A CO₂-tolerant nanostructured layer for oxygen transport membranes

Abstract

Dual-layer membranes with enhanced CO₂ tolerance and unprecedented oxygen permeability under CO₂-containing sweep gas are reported. Specifically, a SrFe_0.8Nb_0.2O_3−δ/Ba_0.5Sr_0.5Co_0.8Fe_0.2O_3−δ (SFN/BSCF) dual-layer membrane structure has been successfully prepared by pulsed laser deposition of SFN thin layer onto polished BSCF membranes. The phase structure and microstructure of the SFN/BSCF membrane are characterized by XRD and TEM, respectively. Two distinct phases originated from SFN and BSCF are both obtained, which suggests that the SFN is in high crystallinity under the as-deposited condition and BSCF maintains its original status. TEM images clearly show that SFN nanostructured layer is compactly coating on the BSCF substrate. Oxygen permeation fluxes of 2.721, 2.276, 1.809 and 1.303 mL cm⁻² min⁻¹ at 900, 850, 800 and 750 °C are attained for a ∼45 nm nanostructured SFN layer decorated on a 1 mm thick BSCF membrane using air as the feed and He as the sweep gas. These high oxygen permeation fluxes are comparable with the pristine BSCF membrane since SFN membrane is also a promising mixed conductor and the coated layer is extremely thin. Under He sweep gas with 10% CO₂, a stable oxygen permeation flux of ∼2.25 mL cm⁻² min⁻¹ at 850 °C is maintained for ∼550 min with the SFN/BSCF membrane, while it is only lower than 0.4 mL cm⁻² min⁻¹ with the uncoated membrane. The results indicate that both high oxygen flux and stability can be simultaneously achieved with adoption a nanostructured protective layer.

---

1. Introduction

Perovskite-type mixed ionic-electronic conducting (MIEC) oxygen transport membranes (OTM), which allow the high permeability and infinite selectivity of oxygen in a single step at relatively high temperatures have drawn increasing attention recently for their great potential in energy saving and environmental protection.^1–6 These OTM can be potentially applied in the oxyfuel process, where O₂ instead of air is utilized for the combustion.⁷ Although oxyfuel combustion has not been commercially applied in practical projects due to the extremely high investment needed, as well as the lacking of suitable OTM materials, the oxyfuel process can produce free-of-N₂ flue gas; therefore, it effectively and greatly facilitates the sequestration of CO₂ in coal-fired power stations, which account for the most part of the total global CO₂ emissions. The reduction of CO₂ emissions into the atmosphere may alleviate the global warming because CO₂ is generally thought to be one of the most significant contributors to global warming.⁸ Extensive efforts for developing OTM during past decades have also been made in the membrane reactors,^2,9–13 e.g. the partial oxidation of methane to produce H₂ and CO and complete combustion of methane for CO₂ capture; however, these applications inevitably contain CO₂ in the sweep gas. Therefore, the OTM that is applied in the above-mentioned applications should not only possess good oxygen permeability but also maintain good stability in CO₂-containing atmosphere.

Unfortunately, most of the typical OTM materials developed with the high oxygen permeation flux are susceptible to the presence of CO₂ due to carbonation processes that result from alkaline earth elements (e.g. Ba or/and Sr) in the A site and structural instability caused by the cobalt element in the B site of the perovskites.^14–20 Among OTM, Ba_0.5Sr_0.5Co_0.8Fe_0.2O_3−δ (BSCF) is supposed to be a state-of-the-art membrane material in view of its highest oxygen permeation flux;^21,22 however, BSCF is also extremely chemically unstable in the presence of CO₂ due to carbonate formation, which may lead to membrane failure gradually. Recently, dual-phase membranes, composed of an electronic conductor (e.g. noble metals or spinel phases) and an oxygen ionic conductor (e.g. solid oxide electrolyte), have been developed to overcome aforementioned issues.^23–29 The dual-phase OTMs were reported to possess stable oxygen permeation flux in CO₂ atmosphere; however, the application of these materials is limited owing to the poor oxygen permeability. Even in a recent concept designed by Liu et al. for oxygen separation with an external short circuiting electrode,^30–32 the oxygen permeability is still much lower than those of perovskite MIECs, and the use of the noble metal may also lead to the increased cost of the investment.

