(609g) Formation of Hydrogen-Selective Carbon Composite Membranes By Incomplete Stabilization of Polyacrylonitrile

Carbon molecular sieve (CMS) membranes offer attractive separation performances and thermal resistance. Nevertheless, their fabrication is typically overly expensive, partly because of the cost of preferred polymeric precursors, like polyimides. And, affordable precursors, such as cellulose derivatives, come with film formation issues. At this point, being the most utilized precursor for carbon fiber production, polyacrylonitrile (PAN) is an appealing option for CMS development. But interestingly, not much has been done for utilizing PAN in CMS membrane fabrication, except for a few early studies and some other failed attempts. Therefore, we tested the utility of PAN for fabricating CMS membranes capable of separating hydrogen (H₂) and carbon dioxide (CO₂). As the membrane substrate, we have exploited gamma-alumina-coated alpha-alumina discs. To obtain uniform membranes with polymer layers of around 2 to 4 micrometers with minimal support infiltration, we used spin coating and a low-boiling-point solvent (hexafluoroisopropanol). And to prevent pinhole formation, we applied dual-layer coating. After achieving defect-free thin-film composite membranes, we optimized the heating protocol based on differential scanning calorimetry, thermogravimetric analysis, and X-ray diffraction studies. We found that the thermal stabilization (cyclization) step should be made prematurely to achieve attractive gas selectivities and permeabilities (exceeding Robeson's 2008 upper bound). Overall, as long as the filmmaking and stabilization steps are optimized, PAN is suitable for developing H₂-selective CMS membranes.