Supported metal catalysts dispersion

Supported metal catalysts are reduced, for example, by treatment in hydrogen at temperatures in the range of 300—500°C. The reduction temperature may influence the stabiUty of the metal dispersion. 

For a supported metal catalyst, the BET method yields the total surface area of support and metal. If we perform our measurements in the chemisorption domain, for example with H2 or CO at room temperature, adsorption is limited to the metallic phase, providing a way to determine the dispersion of the supported phase. 

Table III. Effect of Dilution on the Dispersion of Supported Metal Catalysts...

Supported metal catalysts, M°/S, are typically two-components materials built up with a nanostructured metal component, in which the metal centre is in the zero oxidation state (M°), and with an inorganic support (S), quite various in its chemical and structural features [1], M° is the component typically deputed to the electronic activation of the reagents involved in the catalyzed reactions. S is typically a microstructured component mainly deputed to the physical support and to the dispersion of M° nanoclusters. 

In many cases there is an interaction between the carrier and the active component of the catalyst so that the character of the active surface will change. For example, the electronic character of the supported catalyst may be influenced by the transfer of electrons across the catalyst-carrier interface. In some cases the carrier itself has a catalytic activity for the primary reaction, an intermediate reaction, or a subsequent reaction, and a dual-function catalyst is thereby obtained. Materials of this type are widely employed in reforming processes. There are other cases where the interaction of the catalyst and support are much more subtle and difficult to label. For example, the crystal size and structure of supported metal catalysts as well as the manner in which the metal is dispersed can be influenced by the nature of the support material. 

Recently, ultrathin evaporated films have been used as models for dispersed supported metal catalysts, the main object being the preparation of a catalyst where surface cleanliness and crystallite size and structure could be better controlled than in conventional supported catalysts. In ultrathin films of this type, an average metal density on the substrate equivalent to >0.02 monolayers has been used. The apparatus for this technique is shown schematically in Fig. 8 (27). It was designed to permit use under UHV conditions, and to avoid depositing the working film on top of an outgassing film. ... 

Recently, Chaudhari compared the activity of dispersed nanosized metal particles prepared by chemical or radiolytic reduction and stabilized by various polymers (PVP, PVA or poly(methylvinyl ether)) with the one of conventional supported metal catalysts in the partial hydrogenation of 2-butyne-l,4-diol. Several transition metals (e.g., Pd, Pt, Rh, Ru, Ni) were prepared according to conventional methods and subsequently investigated [89]. In general, the catalysts prepared by chemical reduction methods were more active than those prepared by radiolysis, and in all cases aqueous colloids showed a higher catalytic activity (up to 40-fold) in comparison with corresponding conventional catalysts. The best results were obtained with cubic Pd nanosized particles obtained by chemical reduction (Table 9.13). 

Supported metal catalysts are used in a large number of commercially important processes for chemical and pharmaceutical production, pollution control and abatement, and energy production. In order to maximize catalytic activity it is necessary in most cases to synthesize small metal crystallites, typically less than about 1 to 10 nm, anchored to a thermally stable, high-surface-area support such as alumina, silica, or carbon. The efficiency of metal utilization is commonly defined as dispersion, which is the fraction of metal atoms at the surface of a metal particle (and thus available to interact with adsorbing reaction intermediates), divided by the total number of metal atoms. Metal dispersion and crystallite size are inversely proportional nanoparticles about 1 nm in diameter or smaller have dispersions of 100%, that is, every metal atom on the support is available for catalytic reaction, whereas particles of diameter 10 nm have dispersions of about 10%, with 90% of the metal unavailable for the reaction. 

Y. Yoo, M. Tuck, R. Kondakindi, C.-Y. Seo, Z. Dehouche, K. Belkacemi, Enhanced hydrogen reaction kinetics of nanostructured Mg-based composites with nanoparticle metal catalysts dispersed on supports , J. Alloys and Compounds, 446-447 (2007) 84-89. 

In many catalytic systems, nanoscopic metallic particles are dispersed on ceramic supports and exhibit different stmctures and properties from bulk due to size effect and metal support interaction etc. For very small metal particles, particle size may influence both geometric and electronic structures. For example, gold particles may undergo a metal-semiconductor transition at the size of about 3.5 nm and become active in CO oxidation [10]. Lattice contractions have been observed in metals such as Pt and Pd, when the particle size is smaller than 2-3 nm [11, 12]. Metal support interaction may have drastic effects on the chemisorptive properties of the metal phase [13-15]. Therefore the stmctural features such as particles size and shape, surface stmcture and configuration of metal-substrate interface are of great importance since these features influence the electronic stmctures and hence the catalytic activities. Particle shapes and size distributions of supported metal catalysts were extensively studied by TEM [16-19]. Surface stmctures such as facets and steps were observed by high-resolution surface profile imaging [20-23]. Metal support interaction and other behaviours under various environments were discussed at atomic scale based on the relevant stmctural information accessible by means of TEM [24-29]. 

