Abstract
Oxygen/nitrogen permselective membranes play particularly important roles in fundamental scientific studies and in a number of applications in industrial chemistry, but have not yet fulfilled their full potential. Organic polymers are the main materials used for such membranes because of the possibility of using sophisticated techniques of precise molecular design and their ready processability for making thin and large self-supporting membranes. However, since the difference in the properties of oxygen and nitrogen gas molecules is quite small, for example, their kinetic diameters are 3.46 Å and 3.64 Å, respectively, the architectures of the membrane macromolecules should be designed precisely. It has been reported often that oxygen permeability (P(O(2))) and oxygen permselectivity (α = P(O(2))/P(N(2))) have trade-off relationships for symmetric membranes made from pure polymers. Some empirical upper bound lines have been reported in (ln α - ln P(O(2))) plots since Robeson reported an upper bound line in 1991 for the first time. The main purpose of this review is to discuss suitable macromolecular structures that produce excellent oxygen/nitrogen permselective membranes. For this purpose, we first searched extensively and intensively for papers which had reported α and P(O(2)) values through symmetric dense membranes from pure polymers. Then, we examined the chemical structures of the polymers showing the top performances in (ln α - ln P(O(2))) plots, using their aged performances. Furthermore, we also explored progress in the molecular design in this field by comparing the best polymers reported by 2013 and those subsequently found up to now (2020) because of the rapid outstanding growth in this period. Finally, we discussed how to improve α and P(O(2)) simultaneously on the basis of reported results using not only symmetric membranes of pure organic polymers but also composite asymmetric membranes containing various additives.