Return to search

Reverse-selective zeolite/polymer nanocomposite hollow fiber membranes for pervaporative biofuel/water separation

Pervaporation with a "reverse-selective" (hydrophobic) membrane is a promising technology for the energy-efficient separation of alcohols from dilute alcohol-water streams, such as those formed in the production of biofuels. Pervaporation depends on the selectivity and throughput of the membrane, which in turn is highly dependent on the membrane material. A nanocomposite approach to membrane design is desirable in order to combine the advantages and eliminate the individual limitations of previously-reported polymeric and zeolitic membranes. In this work, a hollow-fiber membrane composed of a thin layer of polymer/zeolite nanocomposite material on a porous polymeric hollow fiber support is developed. The hollow fiber geometry offers considerable advantages in membrane surface area per unit volume, allowing for easier scaling and higher throughput than flat-film membranes.

Poly(dimethyl siloxane) (PDMS) and pure-silica MFI zeolite (silicalite-1) were investigated for these membranes. Iso-octane was used to dilute the dope solution to provide thinner coatings. Previously-spun non-selective Torlon hollow fibers were used as the support layer for the nanocomposite coatings. To determine an acceptable method for coating fibers with uniform, defect-free coatings, flat-film membranes (0 to 60 wt% MFI on a solvent-free basis) and hollow-fiber membranes (0 and 20 wt% MFI) were fabricated using different procedures. Pervaporation experiments were run for all membranes at 65C with a 5 wt% ethanol feed. The effects of membrane thickness, fiber pretreatment, coating method, zeolite loading, and zeolite surface treatment on membrane pervaporation performance were investigated.

Identiferoai:union.ndltd.org:GATECH/oai:smartech.gatech.edu:1853/39538
Date08 April 2010
CreatorsMcFadden, Kathrine D.
PublisherGeorgia Institute of Technology
Source SetsGeorgia Tech Electronic Thesis and Dissertation Archive
Detected LanguageEnglish
TypeThesis

Page generated in 0.0019 seconds