Sammendrag

This paper investigated the possibility of leaving out the traditional clean-up step in the QuEChERS procedure and analysing non-cleaned extracts from fruit, vegetables and cereals with a combination of gas chromatography-tandem mass spectrometry (GC-MS/MS), back-flush technology and large-volume injection. By using calibration standards in cucumber matrix, recovery and precision were calculated in lettuce, orange and wheat for 109 pesticides at 0.01 and 0.1 mg kg−1 in two sets of samples: one with and one without clean-up. For both spiking levels, 80–82% of the pesticides in the non-cleaned extracts and 80–84% of the pesticides in the cleaned extracts were within the acceptable recovery range of 70–120%. Precision data for both levels showed that 95% of the pesticides in the non-cleaned extracts and 93–95% of the pesticides in the cleaned extracts had RSDs below 20%. Recovery and precision data were determined using a two tailed t-test (p = 0.05). By using calibration standards in the respective matrix, we studied if the non-cleaned calibration standards gave an extra matrix effect compared with the cleaned standards by using the slope from calibration graphs and plotting the calculated extra matrix effect minus 100 for each compound. The results showed that for 79% of the pesticides, the extra matrix effect minus 100 was within the acceptable range of −20% to 20%. Five European Union proficiency tests on rye, mandarin, rice, pear and barley, respectively, from 2010 to 2012 were reanalysed omitting the clean-up step and showed satisfactory results. At least 70 injections of non-cleaned extracts were made without detecting any increased need for maintenance during the experimental period. Analysing non-cleaned QuEChERS extracts of lettuce, orange and wheat are possible under the conditions described in this paper because recovery, precision and specificity showed satisfactory results compared with samples subjected to traditional dispersive clean-up.

Sammendrag

Kontroll av vegetabiler for restmengder av plantevernmidler har tradisjonelt vært utført ved hjelp av GC med selektive detektorer som NPD og ECD. En slik instrumentering krever en manuell tolkning av kromatogrammer fra de to detektorene samt at overskridelser av grenseverdier og påvisning av ulovlig bruk må verifiseres med analyse på GC-MS. Tolkningsdelen krever lang erfaring og mange måneders opplæring av personell. I tillegg er den svært tidkrevende (0,5-1 time pr. prøve). Tolkningen av kromatogrammene er såpass komplisert og operatøravhengig at mulighetene for å utgi falske negative prøvesvar alltid er til stede. Full Scan GC-MS analyse kombinert med Deconvolution Reporting Software (DRS) fra Agilent Technologies erstatter den manuelle tolkningen med automatisk søk i to databaser med massespektra. DRS programvaren er en applikasjon som kombinerer resultatene fra MS-chemstation, AMDIS (Automatic Mass Spectral Deconvolution and Identification Software) og NIST (National Institute of Standards and Tecnology) databasesøk. Den automatiske tolkningen tar kun 1-2 minutter pr. prøve og resultatene skrives ut på en rapport. Vår erfaring viser at GC-MS-DRS sparer tid, gir sikrere identifikasjon og avslører flere pesticider i prøvene. I tillegg har det vært mulig å senke bestemmelsesgrensen for en rekke pesticider.