Modelagem e otimização da produção de biossurfactante utilizando melaço de soja

• Ano de defesa: 2018
• Autor(a) principal: Silva, Ana Carolina Borges
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Engenharia Química
• Palavras-chave em Português: Biossurfactante; Modelagem; Otimização; Biosurfactant; Modeling; Optimization; CNPQ::ENGENHARIAS::ENGENHARIA QUIMICA::PROCESSOS INDUSTRIAIS DE ENGENHARIA QUIMICA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/28349; http://dx.doi.org/10.14393/ufu.di.2019.53

Biosurfactants are one of the most important classes of chemicals, being called amphiphilic molecules with hydrophilic (polar and water soluble) and hydrophobic (nonpolar and water insoluble) moieties. There has been a growing increase in the interest of biosurfactants, as these are biodegradable and non-toxic. Among the classes of biosurfactants, the glycolipids stands out, especially the raminolipids, that are produced by bacteria of the Pseudomonas aeruginosa type and have high affinity with organic hydrophobic molecules and less toxicity to the environment, and are able to reduce the surface tension and to emulsify hydrocarbons . Therefore, the objective of this study was to model the production of raminose in a batch-operated process and to optimize it using a fixed-bed continuous process. The modeling of the production of raminose, cell growth and consumption of lipids and sucrose was established using the data obtained in Rodrigues Master's Dissertation (2016). The proposed kinetic models were Saturation by Cells, Monod and Contois. The technique of parameter identification was performed using a non-linear regression algorithm of multiple responses and the integration of the set of differential equations for the calculation of the parameters was performed with the aid of the Runge-Kutta algorithm. The yields were calculated in terms of biomass and product, both for the experimental data and for the kinetic models studied and the qualitative analyzes were performed in order to characterize the biosurfactant produced. The validation of the adjusted parameters was performed by means of a continuous operation column experiment in which the reactor height was 10.9 cm and the flow velocity was 0.288 cm / h. The kinetic models were validated using the PFR and Axial dispersion models. Afterwards, the optimization was performed by graphical visualization of the behavior of the response variable, for the Contois model, being the concentrations of substrates evaluated for different speeds and the response obtained in terms of productivity. The results obtained for the parametric adjustment were satisfactory and the models fitted well to the experimental data, being the Saturation by Cell model what better represented the best experimental points, presenting a value of 19.61 for the sum of the squares of the residues. The Monod and Contois models presented a value of 47.61. The yields Yx / s showed that there is a higher consumption of lipids, being the Contois model the one that presented the best conversion of biomass in raminose. The qualitative analyzes showed good results for surface tension and emulsification index. For the validation, the model that presented the best results for the PFR model was the one of Saturation by the Cell, being necessary to readjust the terms of maintenance. For the Axial Dispersion model, Contois kinetic model showed the best results, predicting the final concentrations for process in an excellent way. The results obtained by the Contois model for the optimization were quite satisfactory and demonstrate less significant improvements in productivity with increasing substrate concentrations and flow velocity.