Measurement uncertainty associated with overall migration testing.

Abstract

The results from a single laboratory that took part in a series of check-sample exercises for overall migration were used to calculate the measurement uncertainty for the overall migration methodology. The results span 10 years of proficiency testing and cover a range of plastic materials tested using a variety of time and temperature test conditions. Twelve sets of results for overall migration into olive oil and 10 sets of results for overall migration into volatile simulants were used. The measurement uncertainty associated with the determination of overall migration from plastics into olive oil was estimated as +/- 2.6 mg dm-2 for results between 1.2 and 15.4 mg dm-2. The measurement uncertainty associated with the determination of overall migration into volatile simulants was estimated as +/- 1.4 mg dm-2 for results between 2.1 and 13 mg dm-2. These estimates are within the analytical tolerances set in European Union regulations, of 3 and 2 mg dm-2, respectively. The performance of all laboratories participating in these 22 check sample exercises was also evaluated. A very large majority (93%) of the participants used official test methods, sometimes with minor modifications as stated. For the simulant olive oil, an average of 81% of laboratories (13 rounds with an average of 21 participants in each) using official methods was within the tolerance specified in regulations. For the volatile food simulants, an average of 87% of laboratories (11 rounds with an average of 38 participants in each) using official methods was within the tolerance specified. It is concluded from this evaluation of within- and between-laboratory data that the official European Standard methods for overall migration testing are suitable for the analysis of routine samples.