Utilização de metodos quimiometricos em cromatografia gasosa com microextração em fase solida / Utilization of chemometric methods in gas chromatography with solid phase microextraction

AUTOR(ES)
DATA DE PUBLICAÇÃO

2007

RESUMO

The purpose of this work was to explore the utilization of gas chromatography with the previous stage of sample preparation through solid phase microextraction in different matrices, innovating and becoming feasible these applications using chemometric methodologies, specially the ones that involve artificial intelligence technique. It was realized the simultaneous optimization for the determination of five pesticides in infuses of Passiflora alata Dryander through the neuron-genetic approach, in other words, through the utilization of genetic algorithm to the optimization of the neural network model. It was obtained for two extraction fibers (PDMS/PVA and PDMS 30 µm) the following parameter values, respectively: 67 and 77 % for the tea dilution, 65,5 and 65,0 ºC for the extraction temperature, 50 and 39 minutes for the extraction time and 0,36 g mL for the NaCl concentration to both fibers. The neuron-genetic approach was also utilized to the simultaneous optimization of the determination of 12 polychlorinated biphenyls in human milk. The genetic algorithm provided the optimized extraction conditions as: NaCl concentration of 0,36 g mL, extraction temperature of 95,0 ºC, extraction time of 60 minutes and addition of 210 µL of methanol. By using beer as sample, it was realized an exploratory study of 20 commercial Brazilian beer labels stored in can, through the Kohonen neural network. It was possible to visualize the formation of six sets and through the mass spectra it was identified the volatile compounds responsible for the similarities and differences among the beer samples, through the fraction volatile analysis, utilizing a PDMS/DVB fiber. Finally, 32 commercial beer labels stored in glass bottles, among these 31 were national and one was international, were evaluated in a sensorial way through the Quantitative Descriptive Analysis to the bitterness and grain taste attributes and it was correlated the sensorial results with the chromatographic data of the beer samples by the utilization of genetic algorithm to the selection of the compounds (identified by the mass spectra) directly correlated with the quality parameters studied

ASSUNTO(S)

solid phase microextraction gas chromatography cromatografia de gas quimiometria chemometrics microextração em fase solida

ACESSO AO ARTIGO

Documentos Relacionados