Internal correction of spectral interferences and mass bias in ICP-MS using isotope pattern deconvolution: Application to the determination of selenium in biological samples by isotope dilution analysis

Abstract

The internal and simultaneous correction of mass bias and spectral interferences (BrH+ and SeH+) in the determination of total selenium in biological samples by isotope dilution analysis is evaluated here using isotope pattern deconvolution (IPD). The procedure takes the measured intensities at masses 76 to 82 in the spiked samples (using 77Se as spike) and calculates the isotopic composition of selenium. This isotopic composition would be a linear function of the isotopic composition of natural abundance selenium and that of enriched 77-selenium if no mass bias or spectral interferences were present or if they were adequately corrected. So, the molar fractions of natural abundance selenium and isotopically enriched selenium are calculated using IPD along with the variance of the multiple linear regression model. The variance of the regression is first calculated when no correction is applied (either mass bias or spectral interferences) and then this value is minimised by applying different factors for mass bias and the ratios SeH/Se and BrH/Br. It is demonstrated that, using the SOLVER application of Excel, the regression variance can be minimized and the correction factors converge to those factor which could have been calculated by external correction. Hence, this procedure eliminates the need for any external correction of mass bias or isobaric interferences in isotope dilution analysis provided that enough isotopes of the element are measured. To illustrate the applicability of this method, raw data obtained by Hinojosa Reyes et al. in a previous paper (J. Anal. At. Spectrom., 2003, 18, 11) were recalculated using the proposed methodology with satisfactory results.