ B., ŠIMON P., BELAJOVÁ E. (2006): Effect of plastic packages on benzo[a]pyrene
concentration in sunflower oil. Czech J. Food Sci., 24: 143–148.
Commercial y available sun?ower oil and the same oil distil ed additional y in a molecular evaporator (to remove natural y
occurring compounds) was spiked with benzo[a]pyrene (BaP) at the levels of 37.1 and 38.6 µg/kg, respectively. Then, it was
?l ed into polyethylene terephtalate (PET) and low density polyethylene (LDPE) receptacles of cylindrical shape, and BaP
concentration was fol owed within 49 h. At the end of the experiments, BaP concentration in the non-distil ed oil packed
into PET decreased to 25.9 µg/kg, and BaP concentration in the distil ed oil decreased to 34.6 µg/kg. The rate and the extent
of BaP removal were evaluated comparing the di?usion and equilibrium coe?cients. The results showed that PET is able
to reduce BaP concentration in sun?ower oil due to BaP sorption on the PET surface, but the rate and the extent of BaP
removal are also a?ected by other compounds present in the oil. As found, LDPE is an inappropriate material for the BaP
removal from sun?ower and rapeseed oils, because BaP concentration in the oils remained at a constant level during the
Keywords: polycyclic aromatic hydrocarbons; benzo[a]pyrene; sunflower oil; polyethylene terephtalate; polyethylene
Polycyclic aromatic hydrocarbons (PAHs) belong of these compounds in various foods to protect
to hazardous contaminants due to their known or the consumers against harmful effects of these
suspected carcinogenicity and/or mutagenicity. compounds. To simplify the difficulties associated
In general, PAHs are formed by incomplete com- with the great variability of PAH composition, BaP
bustion of fossil fuels and other forms of organic has been accepted, in general, as the indicator of
matter. For this reason they are found in al parts the total PAH presence in foods regardless of the
of the environment, including foods (TAMAKAWA fact that BaP constitutes only 1–20% of the total
2004). Moreover, PAHs are also formed in ther- carcinogenic PAHs (ANDELMAN & SUESS 1970).
mal processes during food production such as Relating to PAH legal limits, the situation in EU
baking, grilling, roasting, frying, and smoking is changing now considerably due to the adoption
(TAMAKAWA et al. 1996; CHEN 1997). With regard of regulation 208/2005 limiting the BaP content
to the harmful effects of PAHs on living organ- to the level of 2 µg/kg in oils and fats intended for
isms, there are trends to enact maximum limits the direct human consumption or the use as an
Supported by the Science and Technology Assistance Agency of Slovak Republic, Contract No. APVT-51-011002.
Vol. 24, No. 3: 143–148
Czech J. Food Sci.
ingredient in foods (EC 2005a). This regulation has than 50% in comparison to the initial concentra-
entered into power since 28th February 2005 and tion of 50 µg/kg of each compound tested (CHEN
started to be applied as from 1st April 2005. Apart & CHEN 2005). The ability of PET to lower PAH
from this, EC has also adopted either directive concentrations in polar and non-polar liquid media
2005/10/EC laying down the sampling methods has already been unambiguously proven (ŠIMKO et
and the methods of analysis for the official control al. 2004). However, the extent of the removal proc-
of BaP levels in foodstuffs (EC 2005b), or recom- esses is strongly a?ected by the presence of other
mendation 2005/108/EC on further investigation PAH compounds as wel as other compounds such
into the levels of PAHs in certain foods as fol- as vitamins, sterols, and waxes present in rapeseed
lows: benzo[a]anthracene, benzo[b]fluoranthene, oil (ŠIMKO et al. 2005). With regard to the current
benzo[j]fluoranthene, benzo[k]fluoranthene, knowledge mentioned above, the aim of this work
benzo[g,h,i]perylene, BaP, chrysene, cyclopenta[c, was to study the possible e?ects of PET package on
d]pyrene, dibenzo[a,h]anthracene, dibenzo[a,e]py- BaP concentration in sun?ower oil, and of LDPE
rene, dibenzo[a,c]anthracene, dibenzo[a,i]pyrene, package in sunflower and rapeseed oils.
dibenzo[a,l]pyrene, indeno[1,2,3-c,d]pyrene, and
5-methylchrysene (EC 2005c). Fats and oils belong
MATERIALS AND METHODS
to the most important sources of PAH intake due
to their extensive consumption as well as PAH
Commercial y available sun?ower and rapeseed oils
concentrations which can sometimes reach high were purchased in a local market in Bratislava. The
levels. For example, MORET et al. (1997) analysed oils were packed in PET bottles of the volume of 4 l.
