Towards a Novel Process for the Epoxidation of Propene over Gold/Titania-Based Catalysts

One of problems connected with the use of gold/titania-based catalyst systems is the low yield in propene oxide (up to 2 wt %). It is shown that the low propene oxide yield is due to product inhibition. New options for propene epoxidation with hydrogen and oxygen are discussed. INTRODUCTION The direct gas-phase synthesis of propene oxide (PO), by the use of molecular oxygen, is still aimed for. Agbased catalysis is in principle possible, but selectivity at present is not promising. Highly dispersed gold on titanium containing catalysts show an extraordinary selectivity in the oxidation of propene to the corresponding epoxide (> 99 %), using a combination of H2 and O2 as oxidation mixture [1]. Additional studies of the epoxidation reaction have indicated several drawbacks of the Au/TiO2 catalyst. Firstly, the low conversion (1- 2 %) obtained under typical reaction conditions (atmospheric pressure, 323 - 423 K) cannot be increased by increasing the reaction temperature, because this leads to extensive by-product formation. Secondly, the low efficiency of hydrogen (relatively large amounts of water are formed) is hard to prevent. Finally, significant deactivation of Au/TiO2 catalysts within several hours on stream has been observed. Deactivation can be almost completely prevented by using dispersed TiO2 supports, such as TiO2/SiO2 and Titanium-Silicalite-1 (TS-1) [2]. In recent literature [2,3] a hydroperoxide-like intermediate has been proposed to be responsible for the selective epoxidation. This oxidising species is said to be formed over gold and subsequently used for epoxidation over the Ti-containing support. The support is responsible for the slow desorption of PO. One of the problems for future industrial application might be the low yield in propene oxide. Furthermore, the presence of minor amounts of by-products, not detected on lab-scale catalyst testing, might complicate the separation unit in an industrial process. The reason for the low PO yield over Au/TiO2-based catalysts will be discussed in this paper. Temperature Programmed Desorption (TPD) and micro-flow experiments have been carried out to study PO desorption and catalyst performance, respectively. A catalytic cycle is proposed together with its consequences for the maximum yield in PO. Finally, several process and catalyst options are discussed in view of potential future industrial application.