Surface diffusion, alumina

All the experimental data in Table 6.1 refer to pure gases. Separation experiments, in which surface diffusion is the separation mechanism, are scarcely reported. Feng and Stewart (1973) and Feng, Kostrov and Stewart (1974) report multicomponent diffusion experiments for the system He-Nj-CH in a y-alumina pellet over a wide range of pressures (1-70 bar), temperatures (300-390 K) and composition gradients. A small contribution of surface diffusion (5% of total flow) to total transport could be detected, although it is not clear, which of the gases exhibits surface difiusion. The data could be fitted with the mass-flux model of Mason, Malinauskas and Evans (1967), extended to include surface diffusion. 

There are several major areas of interfacial phenomena to which infrared spectroscopy has been applied that are not treated extensively in this volume. Most of these areas have established bodies of literature of their own. In many of these areas, the replacement of dispersive spectrometers by FT instruments has resulted in continued improvement in sensitivity, and in the interpretation of phenomena at the molecular level. Among these areas are the characterization of polymer surfaces with ATR (127-129) and diffuse reflectance (130) sampling techniques transmission IR studies of the surfaces of powdered samples with adsorbed gases (131-136) alumina(137.138). silica (139). and catalyst (140) surfaces diffuse reflectance studies of organo- modified mineral and glass fiber surfaces (141-143) metal overlayer enhanced ATR (144) and spectroelectrochemistry (145-149). 

Schaper et al. have recently examined the effect of lanthanum additions on the behaviour of alumina.820 They have shown that lanthanum ions prevent the conversion of 7-Al203 to a-Al203, a process which they believe occurs by surface diffusion. They found also that the stability of the 7-Al203 was improved by the presence of La in steam-containing atmospheres at high pressures and temperatures.826... 

The migration mechanism of V2O5 on the surfaces of alumina and titania is not yet understood in detail. As already mentioned, very recent results reported by Haber et al. [40] seem to indicate that a process of defect diffusion through the vanadia monolayer is involved. 

The kinetics of the chemisorption of spiltover hydrogen species have been fully detailed from a mathematical and physical point of view for carbon as well as for alumina by Robell et al. (17) and Kramer and Andre (18), respectively. Both groups of authors concluded that surface diffusion is the rate-determining step in the overall process of hydrogen spillover. This is shown in Fig. 3 for platinized carbon and Fig. 4 for platinized alumina. 

Although explaining the initial formation of inter-particle bridges, subsequent consolidation of the neck must involve other than Links treated through hydroxyl groups. In view of the fact that the temperatures at which this occurs are low as compared to the melting temperature of alumina surface diffusion as suggested by Frochazka end Coble [22] seems the more probable route for neck consolidation. 

The support plays an important effect in the adsorption kinetics of CO on supported clusters. Indeed CO physisorbed on the support is captured by surface diffusion on the periphery of the metal clusters where it becomes chemisorbed. The role of a precursor state played by CO adsorbed on the support is a rather general phenomenon. It has been observed first on Pd/mica [173] then on Pd/alumina [174,175], on Pd/MgO [176], on Pd/silica [177], and on Rh/alumina [178]. This effect has been theoretically modeled assuming the clusters are distributed on a regular lattice [179] and more recently on a random distribution of clusters [180]. The basic features of this phenomenon are the following. One can define around each cluster a capture zone of width Xg, where is the mean diffusion length of a CO molecule on the support. Each molecule physisorbed in the capture zone will be chemisorbed (via surface diffusion) on the metal cluster. When the temperature decreases, Xg increases, then the capture zone increases to the point where the capture zones overlap. Thus the adsorption rate increases when temperature decreases before the overlap of the capture zones that occurs earlier when the density of clusters increases. Another interesting feature is that the adsorption flux increases when cluster size decreases. It is worth mentioning that this effect (often called reverse spillover) can increase the adsorption rate by a factor of 10. We later see the consequences for catalytic reactions. 

The kinetics of actual adsorption of water on the sites of alumina is very fast. However, a substantial resistance to mass transport can be exhibited by the finite diffusivity of water molecules from the external gas phase to the adsorption sites through the porous network of the adsorbent particle. Diffusion of water vapour (molecular and Knudsen) through the pores of the alumina particle as well as the surface diffusion of adsorbed water on the pore walls [ 11-13] can contribute to the overall transport process. The presence of other non-adsorbing or adsorbing components can significantly influence both pore and surface diffusivity values for water. Table 3 shows a family of water vapour diffusivity data on Rhone-Poulenc grade A alumina in presence of N2 and He as carrier gases at a total gas pressure of 1.0 atmosphere. The water isotherm has a type IV shape [ 9,11]. Pore diffusion... 

The resistance to mass transport for adsorption of water into alumina particles can be governed by diffusion of water molecules through the liquid filled pores as well as by surface diffusion of adsorbed water molecules on the pore walls. A surface excess linear driving force model [SELDF] has been successfully used to describe the adsorption of water from liquid mixtures [27]. For isothermal adsorption of water from a bulk liquid mixture from a constant water composition (xj) batch adsorption system, the uptake profile is given by ... 

The extent to which surface transport affects global rates of reaction has not been established. For it to be important, adsorption must occur, but this is also a requirement for catalytic activity. Indirect evidence suggests that in some cases the effect is considerable. For example. Miller and Kirk found higher rates of dehydration of alcohols on silica-alumina than could be explained with only pore-volume diffusion to account for intraparticle resistances. They attributed the discrepancy to surface diffusion. Masamune and Smith found that surface transport of ethanol on silica gel at temperatures as high as 175°C predominated over gas-phase diffusion in the pore. In view of the data available, it seems wise at least to consider the possibility of surface migration in any evaluation of intraparticle effects. This can be done by adding a surface-diffusion contribution to the effective diffusivity considered in the previous section. The method of doing this is presented below, but its usefulness is still limited because of inadequate experimental and theoretical aspects of surface transport. 

These results are in good agreement with those of Schaper et al. who invoked surface diffusion to account for the sintering mechanism of active alumina, the effect of lanthanum being interpreted as a surface interaction (Ref. 5, 6). Our work confirms that lanthanum aluminate LaA103 is the lanthanum active species, and suggests a model for the thermal stabilization of alumina which takes into account the structural interface between lanthanum aluminate and 5-alumina. 

---