CO2 SEPARATIONS USING ZEOLITE MEMBRANES

Abstract

Zeolite and other inorganic molecular sieve membranes have shown potential for separations based on molecular size and shape because of their small pore sized, typically less than 1 nm, and their narrow pore size distribution. The high thermal and chemical stability of these inorganic crystals make them ideal materials for use in high temperature applications such as catalytic membrane reactors. Most of the progress with zeolite membranes has been with MFI zeolites prepared on porous disks and tubes. The MFI zeolite is a medium pore size structure having nearly circular pores with diameters between .53 and .56 nm. Separation experiments through MFI membranes indicate that competitive adsorption separates light gas mixtures. Light gas selectivities are typically small, however, owing to small differences in adsorption strengths and their small sizes relative to the MFI pore opening. Furthermore, competitive adsorption does not work well at high temperature where zeolite membranes are stable and have potential application. Separation by differences in size has a greater potential to work at high temperature than competitive adsorption, but pores smaller than those in MFI zeolites are required. Therefore, some studies focused on the synthesis of a small, 8-membered-pore structures such as zeolite A (0.41-nm pore diameter)more » and SAPO-34, a chabazite (about .4-nm pore diameter with about 1.4 nm cages) analog. The small pore size of the zeolite A and SAPO-34 structures made the separation of smaller molecules by differences in size possible. Zeolite MFI and SAPO-34 membranes were prepared on the inside surface of porous alumina tubes by hydrothermal synthesis, and single gas and binary mixture permeances were measured to characterize the membrane's performance. A mathematical diffusion model was developed to determine the relative quantities of zeolite and non-zeolite pores in different membranes by modeling the permeation date of CO{sub 2}. This model expresses the total flux through the membrane as the sum of surface diffusion through zeolite pores and viscous flow and Knudsen diffusion through non-zeolite pores. As predicted by the model, the permeance of CO{sub 2} decrease with increasing pressure at constant pressure drop for membranes with few non-zeolite pores, but the permeance increased through viscous flow pores and was constant through pores allowing Knudsen diffusion. Membranes having more non-zeolite pores had lower CO{sub 2}/CH{sub 4} selectivities. The SAPO-34 membranes were characterized for light gas separation applications, and the separation mechanisms were identified. Single gas permeances of CO{sub 2}, N{sub 2}, and CH{sub 4} decreased with increasing kinetic diameter. For the best membrane at 300K, the He and H{sub 2} permeances were less than that of CO{sub 2}, because He, H{sub 2} and CO{sub 2} were small compared to the SAPO-34 pore, and differences in the heat of adsorption determined the permeance order. The small component permeated the fastest in CO{sub 2}/CH{sub 4}, CO{sub 2}/N{sub 2}, N{sub 2}/CH{sub 4}, H{sub 2}/CH{sub 4}, and H{sub 2}/N{sub 2} mixtures between 300 and 470 K. For H{sub 2}/CO{sub 2} mixtures, which were separated by competitive adsorption at room temperature, the larger component permeated faster below 400K. The room temperature CO{sub 2}/CH{sub 4} selectivity was 36 and decreased with temperature. The H{sub 2}/CH{sub 4} mixture selectivity was 8 and constant with temperature up to 480 K. Calcination, slow temperature cycles, and exposure to water vapor had no permanent effect on membrane performance, but temperature changes of approximately 30 K/min decreased the membrane's effectiveness. The effects of humidity on gas permeation were studied with SAPO-34 membranes of different qualities. Membranes with high CO{sub 2}/CH{sub 4} selectivities (greater than 20) were stable in water vapor under controlled conditions, but degradation was seen for some membranes. The degradation opened non-SAPO-34 pores that were larger than SAPO-34 pores as shown by the IC{sub 4}H{sub 10} permeance, CO{sub 2}/CH{sub 4} selectivity, and CO{sub 2} flux pressure dependence. In SAPO-34 pores, water apparently adsorbed and increased the gas permeances. Thus, the effect of water on gas permeation is a useful indicator of the membrane quality.« less