Application of statistical experimental design to optimize the photocatalytic degradation of a thiazin dye using silver ion-doped titanium dioxide.

A three-factor, three-levels Box-Behnken design combined with response surface methodology (RSM) was used to optimize the photocatalytic degradation process of Methylene blue (MB), a cationic thiazin dye, using Agdoped TiO₂ under UV irradiation in a batch reactor. The individual and interaction effects of three operational parameters, photocatalyst dose (1.5-2.5 g/L), initial dye concentration (10-50 ppm) and pH (5-9), selected based on single factor study, on the colour removal (decolourization) and COD removal (mineralization) of the dye were determined by fitting the results of the experiments to two quadratic polynomial models relating the parameters to the response variables. Design Expert software version 8.0.6.1 was used to determine the effects of the parameters on the two responses and the optimum values of the parameters. The initial dye concentration was found to have a higher negative effect on the two responses, while photocatalyst dose and pH had lower positive effect. Under the optimum conditions (dose of Ag⁺ doped TiO₂ 1.97 g/L, initial concentration of MB 22.86 ppm and pH of reaction mixture 6.68) the decolourization and mineralization rate of MB were predicted as more than 96% and 71%, respectively. The high correlation between adjusted and predicted R values indicates goodness of fit of the model.