51 samples of olive oils and found the total PAH Free fatty acids were removed during the production
concentrations from 2.94 to 143.12 µg/kg. HOPIA et of the oils by chemical neutralisation.
al. (1986) found various PAHs in Finnish butters,
A part of sun?ower oil mentioned above was re-
margarines, and vegetable oils, and in some raw distil ed using molecular distillation equipment
vegetable oil materials where total PAH concentra- under lowered pressure of 10–20 Pa, and at the
tions in 25 samples varied from 0.17 µg/kg (corn oil) temperature of surface heating 200°C to remove
to 4600 µg/kg (crude coconut oil). It was concluded natural y occurring compounds.
that the enormous PAH concentration in coconut
In the experiment, pre-bubbled PET receptacles
oil could be caused by the direct drying of copra of cylindrical shape with i.d. of 21.4 mm were used.
with smoke. BARRANCO et al. (2004) pointed out to The receptacles were provided by Palma-Tumys
the technological procedures of deodorisation and Ltd., Slovakia. The company uses them for oil and
bleaching which are able to reduce considerably PAH fruit syrup packaging after blowing to the volume
concentrations in oils during the production. How- of 2 litre.
ever, over-limit concentrations of BaP in Spanish olive
LDPE cylindrical shape receptacles having i.d. of
oils was the reason to withdraw them from central 32 mm were supplied by Cechvalab Ltd., Slovakia.
European markets (ŠIMKO et al. 2004). Polymers
BaP of analytical grade was purchased from Su-
could play an important protective role with regard pelco in the solid state. The solution for spiking was
to the high a?nity of PAH to some plastic materials prepared by dissolving BaP in acetonitrile to the
(ŠIMKO 2005). For example, PAH concentrations initial concentration of 500 mg/l.
were e?ectively reduced in the liquid smoke ?avour
Acetonitrile was of gradient grade (Merck, Ger-
(by two orders) during 14 days of storage in LDPE many), methanol and hexane (Slavus, Slovak Re-
?asks. As found, the rate-limiting step of the PAH public) were of analytical grade. The solvents were
sorption from a liquid into LDPE is the di?usion in rectified immediately before use in a distil ation
liquid media (ŠIMKO et al. 1994). During the inter- apparatus.
action with LDPE, PAHs are primarily adsorbed on
Anhydrous Na2SO4 and alumina were purchased
the LDPE surface with subsequent di?usion into the from Merck, Germany.
polymer bulk (ŠIMKO et al. 1999). The most intense
processes of PAHs removal from liquid media into
LDPE take place within 24 h, where the polarity and
viscosity of the liquid media also play an important
First of al , the oils were analysed for the pres-
role. During this period, the PAH concentration ence of BaP. Consequently, 100 g of oil was spiked
dropped in three various liquid systems by more with BaP solution in 2 l volume glass flask and the
Czech J. Food Sci.
Vol. 24, No. 3: 143–148
solvent was al owed to evaporate spontaneously. ised water, B – acetonitrile. Gradient programme:
To accelerate the evaporation, the oil was stirred from 0 to 2 min elution with 30% A and 70% B,
intermittently. Then, roughly 900 g of the oil was then from 2 to 5 min with from 70% B to 100% B
added and the content of the flask was mixed thor- linearly, then the elution from 5 to 10 min isocrati-
oughly. At this stage, a sample of spiked oil was cally, then from 10 to 15 min with from 100% B
taken to determine the initial BaP concentration. to 70% B linearly. The equilibration time between
Then, the PET and LDPE receptacles were fil ed each run was 5 min. The samples were applied
with the spiked oil and placed into a polystyrene using an autosampler needle with 20 µl volume.
box to protect them from light and to keep then Al analyses were carried out in duplicates with an
at a constant temperature. The samples for the average relative standard deviation of 8.4%.
analysis were taken after 1, 3, 5, 7, 11, 24, and
49 h. To maintain the same static conditions and
RESULTS AND DISCUSSION
sampling during the experiments, a new set of
receptacles was taken for each analysis.
