Using RBF and MLP models in predicting Reactive Red 198 removal from aqueous solution by SBA-15/CTAB composite

In this study, the mesoporous material SBA-15 were synthesized and then, the surface was modified by the surfactant cetyl trimethyl ammonium bromide (CTAB). Finally, the obtained adsorbent was used in order to remove Reactive Red 198 (RR 198) from aqueous solution. transmission electron microscope (TEM), thermogravimetric analysis (TGA), Fourier transform infra-red spectroscopy (FTIR), X-ray diffraction (XRD), and BET were utilized for the purpose of examining the structural characteristics of the obtained adsorbent. Parameters affecting the removal of RR 198 such as pH, the amount of adsorbent, and contact time were investigated at various temperatures and were also optimized. The obtained optimized condition is as follows: pH = 2, time = 60 min and adsorbent dose = 1 g/l. Moreover, predictive models based on multi-layer perceptron (MLP) and radial basis function (RBF) networks used for predicting the adsorption amount according to the input parameters including pH, dosage, temperature, time and concentration are presented. Two criterions, namely, CC and RMSE between the observed and predicted amounts are used to validate the models. Comparison of the obtained results using these two models showed that the prediction based on the MLP network model is better than the RBF one.