Zeolite ZSM-5 membranes grown on porous α-Al2O3
Yushan Yan, Michael Tsapatsis, George R. Gavalas, Mark E. Davis
1995
Journal of the Chemical Society Chemical Communications
ZSM-5 membranes with good permeation selectivity for n-butane-isobutane are grown hydrothermally from a clear solution on horizontally-held porous a-A1203 disks. Zeolites have found widespread applications as catalysts and separation media. A L present, all technologies exploiting these materials use the zeolite in powder composite form. The desire to translate the molecular sieving properties of zeolite powder composites into a membrane configuration has long been recognized. To prepare such a
more »
... membrane, the crystals must grow in an interlockng fashion to form a continuous layer on a porous substrate so that the only available transport pathways are through the zeolite pores. This layer must be mechanically durable but thin to provide good flux. Over the past two years great progress has been achieved towards this goal. Most reported membrane\ were silicalite or ZSM-51-5 [membrane growth on macroporous steel WOO^,^,^ mesoporous (ca. 5 nm) y-A120-3,3,4 and macroporous a-A12035 supports], but a ferrierite membrane has also heen reported6 (growth on macroporous a-A1203). In all cases, the syntheses were accomplished by contacting the support with a synthesis gel, but composition and location or orientation of the support varied among the different preparations. The membranes were tested in permeation of pure gases, separation of gas mixtures, or separation of liquid mixtures. The pure gas permeation ratio of n-butane to isobutane was somctimes used as a yardstick of membrane quality because the kinetic Giameters of these two gases are approximately 4.3 dpd 5 A, respectively, compared to the approximately 5.5 A7 pore size of ZSM-5. Some of the membranes prepared had n-butane : isobutane ratios as high as 60 at room temperature but declined to 1 at 200 0C.1,2 The present study differs from the previous ones1-6 in the use of a clear synthesis solution rather than a gel and in the locationorientation of the support during the synthesis. Perhaps these differences are responsible for the observed different dependence of gas fluxes wli th temperature and particularly the increase rather than decrease of the n-butane : isobutane flux ratio with increasing temperature. Our synthesis eniployed porous disks of 99.8% a-A1203 (Coors Ceramic Conipany, 5 cm diameter, 0.6 cm thickness, 0.5 pm pore size) fixed horizontally in a PTFE holder that was positioned just below the surface of a clear liquid of composition TPAOH4NaOH-0.005A1203-6Si02-57 1 H20 (TPAOH = tetrapropylammonium hydroxide). The reaction system was enclosed in a PTFE I essel, placed into a stainless steel autoclave and heated at 175 "C for 16 h at autogenous pressure. The ZSM-5 film grew on the bottom face of the a-A1203 disk. Numerous Table 1 Pure gas permeation measurements synthetic variations were attempted and it appears that the placement of the disk and the liquid composition are important for the synthesis of good films. At this time we are continuing to explore the relationships between the preparation procedure and the resulting membrane fluxes. Therefore, we present here both our best results and the range of values that we have observed over several membranes (
doi:10.1039/c39950000227
fatcat:5pwrk3nbtvckpmtsrt6glmav2a