SBIR Phase II: Chemically Resistant Gas Separation Perfluoromembranes

Compact Membrane Systems, Inc, New Castle DE

Investigators

Abstract

This Small Business Innovative Research Phase II project will optimize and scale up the system developed in Phase I (a nonporous perfluoromembrane system for harsh gas separations). These nonporous perfluoromembrane systems provide industry for the first time with a system (membrane module, glue lines, potting, valves, etc.) that has good gas transport rates and separation capabilities composed totally of perfluorocomponents. In Phase I, laboratory testing and economic evaluations showed these membranes could economically remove hydrogen, carbon dioxide, and key non-condensable gases from chlor-alkali tail gases and in so doing dramatically enhance the recovery of chlorine. Analysis comparing the Compact Membrane Systems, Inc. (CMS) technology to alternative membrane and other unit operations (e.g. absorption) technologies, indicated the CMS technology is significantly superior. Large sheet nonporous perfluoromembrane fabrication has been demonstrated in Phase I. All the key components are in place for large scale module fabrication in Phase II. In Phase II we will optimize and scale up the system. Detailed and representative (-20oC) end use testing and long term testing will be conducted in the laboratory prior to field testing. While the focus of this program is chlor-alkali harsh chemical separations, other harsh chemical processes (e.g. fluorochemical synthesis) will be considered. Our close working relationship with a number of large membrane manufacturers and end users allows us to rapidly and effectively drive this program. Phase I testing was done using both single gas testing and mixed gas testing. Materials evaluated include chlorine, Cl2CF2, SF6, hydrogen, oxygen, nitrogen, carbon dioxide, and helium. Results showed mixed gas results were equal or superior to single gas results. This suggests that minimal plasticization or other anomalies are occurring within the system. This would suggest we can project actual end use performance accurately.