At first, the experiment was carried out with
Sample preparation. The sample preparation commercial y available sunflower oil to be fil ed
was performed according to ISO 15302 as fol ows: in PET receptacles. The experimental y obtained
2 g of oil was weighed with the precision of 0.001 g dependences of BaP concentrations vs. time were
into a 10 ml graduated flask, dissolved in hexane used for the calculation of the diffusion coef-
and diluted to the mark. Then, 22 g of alumina ficient D using a kinetic equation (1), which was
was transferred immediately to a chromatography derived for the diffusion of PAHs in non-stirred
column fil ed with hexane and anhydrous Na2SO4 liquids placed into cylindrically shaped plastic
was added to the top of the column to form a layer bulks (ŠIMKO et al. 2004):
about 30 mm thick. Hexane was let to fal to the
level of the top of Na
t = c ? + ? c 0 ? c ? ? ?
exp ?? D? nt ?
2SO4 layer, and 2 ml of the
graduated flask content was then applied onto
n=1 a ? n
the column. The column was eluted with hexane D was calculated by the non-linear least squares
at a flow of about 1 ml/min, the first 20 ml of the method by minimising the sum of squares of differ-
eluate were discarded and then 60 ml of the eluate ences between the BaP concentrations measured
were col ected in a 100 ml round-bottomed flask. experimental y and those calculated by equation
The eluate was evaporated to about 0.5 ml and (1). BaP equilibrium concentration between the oil
transferred into a vial. The evaporation continued and PET expresses the equilibrium coefficient ?:
under nitrogen until the residue was nearly dry.
c ? c?
The round-bottomed flask was rinsed twice with ? 0
about 1 ml of hexane and transferred quantitatively
to a mini-vial where the evaporation continued where:
under nitrogen. The evaporation was carried out c0 – initial BaP concentration in the oil
to dryness, the residuum was then dissolved in c? – stands for the equilibrium BaP concentration in the
methanol and analysed using HPLC.
oil in in?nite time
HPLC analysis. The analyses were performed
The higher value of ? coefficient corresponds to a
on the liquid chromatograph Agilent Technolo- greater decrease of BaP concentration in the liquid
gies 1100 Series (Halbron, Germany) consisting media. As fol ows from Figure 1, BaP concentra-
of a quaternary pump, micro vacuum degasser, tion in the oil began to decrease immediately after
autosampler, and fluorescence detector, which fil ing the PET receptacles due to BaP sorption on
operated at 300 nm excitation and 410 nm emis- the PET surface. The equilibrium concentration of
sion wavelengths. The separations were carried 25.9 µg/kg was reached within about 24 h of the
out at ambient temperature on LiChrolut column experiment which is the value already observed in
(25 cm × 0.4 cm i.d. packed with Lichrospher rapeseed oil packed in PET (ŠIMKO et al. 2005). As
PAH, particle diameter 5 µm), when pre-column the value of the equilibrium coefficient ? shows
LiChroCart (4 cm × 0.4 cm) with the same particle (Table 1), the elimination of BaP from the oil was
diameter was also used. The flow rate of the mobile quite efficient. Its extent is comparable with that
phase was 1 ml/min. For the determination, the one determined for rapeseed oil which confirms
gradient elution was used as fol ows: A – deion- that the same removal processes took place also in
Vol. 24, No. 3: 143–148
Czech J. Food Sci.
Figure 1. Changes in BaP concentration in sun?ower oil
Figure 2. Changes in BaP concentration in additional y
stored in PET receptacles
re-distil ed sun?ower oil stored in PET receptacles
? – experimental y obtained data (every point is average ? – experimental y obtained data (every point is average
value of four BaP determinations with standard deviation = value of four BaP determinations with standard deviation =
1.2 µg/kg); ? – calculated data using kinetic equation (1)
0.9 µg/kg); ? – calculated data using kinetic equation (1)
the system PET-sun?ower oil. However, the results a dominant fatty acid in rapeseed oil (62%) while
obtained in the experiments with the re-distil ed linoleic acid is dominant in sunflower oil (63%).
oil are very surprising (Figure 2). In general, the While oleic acid is mono-unsaturated, linoleic
distillation (called also physical neutralisation) acid is di-unsaturated fatty acid which means that
is used in oil industry as an alternative to the sunflower oil contains more double bonds, e.g.
chemical neutralisation to remove free fatty acids ?-electron pairs, in comparison to rapeseed oil.
