Use of Neural Network and Genetic Algorithm in Modeling of Dye Separation from Aqueous Solutions by Adsorption onto Carbon Nanotubes

In this study, adsorption of methylene blue from aqueous solution on carbon nanotubes was investigated. Experimental data at three temperatures of 290, 300 and 310 K were fitted to equilibrium models and at 300K with kinetics models. It was found that the Langmuir isotherm is the best fit with isothermal data and the kinetic data could be successfully fitted by McKay and Ho as a pseudo-second order model. The effects of different experimental parameters, such as contact time, solution pH, initial concentration of methylene blue and dose of adsorbent were investigated as effective parameters. Ninety groups of the experimental data for developing a neural network were used. The inputs data were initial solution concentration, dose of adsorbent, initial pH of the solution, contact time and temperature while the final methylene blue concentration in the solution was the output of the network. Finally, the network was optimized using the genetic algorithm. The results were compared with the experimental data and a very good agreement was observed.