Aplicação de desenho experimental estatístico para otimização de método para determinação de ditiocarbamato em erva-mate (Ilex paraguariensis)

Abstract

Dithiocarbamates (DTC) are one of the main classes of fungicides applied to the control of fungal diseases on a large variety of crops. In this work, a fast and cheap sample preparation procedure for DTC analysis in yerba mate using gas chromatography-mass spectrometry (GC-MS) was developed. A multivariate approach by Placket-Burmann Design (PBD) was carried out for screening of DTC extraction variables (concentration and volume of hydrolysis acid solution, isooctane volume, and water bath shaking time and temperature). The significant variables were studied by a Central Composite Rotational Design (CCRD) to optimize the DTC recovery. A complete validation study was performed according to the European Union guideline -SANTE document. The screening experiment obtained by PBD indicates that three experimental parameters were significant: concentration and volume of hydrolysis solution, and time of hydrolysis. The final optimized and validated method proved to be reproducible and robust in routine analysis. For recovery studies, blank yerba mate samples were spiked with a thiram solution at the 3 spike concentrations of 0.1; 0.5 and 2.0 mg kg-¹, as CS2, and analyzed by 3 analysts (n = 7, per analyst). Excellent recoveries, in the range of 81.4 to 116.2%, with good precision (all RSDs ≤ 16.1%), were obtained. The intermediate precision (RSDR) was evaluated by different analysts (1 and 2) and a RSDR≤ 8.2% was obtained, indicating the very good repeatability of the method. The interlaboratory reproducibility also showed recovery and precision values within the acceptable limits (70-120%, and <20%, respectively). The robustness of the method was proven to be good, based on the acceptable deviations with changes of most conditions of the critical parameters. The analytical curves were linear in the concentration range of 0.05-10 mg L-¹ CS2, r²> 0.999, and the matrix effect was not significant (-14.9%). The LOD and LOQ were 0.05 and 0.1 mg kg-¹ CS2, respectively. The method presented important advantages: a five-fold reduction of the isooctane extraction volume; an increase of sample throughput from 1 to 30 sample extractions, due to sample mass reduction without HCl addition; a two-fold reduction of extraction time; a smaller sample amount to process, and less waste volume aiming at green chemistry principles. The application of the multivariate approach proved to be of influence in improving the extraction efficiency.