@phdthesis{516452,
  abstract     = {{Even though acidic zeolites form a crucial catalyst for many petrochemical processes, much of their fundamental reactive behavior is only superficially understood. Most often, catalysts are proposed on an 'ad hoc' basis, without a detailed understanding of their functioning on an atomic scale. It can indeed be difficult to identify the elementary steps of complex reaction networks from a purely experimental basis. For these issues, quantum chemical molecular modeling techniques provide an excellent complementary tool to laboratory data. This relatively new field of research has seen an enormous surge in popularity, mainly because of the rapid increase in computer power and the development of sufficiently accurate theoretical methods, which together make it possible now to model complex industrial processes. In this thesis, we use these modeling techniques for a detailed study on elementary reaction steps in zeolite catalysis.This summary gives only a very short overview of the work, and the interested reader is referred to the more elaborate full text or, for even more detail, to the research articles on which it is based, which are also included at the end of each relevant chapter.

In a preparatory chapter, several general terms and methods used throughout the thesis are introduced. First, two fundamental characteristics of zeolites that are vital in industrial catalysis - the topologically induced shape selectivity and the isomorphic substitution leading to a Bronsted acid site - are briefly explained. Then, the practical aspects of quantum chemical modeling of zeolites are discussed, with special attention given to the model space approximations that are necessary for such extended systems. Chemical reactions need to be modeled by computationally very demanding quantum chemical methods if we are to describe the changes in electronic binding pattern appropriately. Different approximations are possible, with an increase in accuracy usually accompanied by an increase in computational cost. Since zeolites are extended materials with a large number of atoms, a complete and accurate quantum chemical description of the entire system is not only extraordinarily demanding but also, at the moment at least, simply not feasible. This issue has, however, led to the development recently of some advanced techniques that do allow an accurate description of at least the chemically active part of the system. Finally, since in this thesis the most important conclusions are based on rate coefficients, the basics of chemical kinetics are also introduced, describing the molecular-scale calculation of macroscopic quantities using transition state theory.

Subsequently one of the most intriguing substantive problems in heterogeneous catalysis is tackled: the reaction mechanism of the methanol-to-olefin process (MTO). First, a whole class of reaction mechanisms, the so-called direct mechanisms, are investigated, for which initial C-C coupling is taken to occur from C1 species only. Earlier theoretical studies tended to be fragmentary, typically investigating only a single reaction step rather than a complete pathway. Nevertheless, the existence of these individual reaction steps was often considered theoretical evidence for the direct proposal, even though no one had succeeded in defining a complete low-energy pathway. To resolve this complex issue, an extensive reaction scheme is presented in this thesis, including all the possible pathways and their constituent elementary reaction steps on a consistent basis. By combining the individual steps, it is demonstrated that the direct mechanism concept cannot explain the initial C-C coupling. Three bottlenecks are identified:

- the instability of ylide and carbene intermediates,
- the extremely slow conversion of a methane/formaldehyde mixture to
ethanol, and
- the excessively high energy barriers for concerted C-C coupling steps.

Any alternative proposal, like the up-and-coming 'hydrocarbon pool' hypothesis, needs to provide C-C coupling steps that circumvent these bottlenecks.

The hydrocarbon pool model states that organic species trapped in the zeolite pores serve as building platforms, to which C1 species can attach methyl groups. The methylated species subsequently undergoes specific rearrangements and/or additional methylation steps, to finally split off light olefins. The original molecule is then regenerated by additional methylation steps. This way, the highly activated steps of the direct mechanisms could be bypassed. In this thesis, the initiating methylation (and at the same time C-C coupling) step is investigated. The results shed new light on the role of the zeolite framework in this process, and also in how the organic species and the inorganic zeolite cooperate as a supramolecular catalyst. The supramolecular picture is extended here by the explicit inclusion of previously omitted aspects like transition state shape selectivity and electronic stabilization of vital cationic intermediates by the zeolite framework. We should definitely look beyond pure geometrical aspects since electronic embedding plays an equally important role.

Additional insight into the hydrocarbon pool hypothesis is, however, required for a guided optimization of the catalyst. A first step to catalyst improvement has already been made by investigating the effect that small organic groups built into the catalyst might have on the elementary reaction steps. Two such modifications - methylene and amine moieties that are iso-electronic with oxygen - are theoretically investigated here. The methylene moiety is one of the simplest organic groups that fits perfectly as a bridge between two silicon atoms to form the functional Si-CH2-Si group. Even though such mesoporous organosilicate materials have been successfully synthesized before, only recently has a research team been able to synthesize methylene-substituted alumino-silicate zeolites. They failed to explain the observed framework defects, though, like the presence of end-standing Si-CH3 groups. In this thesis the influence of the methylene moiety on fundamental adsorption properties is discussed for both neutral probe molecules and charge compensating cations. Additionally, we demonstrate how the combination of aluminum atoms (plus a Bronsted acid proton) with a methylene moiety will inevitably lead to protonation of the organic group and subsequent cleavage of the framework.

For similar amine-functionalized zeolites, this thesis also shows that protonation of the amine group will not necessarily lead to cleavage of the zeolite structure. Furthermore, Si-NH-Si moieties will provide additional basic sites, comparable to traditional Al-O-Si sites but not constrained to the aluminum tetrahedron. This enables more proton locations as well as the possibility of more favorable transition state geometries. This can result in a drastic reduction in energy barrier for those reactions which would otherwise have a highly strained transition state. Summarizing, we demonstrate how small organic modifications to the zeolite framework can have a considerable effect on the fundamental catalytic properties and MTO-related reactivity. However, neither methylene nor amine groups can be located on the aluminum tetrahedron without being automatically protonated, which in the case of methylene-modified zeolites even results in cleavage of the framework.

This thesis shows very clearly how theoretical modeling is capable of providing new insights into zeolite catalysis. The applications presented here are already located near the limits of what is currently feasible, considering computer power, method development and the current lack of insights into the possible supramolecular character of the system. The rapid evolution in this field of research, even within the timescale of this thesis, makes it as good as certain that further significant advances will soon be within reach, and the thesis closes with the identification of our high-priority research goals for the immediate future. Especially in identification of elementary reaction steps and optimization of the catalyst, there are still quite some challenges ahead.}},
  author       = {{Lesthaeghe, David}},
  isbn         = {{978-90-8578-144-8}},
  language     = {{eng}},
  pages        = {{205}},
  school       = {{Ghent University}},
  title        = {{Unraveling the Reaction Mechanism of Industrial Processes in Zeolite Catalysis: a Quantum Chemical Approach}},
  url          = {{http://lib.ugent.be/fulltxt/RUG01/001/215/945/RUG01-001215945_2010_0001_AC.pdf}},
  year         = {{2007}},
}