Sequential experimental designs for discrimination among adsorption equilibrium models

Abstract

Sequential experimental designs for model discrimination were used for the selection of the most appropriate adsorption isotherm for describing the adsorption of tartrazine onto magnesium oxide and of methyl orange onto activated carbon. Typical adsorption equilibrium models were considered. For the adsorption of tartrazine, it was observed that the sequential approach permitted the re-evaluation of the considered models, as new information was acquired, allowing the inclusion of models not considered initially. For the adsorption of methyl orange, it was observed that model discrimination was achieved only after five additional experiments. These results showed that the use of sequential experimental designs allowed the effective selection of experimental conditions for model discrimination, which might not be possible without the optimization of the design criterion. In both cases, the BET isotherm was the selected model, indicating that adsorption of both adsorbates onto different solids occurs with multilayer formation.