Modeling of decolorization of synthetic reactive dyestuff solutions with response surface methodology by a rapid and efficient process of ultrasound-assisted ozone oxidation

Abstract

The present study investigates the results of decolorization of Malachite Green (MG), Reactive Black 5 (RB5), and Reactive Yellow 145 (RY145) in aqueous solutions based on a rapid and novel process of ultrasound-assisted ozonation. A Placket–Burman design (PBD) as a factorial design was used to quantify and screen the significant effects of the seven factors on decolorization efficiency: temperature (^oC), initial pH, probe position (height from bottom of reactor, mm), reaction time (min), ozone concentration (g/L), mixing speed (rpm), and ultrasonic power (W). Probe position and mixing speed were not found as significant after considering the regression and ANOVA results of PBD. A Box–Behnken design (BBD) as a kind of response surface methodology, with remaining five factors at three levels was set to demonsrate the interactions. The best-fit multi non-linear regression (MNLR) models were derived by using the results of BBD. According to BBD, the maximum decolorization efficiency of 99.31, 99.86, and 99.52% were obtained consistently at the lowest initial pH of 2, the highest reaction time of 30 min, and ozone concentration of 0.15 g/L for MG, RB5, and RY145, respectively. The best-fit MNLR models were cross-validated (R²_pred) accounting for 81.02–88.25% and were expressed (R²_adj) accounting for 93.01–95.70% of variation in decolorization efficiency.

Keywords:

• Decolorization
• RSM, Anova
• Non-linear regression
• Ultrasonic irradiation
• Ozone oxidation

Acknowledgments

The author would like to thank Dr G.A. Evrendilek for her help with laboratory analysis under her supervision of YENIGIDAM project financially supported by the Turkish State Planning Organization (DPT 2009 K 120 410).