Oxygen permeation performance of BaBiO3−δ ceramic membranes
Introduction
Dense metal oxide membranes belong to a separate class of inorganic membrane materials that rely on both ionic and electrical conductivity, the so-called mixed ionic–electronic conducting (MIEC) membranes. These membranes have demonstrated significant potential for delivering tonnage oxygen for clean energy delivery, in processes such as coal gasification or oxy-fuel combustion coal power plants [1]. A large number of MIEC compounds have been investigated in the literature, mainly based on perovskites (ABO3−δ) and fluorites (AδB1−δO2−δ and A2δB2−2δO3) or dual-phases by the introduction of metal or ceramic elements. Among these ceramic structures which exhibit MIEC properties, perovskite materials have been intensively studied due to their high (oxygen) ionic and electronic conductivity as well as their favorable phase stability [2], [3].
MIEC compounds can be fabricated into dense ceramic membranes. MIEC membranes require a driving force for the transport of oxygen, which is supplied by the oxygen partial pressure gradient from the feed and permeate side of the membrane. As this is an ionic process, the electrons involved in the electrochemical oxidation and reduction of oxygen ions and oxygen molecules, respectively, are transported in opposite directions, ensuring overall electrical neutrality [2]. The rate of oxygen transport is limited by surface exchange and/or bulk diffusion steps. Surface exchange kinetics, whereby the oxygen molecule is broken down into its ionic form, is controlled by the material and temperature [4], [5], [6]. Bulk-diffusion kinetics are regimented by oxygen vacancies in the material structure, temperature and membrane thickness [3]. When oxygen permeation is bulk-diffusion limited, significant improvement of oxygen flux can be obtained by reducing the thickness of the membrane, while catalytic surface modification enhances oxygen flux when surface exchange is the limiting step [7], [8]. In industrial applications, oxygen transport limitations must be reduced in order to optimize membrane performance.
MIEC membranes are generally derived from perovskite structures, such as ABO3−δ which contains an equivalent number of A-site and B-site cations as well as oxygen vacancies. The ionic conduction occurs by oxygen ions ‘hopping’ from one vacant site to a neighboring vacant site, while the electronic conduction occurs via Bn+–O–B(n+1)+ conduction pairs. In most instances electronic conductivity is more dominant than ionic conductivity [2], [3]. Of the many perovskite formulations reported in the literature, barium bismuth oxides demonstrated high conductivity (electronic and ionic) and high oxygen vacancy concentration. However, most studies were focused on the structure and conductivity of BaBiO3 [9], [10], [11], [12], [13], [14] which contains bismuth ions with different valence states (+3 and +5) [11], [15]. Klinkova et al. [11] developed detailed phase diagrams and elucidated the structure of BaO:BiO1.5 using thermal gravimetric analysis, differential thermal analysis, X-ray diffraction and transmission electron microscopy. They reported that oxygen losses within the perovskite were enhanced with increasing temperature from 600 to 1000 °C through the formation of oxygen deficient compounds of BaBiO2.88, BaBiO2.83, BaBiO2.75 and BaBiO2.55. Takahashi et al. [14] studied the perovskite composition of (BaO)x(Bi2O3)1−x (x = 0.67) and reported that the conductivity was predominantly electronic. Accordingly, barium bismuth perovskite compounds can give enhanced electrical conductivity and oxygen vacancy formation.
Nevertheless, there are very limited studies reported in the literature on the oxygen permeation of barium bismuth oxides. In this work, we investigate the performance of BaBiO3−δ perovskite membranes based on BaO–Bi2O3 formulation. The molar ratio (z) of BiO1.5 to BaO was varied and the structure of the materials was characterized by X-ray diffraction to evaluate the effect of different perovskite phases on oxygen permeation. The morphology of the samples was analyzed by scanning electron microscopy and energy dispersive X-ray spectroscopy. In addition, the perovskite materials were exposed to thermal cycling in air and long term stability testing at an elevated temperature under a nitrogen atmosphere.
