REPOSITORY > RESULTS

Abstract

Propylene oxide (PO) is a very versatile industrial raw material. It is used in the production of many consumer goods. With the improving living standards and an increase in global population, the global demand for PO is steadily increasing. Currently implemented production methods with organic peroxides or chlorohydrin have several drawbacks such as pollution, high costs of production, and low demand for side products of the reaction. For decades researchers have studied the direct propylene epoxidation reaction using molecular oxygen as the oxidant. So far, the tested catalysts have fallen short of the industrially applicable conditions (activity, stability, and selectivity).
In my work, I focused on copper-based catalysts, where amorphous silica was used as the support. I subsequently modified these catalysts with different alkali (Na, K, and Cs) and alkaline earth metals (Ca). The goal was to gain new understanding on the requirements and reaction pathway for molecular oxygen and propylene activation in the direct propylene oxidation reaction using molecular oxygen. Furthermore, we wanted to understand the underlying factors that determine catalyst selectivity and stability. To examine the viability of the materials in propylene epoxidation, the catalysts were tested in a fixed-bed quartz reactor at atmospheric pressure. Additional in-situ and ex-situ techniques were employed to acquire more information about the mechanism of the reaction. Active phase properties were studied with ex- and in-situ UV/vis, Diffuse Reflectance Infrared Fourier transform (DRIFT) spectroscopy, Temperature Programmed Desorption of CO2 (CO2-TPD), and pyridine saturation were employed to probe the acidic/basic properties of the surface. The materials were imaged with the help of scanning and transmission electron microscopy. Local structure and oxidation state of copper during reaction was observed with in-situ XAS. All the experimental data was complemented by theoretical techniques.
We successfully synthesized highly dispersed modified and unmodified CuOx nanoparticles that were, on average, smaller than 1 nm. With unmodified catalysts we observed significant deactivation as a result of sintering. This was greatly reduced by alkali and earth alkaline modification, and in the case of Cs modification completely eliminated. We found that with our materials, Lewis acid sites have negligible impact on PO selectivity. Catalytic tests showed that only the modification with alkali metals increases PO selectivity. This modification generates additional basic sites of moderate strength on the catalyst surface. We propose these sites are oxygen species with a more electrophilic character, which are selective for propylene epoxidation. Different reduction dynamics were observed depending on the modifying adatom. This indicates altered kinetics of oxygen abstraction and replenishment. A similar amount of Cu+ was seen for the unmodified, Na+, and Ca2+ modified catalysts, which have very different selectivities. This shows that Cu+ amount is not the only factor determining PO selectivity. We found that, in general, alkali metal modification (Na) reduces the average Cu-O bond length, and earth alkaline metals (Ca) have the opposite effect.
We propose that the Cu-O bond length reduction shows a reduction in the nucleophilic character of surface-bound oxygen species. Consequently, allylic hydrogen stripping (AHS) is prevented and oxametallacycle (OMC) formation is promoted. From this we constructed a tentative reaction mechanism. This was additionally confirmed by theoretical techniques, which disclosed a drastic increase in the activation barrier (AB) of AHS. Additionally, a slight increase in the AB for OMC formation was noted, however it was energetically favorable compared to AHS. A decrease in the AB for ring closure and PO desorption was observed with alkali modification.