Determination and Migration of PCBs from the Food Packaging Materials of Paper-Plastic
|School||Central South University of Forestry Science and Technology|
|Course||Of Food Science|
|Keywords||PCBs Paper and plastic packaging GC/ECD migration model|
Polychlorinated biphenyls (PCBs) are a group of highly lipophilic, chemical and thermal stable, persistent contaminants. The control of oral intake seems to be very important, and it’s one of these vital influence factors that the migration of PCBs in the food packaging materials which contact with food directly or indirectly. This project mainly studies the determination and migration of PCBs in the food packaging materials, and makes the migration model to simulate the real migratory behaviour.This paper mainly analyzes PCBs in food packaging materials with accelerated solvent extraction (ASE), sulfuric acid and silica gel column cleanup with gas chromatography-electron capture detector (GC-ECD). This method is applied to test8kinds of samples known information, and some random samples in the market. The migration of PCBs in food packaging materials is experimented to estimate the extent of migration.The migration of PCB138and PCB153from food packaging materials with different PE film thickness to food simulant (A:water, B:3%(w/v) acetic acid, C:15%(V/V) ethanol; D:95%(V/V) ethanol) at4℃,35℃,60℃were studied. Migration modeled based on the model of "limited packaging-limited food"by taking into account diffusion coefficients of paper and plastic and ignoring the discrepancy between their partition coefficients.After some optimized experiments, it’s confirmed that extractive solvent of ASE is acetonitrile, static cycle is twice at80℃, the dosage of concentrated sulfuric acid is in the amount of0.8～1.3mL. The supernatant of hexane phase is combined and loaded onto the silica gel column to clean up. As the elution solvents,3×10mL aliquots of hexane and a10mL aliquot of hexane:dichloromethane (19:1) are added to the column in turn. The collected eluent (approximately50mL) is concentrated and made up with hexane to an accurate volume of1.0mL and analyzed by GC-ECD. Rtx-5MS capillary column is selected to separate8kinds of PCBs at the same time. The calibration curves of PCB28, PCB52, PCB101, PCB118, PCB138, PCB153, PCB180, PCB198presented a good linearity in the range of4～200μg/L, the coefficients of variation and the limits of determination were in the range of0.9982～0.9999and0.0183-0.03780μg/kg, respectively. The overall relative standard derivations were less than4.11%and the average recoveries were in the range of77.90%～100.08%. Compared with Soxhlet extraction method, the optimal ASE sample preparation can reduce analysis time and solvent consumption, and enable unequivocal identification and quantification.The determination results shown that there is at least one PCBs among almost all the samples contained one PCBs, and PCB138is in65%samples, and PCB118,153,180,198are found partly in the samples but their contents are within15μg/kg.It proved that the PCBs contaminants were introduced from the process of the press forming, which might be from the adhesives that are utilized when pressing, or from the lubricating oils used on the equipment and so on, maybe also come from the environmental contamination of the materials just like paper pulp.The results of migration shown that special migration of PCBs increased with the temperature increasing and the migration time extanded, and the migration rate was accelerated with temperature increasing. The migration of non-polar contaminants had a negative correlation with the polarity of similant. The migration had a positive correlation with the initial migratory concentration. True migration rate had a negative correlation with the thickness of PE films and was independent of the initial migratory concentration.In the migration model, diffusion coefficients were estimated with Brandsch Model and partition coefficient was rough estimated with its definition expression. The difference of migration experiment value and simulation was analysed by Maple. The results showed that the overall variation trend of the experiment value was in accord with the model prediction, that is to say, migration increased with the prolong of time and added with the raise of temperture. Most of predicted values were slightly higher than migration experiment values.