Extração de óleo de soja utilizando solventes verdes: cinética, termodinâmica e transferência de massa

• Ano de defesa: 2022
• Autor(a) principal: Gasparetto, Henrique
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• dARK ID: ark:/26339/001300000z0cc
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Engenharia Química; UFSM; Programa de Pós-Graduação em Engenharia Química; Centro de Tecnologia
• Palavras-chave em Português: Acetato de etila; 1-Butanol; COSMO-SAC; P-cimeno; Ethyl acetat; P-Cymene; CNPQ::ENGENHARIAS::ENGENHARIA QUIMICA
• Link de acesso: http://repositorio.ufsm.br/handle/1/27589

This dissertation focuses on the kinetic, thermodynamic, and mass transfer study on applying green solvents in soybean oil extraction. The first work discusses ethyl acetate and 1-butanol solvents application. Temperature and solvent-to-solid mass ratio variables were optimized using the response surface methodology. The So and Macdonald and the mass transfer kinetic models were applied for modeling extraction kinetics. This study shows that ethyl acetate is the best candidate for directly replacing hexane at an industrial level, even though 1-butanol is better for the solubilization of the triacylglycerols. However, both solvents did not present better results than n-hexane. The second study evaluated the application of p-cymene in soybean oil extraction. Temperature and solvent-to-solid mass ratio variables were also optimized, and an in-depth statistical approach was used to compare kinetic models to describe vegetable oil extraction. So and Macdonald's kinetic model was considered the best in describing the extraction process, which was further associated with bootstrap resampling to reduce the bias of the estimators. P-cymene resulted in better extraction yields than nhexane at 55°C. The COSMO-SAC theory could predict the behavior of the solvents mentioned above through the activity coefficient at infinite dilution compared to the Hansen solubility parameters. The last work used experimental data from the literature and those obtained experimentally for ethyl acetate, 1-butanol, and p-cymene, to evaluate the application of a generalized fractional derivative power order model. Better statistical results were obtained compared to the integer order derivative model through the analysis of the coefficient of determination, root-mean-square error, chi-square distribution, Akaike information criterion, and Fisher's exact test. For all solvents, the thermodynamic assessment indicated endothermic, irreversible, and spontaneous processes; the extracted oil's fatty acid profile and infrared spectrum agree with what the literature present.