In supported metallic catalysts, the metals are usually from Groups VIII and VB of the Periodic Table. For highly dispersed metallic catalysts, the support or the carrier is usually a ceramic oxide (silica or alumina) or carbon with a high surface area, as described in chapter 2. Supported metallic catalysts can be prepared in a number of ways as described by Anderson (1975). A description of some of the methods used to prepare representative model (thin film) and practical (technological) powder systems follows. 

HREM methods are powerful in the study of nanometre-sized metal particles dispersed on ceramic oxides or any other suitable substrate. In many catalytic processes employing supported metallic catalysts, it has been established that the catalytic properties of some structure-sensitive catalysts are enhanced with a decrease in particle size. For example, the rate of CO decomposition on Pd/mica is shown to increase five-fold when the Pd particle sizes are reduced from 5 to 2 nm. A similar size dependence has been observed for Ni/mica. It is, therefore, necessary to observe the particles at very high resolution, coupled with a small-probe high-precision micro- or nanocomposition analysis and micro- or nanodiffraction where possible. Advanced FE-(S)TEM instruments are particularly effective for composition analysis and diffraction on the nanoscale. ED patterns from particles of diameter of 1 nm or less are now possible. 

It appears that supported metal catalysts can be used to promote synthetically useful organometallic reactions. The utilization of such reactions can be of practical, economic, and environmental importance to the fine chemical industry. Frontier Molecular Orbital and mechanistic considerations indicate that these reactions, along with hydrogenations and, presumably, oxygenations, take place on the coordinately unsaturated comer atoms present on the surface of these dispersed metal catalysts. 

In addition, the same studies that were carried out on the Pt(lll) crystal face result in reaction rates identical to those found on stepped crystal surfaces of platinum. These observations support the contention that well-defined crystal surfaces can be excellent models for polycrystalline supported metal catalysts. It also tends to verify Boudart s hypothesis that cyclopropane hydrogenolysis is an example of a structure-insensitive reaction. The initial specific reaction rates, which were reproducible.within 10%, are within a factor of two identical to published values for this reaction on highly dispersed platinum catalysts. The activation energies that were observed for this reaction, in addition to the turnover number, are similar enough on the various platinum surfaces so that we may call the agreement excellent. 

The performance of a catalyst is well known to be sensitive to its preparation procedure. For this reason, ideally an oxide-supported metal catalyst should be subjected to a number of characterization procedures. These may include measurements of the metal loading within the overall catalyst (usually expressed in wt%), the degree of metal dispersion (the proportion of metal atoms in the particle surfaces), the mean value and the distribution of metal particle diameters, and qualitative assessments of morphology including the particle shapes and evidence for crystallinity. These properties in turn can depend on experimental variables used in the preparation, such as the choice and amounts of originating metal salts, prereduction, calcination or oxygen treatments, and the temperature and duration of hydrogen reduction procedures. 

The next question is Where do supported metal catalysts fit into this pattern of co-ordination numbers Most platinum group metal catalysts can be prepared in supported forms in which the dispersion (defined as the % of metal atoms exposed at the surface of the particles) approaches 100%. While there may be good grounds for doubting the accuracy of calculations of dispersions, depending as they do on arbitrary assumptions about particle shapes,14 adsorption ratios, etc., it is certain that dispersions greater than, say, 50% are frequently obtained. Table 1 shows how the dispersion relates to particle diameter and to number of atoms for a simple octahedral structure. From this we see that 50% dispersion corresponds to a particle diameter of... 

It is worth mentioning that spontaneous monolayer dispersion is also a very useful scientific basis underlying the process of regeneration of deactivated metal catalysts. Supported metal catalysts may sinter during use at elevated temperatures. Sintering will cause the metal catalyst to lose initial activity, and in order to recover it one has to find an effective way to redisperse the metal on the catalyst support. Applying what we have learned from our studies on spontaneous monolayer dispersion to... 

Redispersion through an oxidation-reduction cycle as described previously is, indeed, an effective way to regenerate supported metal catalysts that have been deactivated because of sintering, and the underlying principle is spontaneous monolayer dispersion. 

---