A neural network prediction of conversion of benzothiophene oxidation catalyzed by nano-Ti-beta catalyst

The oxidation of benzothiophene (BT) in isooctane with hydrogen peroxide was performed in a batch reactor in a non-severe condition using nano-crystalline Ti-beta catalyst synthesized by dry gel conversion method. The Ti-beta catalyst was characterized by XRD and HRTEM analysis. A three layered feedforward neural network was employed to predict the BT conversion with respect to reaction time, catalyst concentration, hydrogen peroxide to BT mole ratio, temperature and initial BT concentration. The sensitivity analysis of process parameters showed that the BT conversion is highly sensitive to the reaction time, initial BT concentration and temperature.