An evaluation of M2+ interference correction approaches associated with As and Se in ICP-MS using a multi-day dataset along with ICP-MS/MS/HR-ICP-MS based analysis and hierarchical modeling as a means of assessing bias in fortified drinking waters and single component matrices

Abstract

Three ½ mass oriented rare earth element (REE) M²⁺ correction approaches (fixed factor, a dual internal standard, and an in-sample) are evaluated for use in an ICP-MS environmental method update. The multi-variant-based evaluation includes analyzing the same 19 REE-fortified matrices on eight different days over a two-month period using two instrument tunes. These REE-fortified matrices were also analyzed using HR-ICP-MS and ICP-MS/MS to estimate the reference value for use in the principal component analysis (PCA) and hierarchical modeling evaluation. A fixed factor is unable to compensate for matrix and mass dependent drift and because of this it generates the largest across matrix, tune, and day 95^th percent confidence bounds for the REE corrections on both As (1.1 ppb) and Se (23 ppb) using samples fortified with 100 ppb Nd, Sm & Gd. The PCA analysis indicated that M²⁺ ions cluster together across matrix, tune and day better than M¹⁺ and these tighter correlations are reflected in reduced 95^th percentile confidence bounds for dual M²⁺ internal standards (M²⁺; As = 0.3 ppb; Se = 5.4 ppb; n = 704) relative to M¹⁺ internal standards (M¹⁺; As = 0.6 ppb; Se = 12.0 ppb; n = 1056). The use of an in-sample M²⁺ correction produced comparable 95^th percent confidence bounds (As = 0.2 ppb; Se = 3.4 ppb; n = 352) relative to the M²⁺ internal standard approaches. Finally, the hierarchical modeling indicated M²⁺ ions as internal standards tend to minimize the across day variability induced by cone changes and the daily reoccurring matrix shifts in the M²⁺/M¹⁺ ratio associated with 250 ppm matrices of Na, Ca, and Mg. This internal standard driven reduction in variability can be beneficial in compliance monitoring methods.