Hierarchical ZSM-5 zeolite catalysts for the selective oxidation of benzene

Zeolites are widely used as catalysts, especially in oil refining and the petrochemical industries. Nowadays the cracking of heavy oil feeds as well as the processing of larger (bio)molecules demands for improved catalysts that can overcome the pore size constraints and diffusion limitations of the conventional zeolite based microporous catalyst. The challenge is to develop new catalyst which have pores in the mesopore size region combined with the catalytic activity and stability of the conventional zeolite. A relatively easy method to increase the pore size from the micropore to the mesopore region can be found in the synthesis of the ordered mesoporous silicates. These materials can be made in a broad range of compositions with a variety of pore sizes. Nevertheless they are amorphous and therefore lack the typical features as stability and strong acidity found in crystalline zeolites. A set of SBA-15 type mesoporous silicates was synthesized with varying silicon to aluminium ratios in an attempt to increase the Brønsted acidity of the mesoporous silicate. Decreasing the silicon to aluminium ratio below 10 lead to a loss of the ordered pore structure. When the ratio falls below 5 it was not possible to obtain any ordered mesoporous structure despite the use of a single source molecular precursor or a different solvent to match hydrolysis rates. The aluminium structure and coordination in these SBA-15 catalysts resembles that of conventional amorphous silica alumina (ASA). This nature is reflected in the activity of the SBA-15 catalysts in the hydroisomerisation of n-heptane, where the activity for SBA-15 type catalysts is only marginally higher when compared to other ASA’s. The small difference in activity can be explained by the presence of an increased amount of weakly acidic hydroxyl groups on the surface of the SBA-15. Apparently ordered mesoporous aluminosilicates resemble the conventional ASA’s, and therefore lack stability and strong Brønsted acidity needed for catalysis. To systematically test the effect of mesoporosity on the catalytic activity of zeolites several synthetic routes including carbon black templating, desilication and templating with organosilanes were employed to synthesize Fe/ZSM-5 catalysts. The hierarchical Fe/ZSM-5 catalysts show similar crystallinities, which is substantially lower than the conventional microporous Fe/ZSM-5. Hierarchical catalysts prepared by either desilication or organosilane templating show an number of active Fe2+ centers comparable to the conventional catalyst. The carbon black templating method, which can be used to obtain high mesoporosity, is not conductive to the formation well dispersed iron and therefore shows an low active site density. Hierarchical Fe/ZSM-5 catalysts show an increased activity in the selective hydroxylation of benzene by nitrous oxide. These increased catalytic properties are a result of the hierarchical structure. First of all the open framework enhances diffusion, which results in a higher initial activity for the catalysts, secondly there is an increased accessibility of the microporous space. As a consequence of the latter the catalyst is less prone to deactivation as a smaller part of the microporous space becomes clogged upon formation of carbonaceous deposits. The best performance in terms of activity and stability is obtained for the organosilane templated catalysts, as these combine a high mesopore volume well integrated into the crystal with an amount of active centers comparable to that of the conventional Fe/ZSM-5 catalyst. As the organosilane templated Fe/ZSM-5 catalyst outperforms the other hierarchical Fe/ZSM-5 catalysts in terms of activity and stability, the mechanism of formation of this hierarchical ZSM-5 zeolite was studied as function of time, temperature and template amount. It is shown that the crystallization of the hierarchical ZSM-5 is a two stage process. Firstly an amorphous highly mesoporous aluminosilicate phase is formed. In the second step this phase is consumed to form the final hierarchical zeolite, which consists of large globular particles. Under the reaction conditions tested, complete consumption of the amorphous phase is not possible. As a result an aluminium rich amorphous phase remains after the crystallization process and the Brønsted acidity of the hierarchical zeolite is low compared to conventional ZSM-5 with a similar Si/Al ratio. Increasing the synthesis temperature for the hierarchical zeolite leads to a more crystalline material, however this comes at the expense of the mesoporosity due to the limited stability of the organosilane mesoporogen at elevated temperatures. A decrease of the synthesis temperature on the other hand favors the formation of the mesoporous amorphous phase at the expense of the hierarchical zeolite. Addition of iron to the synthesis gel enhances the crystallization rate yielding hierarchical Fe/ZSM-5 zeolites, which show higher crystallinity as their hierarchical ZSM-5 analogues. The hierarchical ZSM-5 catalysts with various crystallinities were tested for their activity in the catalytic hydroconversion of n-heptane. Due to the low incorporation of aluminium in the framework the hierarchical ZSM-5, the catalysts show low activity compared to conventional ZSM-5. Where the hydroisomerisation is highly dependent