Uncertainty of determination of palladium in road dust sample by inductively coupled plasma mass spectrometry

The determination of palladium in a road dust sample taken close to the highway (w (Pd)=450 ng g(-1)) was carried out by the ICP-MS method after sample decomposition by aqua-regia. Analyses were evaluated by two methods: external calibration accompanied with mathematical correction of spectral interferences (EC) and isotope dilution measurement after separation of Pd by extraction to dibutyl sulfide solution (ID). In both cases, the uncertainties and accuracy of results were investigated. Although in the case of ideally homogeneous sample the repeatability of EC results (11 ng g(-1)Pd) was somewhat lower than those of ID results (16 ng g(-1)Pd), the uncertainties of results of both techniques were almost the same and they reached the level of 19 ng g(-1)Pd. The main uncertainty source of the EC method is represented by the correction of spectral interferences. In case of real non-homogeneous sample, the main uncertainty component represents the soil sampling. The uncertainty of results (approx. 75 ng g(-1)Pd) only slightly exceeded the repeatability (approx. 70 ng g(-1)Pd). The accuracy of results was proven by analyses of CRM TDB-1 Diabas Rock (in case of ID) and by the standard addition method (in case of EC).