Toward Supported Oxide Catalysts via Solid-Solid Wetting

BY HELMUT KNOZINGER AND EDMUND TAGLAUER

1 Introduction

Supported oxides of transition metals, particularly of groups Vb (V), Vlb (Cr, Mo, W), and Vllb (Re) are widely used as catalysts for various reactions. These so-called "monolayer-type" catalysts are formed when one metal-oxide phase is dispersed on the surface of a second metal-oxide support. Typical catalyst supports in industrial applications are transition aluminas, silica, and titania. Alumina-supported molybdenum and tungsten-based catalyst precursors are extensively used in the petroleum industry in hydrotreating processes. Consequently, numerous studies have been carried out to analyze their function in hydrodesulfurization (HDS), hydrodenitrogenation (HDN), and hydrodemetalization (HDM) of petroleum and coal products. The oxidation of hydrocarbons, carbon monoxide hydrogenation and the water gas shift reaction are also catalyzed by supported molybdena and tungsta. TiO2-supported vanadium, molybdenum, and tungsten oxide catalysts were found to be highly active for the selective catalytic reduction (SCR) of NOx with NH3. The vanadium oxide/TiO2 system is also widely used for selective catalytic oxidations of hydrocarbons. Supported Re2O7 effectively catalyzes the metathesis reaction and chromia-based catalysts are active for polymerizations (SiO2 supported) or redox reactions (Al2O3 supported).

Typically, this class of catalysts is prepared by impregnation of the support from an aqueous solution containing a suitable precursor compound or (less frequently) by gas-phase chemisorption of a volatile metal compound (e.g., Mo(CO)6) on a carrier. When catalysts are prepared by impregnation on an industrial scale, large volumes of solutions must be handled and eventually large volumes of wastewater must be disposed. As a consequence, there might be an interest to synthesize catalysts via alternative routes that would not require impregnation and precipitation steps. Solid-state reactions provide a significant potential in this context, since reactions between two (or more) solids necessarily must involve the interfaces between them. Several processes can occur when an active solid component undergoes reactive interactions with another solid, the support. The active component may (1) retain its chemical identity, the support simply acting as a dispersing agent, (2) dissolve in the support matrix to form a solid solution, or (3) form new surface and/or stoichiometric bulk compounds. Haber has strongly emphasized the role which surfaces and interfaces play in the reactivity of solids. In powder mixtures, depending on the relative rates of nucleation and nuclei growth on one hand and surface migration or gas-phase transport on the other hand, two principal routes for the reaction progress can be envisaged. If the nucleation and nuclei growth rates are much higher than migration rates, a solid-state reaction can only occur at intergranular contacts and will lead to the formation of a bulk compound (Route I). If, in contrast, the migration of one mobile component across the surface of another less mobile component is very fast, grains of the latter will be encapsulated by a thin layer of the former, so that the entire surface becomes the reaction interface (Route II). A schematic representation of the propagation of the reaction interface via routes I and II is given in Figure 1. Several examples of solid-state reactions proceeding via route II have been reported in the literature. If the rate of formation of a bulk compound across the reaction interface is negligibly small, the process may come to a close once the surface layer has formed.

The migration of one solid over the surface of another solid is frequently described as surface diffusion of constituents of the lattice in a concentration gradient. Haber and coworkers suggested the wetting of one solid by a second solid under the action of forces of surface tension as an alternative mechanism.

It is tempting to take advantage of these phenomena known from solid-state chemistry in the preparation of supported oxide catalysts, although this has been realized in practice only in exceptional cases. The increasing interest in this area is in fact documented by a recent review by Xie and Tang on spontaneous spreading, which covers the literature up to 1987. In the present review we are reporting on wetting and spreading phenomena in systems of particular interest for catalyst preparation, where mixtures of oxides will play a central role.

2 Theoretical Considerations

2.1 Thermodynamics of Wetting and Spreading. - The thermodynamics of wetting of a solid by a liquid is well established and discussed in detail in relevant textbooks. The same principles can be applied in the phenomenological treatment of the wetting of one solid by another solid, a phenomenon that also plays a major role in the redispersion of supported particles on the surface of an oxide carrier (e.g., supported catalysts). Sintering and redispersion in supported metal catalysts have been discussed by Ruckenstein in several papers and excellent review articles.

Redispersion of particles on the surface of a carrier is a phenomenon that has much in common with the spreading of one solid component over the surface of a second solid in the course of solid- state reactions as discussed in the introduction. In this case, grains of both components are contacting each other in powder mixtures and the spreading will be initiated from the contact zones. This same situation is apparent when supported catalysts are to be prepared by spreading from powder mixtures containing the support and the precursor of the final supported active phase, where the active phase is formed by spreading of the precursor. It is therefore important to define the conditions under which solid-solid wetting and spreading can be expected to occur. A schematic representation of wetting and spreading is shown in Figure 2.

The overall change in interfacial-free energy ΔF is given by Equation (1):

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where γij denotes the specific surface-free energy between phases i and j, ΔA the change in surface/interface area, and subscripts a, s, and g denote active phase, support and gas phases, respectively. For wetting of the support by the active phase to occur, the interfacial-free energy change must be negative (ΔF<0), hence, the condition

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

or

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

if /ΔAa/ = /ΔAas/ = /ΔAs/ must be fulfilled. Hence, for predictions to be made of whether or not solid-solid wetting can principally occur in a given...