Oxyfuel combustion involves the combustion of carbon-based fuels in a pure O2 stream, however, the limiting factor in the industrial implementation of these methods is the large amount of pure O2 that is required to be generated from air (O2/N2 separation). Microporous solids that are able to efficiently perform this separation have the potential to significantly reduce the large energy-costs currently associated with oxyfuel combustion. Small-pore zeolites have been employed for O2/N2 separations by exploitation of the difference in the kinetic diameter between the two gases through physical separation involving molecular sieving. The chemical tunability of the pore space of framework materials, however, facilitates the separation of O2 and N2 by taking advantage of the electronic differences between the two gases. In particular, MOFs containing electron-rich redox-active sites, such as Cr3(btc)2 [77] and Fe2(dobdc) [78], have been shown to reversibly bind O2 selectively over N2 via electron transfer from the metal centre to the O2.

The Fe2(dobdc) material binds O2 preferentially over N2 at 298 K with an irreversible capacity of 9.3 wt%, corresponding to the adsorption of one O2 per two Fe centres [78]. Remarkably, at 211 K the O2 uptake is fully reversible and the capacity increases to 18.2 wt%, corresponding to the adsorption of one O2 per Fe centre. Mossbauer and infrared spectroscopy measurements indicated partial charge-transfer from the FeII to the O2 at low temperature and complete charge — transfer to form FeIn and O22- at room temperature. NPD data (4 K) confirm this interpretation, revealing O2 bound to Fe in a symmetric side-on mode with an O2 intranuclear separation of 1.25(1) A at low temperature and of 1.6(1) A in a slipped side-on mode when oxidized at room temperature (Fig. 3.7).

Similar work reported highly selective and reversible O2 binding in Cr3(btc)2 [77], with infrared and X-ray absorption spectra suggesting the formation of an O2 adduct with partial charge-transfer from the CrII centres exposed on the surface of the framework. NPD data confirm this mechanism of O2 binding and indicate a lengthening of the Cr-Cr distance within the “paddle-wheel” units of the frame­work from 2.06(2) to 2.8(1) A.

Selectivity for O2 over N2 was also achieved in polymer/selective-flake nano­composite membranes fabricated with a polyimide and a porous layered alumino — phosphate. Using SANS to probe the large-scale structure of the O2/N2 host material, the substantially improved selectivities of O2 over N2 was shown to occur within only 10 wt% of the AlPO layers [79].