Surface molecular exchange

The situation is completely different for mass transfer within the pore network of monolithic compounds. Here mass transfer can occur both on the pore surface or in the pore volume and molecular exchange between these two states of mobility can occur anywhere within the pore system, being completely uncorrelated with the respective diffusion paths. As a consequence, Eq. (3.1.11) is applicable, without any restrictions, to describing long-range diffusion in the pore space. Equation (3.1.14) is thus obtained,... 

SPAN module. It was mentioned at the beginning that the special polyacrylonitrile fibers of SPAN have a wall thickness of 30 gm, which is considerably thicker than the 8 gm wall thickness of the SMC modules [19]. As a consequence, the presence of stronger capillary effects from the special porous fiber material of the SPAN module would be a reasonable conclusion. Furthermore, the texture of the special polyacrylonitrile fibers is expected to have better surface properties, supporting the permeation of molecules as compared with synthetically modified cellulose. In conclusion, both convection and diffusion effectively contribute to the filtration efficiency in a SPAN module, whereas for the SMC membrane, diffusion is the driving force for molecular exchange, the efficiency of which is also considerable and benefits from the large surface-to-volume ratio. 

The co-existence of at least two modes of ethylene adsorption has been clearly demonstrated in studies of 14C-ethylene adsorption on nickel films [62] and various alumina- and silica-supported metals [53,63—65] at ambient temperature and above. When 14C-ethylene is adsorbed on to alumina-supported palladium, platinum, ruthenium, rhodium, nickel and iridium catalysts [63], it is observed that only a fraction of the initially adsorbed ethylene can be removed by molecular exchange with non-radioactive ethylene, by evacuation or during the subsequent hydrogenation of ethylene—hydrogen mixtures (Fig. 6). While the adsorptive capacity of the catalysts decreases in the order Ni > Rh > Ru > Ir > Pt > Pd, the percentage of the initially adsorbed ethylene retained by the surface which was the same for each of the processes, decreased in the order... 

C-Tracer studies of acetylene adsorption on alumina- and silica-sup-ported palladium [53,65], platinum [66] and rhodium [53] show the coexistence of at least two adsorbed states, one of which is retained on the surface, the other being reactive undergoing molecular exchange and reaction with hydrogen. Acetylene adsorption exhibits the same general characteristics as those observed with ethylene (see Sect. 3.2). However, there are important differences. The extent of adsorption and retention is substantially greater with acetylene than with ethylene. Furthermore, the amounts of acetylene retained by clean and ethylene-precovered sur-... 

If the calculated value of is equal to the measured intracrystalline lifetime, Tinira, the rate of molecular exchange between different crystals is controlled by the intracrystalline self-diffusion as the rate-limiting process. Any increase of Timn, in comparison with Tf,j L indicates the existence of transport resistances different from intracrystalline mass transport. Under the conditions of TD NMR one has A r. > Antra, thus these resistances can only be brought about by sur ce barriers. The ratio Timra/Tfn L represents, therefore, a direct measure of the influence of surface barriers on molecular transport. 

From the NMR tracer desorption and self-diffusion data (second and third lines of Table I), one obtains the relation Timm > TmlL. In the example given, intercrystalline molecular exchange is limited, therefore, by transport resistances at the surface of the individual crystals. Combined NMR and high-resolution electron microscopy studies 54) suggest that such surface barriers are caused by a layer of reduced permeability rather than by a mere deposit of impenetrable material on the crystal surface, although that must not be the case in general. 

The presence of two signals of water adsorbed on the surface of the CSS sample corroborates the hypothesis that there are two kinds of active centers on the surface, and in view of the time scale of NMR the molecular exchange between these centers is a slow process (Section II). Thus the character of active centers responsible for the appearance of these signals is of interest. As the carbon concentration in the sample reaches 35 wt%, it can be assumed that the whole surface of the initial silica is covered with carbon and that both signals correspond... 

