The polymeric upper bound for N-2/NF3 separation and beyond; ZIF-8 containing mixed matrix membranes

Abstract

Global production for NF3 is continuously increasing, especially due to its heavy consumption in the semiconductor industry. Even though the amount of its emission is relatively small compared to other greenhouse gases, particularly CO2, the relatively long atmospheric lifetime of NF3 makes its emission cumulative, possibly contributing to the global climate change. Membrane-based separation techniques are very promising for the energy-efficient NF3 recovery. It is, therefore, critically important to evaluate the N-2/NF3 separation performance by using commercial polymeric membranes. Here, for the first time, the empirical N-2/NF3 upper bound relationship is established by using a wide variety of commercial polymeric membranes including both glassy and rubbery polymers based on their single gas (i.e. N-2 and NF3) permeation characterization. Among those tested, 6FDA-DAM:DABA (3:2), Teflon (R) AF 2400 and PTMSP exhibited relatively high N-2/NF3 separation performance. The theoretical N-2/NF3 upper bound curve was also defined and found comparable with our empirical upper bound limit. In an effort to improve the N-2/NF3 separation performance, mixed matrix membranes were prepared by incorporating zeolitic imidazolate framework molecular sieves into Matrimid (R) 5218. The effects of solvents, particle sizes, and ligands on the transport properties in mixed matrix membranes were investigated.