Development of a method for the systematic use of bilinear multivariate calibration methods for the correction of interferences in inductively coupled plasma-mass spectrometry

The multivariate bilinear calibration methods Principal Component Regression (PCR) and Partial Least Squares Regression (PLS) are investigated for their capability to correct interferences in ICP-MS analysis. A universally applicable method is developed by optimizing the parameters for experimental design, for data pretreatment by internal standardisation, scaling, and mean centering, for the determination of the optimal number of latent variables, and for variable selection. Validation is performed with an artificial test system containing a wide range of representative analytes and interferences. (C) 1998 Elsevier Science B.V.