Fortunately, with the development of CO₂-tolerant MIEC membranes, some initial experimental attempts have found that a few perovskite MIECs, by proper cation substitution with Nb or other less reducible metal ions, can effectively tolerate CO₂.^33–39 To balance the oxygen permeation flux and stability in the CO₂ atmosphere, Yi et al. proposed a new CO₂-tolerant MIEC that contains Sr²⁺ in the A site and Fe/Nb ions in the B site. Indeed, Nb⁵⁺ has recently been reported as a good stabilized dopant for many MIEC perovskites.^36,40–42 It was found that SrFe_0.8Nb_0.2O_3−δ (SFN) could steadily operate in a pure CO₂ sweep gas for over 210 h at 900 °C and no carbonates were found in the XRD patterns, although the permeability was still relatively low.³⁶

Coated membranes were recently reported to improve activity and/or stabilize the degradation of the MIEC electrodes and membranes.^43–51 For example, a recent report by Serra et al. suggested the adoption of a CO₂ protective layer to overcome the instability issues of membranes in the CO₂ atmosphere.⁴⁶ They deposited a crack-free dense thin layer of Ce_xGd_1−xO_2−δ (GDC) onto the dense La_xSr_1−xCo_yFe_1−yO_3−δ (LSCF) as a CO₂ protective layer by spray pyrolysis technique.⁴⁶ The GDC-protected LSCF membrane exhibited a higher stability than the uncoated LSCF membrane.⁴⁶ The thermal expansion coefficient (TEC) mismatch between GDC and LSCF, as well as the extremely low oxygen permeation flux of the fluorite phase, may drive us to conceive a better assembly according to the present materials development and research progress in the membranes and solid oxide fuel cells.

Here, we adopt SFN as a CO₂ protective layer and BSCF as a monolithic membrane with the aim of obtaining a proper balance in stability and performance. The SFN is a MIEC with CO₂ tolerance, which is coated by pulsed laser deposition (PLD) on the high-oxygen-permeation BSCF membrane. The as-obtained dual-layer membrane possesses both high oxygen permeability and good operational stability under CO₂ atmosphere.

2. Experimental

2.1. Preparation of materials and pellets

The BSCF powders were prepared via a combined EDTA-CA complexing method, which can be found in our previous work.²² SFN powders were synthesized via a standard solid state reaction method. First, raw materials of SrCO₃, Fe₂O₃ and Nb₂O₅ with an analytic grade were mixed thoroughly by ball milling for 60 min at a rotational rate of 400 rpm in an acetone medium. Then, the powder mixture was pressed into pellets and finally calcined at 1280 °C for 20 h. To obtain the desired phase structure and homogeneous grain distributions, the pellets were thoroughly grinded, pressed and sintered at 1300 °C for another 10 h again. The resulted pellets were used as the target for deposition.

2.2. Preparation of dual-layer membranes

BSCF membranes were dry pressed and sintered at 1100 °C for 5 h. The dimensions of membranes were ∼12 mm in diameter and ∼1 mm in thickness. Membranes with the relative densities of higher than 95% were used for the oxygen permeation test. The membranes were mechanically polished and cleaned in acetone and ethanol before pulsed laser deposition. The SFN nanostructured protective layer was fabricated by using a PLD system (Neocera JP-788), which was integrated with a Nd:YAG laser (Continuum Surelite SLIII-10, λ = 266 nm).⁵² Before deposition, the chamber was pumped to a background pressure of ∼9 × 10⁻⁶ mTorr and the substrate was heated to 700 °C. Ablation was performed with a 10 Hz repetition rate under an oxygen partial pressure of 160 mTorr. The energy density per pulse at the target site was ∼5 J cm⁻². A total of 5000 laser shots were employed during deposition. The as-grown SFN protective layers were then annealed in an oxygen pressure of 200 Torr at 700 °C for 30 min. For the characterizations and measurements, the SFN coated disk sample were further annealed at 900 °C for 2 h.

