Chemical interactions in multimetal/zeolite catalysts

Abstract

This two-year project has led to a significant improvement in the fundamental understanding of the catalytic action of zeolite-supported redox catalysts. It turned out to be essential that we could combine four strategies for the preparation of catalysts containing transition metal (TM) ions in zeolite cavities: (1) ion exchange from aqueous solution; (2) chemical vapor deposition (CVD) of a volatile halide onto a zeolite in its acidic form; (3) solid state ion exchange; and (4) hydrothermal synthesis of a zeolite having TM ions in its lattice, followed by a treatment transporting these ions to ''guest positions''. Technique (2) enables us to position more TM ions into cavities than permitted by the conventional technique (1).viz one positive charge per Al centered tetrahedron in the zeolite lattice. The additional charge is compensated by ligands to the TM ions, for instance in oxo-ions such as (GaO){sup +} or dinuclear [Cu-O-Cu]{sup 2+}. While technique (3) is preferred over CVD where volatile halides are not available, technique (4) leads to rather isolated ''ex lattice'' oxo-ions. Such oxo-ions tend to be mono-nuclear, in contrast to technique (2) which preferentially creates dinuclear oxo-ions of the same TM element. A favorable element for the present research was thatmore » the PI is also actively engaged in a project on the reduction of nitrogen oxides, sponsored by EMSI program of the National Science Foundation and the US Department of Energy, Office of Science. This combination created a unique opportunity to test and analyze catalysts for the one step oxidation of benzene to phenol and compare them with catalysts for the reduction of nitrogen oxides, using hydrocarbons as the reductant. In both projects catalysts have been used which contain Fe ions or oxo-ions in the cavities the zeolite MFI, often called ZSM-5. With Fe as the TM-element and MFI as the host zeolite we found that catalysts with high Fe content, prepared by technique (2) were optimal for the De-NO{sub x} reaction, but extremely unselective for benzene oxidation to phenol. Conversely, the catalysts prepared with (4) had the highest turnover frequency for benzene oxidation, but performed very poorly for NO{sub x} reduction with so-butane. In fact the Fe concentration in the former catalysts were so low that it was necessary to design a special experimental program for the sole purpose of showing that it is really the Fe which catalyzes the benzene oxidation, not some acid center as has been proposed by other authors. For this purpose we used hydrogen sulfide to selectively poison the Fe sites, without deactivating the acidic sites. In addition we could show that the hydrothermal treatment of catalysts prepared by technique (4) is essential to transform iron ions in the zeolite lattice to ''ex lattice ions'' in guest positions. That line of the work required very careful experimentation, because a hydrothermal treatment of a zeolite containing Fe ions in its cavities can also lead to agglomeration of such ions to nano-particles of iron oxide which lowers the selectivity for the desired formation of phenol. This part of the program showed convincingly that indeed Fe is responsible for the benzene oxidation catalysis. The results and conclusion of this work, including the comparison of different catalysts, was published in a number of papers in the scientific literature, listed in the attached list. In these papers also our analysis of the reaction orders and the possible mechanism of the used test reaction are given.« less