Section snippets
Synthesis and preparation
Powders with varying BiO1.5 (as Bi2O3) to BaO molar ratios of z = 0.5–3 were prepared by solid-state reactions at high temperatures. Stoichiometric quantities of Ba(NO3)2 (99.0+%, All-tech) and Bi2O3 (99.9+%, Sigma–Aldrich) were ball-milled with ethanol in a planetary mill (Pulverisette 5, Fritsch) at 500 rpm for 3 h. As-mixed powders were initially calcined at 600 °C for 3 h in order to decompose Ba(NO3)2 into BaO, followed by a further increase in temperature to 700 °C for 9 h. As-calcined powders
Results and discussion
Fig. 2(a) and (b) shows typical SEM images of barium bismuth oxide disk membranes z = 0.86 sintered at 1000 and 1080 °C, respectively. The lower sintering temperature resulted in the formation of porous structure. However, the sintering effect is clearly evidenced when the sintering temperature was raised to 1080 °C, delivering a dense surface with large grains of approximately 50 μm in size. EDX results for two representative samples presented in Table 1 indicated the formation of a homogenous
Conclusions
In this work, perovskite disk membranes were synthesized with the molar ratio (z) of BiO1.5 to BaO between 0.5 and 3 at varying sintering temperatures. All XRD patterns predominantly exhibited peaks characteristic of BaBiO3 perovskite structures, though z = 3 sample showed a rhombohedral phase. Powders prepared with z = 1.17–3 revealed low intensity peaks at 2θ = 40.2 and 50.5° which were associated with Bi-rich perovskite phase and lack of mechanical stability for testing membranes in excess of 800
Acknowledgments
The authors would like to thank the Australian Research Council for financial support on this research project. Jaka Sunarso also acknowledges Ph.D. scholarship (UQIRA and UQILAS) provided by the University of Queensland. The authors would like to thank Dr. Kevin Jack for his assistance in the XRD analysis.
References (24)
- et al.
Development of mixed conducting membranes for clean coal energy delivery
Int. J. Green Gas Cont.
(2009) - et al.
Mixed ionic–electronic conducting (MIEC) ceramic-based membranes for oxygen separation
J. Membr. Sci.
(2008) - et al.
Microstructural development, electrical properties and oxygen permeation of zirconia–palladium composites
Solid State Ionics
(1995) - et al.
Increasing oxygen flux through a dense oxygen permeable membrane by photolithographic patterning of platinum
J. Membr. Sci.
(2006) - et al.
Catalytic effects in oxygen permeation through mixed-conductive LSCF perovskite membranes
Solid State Ionics
(2002) - et al.
Research on the BaBiO3−δ system (0 ≤ δ ≤ 0.5)
J. Solid State Chem.
(1995) - et al.
Bismuth valence order–disorder study in BaBiO3 by powder neutron diffraction
Solid State Commun.
(1988) - et al.
Thermal stability of the perovskite BaBiO3
J. Solid State Chem.
(1999) - et al.
Ba2BiO4 surprisingly found as a cubic double perovskite Ba2(Ba2/3Bi1/3)BiO6−δ
Solid State Commun.
(1990) - et al.
BaBiO2.5, a new bismuth oxide with a layered structure
J. Solid State Chem.
(1991)
Electrical conduction in the sintered oxides of the system Bi2O3–BaO
J. Solid State Chem.
Infrared signature of charge disproportionation in BaBiO3 and related compounds
Solid State Commun.
Cited by (33)
A comprehensive review on oxygen transport membranes: Development history, current status, and future directions
2021, International Journal of Hydrogen EnergyNumerical modeling of heat transfer characteristics in a two-pass oxygen transport reactor for fire tube boilers under oxy-fuel combustion
2021, Applied Thermal EngineeringCitation Excerpt :The critical membrane thickness was found to be about 1.0 mm within the operating temperature range of the membrane between 700-1000 °C [22]. The effect of temperature on oxygen permeation flux has been reported to be directly proportional, thus, oxygen permeation increases when the temperature rises [21,26–29]. The oxygen permeation flux was found to be inconsequential at temperatures lower than 600 °C.
Oxy-fuel combustion in a two-pass oxygen transport reactor for fire tube boiler application
2018, Applied EnergyCitation Excerpt :Temperature has been among the most sensitive factors influencing oxygen permeation flux of membranes. Among the countless literatures that reported increase in O2 permeation due to temperature increase include [18,23–26]. The oxygen permeation flux was insignificant at temperatures below 600 °C.
New high temperature sealing technique and permeability data for hollow fiber BSCF perovskite membranes
2016, Chemical Engineering and Processing: Process IntensificationCitation Excerpt :A commonly adopted procedure is to measure the amount of N2 at the permeate side, where only sweep (often He or Ar) and permeating oxygen should be present [29]. This allows for subtracting the contribution of the molecular oxygen flux caused by leakage, as commonly proposed [30,31]. However, if a set-up configuration such as the one presented in this work is used, the size of the leak detected by this method is intrinsically low, because the permeate side is usually slightly over-pressurized relative to the retentate side – which in general is an outer shell or the open air at 1 bar.
Synthesis of BaBi<inf>0.2</inf>Co<inf>0.35</inf>Fe<inf>0.45</inf>O <inf>3-δ</inf> by a novel aqueous soft chemical method and its characterizations
2014, Journal of Alloys and Compounds