2.3. Basic characterizations

The single phase perovskite structure of the BSCF membrane and SFN target was first confirmed by X-ray diffraction (XRD, PANalytical Empyrean) in the 2θ range between 20° and 80° using Cu Kα radiation at 40 kV and 40 mA (λ = 1.5406 Å). Then, the phase structure of SFN/BSCF was characterized by grazing incidence XRD (PANalytical Empyrean) to obtain the structure information of the coated thin film. Microstructural features of the dual-layer SFN/BSCF membrane were characterized by transmission electron microscope and selected area electron diffraction (TEM/SAED, JEM 2010F). The resistance of BSCF and SFN oxides to CO₂ was characterized by XRD, Fourier transform infrared spectroscopy (FT-IR, AVATAR-360) and CO₂-temperature programmed desorption (TPD) with mass spectrometer (MS, Hiden QIC-20). The electronic conductivity was conducted by the DC four-probe method by using a digital source meter (Keithley 2420). The thermal expansion behavior of the BSCF and SFN was studied by using a dilatometer (NETZSCH DIL 402C). Oxygen bulk diffusion coefficient (D_chem) and the oxygen surface exchange coefficient (k_chem) of the SFN bar sample were calculated from the electrical conductivity relaxation (ECR) curves by ECRTOOLS.⁵³ For details of some applied measurements, please refer to the previously published work.¹⁴

2.4. Oxygen permeation

A high-temperature oxygen permeation test station equipped with a gas chromatograph (GC, Varian CP 3800) was constructed for oxygen permeation evaluation. The SFN/BSCF dual-layer membrane was sealed onto a dense alumina tube with the silver glue for oxygen permeation experiments. The edge of the membrane disk was also sealed with silver glue to avoid the oxygen flux contribution from radical direction and the exposed effective surface area of ∼0.57 cm² was obtained. Oxygen was separated from ambient air (the feeding gas) using high purity He or a mixture of He and CO₂ with the total flow rate of 100 mL min⁻¹ as sweep gas, which were adjusted by mass flow controllers. The separated oxygen was analyzed by using GC equipped with a thermal conductivity detector (TCD) and a capillary molecular sieve 5 Å column. The total oxygen permeation fluxes (mL min⁻¹ cm⁻²) were calculated by as explained elsewhere. Here C_O and C_N are the measured oxygen and nitrogen gas concentrations in the sweep side, respectively. F (mL min⁻¹) is the measured gas low rate of the in the sweep side. S (cm²) is the effective surface area of the membrane. (28/32)^1/2 is related to the Knudsen diffusion coefficient according to the N₂ and O₂ molecular weight.

3. Results and discussion

The phase reactions between two layers may lead to the degradation of the oxygen permeation due to the formation of the insulator layer in the interface. Therefore, two layers without forming the insulator layer is of critical importance for the development of the dual-layer membranes. Moreover, the deposited layer should form the desired phase structure to ensure the protective function of the thin film. The XRD patterns of SFN/BSCF dual-layer membrane are presented in Fig. 1. For comparison, separate XRD patterns of SFN and BSCF are also shown in Fig. 1. Both phase structures can be identified without detecting any impurities, suggesting that the surface of the SFN forms the perovskite structure with fully crystalline through the PLD method. No preferred orientation was observed due to the use of the polycrystalline BSCF membrane as the substrates. The relative intensity of the (100) plane from the BSCF peaks increases while others decrease in the SFN/BSCF membrane. This negligible difference may be due to the slightly preferred orientation of the polished surface of the BSCF since all the substrate for the deposition were mechanically polished. In this study, the highest peak was achieved at the (110) plane similar to the uncoated membrane. On the contrary, a recent report observed that the highest reflections of CGO was at (111) plane. This difference may result from the substrate that they had applied.⁴⁶ In addition, in this study, no obvious changes in 2θ position are observed, and the calculated unit cell parameters of SFN (a = 3.9165 Å) and BSCF (a = 3.9863 Å) are very similar to the separated single phases (a = 3.9144 Å for SFN and a = 3.9855 Å for BSCF), also in good agreement with the values determined previously,³⁶ which is indicative of no element inter-diffusion.

Fig. 1 XRD patterns of BSCF, SFN and SFN/BSCF dual-layer membranes.