from raw oils. Nonetheless, during the distil ation PAHs are compounds rich in ?-electrons localised
other compounds are also removed (e.g. waxes, in the condensed aromatic rings which invokes
vitamins, sterols, etc.) which gives rise to almost an idea of their interaction with ?-electrons of
a pure triacylglycerol mixture formation. The linoleic acid. In non-distil ed sunflower oil, ?-elec-
removal of these compounds from rapeseed oil trons probably interact with natural y occurring
increased the equilibrium coefficient as shown in compounds also rich in ?-electrons, e.g. vitamin
Table 1 (ŠIMKO et al. 2005). However, the distil- E. These compounds were removed by molecular
lation of sunflower oil brought about a consider- distil ation, and then the liberated ?-electrons of
able decrement of the equilibrium coefficient linoleic acid could interact with ?-electrons of
although the diffusion coefficient had an almost BaP, which was revealed by the decreased value
identical value. This fact was really unexpected of the equilibrium coefficient. This assumption
because our aim was to remove the components can be supported by the great similarity of the
competing with BaP for the adsorption centres diffusion coefficient values for both sunflower
on the PET surface. However, the decrease of oils (Table 1), which evokes that the main part of
the equilibrium coefficient observed could be BaP interacted with ?-electron pairs of linoleic
explained by a matrix effect due to the different acid, and that only a minor part of BaP interacted
compositions of the vegetable oils. Oleic acid is with PET. Overal , this “double behaviour” of BaP
Table 1. Parameters calculated from the experimental y measured values
Di?usion coe?cient D
Equilibrium coe?cient ?
3.9 × 10–2
23.2 ± 0.3
Re-distilled sun?ower oil
6.6 × 10–2
22.4 ± 0.3
2.5 × 10–2
18.3 ± 0.4
Re-distilled rapeseed oil*
1.8 × 10–2
20.4 ± 0.3
*values taken from ŠIMKO et al. (2005)
Czech J. Food Sci.
Vol. 24, No. 3: 143–148
Figure 3. BaP concentration in sun?ower (?) and
rapeseed (?) oils stored in PE receptacles (every
point is average value of four BaP determinations with
standard deviation = 0.3 µg/kg)
resulted in the lowered efficiency of the proc-
esses of BaP removal from distilled sunflower
oil. Because the removal of BaP is limited by the
The results and findings of this work lead to the
surface adsorption and the di?usion of BaP into fol owing conclusions:
LDPE bulk had been already proven, LDPE was 1. The BaP concentration in the sunflower oil fil ed
studied for BaP removal from the oil. However,
into PET can decrease due to the interaction
BaP concentration in sunflower oil remained at
between BaP contained in the oil and PET.
a constant level, as shown in Figure 3, which was 2. The rate of BaP diffusion in the liquid phase and
quite surprising. To assure this finding, the experi-
the extent of its adsorption onto PET depend
ment was repeated with sunflower oil having been
on the presence of other compounds contained
replaced with rapeseed oil. In spite of this change,
in the oil.
the concentration remained at the constant level 3. The removal of these compounds by distil ation
again, as shown in Figure 3. This observation is
brings about a decrease in the extent of the BaP
real y surprising with regard to the previous find-
removal, probably due to the intensification of
ings because LDPE functioned effectively and it
BaP and the oil matrix interactions through
was able to decrease PAH concentration by one
(CHEN & CHEN 2005) or two orders (ŠIMKO et al. 4. From this point of view, the chemical neutrali-
1994) in liquid media. One explanation may reside
sation for free acids removal from sunflower
in the fact that LDPE was used effectively for the
oil should be preferred to the physical process
removal of PAHs from polar or semi-polar liquid
of distil ation.
media. Oils, however, are either a non-polar ma- 5. LDPE is not an appropriate material for the
trix or contain double bonds able to interact via
elimination of PAHs from a non-polar matrix,
?-electron pairs with delocalised ?-electrons of
e.g. sunflower and rapeseed oils.
PAHs. As these interactions are probably stronger 6. The dependence of BaP concentrations on time
in comparison to the interactions of PAHs with
exhibits oscil ations; the reason for their exist-
LDPE (which is free of ?-electron pairs), they
ence is so far unknown.
are able to hinder the migration from PAH into 7. This knowledge may be utilised in the indus-
LDPE and maintain the BaP concentration at a
trial production of vegetable oils inserting an
constant level. As can be seen, Figures 1 and 2,
additional operation of PAH sorption on the
the dependences of BaP concentrations on time
surface of the PET particles just after the bleach-
exhibit oscil ations, especial y at the beginning of
ing procedures to remove the residual PAHs,
the adsorption processes. These oscil ations occur
and in such way to protect human organism
not only in both types of sunflower oil but they
against the exposure to these carcinogenic
have already been observed also in rapeseed oil.
