Reduction of matrix effects in quantitative and semi-quantitative inductively coupled plasma-optical emission spectrometry using a partial local thermodynamic equilibrium model and an internal standard

Abstract

A means to reduce matrix effects by using a simple pLTE model to account for changes in sensitivity due to matrix induced changes in plasma temperature and an internal standard to account for matrix induced changes in sample transport rate is described. For quantitative analysis where standards contain all of the elements of interest the changes in sensitivity due to 1.0M Ba and Na and 1.5M Li and Ca are reduced to less than 20% for most analyte lines investigated using this method. The maximum error due to the presence of 1.5M La is reduced from more than a factor of 14 to less than 40%. Semi-quantitative concentrations from the pLTE model based single element calibration approach described previously are essentially independent of matrix composition (within the uncertainty of semi-quantitative concentrations in the absence of a high matrix concentration) for the concomitant elements tested. However, the pLTE based approach is less accurate for semi-quantitative analysis than a one-time multi-element calibration, except when the most severe matrix effects are observed (such as those due to 1.5M La).