To further confirm the high chemical compatibility between SFN and BSCF, the materials in powder form in the weight ratio of 50 : 50 were mixed and calcined at various temperatures in air for 5 h. The mixed powders are effective for the determination of the phase reactions between materials.^54,55 The corresponding XRD patterns of the mixtures after the calcinations are shown in Fig. 2. For comparison, the XRD patterns of BSCF and SFN powders, BSCF and SFN mixture at room temperature (RT) are also presented in Fig. 2. No extra peaks are observed in the XRD patterns for the mixtures calcined at 800 or 900 °C, indicating no apparent chemical reactions between SFN and BSCF powders below 900 °C. The highest temperature used for the preparation and testing in this work is 900 °C; thus, the high chemical compatibility can be maintained in the investigations. Moreover, it should be pointed out that chemical reactions or inter-diffusion between solids are highly dependent on the calcination temperature; thus, the two materials may be not compatible at higher temperature, and determining the compatibility is beyond the scope of our investigation.

Fig. 2 XRD patterns of BSCF, SFN and 50 wt% BSCF + 50 wt% SFN powder mixtures calcined at different temperatures in air for 5 h.

The information based on XRD detection is solely attributed to the phase composition, while the interface situation between membranes from XRD are naturally restricted; therefore the interface between the SFN protective layer and BSCF membrane and phase structure were further investigated with the high-resolution transmission electron microscopy (HRTEM) and selected area electronic diffraction (SAED). From HRTEM, the thermal compatibility between the BSCF membrane and SFN protective layer can also be observed. As shown in Fig. 3, two major parts (I & II, IV & V), separated by the glue (III), can be observed since two slices of the bulk sample are glued together such that the layers face each other. The dual layer is clearly displayed in both I & II and IV & V parts, where a ∼45 nm SFN perovskite is coated on the top of the BSCF membrane with good attachment. The columnar nanostructured features with almost no open pores in the layer and tight interface suggest that the CO₂ adsorption occurs only at the surface of the SFN protective layer. The nanostructured layer is sufficiently dense for this application and is of great benefit to the oxygen reduction reactions. The observation of SAED patterns of the SFN and BSCF layer can be indexed according to a cubic structure, where the lattice plane distances are found to be quite similar to those obtained from XRD patterns. The rings and lattice patterns in SAED are observed for SFN and BSCF layer, respectively. The lattice pattern of BSCF layer that appears may be due to the local observation of the technique since the patterns in XRD of both SFN and BSCF are confirmed to be polycrystalline cubic structures.

Fig. 3 Cross-sectional TEM image of the SFN/BSCF dual-layer membrane (a) and SAED patterns of the corresponding BSCF (b) and SFN (c) layers. The area I, II, III, IV and V denote the BSCF substrate, SFN film, glue, SFN film and BSCF substrate, respectively. The areas IV and V are not artificially separated in order to show the interface, otherwise they would be covered by the dash line.

For such a dual-layer OTM, the mismatch in thermal expansion between components may be a major problem causing degradation and decrease in the lifetime of the membranes due to the delamination and/or cracks. To avoid delamination in this configuration, similar TECs are required. As shown in Fig. 4, linear thermal expansion curves of SFN and BSCF disk samples in the temperature range of 30–900 °C in air are shown, where simultaneous change at ∼500 °C for both SFN and BSCF can be observed. The larger expansions of SFN and BSCF at the temperature higher than 500 °C may be due to the chemical expansion from thermally induced reduction of transition metal ions. The average TECs of SFN and BSCF at selected temperature ranges are calculated and the same tendency is confirmed. For example, the TECs of SFN and BSCF are 13.9 × 10⁻⁶ K⁻¹ and 13.5 × 10⁻⁶ K⁻¹ between 30 and 500 °C, whereas the TECs of SFN and BSCF are 22.1 × 10⁻⁶ K⁻¹ and 21.7 × 10⁻⁶ K⁻¹ between 500 and 900 °C. The results of the thermal expansion study indicate that an excellent match and a good adhesion between two membranes can be realized; thus, SFN may be compatible with BSCF in the dual-layer membrane.

Fig. 4 Thermal expansion curves of BSCF and SFN disk samples.

