학술논문

Toward a generic extraction method for simultaneous determination of pesticides, mycotoxins, plant toxins, and veterinary drugs in feed and food matrixes
Document Type
Author abstract
Report
Source
Analytical Chemistry. Dec 15, 2008, Vol. 80 Issue 24, p9450, 10 p.
Subject
United States
Language
English
ISSN
0003-2700
Abstract
A fast and straightforward generic procedure for the simultaneous extraction of various classes of pesticides, mycotoxins, plant toxins, and veterinary drugs in various matrixes has been developed, for subsequent analysis by liquid chromatography with mass spectrometric detection. As a first step, four existing multianalyte procedures and three newly proposed methods were compared for a test set of 172 pesticides, mycotoxins, and plant toxins spiked to a feed matrix. The new procedures, which basically involved extraction/dilution of the sample with water and an acidified organic solvent (methanol, acetonitrile, or acetone), were most promising. The three new generic extraction methods were further tested for applicability to other matrixes (maize, honey, milk, egg, meat). Overall, the best recoveries were obtained for acetone, followed by acetonitrile. With respect to matrix effects, acetonitrile was the most favorable solvent and methanol was the worst. The occurrence of matrix effects decreased for the matrixes in the order of feed > maize > meat > milk > egg > honey. The extraction method selected as the default procedure (water/acetonitrile/1% formic acid) was also evaluated for applicability to multiple classes of veterinary drugs in all six matrixes, with satisfactory results. Finally, the generic extraction procedure was validated for 136 pesticides, 36 natural toxins, and 86 veterinary drugs in compound feed and honey at three levels (0.01, 0.02, and 0.05 mg/kg) using ultraperformance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) for analysis of the extracts. For over 80% of the analytes, recoveries were between 70 and 120% and precision (expressed as relative standard deviation) was mostly in the range of 5-10% (except for feed at 0.01 mg/kg; adequate recoveries for 62% of the analytes). The limits of detection were from