Suspected-target pesticide screening using gas chromatography–quadrupole time-of-flight mass spectrometry with high resolution deconvolution and retention index/mass spectrum library

A strategy for suspected-target screening of pesticide residues in complicated matrices was exploited using gas chromatography in combination with hybrid quadrupole time-of-flight mass spectrometry (GC–QTOF MS). The screening workflow followed three key steps of, initial detection, preliminary identification, and final confirmation. The initial detection of components in a matrix was done by a high resolution mass spectrum deconvolution; the preliminary identification of suspected pesticides was based on a special retention index/mass spectrum (RI/MS) library that contained both the first-stage mass spectra (MS1 spectra) and retention indices; and the final confirmation was accomplished by accurate mass measurements of representative ions with their response ratios from the MS1 spectra or representative product ions from the second-stage mass spectra (MS2 spectra). To evaluate the applicability of the workflow in real samples, three matrices of apple, spinach, and scallion, each spiked with 165 test pesticides in a set of concentrations, were selected as the models. The results showed that the use of high-resolution TOF enabled effective extractions of spectra from noisy chromatograms, which was based on a narrow mass window (5 mDa) and suspected-target compounds identified by the similarity match of deconvoluted full mass spectra and filtering of linear RIs. On average, over 74% of pesticides at 50 ng/mL could be identified using deconvolution and the RI/MS library. Over 80% of pesticides at 5 ng/mL or lower concentrations could be confirmed in each matrix using at least two representative ions with their response ratios from the MS1 spectra. In addition, the application of product ion spectra was capable of confirming suspected pesticides with specificity for some pesticides in complicated matrices. In conclusion, GC–QTOF MS combined with the RI/MS library seems to be one of the most efficient tools for the analysis of suspected-target pesticide residues in complicated matrices.