A method based on QuEChERS extraction and LC-quadrupole-Orbitrap™ MS detection was established utilising an improved fully non-targeted way of data acquisition with and without fragmentation. A full-scan acquisition event without fragmentation (resolving power 70,000) was followed by five consecutive fragmentation events (variable data independent acquisition - vDIA; resolving power 35,000) where all ions from the full-scan range are fragmented. Compared with fragmentation in a single event (all-ion fragmentation - AIF), this improves both selectivity and sensitivity for the fragment ions, which is beneficial for screening performance and identification capability. The method was validated, using the data from the same measurements, for two types of analysis: quantitation/identification and qualitative screening. The quantitative validation, performed according to the guidelines in SANCO/12571/2013, tested the performance of the method for 184 compounds in lettuce and orange at two spiking levels: 10 and 50 ng g(-1). The validation showed that the vast majority of the compounds met the criteria for trueness and precision set in the SANCO guidance document. In the qualitative validation the same 184 compounds were used to test the untargeted screening capabilities of the method. In this validation the compounds were spiked at three levels into 11 different fruit and vegetable matrices, which were measured twice on separate days. Taking all data from the qualitative validation together, an overall detection rate of 92% was achieved at the 10 ng g(-1) level, increasing to 98% at 200 ng g(-1). A screening detection limit (as defined in the SANCO guidelines) of 10 ng g(-1) could be achieved for 134 compounds. For 39 and two pesticides the SDL was 50 and 200 ng g(-1), respectively. For the other nine compounds no SDL could be established. The identification (ion ratio) criteria as recommended in the SANCO document could be met for 93% of the detected pesticide/matrix/concentration combinations. The outcome of both validations shows that the described method can be used to combine quantitative analysis and the identification of frequently detected pesticides (so far typically done using triple quadrupole MS/MS) with a qualitative screening to be used for a wide range of less frequent detected compounds in one measurement.
Keywords: LC-MS; fruits; method validation; pesticides; vegetables.