ausblenden:
Schlagwörter:
-
Zusammenfassung:
The presented work is focused on quantitative identification of gas permeation in inorganic, micro-porous glass membranes. The pore size diameter of the membranes was in the range between 1.5±5nm. The complex mechanism of gas transport in these membranes is still not well understood, however, Knudsen diffusion, surface diffusion and configurational diffusion can be considered as rate controlling mechanisms.
Two different versions of experiments were used for measuring gas permeabilities through the membranes: (a) dynamic and (b) steady state ones.
(a) The round, flat membrane was located between the upper and lower gas chamber. The gas stream passed the upper chamber with constant flow rate at atmospheric pressure. The lower chamber initially was at approximately zero pressure and as gas penetrated through the membrane, the pressure was increasing and was recorded with a sensitive pressure gauge. The temperature had been varied from 295K to 475K.
(b) The upper chamber was fed with gas at constant pressure slightly above atmospheric pressure, while from the lower chamber gas was pumped off with a strong turbo-molecular pump connected to a mass spectrometer. The steady state flux was established very quickly. Conditions allowed measurements at temperature from 295 to 413K.
The dynamic and steady state experiments gave similar values of the effective permeabilities. Permeabilities of He, Ar, N2, CO2 and C3H8 have been measured. In order to describe the observations, anticipating ideal Knudsen diffusion was a first attempt. But it turned out that the separation factors between the various gases were much bigger than expected. Based on the observed permeability data, ideal selectivity coefficients were calculated for each of gas in comparison with He. Comparing the selectivities from Fig. 1 one may see that the selectivity factors are much higher than those for Knudsen separation at lower temperature and with increasing the temperature, factors are decreasing to Knudsen ratio.
Consequently, activated diffusion [1] had also to be considered. For membranes of small pore sizes, so-called surface flow due to activated diffusion can be another additional mechanism, because of strong interactions between gas molecules and the pore walls. Therefore, also the kinetic diameter of the gas molecules became an additional parameter. Even inert gases can be adsorbed on these membranes [2]. This statement was confirmed by experimental results showing that permeabilities at higher temperatures become Knudsen-diffusion controlled, while at lower temperatures strong deviations from Knudsen-behavior were observed. It seems that gases have been really adsorbed at the surface. These data could be helpful to clarify the mechanisms by which gases may be separated by micro-porous glass membranes.
To describe these observations the concept of potential barrier between the gas molecules and the solid surface was applied. That means if the molecule after collision has a kinetic energy bigger than the surface potential energy, it passes the potential surface field, and such a flow is called gas-phase flow. When molecules can not pass through the force field they again collide with the surface. The flow of these molecules is called surface flow [3].
To eliminate surface flow for extracting the tortuosity factor free from effects of surface diffusion and adsorption, experiments were performed at higher temperatures. Based on the differences between permeabilities of different gases, tortuosity and potential barrier can be fitted. At lower temperatures the adsorption and surface diffusion can be obtained as the difference between total flux and gas flux [4].
Both versions of experiments give us a lot of information aiming at to quantify mass transfer processes in micro-porous media, better understanding of surface flow in nano-separation facilities and to predict the separation effects to be expected.
Further work is currently focused on studying membranes with different pore sizes in the nanometer range or different type and magnitude of surface area in order to distinguish various adsorption effects. These micro-porous membranes might have a greater potential for gas separation compared to membranes with larger pores.
[1] SHELEKIN A. B., DIXON A. G., MA Y. H., AIChE J., 41, 58 (1995)
[2] BHANDARKAR M., SHELEKIN A. B., DIXON A. G., MA Y. H., J. Membrane Sci., 75, 221 (1992)
[3] HWANG S. T. AND KAMMERMEYER K., Ind. Eng. Chem., Fundam., 7, 671 (1968)
[4] DO D. D., PARK I. S., RODRIGUES A., Catal .rev.- Sci. Eng., 38(2), 189 (1996)