Interlaboratory test on polymers: determination of heavy metals in polymer metrices

Abstract

This report presents results obtained from interlaboratory tests carried out in 2002. Twelve participants, mainly from industry and research institutes, measured three different heavy metals (Pb, Cr and Cd) in two polymer matrices. The measured data were collected by EMPA and evaluated using a robust statistical method. Repeatability and reproducibility of data were of special interest. In a special procedure, a sample of polyvinylchloride and a sample of polyurethane were contaminated homogeneously with precise quantities of lead, cadmium and chromium. Since there are no generally applicable or universal ISO standards for the determination of heavy metals in plastics, participants were at liberty to choose their own method of analysis. Almost all laboratories opted to determine the heavy metal contents with ICP-OES or AAS following different microwave digestions. The repeatability and reproducibility limits of the heavy metal determinations differed significantly from each other. While the relative repeatability limits r, relative, for all the samples examined in the internal laboratory comparison were found to lie within a range of 4–8%, the relative reproducibility limits R, relative were higher by a factor of ca 10 (range: 12–65%). Even where contaminant concentrations are identical, the determination of heavy metals in plastics strongly depends on the polymer matrix. This matrix effect cannot be predicted, it has to be determined for each polymer type, because chemical interactions between contaminant and polymer matrix can interfere. The resulting combined uncertainty of measurement ( u c), for a single laboratory, a most important statistical term estimated from the statistical interlaboratory values, is thus more affected by matrix effects and less by sampling and measuring methods.