Report on the collaborative trial organised by the JRC on the determination of PVC and phthalates in textile products

摘要

On behalf of CEN/TC 248/WG 26 and ISO/TC 38/WG 22, the European Commission’s Joint Research Center (JRC) organised a collaborative study on the determination of PVC and phthalates in textiles products.The purpose of the study was the comparison and validation of four methods for the determination of phthalates and the validation of one method to determine the content of PVC in textile products. Methods 1, 2 and 3 were based on ultrasonic extraction of phthalates with n-hexane/acetone 80/20 v/v, n-hexane, tert-butyl methyl ether, respectively, and method 4 foresaw PVC dissolution in tetrahydrofuran and re-precipitation with acetonitrile. One method for the quantification of PVC in textile products based on the dissolution of PVC with tetrahydrofuran, followed by the washing of the residue and its gravimetric determination was also investigated (method 5).The collaborative exercise was organised, according to ISO 5725-2, as a balanced uniform-level experiment with the same number of test results in each laboratory, which each laboratory analysing the same levels of test samples. Thirteen laboratories both European and non-European participated to this study. The Italian company MP S.p.A produced both the PVC samples and the textile ones, made by cotton spread with PVC layer. Four textile and one PVC samples containing in total 7 phthalates (DEHP, DBP, BBP, DINP, DIDP, DNOP and DIBP) at 3 concentration levels and one sample, in which the PVC mass per cent had to be measured were analysed in triplicates. Levels I, II and III refer to samples containing a specific phthalate in concentrations of approximately 200, 1000 and 5000 mg/kg. In the case of DIDP and DINP, level I corresponded to approximately 500 mg/kg of PVC. These phthalate concentrations were selected in order to assess the precision of the analytical methods in the range of the current limits for toys and childcare articles.The homogeneity study was carried out by the JRC and all samples could be considered ‘sufficiently homogeneous’ according to the IUPAC harmonised protocol for proficiency testing. Results were statistically evaluated following the rules laid down in ISO 5725 parts 2 and 5. The consensus values and the precisions of the various methods, in terms of repeatability and reproducibility limits as well as repeatability and reproducibility relative standard deviations, were evaluated. Applying ISO 5725-2, the statistical outliers identified with Cochran’s and Grubbs’ tests were rejected, together with the results of LC0004 for method 4 and the ones of LC0005 for DIDP in methods 1-4, which were identified as outliers with Mandel’s h statistics. On the contrary, considering ISO 5725-5, all test results were retained and robust statistics was used. These two alternative approaches gave results that could be considered in good agreement. Generally, the differences were always lower than 35 %, except in few cases.Concerning the extraction efficiency, method 4 proved to be the best one in terms of phthalates’ recovery, whereas method 2 was the worst one. Practically the same recovery rate was shown by methods 1 and 3. Relative standard deviations of repeatability and reproducibility ranged from 3.0 to 23.5 % and from 19.4 and 189.9 % respectively. This means that both the four methods and the laboratories’ performance have to be drastically improved. Poor repeatability was observed in the case of several laboratories and the large spread in the mean values calculated in the 13 laboratories is responsible for the high observed relative standard deviation of reproducibility. Regarding method 5 for the quantification of PVC, repeatability and reproducibility relative standard deviations were 0.6 and 1.4% respectively. Considering that these values are in the same range of the values obtained with similar dissolution methods validated in the context of quantification of fibre binary mixtures, this method can be considered validated.