Comparing the catalytic performances of mixed molybdenum with cerium and lanthanide oxides supported on HZSM-5 by multiobjective optimization of catalyst compositions using nondominated sorting genetic algorithm

Mixed oxides, La2O3–MoO3 and CeO2–MoO3, were being used as catalysts for thermal catalytic cracking of naphtha. Effect of catalyst compositions on the yield of light olefins (ethylene and propylene) were systematically examined using central composite design (CCD) coupled with the response surface methodology. Two empirical models for each system, based on these preparation variables for the yields of ethylene and propylene were constructed in two CCD studies. These models, shown as contour diagrams, indicated that the yields of ethylene and propylene gradually increased up to a certain amount of loadings. In fact, the yield of light olefins has been decreased by the high loadings of rare earth elements (La and Ce) and molybdenum. In order to compare the catalytic performance of these two systems, a multiobjective optimization (MOO) for simultaneous maximization of ethylene and propylene was carried out by the elitist nondominated sorting optimization algorithm or NSGA-II. This algorithm resulted in Pareto-optimal solutions and an additional criterion was proposed over the solutions to obtain a final unique optimal solution. It was found that the supported CeO2–MoO3 increased the yield of light olefins slightly more than supported La2O3–MoO3 at the optimum point.