Moreover, they were also observed in paraffin oil
where the concentration was measured directly
without the sample preparation procedures, just to
exclude the possible experimental errors (ŠIMKO ANDELMAN J.B., SUESS M.J. (1970): PAH in the water
et al. 2005). This indicates that the adsorption
environment. Bul etin WHO, 43: 479–508.
processes in the systems PET-vegetable oils are BARRANCO A., ALONSO-SALCES R.M., CRESPO I., BER-
complicated or even combined with other proc-
RUETA B., GALLO B., VICENTE F., SAROBE M. (2004):
esses unknown by this time.
Polycyclic aromatic hydrocarbon content in commer-
Vol. 24, No. 3: 143–148
Czech J. Food Sci.
cial Spanish fatty foods. Journal of Food Protection, ŠIMKO P., SKLÁRŠOVÁ B., ŠIMON P., BELAJOVÁ E. (2005):
Decrease of benzo[a]pyrene concentration in rapeseed
CHEN B.H. (1997): Analysis, formation and inhibition of
oil packed in polyethylene terephtalate. European Journal
polycyclic aromatic hydrocarbons in foods: an overview.
of Lipid Science and Technology, 107: 187–192.
Journal of Food and Drug Analysis, 5: 25–42.
ŠIMKO P., ŠIMON P., KHUNOVÁ V., BRUNCKOVÁ B., DRDÁK
CHEN J., CHEN S. (2005): Removal of polycyclic aro-
M. (1994): Kinetics of polycyclic aromatic hydrocarbons
matic hydrocarbons by low density polyethylene from
sorption from liquid smoke ?avour into low density
liquid model and roasted meat. Food Chemistry, 90:
polyethylene packaging. Food Chemistry, 50: 65–68.
ŠIMKO P., ŠIMON P., KHUNOVÁ V. (1999): Removal of
European Commision (2005a): O?cial Journal of the
polycyclic aromatic hydrocarbons from water by migra-
European Union, L 34: 3–5.
tion into polyethylene. Food Chemistry, 64: 157–161.
European Commision (2005b): O?cial Journal of the
ŠIMKO P., ŠIMON P., BELAJOVÁ E. (2004): Lowering of
European Union, L 34: 15–20.
concentration of polycyclic aromatic hydrocarbons in
European Commision (2005c): O?cial Journal of the
liquid media by sorption into polyethylene terephtalate
European Union, L 34: 43–45.
– a model study. European Food Research and Tech-
HOPIA A., PYYSALO H., WICKSTRÖM K. (1986): Marga-
nology, 219: 273–276.
rines, butter and vegetable oils as sources of polycyclic TAMAKAWA K. (2004): Polycyclic aromatic hydrocarbons in
aromatic hydrocarbons. Journal of American Oil Che-
foods. In: NOLLET L. (ed.): Handbook of Food Analysis.
mical Society, 63: 889–893.
Marcel Dekker, Inc., New York: 1449–1483.
MORET S., PIANI B., BORTOLOMEAZZI R., CONTE L.S. TAMAKAWA K., KATO T., OBA M. (1996): Polycyclic
(1997): HPLC determination of polycyclic aromatic
aromatic hydrocarbons. In: NOLLET L. (ed.): Hand-
hydrocarbons in olive oils. Zeitschrift für Lebensmittel-
book of Food Analysis. Marcel Dekker, Inc., New York:
Untersuchung und Forschung, 205: 116–120.
ŠIMKO P. (2005). Factors a?ecting elimination of poly-
cyclic aromatic hydrocarbons in smoked meat foods
Received for publication August 18, 2005
and liquid smoke ?avours. Molecular Nutrition & Food
Accepted after corrections February 6, 2006
Research, 49: 637–647.
Ing. PETER ŠIMKO, Výskumný ústav potravinársky, Priemyselná 4, P.O. Box 25, 824 75 Bratislava, Slovenská republika
tel.: + 421 255 574 622, fax: + 421 255 571 417, e-mail: firstname.lastname@example.org