Alloy films phase separation

Fig. 29. CO oxidation at 240°C over Pd-Rh alloy films as a function of apparent surface composition homogeneous alloys and pure metals ( ) alloys showing phase separation (3) (34).

This conclusion was additionally confirmed by Palczewska and Janko (67) in separate experiments, where under the same conditions nickel-copper alloy films rich in nickel (and nickel films as well) were transformed into their respective hydride phases, which were proved by X-ray diffraction. The additional argument in favor of the transformation of the metal film into hydride in the side-arm of the Smith-Linnett apparatus consists of the observed increase of the roughness factor ( 70%) of the film and the decrease of its crystallite size ( 30%) after coming back from low to high temperatures for desorbing hydrogen. The effect is quite similar to that observed by Scholten and Konvalinka (9) for their palladium catalyst samples undergoing the (a — j8) -phase transformation. 

The overall compositions of films 2 and 4 was such that homogeneous alloys were not expected, and the X-ray diffraction peaks showed evidence of phase-separation with two maxima corresponding to the compositions recorded in column 8, Table II. When palladium was deposited on top (and the film was rather light) then the apparent surface composition after annealing was 10% Rh and X-ray diffraction indicated a phase also containing 10% Rh (film 2). This convenient result was not observed when the order of deposition was reversed (film 4). These differences between apparent surface composition and the overall composition of the homogeneous alloy (or one of the phases in the miscibility gap) are discussed in Moss and Gibbens (34), with further examples. The main point to be made here is the rather variable nature of the surface composition compared with that expected, due to the operation of a number of factors. 

Composition range 30-80% Rh. In this composition range phase separation occurs, and the structure of such Pd-Rh alloy films has been reviewed (Section II). Phase I varied in composition and phase II contained 88 5% Rh. It was proposed that these results could be explained by the preferential nucleation of rhodium so that the crystallites consisted of a phase II kernel surrounded by an outer shell (phase I), the Rh content of which increased with an overall increase in the Rh content of the alloy film. Note the essential difference to the Cu-Ni films (38, 33) discussed in Section IV.A where complete separation into two phases of fixed equilibrium composition is envisaged, and over a wide composition range the crystallite surfaces have the same composition. 

There is now available a substantial amount of information on the principles and techniques involved in preparing evaporated alloy films suitable for adsorption or catalytic work, although some preparative methods, e.g., vapor quenching, used in other research fields have not yet been adopted. Alloy films have been characterized with respect to bulk properties, e.g., uniformity of composition, phase separation, crystallite orientation, and surface areas have been measured. Direct quantitative measurements of surface composition have not been made on alloy films prepared for catalytic studies, but techniques, e.g., Auger electron spectroscopy, are available. 

The subject of Adsorption and Catalysis on Evaporated Alloy Films is reviewed and Moss and Whalley conclude that phase separation caused a variety of complications which makes it difficult to define the nature of catalytic activity. 

Burton et al. [26] also studied thin films of various thickness for alloys in which there is a miscibility gap (12 > 0). Figure 3.8 shows the results of calculations for a 50 atom% Au-Ni alloy. The Au-Ni bulk phase diagram has a miscibility gap below r,. = 1000 K. For T > T,. the segregation of gold takes place only in the surface region, with a core that approaches the bulk composition as the film thickness increases. For T < T,. the films exhibit phase separation, with the Au-rich phase accumulating at the surface and the Ni-rich phase accumulating at the center. 

Moreover, thin (less than 2 pm thickness) and pinhole-free Pd-Cu alloy composite membranes with a diffusion barrier have been fabricated on mesoporous stainless steel supports (MSSS) by vacuum electro-deposition (Nam and Lee, 2001). The deposition film was fabricated by multilayer coating and diffusion treatment and the formation of Pd-Cu alloys was achieved by annealing the as-deposited membranes at 450°C in nitrogen atmosphere. To improve the structural stability of Pd alloy/Ni-MSSS composite membranes, a thin intermediate layer of silica by sol-gel method was introduced as a diffusion barrier between Pd-Cu active layer and a modified MSSS substrate. The composition and phase structures of the alloy film were studied by energy dispersive analysis (EDS) and XRD the typical Pd-Cu plating had a composition of 63% Pd and 37%Cu and the atomic inter-diffusion of Pd and Cu resulted in Pd-Cu alloys in an fee structure. The electron probe microanalyser (EPMA) profiling analysis indicated that the improved membranes were structurally stable. The Pd-Cu alloy composite membrane obtained in this study yielded excellent separation performance with H2 permeance of 2.5 x 10 cm /(cm cmHg s) and Hj/Nj selectivity above 70000 at 450°C. 

Dual sputtering deposition technique was used to prepare submicron thin Pd-Cu alloy films, which allowed a high composition control of the layer (Hoang et al, 2004). The composition, surface morphology and phase structure of the sputtered layers were investigated by EDS, X-ray, XPS, SEM, TEM and XRD. For example, the XRD data proved that the Pd-Cu layers were an alloy of Pd and Cu. Subsequently, the characterized Pd-Cu alloy layers were deposited on a silicon support structure to create a 750 nm thin Pd-Cu membrane for H2 separation. The reported membrane obtained a high H2 flux of 1.6 moF(m s) at a temperature of452°C, while the selectivity was at least 500 for H2/He. 

Figure 11.35 Shows the x-ray diffraction spectra for an as-deposited nanograined Cu-Mo alloy film and the separation into Cu and Mo phases resulting from a 60 minute anneal at 600°C. Note the peak shift visible in the 211 peak for the as-deposited film compared to the alloy film. The 110 peak also shifted but the scale of the shift is too small to be visible on these axes. Note also the absence of either the Cu 111 or the Cu 200 peaks in the as-deposited film. The combination of shifted Mo bcc peaks and the absence of Cu peaks shows that a single phase alloy was created. The texture of the film can be observed in the much greater size of the 211 peak in the as deposited film compared to the normally-stronger 110 peak. After annealing the behavior is reversed. [29,30]...

---