Molecular exchange between the crystallites and the intercrystalline space may, however, be controlled by processes other than ordinary diffusion. A substantial retardation of molecular exchange may be caused by transport resistances on the external surface of the crystallites. It has been shown in PFG NMR studies that such surface barriers may be brought about during the process of zeolite manufacturing (e.g. by hydrothermal treatment) [1,6] and by coke depositions [1,7]. In this case, irrespective of possibly large rates of molecular redistribution within the crystallites, the rate of molecular escape out of the crystallites may be slowed down dramatically. In effect, in this case, the product molecules should be distributed essentially homogeneously over the whole space of the individual crystallites. 

Tertiary carbocations may be conveniently prepared in such media from alkyl halides, alcohols, and alkenes. Secondary cations can be observed at low temperatures, but they rearrange readily to more stable tertiary ions. For such cases, special techniques of mixing the reactants, e.g. cocondensation on a cold surface ( molecular beam technique ), have been developed43. Attempts to prepare primary ions in the same manner have not been successful. Methyl and ethyl fluorides exchange halogen but do not generate observable concentrations of cations. All other simple... 

Temp, of treatment (°C) Surface (meter /gm) Chemical composition Phase composition Bate of molecular exchange of oxygen K X 10 (moleerdes/cm sec at 300°C, 10 torr Bate of exchange with oxygen of oxide R X 10 molecules/em sec at 300°C, 10 torr... 

Covering temperatures from -140 up to 200 C and chain lengths from one to six carbon atoms, the intracrystalline mean life times are found to coincide with the values oi rV 1 r calculated from the nmr self-diffusion coefficients. This clearly indicates that molecular exchange is controlled by intracrystalline self-diffusion, and that for the considered adsorbate-adsorbent systems there are no perceptible surface barriers. 

Finally, we consider the situation at the ocean surface. The exchange of gases across the gas-liquid interface is often treated in terms of the thin-film model depicted in Fig. 1-16 (Danckwerts, 1970 Liss and Slater, 1974). The resistances due to turbulent transport in both media are here considered small compared with those in the laminar layers, where the transfer must occur by molecular diffusion. Accordingly, assuming steady-state conditions, the flux through the interface is given by... 

If molecular exchange is controlled by intracrystalline diffusion, then the intracrystalline mean lifetime is given by Eq. (2), where it is assumed that the crystallites may be approximated by spheres (Sec. II.A.). Clearly, coincides with the directly measured Tj ,ra if desorption is controlled by intracrystalline diffusion. If, however, the rate of molecular exchange is additionally reduced by transport resistances at the crystallite boundary (so-called surface barriers), Tji,ra may be much greater than ... 

From Eq. (2), the measured diffusivities may be used to determine the mean lifetime of the reactant and product molecules within the individual crystallites under the assumption that the molecular exchange is exclusively controlled by intracrystalline diffusion. These values, being of the order of 30 ms, are found to agree with the real intracrystalline mean lifetime directly determined by NMR tracer desorption studies (208], so that any influence of crystallite surface barriers may be excluded. From an analysis of the time dependence of the intracrystalline concentration of the reactant and product molecules, the intrinsic reaction time constant is found to be on the order of 10 s. This value is much larger than the intracrystalline mean lifetimes determined by PFG NMR, and thus any limiting influence of mass transfer for the considered reaction may be excluded. In agreement with this conclusion, the size of the applied crystallites was found to have no influence on the conversion rates in measurements with a flow reactor (208]. 

Extending our molecular picture to include kinetic considerations, we can show that the time for a liquid surface to take up its equilibrium value of surface tension is very short. The time may be estimated from the kinetic theory relation t = I ID, where D is the diffusion coefficient and / is approximately equal to the lattice spacing. This time characterizes the time for molecules at the surface to exchange with those in the immediate bulk below. With D 10 m and / 0.3 nm (the diameter of a water molecule), we find t 10 s, which is indeed very small. In practice, the time is somewhat larger than this, but nevertheless extremely small. 

---