Photocatalytic removal of C.I. Basic Red 46 on immobilized TiO2 nanoparticles: artificial neural network modelling.

C.I. Basic Red 46, commonly used as a textile dye, was photocatalytically removed using supported TiO2 nanoparticles irradiated by a 30 W UV-C lamp in a batch reactor. The investigated photocatalyst was industrial Degussa P25 (crystallite mean size 21 nm) immobilized on glass beads by a heat attachment method. The catalyst was characterized by XRD, SEM, TEM and BET techniques. The process of the dye decolorization in the presence of TiO2 nanoparticles was experimentally studied through changing the initial dye concentration, UV light intensity and initial pH. The influence of inorganic anions such as chloride, sulphate, bicarbonate, carbonate and phosphate on the photocatalytic decolorization of BR46 was investigated. The decolorization of BR46 follows the pseudo-first-order kinetic according to the Langmuir-Hinshelwood model (k1 = 0.273 mg L(-1) min(-1), 2 = 0.313 (mg L(-1))(-1)). The efficiency parameters such as apparent quantum yield and electrical energy per order (EEO) were estimated. An artificial neural network model (ANN) was developed to predict the photocatalytic decolorization of BR46 solution. The findings indicated that the ANN provided reasonable predictive performance (R2 = 0.96). The influence of each parameter on the variable studied was assessed: initial concentration of the dye being the most significant factor, followed by the initial pH and reaction time.