Propane Dehydrogenation Over Pt Nanocluster Catalysts

Propylene is an important intermediate material in petrochemical industry. Nowadays, most of the propylene is obtained from steam pyrolysis processes with gasoline,light diesel and light hydrocarbons used as feed, thus, the increasing vehicle fuel demands limits the production of propylene. Due to the growing demand for propylene, the research on propane dehydrogenation to propylene has been becoming a key issue in petrochemical industry.Supported Pt nanocluster catalysts for propane dehydrogenation were prepared by combining sol-gel method and chemical reduction method. Based on the performance evaluation and BET, XRD, pyridine IR absorption spectrum, thermogravimetry characterizations of the catalysts, effects of the support dry time, the preparation method, the promoters Sn,K added were investigated. The results indicated that the catalysts exhibited high catalytic activity, used silica sol-gel as support and under the conditions of PVP/Pt ratio 30 (mass), impregnation time 8 h. The addition of K promoter and the Sn, K promoters can greatly increase the dispersion of Pt nanocluster particles, it improves the catalytic activity in a certain extent. The addition of Sn had significant influence on catalytic activity of the catalysts. A suitable amount of Sn added not only increased the Br?nsted acid sites and moderate/strong Lewis acid sites, but also prevented the catalyst from coking and improved significantly the catalyst stability and activity. Propane conversion and propane selectivity were still maintained at 18.8 % and 93 %, respectively after the catalyst with Sn/Pt 25 on stream for 30 h.The intrinsic kinetics of propane dehydrogenation over the catalysts was studied in a continuous flow fixed bed reactor. Eliminating both internal and external diffusion influence, intrinsic kinetic data were obtained under the conditions of 590~630℃, GHSV 300~1000 h-1, 0.1MP and C3H8/H2 molar ratio 1.0. Based on Langmuir-Hinshelwood mechanism, an intrinsic kinetic model was developed and the model parameters were estimated using the simplex algorithm, and also the model was validated by statistic analysis. The results show that the model which supposed the C-H bond fission as the rate-determining step is the best one to represent the experimental data, and can provide important information for industrial reactor design.