Matrix-effects of vegetable commodities in electron-capture detection applied to pesticide multiresidue analysis.

The influence of the sample matrix in the analysis of pesticides in vegetable samples has been studied in order to determine if the matrix content introduces a systematic or proportional (or both) bias in the measurements. Experiments have been carried out during a 4-month period, in which calibration curves, prepared in solvent and in vegetable matrix, were prepared and analysed. A statistical treatment has been applied in order to: (i) check the stability of such calibrations during the period studied; (ii) compare both solvent and matrix-matched calibrations; and (iii) obtain a correction function. Applying the correction function to the results obtained with a solvent calibration it is possible to make a prediction of the values obtained applying a matrix-matched calibration. The performance of the correction function has been validated with recovery data. Finally the uncertainty derived from the use of each calibration plot and the correction function has been calculated.