Influence of the number of calibration points on the quality of results in inductively coupled plasma mass spectrometry

Abstract

Quantification methods based on multiple-point calibration by weighted linear regression, and double-point calibration (measurement of the blank and one standard) were investigated under multielement routine conditions for trace analysis. Proper quantification by both calibration methods, in terms of bias and uncertainty, was achieved when compared to reference materials. The expanded uncertainty of the results was not influenced by the number of calibration standards, since a coverage factor of 2 (95% confidence level) was adopted both for double and multiple-point calibration. The use of this factor in multiple-point calibration instead of Student's t, which depends on the number of calibration points, is justified because the calculated uncertainty related to the calibration is a good estimation of the calibration standard uncertainty, due to the highly linear behaviour of the ICP-MS technique. Relative expanded uncertainties ranged from 10–15% for concentrations around the LOQ to 3–5% for concentrations higher than 100 times the LOQ.