Search

Permeation of binary gaseous mixtures through a silicalite-1 membrane was studied. Results are reported for the fluxes and separation factors as a function of the feed composition, the pressure and the temperature. For the various cases a comparison is made between the binary and the single-component permeation behavior. For the permeation of the mixtures the stronger adsorbed component is little affected by the presence of the weaker adsorbed component. On the other hand, for the weaker adsorbed component a clear reduction in the flux is observed. As a consequence, the separation factor for mixtures differs considerably from the so-called ideal separation factor, which is the ratio of the one-component permeances. As expected, the separation factor depends on the temperature. The separation factor is also a function of the composition and the feed pressure. This means that the binary equilibrium adsorption cannot be described with the extended Langmuir model. Binary permeation and separation behavior is described using the ideal adsorption solution theory for equilibrium adsorption.

The one-component steady-state permeation of gases through a silicalite-1 zeolite composite membrane as a function of the temperature is studied from 190 to 680 K for light hydrocarbons, noble gases, and some inorganic gases. In general, with increasing temperature the permeance shows a maximum followed by a minimum. For gases weakly adsorbed the permeance has only a minimum and for gases strongly adsorbed only a maximum is observed in the permeance. The permeance for various gases, for a feed pressure of 101 kPa, span four orders of magnitude. The lowest permeation is for i-butane at 300 K: a permeance of 0.07 × 10−8 mol. m−2.s−1.Pa−1. The highest value is observed for methane: a permeance of 70 × 10−8 mol. m−2.s−1.Pa−1 at about 240 K. A comparison between the isobars and the temperature dependence of the steady-state permeance, both at 101 kPa, shows that at the temperature where the amount adsorbed vanishes the permeance starts to increase. The temperature dependence of the steady-state fluxes through the silicalite-1 membrane can be described only if two diffusion mechanisms are taken into account. For high occupancies the mass transport can be described by equilibrium adsorption followed by surface diffusion and for low occupancies the mass transport can be described by activated gaseous diffusion. With increasing temperature the mass-transport mechanism shifts from the surface diffusion regime to the activated gaseous diffusion regime. With these two diffusivities modeling results agree well with experimental results for the one-component flux through the silicalite-1 zeolite membrane.

Using Maxwell-Stefan equations, experimental and computational results of binary diffusion in pore- and cage-type zeolitic structures are described. In the generalized Maxwell-Stefan (GMS) formulation, the Fick diffusivity is written as the product of two separate contributions, the GMS or corrected diffusivity and the thermodynamic factor. The concentration dependence of the GMS diffusivity for one- and two-component diffusion in zeolitic structures is investigated. From the Maxwell - Stefan equations, different models for the Fick diffusion coefficient matrix for the description of binary mass transport in molecular sieve materials are derived. Various models used predict binary diffusion in zeolitic structures. First, theoretical predictions of binary apparent diffusivities as a function of the occupancy are compared to results from Monte Carlo simulations. Second, theoretical results of binary uptake profiles are compared to experimental results for the system ethylbenzene/benzene/ZSM-5. For different zeolitic structures, that is, pore- and cage-type structures, results of the Monte Carlo simulations agree well with the theoretical predictions. In cage-type structures, the effect of counterexchange between sorbed molecules is demonstrated. Experimental results of transient uptake profiles of a mixture of benzene and ethylbenzene in ZSM-5 follow predictions of the theoretical single-file diffusion model.