Métodos quimiométricos aplicados na determinação espectrofotométrica de misturas de aminoácidos

• Ano de defesa: 2007
• Autor(a) principal: Nunes, Lidiane Cristina
• Orientador(a): Reis, Efraim Lázaro
• Banca de defesa: Fidencio, Paulo Henrique , Queiroz, Maria Eliana Lopes Ribeiro de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Mestrado em Agroquímica
• Departamento: Agroquímica analítica; Agroquímica inorgânica e Físico-química; Agroquímica orgânica
• País: BR
• Palavras-chave em Português: Quimiometria; Análise de componentes principais; Aminoácidos; Espectrofotometria
• Palavras-chave em Inglês: Chemometry; Principal component analysis; Amino acids; Spectrophotometry
• Área do conhecimento CNPq: CNPQ::CIENCIAS EXATAS E DA TERRA::QUIMICA::QUIMICA ANALITICA
• Link de acesso: http://locus.ufv.br/handle/123456789/2173

The purpose of this paper was to make simultaneous amino acid determination using methods of multivariated calibration. The method is based on the reaction between amino acids and sodium 1, 2-naphthoquinone 4-sulphonate (NQS). A set of 32 synthetic mixtures with different concentrations of Phenylalanine Histidine, Proline, Tryptophan and Isoleucine was prepared in accordance with an factorial planning 25. The derivative specters were obtained by the reaction in a flow system, the solutions were added continuously, using a peristaltic pump with fixed outflow in 0,40 ml min-1. Once that the NAS decomposes in alkaline way, the reaction was developed on-line for the insertion in confluence of NQS and the tamponade solution (Na2CO3/NaOH) pH of 10. For the development of the methodology, the reaction bobbin was placed in a 70 ºC thermal bath and kept at this temperature. The final solution was collected in quartz bucket and the reading was carried through in a Hitachi model U 2000 UV-VIS spectrophotometer. Two chemometric techniques were applied: Principal Component Analysis (PCA) and Partial Least Squares (PLS). The multivaried models were developed from the original specters (UV-VIS), registered between 290 and 590 nm and preprocessed with first derivative. For the PLS technique, better results were obtained with models developed from the derivated data, involving the use of 9 latent variable model (VLS). In these conditions, forecast errors below to 6% were obtained for all the species in study. In the real samples analysis, some inconsistencies originated mainly for the presence of not shaped substance were observed. In a generalized manner, the results show a good forecast capacity of the multivariated models based on a UV-VIS spectrophotometry. For real samples analysis, however, it has the need of considering the presence of other substances that should be shaped with the interest amino acids.