Removal of direct blue 129 from aqueous medium using surfactant-modified zeolite: a neural network modeling

Conserving water for human survival and providing future security are important issues thatneed to be addressed.Methods: In this study, a zeolite modified with hexadecyl trimethyl ammonium bromide (HDTMA-Br), acationic surfactant, and its application in removing direct blue 129 (DB129) was examined. Fourier transforminfrared spectroscopy (FT-IR) and scanning electron microscopy (SEM) were used to characterize bothmodified and unmodified zeolites. The effects of operational parameters such as the amount of adsorbent,initial dye concentration and pH on the removal efficiency of the dye were examined.Results: The results showed that in the initial dye concentration of 50 mg/L, the optimum amounts ofadsorbent and pH were 0.3 g and 7, respectively. Increasing the dye concentration from 20 to 100 mg/Lresulted in the reduction of the removal efficiency from 100% to 79% in the contact time of 90 minutes. Theresults indicated the highest attracting correlation with Langmuir model. The maximum adsorbent capacityobtained from Langmuir model was 25 mg/g. The kinetics of the dye adsorption on the modified zeolitefollowed pseudo-second-order kinetics model. Calculated thermodynamic parameters showed that Gibbsfree energy changes (DGo) at temperatures of 20 and 45°C were -29.41 and -35.20 kJ/mol, respectively.Enthalpy (DHo) and entropy changes were equal to 41.181 kJ/mol and 0.241 J/mol K, respectively. Theresults showed that the processing was a spontaneous endothermic reaction. The process modeled byartificial neural networks (ANN) showed that the experimental results can be accurately modeled usingneural network model. The correlation coefficient found between the experimental and the model resultswas 0.951.Conclusion: Due to the low cost, high abundance and availability of zeolite, the removal efficiency of thisadsorbent can be increased to desirable levels by modifying.

Keywords:

Zeolite , Adsorption , Kinetics , Thermodynamics , Neural networks