Testing migration from the PVC gaskets in metal closures into oily foods

The official method for testing migration from the gaskets of lids into oily foods was scrutinized. Oil is filled into a glass jar, the lid is mounted and the jar turned on its lid. After initial heating (typically 1 h at 100 °C to simulate pasteurization and 1 h/130 °C for sterilization), the jar is kept on its lid at 40 °C for 10 days. This testing should predict the migration into oily food for the shelf life of the product, which may be as much as 5 years, that is, migration must be accelerated to condense several years into 10 days. However, increase of temperature to 40 °C only accelerated by a factor of 2.4 and 3.9 for ESBO and polyadipate, respectively. Most of the acceleration is achieved by exposing the gasket to a far larger amount of oil than in reality. In the upright position, only some 30 mg oil adheres to the about 350 mg gasket material which is in food contact, that is, rapidly equilibrium is reached. Testing with a large amount of oil prevents this equilibrium being reached. This type of acceleration depends on the migrant, for example, strongly differs for ESBO and polyadipate, and is not suitable to extend the 10-day testing to a prediction for years in general manner. Migration also depends on the oil used for testing, for example, with coconut oil being 2–3 times that with olive oil. Initial heating to 100 or 130 °C caused a substantial initial migration, but for gaskets plasticized with ESBO it left behind a layer of PVC almost free of plasticizer which stopped migration at 40 °C like a skin for the 3 weeks tested. It is concluded that the official testing is not suitable for compliance testing of lids. Testing should occur by simulation of a more realistic scenario and accelerating it the best possible.