Copper aggregation

Fig. 1. Copper aggregates formed by electrodeposition. Top Fractal growth. Bottom Dense-Branching Morphology. The diameter of the cathode wire is 50 pm. 

The surface properties of unsupported ruthenium-copper aggregates are considered in this chapter. In a subsequent chapter on bimetallic cluster catalysts, the properties of supported ruthenium-copper and osmium-copper catalysts are considered in detail. 

The crystal structures of metallic ruthenium and copper are different, ruthenium having a hexagonal close-packed structure and copper a face-centered cubic structure (7). Although the ruthenium-copper system can hardly be considered one which forms alloys, bimetallic ruthenium-copper aggregates can be prepared that are similar in their catalytic behavior to alloys such as nickel-copper (3,4,8). 

After reduction, the ruthenium-copper aggregates prepared by either method are cooled to room temperature in a stream of helium. They are then passivated by gradual admission of air to the helium and stored in closed containers until needed. In general, after these materials are charged to various kinds of experimental apparatus, they are again contacted with hydrogen at elevated temperature as a further step in preparing them for measurements of interest. 

If portions of the materials are exposed to hydrogen at temperatures (500-600°C) higher than the 400°C employed in the initial preparation, the properties of the final ruthenium-copper aggregates are changed substantially, despite the fact that the total surface areas are not very different. The ruthenium-copper aggregates prepared by these procedures have surface areas in the approximate range of 5-7 m2/g. In such aggregates the surface atoms constitute approximately 1 % of the total metal atoms. 

Ruthenium-copper aggregates of the type described have been studied with chemical and physical probes. Chemical probes that have been very informative include hydrogen chemisorption and the hydrogenolysis of ethane to methane. Physical probes useful in these characterizations include X-ray diffraction and electron spectroscopy. 

In a typical hydrogen adsorption experiment, ruthenium-copper aggregates are first contacted with flowing hydrogen in the adsorption cell at 400°C to ensure thorough reduction. The cell is then evacuated to a pressure of approximately 10-6 torr and cooled to room temperature for adsorption measurements. Isotherms for total hydrogen adsorption and for weakly adsorbed hydrogen are then determined in the manner described for nickel-copper catalysts in Chapter 2. 

Typical isotherms for ruthenium-copper aggregates containing 5 at.% copper are shown in Figure 3.1. The dashed line represents strongly adsorbed... 

In the upper field of Figure 3.2 the hydrogen chemisorption data refer to strongly adsorbed hydrogen. The volume adsorbed per square meter of surface is shown for ruthenium-copper aggregates of varying copper content. 

Figure 3.2 Hydrogen chemisorption capacity and ethane hydrogenolysis activity of ruthenium-copper aggregates as a function of copper content (3). (Reprinted with permission from Academic Press, Inc.)...

Thus although ruthenium and copper are immiscible in the bulk, ruthenium-copper aggregates can be prepared that have surface properties very different from those of pure ruthenium. The ruthenium-copper aggregates exhibit chemisorption and catalytic properties which would not be expected for simple physical mixtures of ruthenium and copper granules. The presence of the copper clearly has a marked effect on surface processes occurring on ruthenium. On the basis of the chemisorption and catalytic results, we conclude that the copper tends to cover the ruthenium surface. We thus adopt the view that copper is chemisorbed on the ruthenium. 

In addition to hydrogen chemisorption and ethane hydrogenolysis, the reactions of cyclohexane provide a useful chemical probe for investigating ruthenium-copper aggregates. On pure ruthenium, two reactions of cyclohexane are readily observed dehydrogenation to benzene and hydrogenolysis to lower carbon number alkanes. The product of the latter reaction is predominantly methane, even at very low conversions. 

As shown in Figure 3.5, addition of copper to ruthenium decreases hydrogenolysis activity markedly but has a much smaller effect on dehydrogenation activity. The presence of copper thus improves the selectivity of conversion of cyclohexane to benzene. The data were obtained at 316°C with cyclohexane and hydrogen partial pressures of 0.17 and 0.83 atm, respectively. The ruthenium-copper aggregates were heated to 400°C in hydrogen in their preparation. 

As noted in the previous section, monolayer coverage of ruthenium by copper in ruthenium-copper aggregates with total surface areas in the approximate range of 5-7 m2/g is possible for a copper content in the vicinity of 1.5 at.%, if all the copper is present at the surface. X-ray diffraction patterns on ruthenium-copper aggregates of this type exhibit lines for ruthenium but not for copper. Such observations are consistent with a model of an aggregate in which copper is present as a thin layer or as small clusters at the surface of a crystallite composed essentially of pure ruthenium. 

In ESCA studies of ruthenium-copper aggregates we have used the ratio of the intensity of the Cu 2pin line to that of the Ru 3d5/2 line as a measure of the degree of coverage of ruthenium by copper (14). The kinetic energies... 

It is of interest to relate ESCA data on ruthenium-copper aggregates to data on hydrogen chemisorption on these materials, since changes in the degree of coverage of ruthenium by copper affect both the Cu/Ru intensity ratio and hydrogen chemisorption capacity. Such a relationship is shown in Figure 3.11... 

Pure copper exhibits no strong chemisorption of hydrogen. Data on strong hydrogen chemisorption on ruthenium-copper aggregates have been used as a measure of the amount of ruthenium surface not covered by copper (3,4). As the coverage by copper increases, the amount of strongly chemi-... 

Figure 3.11 Correlation of hydrogen chemisorption capacity of ruthenium-copper aggregates with the ratio /c //Ru of the intensity of electron emission from the copper 2pm level to that from the ruthenium 3dsl2 level (14) (chemisorption data at room temperature). (Reprinted with permission from North-Holland Publishing Company.)...

Recent work conducted in the laboratory of G. Ertl in Munich has extended the investigations on the ruthenium-copper system to include single crystal specimens (18-20). The results of the work are in excellent accord with those obtained in our laboratory on unsupported ruthenium-copper aggregates and on supported ruthenium-copper clusters as well. Our work on supported bimetallic clusters of ruthenium and copper is discussed in detail in the following chapter. 

Figure 4.2 Influence of the state of dispersion of ruthenium-copper catalysts on the relationship between hydrogen chemisorption capacity and catalyst composition (10, 11). (Square points for large ruthenium-copper aggregates represent total hydrogen chemisorption triangular points represent strongly chemisorbed hydrogen, i.e., hydrogen which is not removed from the surface by evacuation of the adsorption cell at room temperature to a pressure of approximately 10 6 torr). (Reprinted with permission from Academic Press, Inc.)...

---