Estratégia para a modelagem de dados de segunda ordem usando cromatografia líquida para a quantificação de pesticidas em hortaliças

• Ano de defesa: 2020
• Autor(a) principal: Sousa, Emanuella Santos
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Química; Programa de Pós-Graduação em Química; UFPB
• Palavras-chave em Português: Resíduos de pesticidas; HPLC-DAD; Calibração multivias; MCR- ALS; Pesticide residues; Multiway calibration; CNPQ::CIENCIAS EXATAS E DA TERRA::QUIMICA
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/18227

Pesticides are substances capable of controlling the spread of pests, increasing the cultivation, production and supply of food. For this reason, they are among the emerging pollutant classes of concern most investigated in the environment, as they can be found in the air, soil, water and, mainly, food. he present work shows a study of the development of a chromatographic methodology and multivariate calibration, using chemometric techniques such as UPLS-RBL (Unfold Partial Least Square with Residual Bilinearization) and MCR-ALS (Multivariate Curve Resolution-Alternating Least Square), for quantification of multidrug residues of pesticides (carbendazim, thiabendazole, fuberidazole, carbofuran, carbaryl, flutriafol and 1-naphtol) in vegetable samples obtained in the city of João Pessoa-PB. The objectives of the research were to optimize a fast and effective method for the determination of these pesticides in vegetable samples using QuEChERS extraction and: Liquid Chromatography-Diode Array Detector (LCDAD). To reduce the amount of solvent spent and the analysis time, a gradient elution was used, however, this elution causes a baseline capable of interfering in the rank estimate of the multivariate calibration data matrix. To solve this problem, caused by the technique, a “chemometric cleanup” was proposed, subtracting the blank from all samples. In addition, the performance of alignment algorithms was evaluated, which also seek to circumvent the shift in retention time that occurs between chromatographic measurements. The performance of the COW alignments (Correlation Optimized Warping), icoshift (Interval Correlation Optimized Shifting) and coshift (Correlation Optimized Shifting) were studied. The latter showed better performance in the calibration samples, however, in the vegetable samples it did not obtain a satisfactory result, causing distortions in the spectral and elution profiles, not being used as a pre-treatment step for the samples. In quantifying the analytes, the two chemometric techniques (UPLS-RBL and MCR-ALS) were evaluated in order to achieve the second order advantage. The MCR-ALS performed better in the calibration samples, with REP ranging from 5.1 to 10.2% and quantification limits (LOQ) ranging from 0.0093 to 0.058 mg L-1 , meeting the maximum residue limits established Brazilian legislation (ANVISA). For the UPLS-RBL, it was not possible to obtain low prediction errors (REP), since the model generated took into account the recovery of the analyte profile itself or the profile of the coeluted constituent. In the samples of vegetables, which consists of fruits, leaves and tubers, it was possible to quantify the presence of carbendazim and carbaryl in the cultures of peppers, lettuce, carrots and beets, with concentrations ranging between 0.02 and 0.07 mg L-1 , however, these pesticides are not allowed in these crops; carbofuran was quantified in four of the six samples, with concentrations ranging from 0.014 to 0.08 mg L-1, this pesticide is no longer authorized in the country. Thiabendazole and flutriafol were also quantified, but within the limits established by the legislation. In short, this research aims to assist in the development of studies of multiresidue pesticides in more complex matrices, considering the obtaining of the second order advantage acquired by multipath calibration. In addition, it is extremely important to alert the population to food security, in order to collect from government institutions all information on research, use and monitoring of pesticide residues.