Abstract
A proficiency test on the determination of polycyclic aromatic hydrocarbons (PAHs) was organized by the German National Reference Laboratory for PAH in 2010. The test samples were produced by spiking cereal-based instant baby food with PAHs at concentration levels between 0.6 and 3.8 μg/kg; homogeneity and stability of the test material were verified before sample dispatch. Twenty-one official laboratories from Germany and Austria participated in the test and the evaluation of the test was done by applying methods of robust statistics. The individual performance was assessed with the help of z-scores. As to the quantitative results, the dispersion of data for the most important group of benzo(a)pyrene (BaP), benz(a)anthracene (BaA), benzo(b)fluoranthene (BbF), and chrysene (CHR) appeared to be acceptable, with a relative robust standard deviation ranging from 13.2% for BbF to 26.7% for BaA. In total, the performance of one laboratory had to be rated as unsatisfactory because of a result for BaP outside the limits of tolerance. The methods applied in the test may be considered to be comparable, as no significant effects in the distribution of data could be attributed to certain analytical procedures.