on strong Brønsted acid sites, the selective hydroxylation of benzene depends on the presence of active iron centers and the structure of the catalyst. This is well reflected by the hierarchical Fe/ZSM-5 catalysts grown as a function of time. Where the initial activity for the hierarchical Fe/ZSM-5 catalyst crystallized only for a short time is lower compared to conventional Fe/ZSM-5, the overall phenol production is already higher due to the increased resistance towards deactivation. With increasing crystallization time for the hierarchical Fe/ZSM-5 both the initial activity and phenol production increase. As formation of the hierarchical Fe/ZSM-5 using an organosilane surfactant as the mesoporogen does lead to the formation of highly active catalysts, the mesopore structure appears random and decreases with crystallization time and temperature. Moreover part of the synthesis gel is not converted to a crystalline material and thus inactive in catalysis. To overcome these problems a dry gel conversion approach was used. Initial a amorphous hierarchical Fe,Al-silicate was grown a low temperature. At the given synthesis conditions surfactants comparable to the organosilane yield MCM-type mesoporous silicates, however only mesoporous silicates with a wormhole pore structure were obtained with the organosilane. Interestingly the "amorphous" Fe,Al-silicate synthesized using an organosilane surfactant as the mesoporogen shows already features of zeolitic ordering. Where the material is amorphous in XRD, the IR and Raman spectra show the presence of zeolitic building blocks. This remarkable feature is well reflected in the hydroxylation of benzene were the amorphous catalyst shows a relatively high activity compared to other mesoporous materials. To increase the activity in the benzene oxidation the as synthesized mesoporous Fe,Al-silicate was dry gel converted to a hierarchical zeolite. Initially the mesoporous framework is destroyed yielding a catalyst, which shows hardly any activity, however with increased dry gel conversion times a hierarchical zeolite grows. The iron on this hierarchical Fe/ZSM-5 zeolite is highly dispersed making the catalyst only moderately active in the catalytic decomposition of N2O. The activity for the selective hydroxylation of benzene on the other hand increases rapidly with increased dry gel conversion times as a result of the formation of well dispersed very small zeolitic domains in a mesoporous matrix. The higher activity of these hierarchical catalysts suggests that the conventional microporous Fe/ZSM-5 reference catalyst suffers from severe diffusion limitations due to its large coherent crystalline domain size. Clearly the decrease of the crystalline micropore domain size is beneficial in terms of longevity where it comes to deactivation due to clogging of the micropore space. In that light it seems worthwhile to develop a catalyst where the crystalline domains are as small as possible. Although small nanoblocks of several unit cells in dimension can be synthesized, their limited stability makes them unsuited for catalysis. Another way of decreasing the domain size to the limits is by using so called nano-sheets. In this case sheets are formed which have a thickness of exactly one unit cell, creating extremely short micropore path lengths along the straight channel of the MFI type zeolite. The effect of the open framework and short micropore channels on the selective hydroxylation of benzene in tremendous. The nanosheets have a higher activity and longevity than all the other Fe/ZSM-5 catalysts, despite the large amount of coke deposits on the catalysts. Physisorption shows that for the nanosheets, in contradiction to the conventional Fe/ZSM-5, the majority of the coke resides in the mesopores. Obviously the short micropore diffusion path lengths prevent secondary hydroxylation of the product phenol in the micropore and thus hamper the clogging of the micropores. The mesopores in the catalyst on the other hand open up the framework and provide a pathway to the micropores making them accessible for catalysis increasing the overall activity of the catalyst. The work in this thesis focuses on the synthesis of hierarchical zeolites and their application as catalysts. Although numerous hierarchical porous materials have become accessibile in the last two decades, there is still a need for improvement of tailoring their (pore) structure and reactivity. Whereas initially research has focussed on the synthesis of novel materials by trial and error, nowadays there is a shift towards rational design of hierarchical porous catalysts. It can be expected that in the near future rational design of templates to target specific zeolite pore topologies including mesopores will become possible. In this respect, it should be stressed that the number of templates accessible by straightforward organic synthesis methods is virtually endless. In order to select the most promising of these new templates thorough understanding of the templating-silicate interactions and zeolite growth are necessary. Therefore, the design of new hierarchical systems in the future will become an area involving the interplay of computer simulations, synthesis of templates and by applying them novel porous structures: ideally, we will be able to predict templates for synthesis of optimal zeolites for specific reactions by computer simulations.