Técnicas de Modelagem Matemática Aplicadas ao Estudo de Células a Combustível de Etanol Direto

Abstract

Direct alcohol fuel cells are a subcategory of proton exchange membrane cells. They have been studied by researchers due to their low operating temperature, solid state of the electrolyte and good efficiency compared to other types of fuel cells. Within this context, this work modeled the kinetics of direct ethanol fuel cells through phenomenological and empirical models. Firstly, ideal models were considered, based on Tafel kinetics, representing the complete oxidation of ethanol (for Pt-Sn anodes with different Sn contents). Such models were not consistent with the phenomenon studied, a fact already expected due to their limitations. Later, realistic models were considered, taking into account the formation of by-products and lower generation of electrons. This type of model describes the catalytic oxidation of ethanol by adsorption of species and formation of partially oxidized products. These models showed good fit, confirming the hypothesis about the degree of alloying of Sn in catalysts and its role in hydroxyl formation during the reaction. For Pt3Sn anodes submitted to different heat treatments, an empirical model was developed, based on an artificial neural network, which mapped the potential/current relationship as a function of the treatment temperature. This type of model has proved to be more suitable for practical applications such as control and prediction systems.