AN INTEGRATED APPROACH OF ARTIFICIAL NEURAL NETWORK (ANN) AND RESPONSE SURFACE METHODOLOGY (RSM) IN OPTIMIZATION OF DYE ADSORPTION ONTO SEPIOLITE
Author(s): Sermin Elevli ,Feza Geyikci, Erdal Kilic, Semra Coruh
J. Ponte - Feb 2016 - Volume 72 - Issue 2
doi: 10.21506/j.ponte.2016.2.1
Abstract:
This study deals with the adsorption of basic dye (crystal violet) onto natural sepiolite. The effects of experimental factors (particle size, adsorbent dosage, initial concentration, and pH) on the adsorption process were examined by an integrated approach of ANN and RSM. At the first step, an ANN was successfully constructed to model dye adsorption using existing data from traditional experiments. The established model was then used as a predictor to achieve better understanding of the adsorption process and to obtain optimal settings of the experimental factors. From the obtained counter graphs of Box-Behnken Design, it was concluded that settings for dosage around 30 g/L, pH around 6, particle size around 0.5 mm and concentration around 100-200 mg/L must be chosen in order to maximize removal efficiency. The results also showed that sepiolite is a suitable material for the adsorption of crystal violet from aqueous solution.
|