Assessment of Test Portion Sizes after Sample Comminution with Liquid Nitrogen in an Improved High-Throughput Method for Analysis of Pesticide Residues in Fruits and Vegetables.

In this study, sample processing of bulk commodities using an efficient one-step comminution procedure with liquid nitrogen (LN2) was devised and assessed in the analysis of pesticide residues in fruits and vegetables. LN2 was added to the fresh samples from a tank by opening a valve, and the standard food chopper was kept in a laboratory hood to reduce safety risks. Test portions of four replicates each of 0.25, 0.5, 1, 2, 5, 10, and 15 g were taken from eight fruits and vegetables (tomato, squash, broccoli, apple, grape, peach, green bean, and cucumber) individually comminuted with LN2. For comparison without comminution, similar test portions of a reconstituted freeze-dried certified reference material of pesticides in cucumber were also analyzed by the same method. More than 100 pesticides were monitored by both ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) and instrument top sample preparation (ITSP) + fast low-pressure gas chromatography-tandem mass spectrometry (LPGC-MS/MS). A new version of QuEChERS-based sample preparation was followed, in which 5 mL of 4:1 (v/v) acetonitrile/water per gram of sample is used for extraction and 200 μL of initial extract is quickly evaporated, reconstituted in water, and ultracentrifuged for UHPLC-MS/MS analysis. For ITSP+LPGC-MS/MS, another portion of the initial extract undergoes salt-out partitioning with 4:1 (w/w) anhydrous MgSO4/NaCl and the upper layer extract is transferred to an autosampler vial for automated cleanup and analysis in parallel. Quality control spikes were made during the comminution, extraction, cleanup, and analysis steps to isolate and estimate the individual and overall measurement uncertainties of the approach. The recommended test portion size is 2 g for routine monitoring by this approach, but results demonstrated that subsamples as low as 0.5 g typically gave overall biases and relative standard deviations of <10% for nearly all pesticides, commodities, and methods, which is 3-5% lower than previously evaluated sample processing and analytical methods. This approach can be used to improve data quality, laboratory efficiency, and sample throughput in routine monitoring programs for regulatory, risk assessment, and other purposes.