Oxygen Selective Barium Bismuth based Perovskite Membranes

This thesis focused on the research and development of novel dense perovskite membranes for oxygen separation from air. Owing to the mixed ionic-electronic conducting (MIEC) properties of perovskite materials, oxygen ionic transport takes place at high temperatures generally in excess of 600oC, resulting in the production of pure oxygen. Despite the attractiveness of this technology, the oxygen fluxes are still limited at temperatures below 850oC, mainly attributed to low oxygen ionic diffusion. To address this problem, the research community has worked intensively on the compositional tailoring of perovskite materials with ABO3-d structure. These perovskite materials have the capability to incorporate different cation types namely A/A’ and/or B/B’ in their A-site and B-site leading to different A1-xA’xB1-xB’xO3-d compounds with specific structures and properties. Examples include (La,Sr)(Co,Fe)O3-d parent compounds and a large array of doping using various different cations, e.g. Ba, Ca, Sr, Na, La onto A-site of LaCo0.8Fe0.2O3-d; Cu, Ni, Co, Fe, Mn, Cr onto B-site of La0.6Sr0.4CoO3-d; Ba onto A-site of SrCo0.8Fe0.2O3-d; La, Nd, Sm, Gd onto A-site of SrCoO3-d and Ti and Bi onto B-site and A-site of SrFeO3-d. In this thesis, barium was selected as the base material for A-site cation in the perovskite lattices due to its fixed valence cation, Ba2+ and large ionic radius, 1.60Å which favours the creation of large lattice spacing and higher freedom of oxygen ionic movements. To complement barium, bismuth was chosen as a cation dopant since it has ionic radius variability at different valence states (Bi3+ and Bi5+) and coordination numbers which allowed flexible incorporation of bismuth in A-site and/or B-site. Hence, this work takes the advantage of these combined features of barium and bismuth to manipulate the perovskite structure towards favourable oxygen transport properties. This can be achieved through the fixed placement of barium in A-site in conjunction with flexible bismuth position in perovskite lattices. The first postulation of this thesis relates to the bismuth incorporation into perovskite lattice of barium-based perovskite oxides aiming at improving the ionic transport and oxygen fluxes at intermediate temperatures (650-850oC). This postulation is based on the superior ionic transport properties of bismuth. To test the first postulation, barium-bismuth perovskite oxides with different molar ratios of BiO1.5 to BaO (z), where z varied between 0.5-3, were investigated. It was found that barium-rich perovskite oxide of z=0.86, BaBi0.86O2.29 with slight deviation from z=1 showed the optimised oxygen fluxes of 1.2 ml cm-2 min-1. This was attributed to a higher sintering temperature of 1080oC for a compound of z=0.86 instead of 1000oC for a compound of z=1. Nevertheless, due to the insufficient amount of defects for high ionic diffusion, barium-bismuth perovskite oxides did not achieve optimum performance below 850oC. To that end, incorporating complementary metal oxides were deemed necessary to improve the membrane performance. The second postulation in this thesis was that iron addition onto barium-bismuth perovskite oxides provides better structural stability for barium-bismuth-iron perovskite oxides due to iron’s reduction tolerant properties. This postulation was verified by investigating barium-bismuth-iron perovskite oxides within the family of [Ba2−3xBi3x−1][Fe2xBi1−2x]O2+3x/2 with x between 0.17-0.60. It was found that upon increasing x from 0.33 to 0.60, the structure of these compounds changed from cubic to tetragonal and then to hexagonal. Compounds with x=0.33-0.40 exhibited the highest oxygen fluxes attributed to the cubic structure formation. While barium-bismuth-iron perovskite oxides delivered better structural stability with respect to barium-bismuth perovskite oxides, the optimised compound (x=0.33, [Ba][Fe0.67Bi0.33]O2.5) delivered low oxygen fluxes of 0.59 ml cm-2 min-1. Hence, the incorporation of iron decreased the oxygen flux by 50.83% as compared to the best composition of barium-bismuth perovskite oxides. In other words, the incorporation of iron in the barium-bismuth perovskite lattice reduced oxygen ionic transport, though relatively better structural stability was obtained. In addition, structure transition phenomena were found for compounds of x=0.55 and 0.66. The third postulation in this thesis was that bismuth doped barium-scandium-cobalt perovskite oxides enhanced the oxygen fluxes below 850oC. This postulation was envisaged based on the partial substitution of barium or cobalt on a perovskite compound BaSc0.1Co0.9O3-d with bismuth. This work aimed at using bismuth to reduce the ionic radius discrepancy between Ba and Sc/Co and thus ensuring the attainment of cubic structure below 850oC. Cobalt was utilized as the main component in the B-site cation to boost the oxygen permeation properties of the resultant perovskite oxides while scandium was added in small proportions to enhance the crystal structural stability and electrical conductivity as well as to counteract cobalt’s reducibility. A notable point discovered here is that a low amount of bismuth doping, e.g. less than or equal to 10 mole % stabilised the cubic structure. As a result, oxygen fluxes in these compounds increased up to two orders of magnitude between 650-850oC when compared to a non-doped compound. It was further demonstrated that nominal B-site doping was more beneficial to oxygen permeability with respect to nominal A-site doping. In particular, the optimised composition of BaBi0.05Sc0.1Co0.85O3-d delivered very high fluxes of 2.17 ml cm-2 min-1 at 950oC. Structure transition phenomena from non-cubic to cubic structure were also observed for non-doped compound (BaSc0.1Co0.9O3-d) and A-site doped compounds with 20 and 30 mole % of bismuth. A final contribution of this thesis was the development of the hollow fibre membranes with the optimised composition of barium-bismuth-scandium-cobalt perovskite oxide, BaBi0.05Sc0.1Co0.85O3-d, using a combined phase inversion and sintering technique. The hollow fibre membranes showed for the first time the achievement of very high oxygen fluxes e.g. 11.34 ml cm-2 min-1 at 950oC, in excess of target values for air separation units of 10 ml cm-2 min-1. This work further showed the effect of sweep gas flow rate on the oxygen permeances of BBSC hollow fibres, particularly above 800oC. This finding demonstrates that high oxygen fluxes were achieved only at high sweep gas flow rates of 150 ml min-1. For temperatures at or below 800oC, the oxygen flux was independent of the sweep gas flow rate, thus indicating the limitations imposed by oxygen ionic bulk diffusion and surface reaction. Through a compositional tailoring approach, this thesis has demonstrated the potential of barium-bismuth based perovskite oxides in advancing oxygen transport membrane technologies