Evaluation of some parameters affecting troublesome pesticide analysis in gas chromatography-ion-trap mass spectrometry

Nowadays, multiresidue methods for pesticides monitoring in food commodities are commonly employed. It is also well known that the presence of several compounds from the matrix introduces a bias during the detection and the quantification steps. The so-called matrix effect phenomenon is related to the masking or formation of active sites. In GC, this phenomenon occurs in the injector port, and in the separative system (retention gap and/or analytical column) and also causes ionization potential modification of analytes. The main consequence of matrix effect is an increasing or decreasing analyte signal in the presence of the matrix (real sample) in respect to the same analyte in solvent (standard solution). In standard mixture, pesticides themselves interact with the active present sites among analytical chain from injector to the detector. Some matrix components, sometimes at trace amounts, are inevitably present in analyzed samples even after numerous and diverse clean-up procedures. In this paper, the influence of some analytical parameters on pesticide signal response is explored using gas chromatography with ion-trap mass-selective detection (GC-IT-MS). Moreover, the responses of characteristic troublesome analytes are analyzed in various kinds of matrices. Finally, matrix compound identification is initiated to study analyte-matrix relationship. Sample acidification with 0.1% acetic acid was the most appropriate for the majority of pesticides, while 0.1% formic acid was more suitable for base-sensitive ones (amitraz, imazalil, thiabendazole). Among tested calibration methods, matrix matched calibration provides the best results. In green bean matrix model, a matrix/pesticide ratio of 1/1 induces the best detected signal for almost every investigated analytes. Presence and quantity of some identified matrix co-extracted compounds like sterols could be partially a cause of signal enhancement. (c) 2008 Elsevier B.V. All rights reserved.