Geochemical matrix differently affects the response of internal standards and target analytes for pesticide...

Electrospray ionization (ESI) is the most common technique in liquid chromatography coupled to tandem mass spectrometry (LC-MS-MS) allowing for sensitive detection of polar compounds with online water concentration. The technique is popular in groundwater monitoring programs and has permitted great progress in the detection and quantification of polar pesticide transformation products (TP) in recent years. However, ESI is also known to be prone to matrix effects. The common solution to this potential bias is the use of labelled internal standards. Unfortunately, these are not available for all target compounds, which leads to the linkage of target compounds to non-homologue internal standards with unknown consequences for quantification in variable geochemical settings. We investigated these matrix effects for polar TP with a molecular mass range of 225-350 Da and logD_pH7 between -0.27 and -1.7 as well as for parent compounds with logD_pH3 between 0.84 and 3.22. The acquired internal standards were tested on a gradient of DOC, anions, conductivity and inorganic carbon with a set of ten carefully chosen groundwater samples. Internal standards that were measured in positive ionization mode proved to be insensitive to geochemical variations while those that were measured in negative ionization mode showed reduced response with increasing anion concentration. All pairs of internal standards and target analytes were investigated for deviating matrix effects using standard addition experiments. Positive ionization compounds and target compounds with deuterated homologues showed little deviation while non-homologue pairs in negative mode proved to be strongly biased. Although bias was up to factor five for some compounds it was remarkably stable over the entire gradient studied, suggesting an identical suppression mode at varying matrix levels for different compounds. We advocate the conduct of standard addition experiments if homologue internal standards are not available.