Abstract

The quality of the results obtained by quantitative and semiquantitative methods in ICP-MS was critically investigated with a certified reference material (SLRS-3 Riverine Water), under multi-element routine analysis conditions. Quantitative methods were based on external calibration by simple and weighted linear regression and the semiquantitative one on the use of a commercial program (TotalQuant III) and internal calibration with Sc, In and Bi. Statistical weights used in weighted linear regressions were calculated from a noise model, which included the different ICP-MS noise sources in order to facilitate the use of this regression mode. Analytical results indicated that the concentrations of elements studied were in agreement with certified values by the three quantification methods. Relative uncertainties obtained by semiquantitative analysis ranged from 20%, for concentrations around the LOQ, to 6% for concentrations higher than 100 times the LOQ. They were similar to or better than those obtained by weighted regression calibration over a wide calibration range (0.01–100 ng ml⁻¹). The uncertainty related to the linear regression calibrations could be reduced by one order of magnitude by adjusting the calibration range to the concentration of the analyte and increasing the analysis time.