Nitrate removal by quaternized mesoporous silica gel in ternary anion solutions: Flow-through column experiments and artificial neural network modeling
The aim of this study was to examine nitrate removal by quaternized silica gel (q-SG) in ternary solutions of nitrate, phosphate, and sulfate under flow-through column conditions. q-SG was synthesized by grafting dimethyloctyl[3-(trimethoxysilyl)propyl] ammonium chloride on silica gel. Fixed-bed column experimental conditions (N = 15) were designed using central composite design to examine dynamic removal behaviors of competing anions in columns containing q-SG. During the experiments, influent solution containing ternary anions of nitrate, phosphate, and sulfate was injected into flow-through columns. In the effluent, the ternary anions along with chloride were monitored to obtain competitive breakthrough curves. Column experiments demonstrated the dynamic and competitive removal behaviors of anions during adsorption and leaching in the columns. Artificial neural network (ANN) model was developed based on the column experimental data to predict the removal rates of anions in the column experiments. In the model development, influent concentrations of nitrate, phosphate, and sulfate were selected as three variables in the input layer, whereas removal rates of nitrate, phosphate, and sulfate were chosen as three variables in the output layer. The developed ANN model with topology 3:8:9:3 (three input variables, eight neurons in the first hidden layer, nine neurons in the second hidden layer, and three output variables) could simultaneously predict the removal rates of anions in column experiments.
