Development of Experimental Results by Artificial Neural Network Model for Adsorption of Cu2+ Using Single Wall Carbon Nanotubes

• Published in: Separation science and technology 2013-05, Vol.48 (10), p.1490-1499
• Main Authors: GeyikASCi, Feza, Aoruh, Semra, KA-lA-ASC, Erdal
• Format: Article
• Language: English
• Subjects: Adsorption; artificial neural network; carbon nanotubes; Construction; copper; isotherm; kinetic; Learning theory; Mathematical models; Neural networks; Neurons; Single wall carbon nanotubes
• ISSN: 0149-6395; 1520-5754
• DOI: 10.1080/01496395.2012.738276

Removal of copper ions from aqueous solution using single wall carbon nanotubes (SWCNTs) as a function on pH was studied using batch technique. The results indicate that adsorption is strongly dependent on pH. The adsorption of Cu 2+ on SWCNTs increases slowly with increasing pH value at pH  7.0. The equilibrium adsorption data were analyzed by the Langmuir, Freundlich, and Temkin adsorption isotherm models. The Freundlich adsorption model agrees well with experimental data. The pseudo-second order kinetic was the best fit kinetic model for the experimental data. The experimental results were also constructed an artificial neural network (ANN) to predict removal of copper ions. A four-layer ANN, an input layer with four neurons, two hidden layers with 13 neurons, and an output layer with one neuron (4-8-5-1) is constructed. Different training algorithms are tested on the model proposed to obtain the best weights and bias values for ANN. Our results suggest that SWCNTs have a good potential application in environmental protection. This novel modeling tool is newly grown and has been used yet to model the above-mentioned experiments for SWCNTs.