Determination of glyphosate, AMPA and glufosinate in honey: Comparative tests on the sample preparation stage

Abstract

Beekeeping is an activity of great economic and eco-systemic importance, and Argentina is a major country in the world in this activity. Regarding the cleaning and safety of honey, residues and chemical contaminants´ unwanted presence is of great concern, both by residues originated in the treatments applied to the hive and by the use of agrochemicals in agriculture that potentially contaminate the beekeeping chain. To meet today´s quality and safety requirements is necessary to realize more pesticide control analysis.Argentina, The United States and Brazil use 80% of glyphosate´s world total. In our country since 1996 has been applied to various transgenic crops headed by soy, with a significant annual load (200 thousand tons). Therefore, the main objective was to develop an analytical methodology to determine glyphosate, AMPA and glufosinate in honey and related matrices such as nectar. Based on the pre-column derivatization strategy with FMOC-Cl and analysis by UHPLC-MS/MS. Different alternatives were evaluated and adapted to the objective. For methodology optimization: extraction solvents, volume and concentration of the derivatizing reagent, volume and concentration of the borate buffer, different reagents to achieve the optimum pH (pH 9), reaction time and cleaning alternatives of the extracts also were evaluated (L-L partition and SPE).Recovery tests at two concentration levels (10 and 100 ug kg-1) in triplicate were performed to evaluate the methodology. First, honey samples weighed in 50 mL centrifuge tubes were placed in a thermostated bath at 40-50°C for 5 min to achieve fluidity. They were then spiked with the standard mixture (glyphosate, AMPA and glufosinate) and vortexed to homogenize. After that, the extraction, derivatization and cleaning stage were performed for each of the evaluated alternatives. The final methodology involves an extraction step with water and mechanical agitation for 30 min followed by the derivatization reaction and finally L-L partition with dichloromethane as a clean-up stage1. Satisfactory results have been obtained, showing that this method is consistent and reliable, with RSDs < 20% and recoveries between 70-105% at 5, 10, 25 and 50 ug kg-1.Also, together with the optimization of the sample preparation, to achieve a better response of the analytes as a whole, the instrumental conditions were optimized based on experiments´ statistical designs to increase detection sensitivity to obtain low detection and quantification limits.In this way, it was possible to achieve an acceptable analytical method for determining glyphosate, AMPA and glufosinate that was applied to analyze raw and commercial honey samples.