Modeling of reactive orange 16 dye removal from aqueous media by mesoporous silica/ crosslinked polymer hybrid using RBF, MLP and GMDH neural network models

Abstract

In this study, SBA-15 mesoporous silica was synthesized and functionalized with cross-linked polyacrylic acid and used to remove Reactive Orange 16 (RO16) from aqueous media. X-ray diffraction (XRD), filed emission scanning electron microscope (FE-SEM), fourier transform infrared spectrometer (FT-IR), thermo gravimetry analysis (TGA) and BET method were used to study the characteristics of the resulting nanocomposite (SBA-15/CPAA). The effective parameters on dye removal, including pH, contact time, temperature and adsorbent dosage were investigated and optimized. The optimum condition was: pH = 2, adsorbent dosage of 0.03 g and contact time of 60 min at 25 °C. Uv–Vis spectrophotometer was used to determine the amount of residual dye in solution. Multilayer Perceptron (MLP), Redial Basis Function (RBF) and Group Method of Data Handling (GMDH) models have been applied for predicting adsorption value according to the amount of pH, dosage, temperature, concentration and contact time. Comparison of these models revealed that the MLP model was more accurate than the other two models.