Quantitative composite testing model based on measurement uncertainty and its application for the detection of phthalate esters

Abstract

To improve the quantitative detection efficiency of chemical analysis and reduce the detection cost, the sample pass rate was estimated and mathematical statistics were used to calculate the optimal group size (Kopt) of the composite testing to save on the maximum workload. A quantitative composite testing model was developed based on chemical analysis measurement uncertainty. Using this model, the maximum allowable number of composited samples (Kmax) is first calculated using parameters of regulated limits (L), limit of quantification (LOQ), and method measured uncertainty (Urel) to ensure that the sensitivity of the composite testing can meet the limit requirements. Finally, the appropriate composite group size (Ka) can be obtained by creating a balance between Kopt, Kmax, and the practical information used for that particular test. Furthermore, based on a constructed model, a practical quantitative composite testing method of 3–10 samples was established for the routine detection of toy phthalates (PAEs). The experimental results showed that the quantitative limits of 7 PAEs were 9.1–41.8 mg/kg, the relative expansion uncertainties were 16.6%–23.2%, and the recovery rates were 91.0%–112.3%, with a relative deviation of less than 10%. All these meet international PAEs standards. Compared with the traditional individual and qualitative composite testing, this model will not decrease the detection sensitivity, but can save up to 17.9%–80.4% of the workload when it is employed in toy PAEs testing with the pass rate of 80%–99%. This quantitative composite testing method will be implemented in the coming revision of ISO 8124-6 toy PAEs standards.