Qualitative Aspects in the Analysis of Pesticide Residues in Fruits and Vegetables Using Fast, Low-Pressure Gas Chromatography−Time-of-Flight Mass Spectrometry

Quantitative method validation is a well-established process to demonstrate trueness and precision of the results with a given method. However, an assessment of qualitative results is also an important need to estimate selectivity and devise criteria for chemical identification when using the method, particularly for mass spectrometric analysis. For multianalyte analysis, automatic instrument software is commonly used to make initial qualitative identifications of the target analytes by comparison of their mass spectra against a database library. Especially at low residue levels in complex matrices, manual checking of results is typically needed to correct the peak assignments and integration errors, which is very time-consuming. Low-pressure gas chromatography-mass spectrometry (LP-GC-MS) has been demonstrated to increase the speed of analysis for GC-amenable residues in various foods and provide more advantages over the traditional GC-MS approach. LP-GC-MS on a time-of-flight (ToF) instrument was used, which provided high sample throughput with <10 min analysis time. The method had already been validated to be acceptable quantitatively for nearly 150 pesticides, and in this study of qualitative performance, 90 samples in total of strawberry, tomato, potato, orange, and lettuce extracts from the QuEChERS sample preparation approach were analyzed. The extracts were randomly spiked with different pesticides at different levels, both unknown to the analyst, in the different matrices. Automated software evaluation was compared with human assessments in terms of false-positive and -negative results. Among the 13590 possible permutations with 696 blind additions made, the automated software approach yielded 1.2% false presumptive positives with 23% false negatives, whereas the analyst achieved 0.8% false presumptive positives and 17% false negatives for the same analytical data files. False negatives frequently occurred due to challenges at the lowest concentrations, but 70% of them involved certain pesticides that degraded (e.g., captafol, folpet) or otherwise could not be detected. The false-negative rate was reduced to 5-10% if the problematic analytes were excluded. Despite its somewhat better performance in this study, the analyst approach was extremely time-consuming and would not be practical in high sample throughput applications for so many analytes in complicated matrices.