Uso de Python para geração de gráficos de energia de Gibbs a partir de parâmetros do modelo NRTL, para sistemas ternários de equilíbrio líquido-líquido

• Autor(a) principal: Quinto, Francisco Hiago Barbosa
• Data de Publicação: 2025
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Ceará (UFC)
• Texto Completo: http://repositorio.ufc.br/handle/riufc/80634
• topic: CNPQ::ENGENHARIAS::ENGENHARIA QUIMICA; Extração líquido-líquido; Equilíbrio de fases; Modelos termodinâmicos; Ferramentas computacionais; Aplicativo online; Liquid-liquid extraction; Phase equilibrium; Thermodynamic models; Computational tools; Online application

Several industrial and laboratory processes use liquid-liquid extraction as a basis, either to separate or purify a given compound of interest, since for systems that form azeotropes, or that require a lot of energy to be distilled in the traditional way, this technique proves to be more efficient. The design of liquid-liquid extraction processes requires knowledge of phase equilibrium that can be obtained through thermodynamic models, such as the Non-Random Two-Liquid (NRTL) model, which have parameters adjustable to the experimental data. However, these parameters may present problems that, if not identified, can lead to incorrect design of industrial processes. The generation of Gibbs energy curves and surfaces, which is an indication of the stability of the system, through the adjustable parameters of the models, can help in the identification of inconsistencies and in the understanding of the system. Thus, this work aimed to develop free tools to facilitate the generation of these graphs using the Python language, and comparing the results obtained for two systems previously analyzed in the literature, in addition to obtaining the deviations between the values obtained by the algorithm and those obtained by XSEOS, an Excel add-in. As a result, two Python notebooks were developed, which allow direct access to the source code, and an online application, for quick and easy use by users, with small deviations in the comparison of the cases available in the literature, allowing us to conclude that the developed tools proved to be effective in generating Gibbs energy graphs, from adjustable parameters of the NRTL model.