Teor de óleo de alho em cápsulas gelatinosas por espectroscopia no infravermelho médio e técnicas quimiométricas

• Ano de defesa: 2017
• Autor(a) principal: Visani, Valeria
• 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: Óleo de alho; Infravermelho médio; Técnicas quimiométricas de classificação e calibração multivariada; SPA-LDA; MCR-ALS; Garlic oil; Medium infrared; Chemometric techniques of classification and multivariate calibration; CNPQ::CIENCIAS EXATAS E DA TERRA::QUIMICA
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/12028

Garlic oil, in the form of gelatin capsules, contains various substances considered beneficial to health, which act to lower cholesterol and blood pressure, in the treatment of flu and diabetes, and to prevent the development of tumors and cardiovascular diseases. However, disagreements over its functionality have been reported and, one possible reason is the lack of standardization of these products. Considering that there is no established legislation on the quality control of garlic oil, commercialized in capsules, it is important to develop simple and fast methods for the analysis of these products. In this work, we propose screening and quantification analysis methods, using attenuated total reflectance (ATR) spectroscopy and chemometric methods for the classification and quantification of garlic oil capsules as a function of their concentrations. The Successive Projections Algorithm (SPA), together with Linear Discriminant Analysis (LDA), obtained the best results of classifying capsules in three classes, using models constructed using only the selected variable 924 cm-1, where the existing band is assigned to stretching vibrations of C-S-C and deformation of =CH2, present in the major constituents of garlic oil. For the 136 capsules tested, we obtained a rate of 88.2% of agreement with the information contained in their labels. In the quantification of the garlic oil content of the capsules, the MCR-ALS (Multivariate Curve Resolution with Alternating Least Squares) method, using a column wise configuration, was able to solve the mixtures, recovering the spectra of garlic and soybean oil with correlations of 0.8914 and 0.9997, respectively. As a result, MCR-ALS obtained a lack of fit of 3.63% and explained variance of 99.87%, limits of detection and quantification equals to 0,09% and 0,27%, respectively, RPD (Residual Prediction Deviation) equal to 8,89 and the recovery trials provided indexes of 98% to 99%, indicating good accuracy of the method. As in the classification, the concentrations predicted for some brands of capsules, proved to be in disagreement with the values declared on their labels, including between different batches of the same brand. In this way, it is justified the need of control of the inspection organs with the manufacturers, to guarantee the quality of these products. The results presented suggest that the proposed methods are effective in investigating gelatinous capsules of garlic oil as to their content.