Statistical modelling and optimization of dibenzothiophene (DBT) oxidation in deep desulfurization of diesel

In this research, the effects of process variables on the efficiency of DBT oxidation in the formicacid/H2O2 system for oxidative deep desulfurization of diesel are systematically evaluated by statistical modelling, analysis and optimization using response surface methodology (RSM) by implementing the Central Composite Design (CCD). Three control variables including temperature, H2O2/sufur ratio, and catalyst dosage are investigated. A quadratic regression model is developed to predict the yield of sulfur elimination as the model response. Analysis of variance confirmed that the developed model is in good agreement with the experimental results. The model indicates that three studied variables have significant effects on the response; however, temperature is the most significant factor. Moreover, the model suggests an important interaction between temperature and H2O2/sulfur ratio contributed to the response, which can be attributed to the thermal decomposition of H2O2 at higher temperatures and water hindrances which produced from oxidative desulfurization reaction. The optimization accomplished by the model shows that the optimal condition for maximum yield of desulfurization is obtained at high temperature (57 °C), minimum H2O2/sulfur ratio (2.5 mol/mol) and catalyst dosage of 0.82 mL in the reaction system (50 mL solution of DBT in n-hexane including 500 ppmw concentration of sulfur). Using these optimal values, the maximum yield of desulfurization is predicted 95% after 1 hr reaction. In the optimization process, minimizing H2O2/sulfur ratio and catalyst dosage for the maximum yield of desulfurization is economically considerable. The results indicate that RSM can be applied effectively for the modelling of DBT oxidation and economical optimization for higher efficiency of deep desulfurization of model fuel.