Dynamic fuzzy neural network for simulating the fixed-bed adsorption of cadmium,nickel, and zinc on bone char |
| |
| Authors: | F A Gordillo-Ruíz F J Sánchez-Ruíz D I Mendoza-Castillo H E Reynel-Ávila A Bonilla-Petriciolet |
| |
| Institution: | 1.Instituto Tecnológico de Aguascalientes,Aguascalientes,Mexico;2.Consejo Nacional de Ciencia y Tecnología, Cátedras CONACYT,Mexico,Mexico |
| |
| Abstract: | This study introduces the application of a dynamic fuzzy neural network for fitting and simulating the adsorption of nickel, cadmium, and zinc ions in mono- and bi-metallic solutions (nickel–cadmium, nickel–zinc, and cadmium–zinc) using packed-bed columns with bone char. This neural network model has shown a flexible and self-adaptive architecture with a faster learning speed than that of traditional artificial neural approaches. Results showed that this neural network model was reliable for representing the high asymmetry behavior of concentration profiles in both mono- and bi-metallic breakthrough curves where its accuracy was quite reasonable. Breakthrough parameters for mono-component and binary systems of tested heavy metals were calculated and compared. This analysis showed that the removal of these heavy metal ions in binary systems was a strong competitive adsorption process where the presence of co-ions reduced the removal performance of bone char at fixed-bed adsorbers. Results of surface characterization of adsorbent samples with X-ray photoelectron and infrared spectroscopy supported a removal mechanism based on an ion exchange between calcium from hydroxyapatite of bone char and heavy metal ions in the solution forming new metal–phosphate interactions in the adsorbent surface. |
| |
| Keywords: | |
| 本文献已被 SpringerLink 等数据库收录！ |
|