It is known that the alkaline earth metal ions in the perovskite structure can react with CO₂. Here XRD, CO₂-TPD and FT-IR were applied to check the resistivity of SFN and BSCF to the CO₂. Before subjecting to the characterizations, both SFN and BSCF samples were treated at 850 °C in pure CO₂. As shown in Fig. 5a, pure phase with cubic perovskite structure of the SFN disk sample was preserved, and no impurities or other phases are observed in XRD patterns from the 100 h CO₂-treated SFN disk sample, while carbonate formation was clearly observed and perovskite structure was partially decomposed for the CO₂-treated BSCF disk sample. This indicates that presence of CO₂ could erode the BSCF material strongly by the interaction between CO₂ and alkaline earth metal ions. These results are in good agreement with previous research for SFN and BSCF.^36,56 As also indicated in Fig. 5b, CO₂-TPD profile of the CO₂-treated (850 °C for 10 h) BSCF and SFN shows distinctive behavior. The CO₂ desorption peak of BSCF and SFN powder samples from the decomposition of the carbonates appeared at ∼700 °C since the CO₂-TPD technique was carried out using pure Ar inert gas as the carrier gas with the flow rate of 20 mL min⁻¹. It can be clearly seen that the area of the desorption peak of BSCF is obviously larger than that of the SFN powder sample. The small desorption area of the SFN powder sample indicates poor ability to adsorb CO₂, suggesting good tolerance to the CO₂. This is also in accordance with the FT-IR results (Fig. 5c). Several characteristic peaks (1755, 1063 and 857 cm⁻¹) from the vibration bands for the CO₃²⁻ can be observed from the CO₂-treated (850 °C for 10 h) BSCF, while CO₂-treated (850 °C for 10 h) SFN are completely resistive to CO₂ since no vibration bands of CO₃²⁻ can be observed. Although the mechanism of CO₂ adsorption on perovskite is not clear, it may be concluded that Nb element in the B site can suppress the diffusion of elements in A site,^57,58 or suppress oxygen vacancies in the surface since oxygen vacancies in the perovskite structure are also thought to contribute to the formation of the carbonate.⁵⁹ Thus, these results indicate that SFN can be applied as a protective layer to tolerate CO₂ in the membrane process.

Fig. 5 XRD patterns, CO₂-TPD curves and FT-IR spectra of CO₂-treated SFN and BSCF samples.

The temperature dependence of the oxygen permeation fluxes through SFN membrane, BSCF membrane and SFN-coated BSCF membrane were first evaluated in a CO₂-free pure He sweep gas between 750 and 900 °C to demonstrate the dual-layer configuration. As shown in Fig. 6, relatively high oxygen permeation fluxes of 2.721, 2.276, 1.809 and 1.303 mL cm⁻² min⁻¹ were achieved at 900, 850, 800 and 750 °C for a 1 mm thick SFN/BSCF membrane, which are only slightly lower than those from single phase BSCF membrane (3.118, 2.657, 2.210 and 1.746 mL cm⁻² min⁻¹ at 900, 850, 800 and 750 °C). Although the oxygen permeation fluxes of SFN are higher than those of doped ceria (see, e.g., ref. 32) due to the mixed conductivity of the SFN material, the SFN membrane itself shows relatively low oxygen permeation fluxes of 0.397, 0.272, 0.168 and 0.085 mL cm⁻² min⁻¹ at 900, 850, 800 and 750 °C for a 1 mm thick membrane. Note that the relatively high oxygen permeation flux of SFN-coated BSCF membrane in our configuration can be ascribed to the nanostructured features of the SFN layer.

Fig. 6 Temperature dependence of the oxygen permeation fluxes of BSCF, SFN/BSCF and SFN membranes in the He sweeping gas.

To further interpret the relatively high oxygen permeation fluxes, the oxygen exchange kinetics and electrical conductivity of SFN was investigated. We applied ECR method to derive D_chem and k_chem values of SFN based on an abrupt change of oxygen partial pressure from 0.21 to 0.1 atm in the investigated temperature range (750–900 °C). As shown in Fig. 7, the ECR curves at different temperatures were recorded. ECRTOOLS developed by Ciucci⁵³ was adopted to derive reliable D_chem and k_chem values of SFN bar sample. The obtained values of both D_chem and k_chem were also double checked by sensitivity analysis in order to present a reasonable comparison. Relatively high D_chem values of 8.46 × 10⁻⁵, 2.97 × 10⁻⁵, 2.56 × 10⁻⁵ and 1.74 × 10⁻⁵ cm² s⁻¹, as well as k_chem values of 5.50 × 10⁻⁴, 3.88 × 10⁻⁴, 2.69 × 1 0⁻⁴ and 2.48 × 10⁻⁴ cm s⁻¹, are obtained at 900, 850, 800 and 750 °C, respectively. The obtained values of SFN for both D_chem and k_chem are approximately one order of magnitude lower than those of BSCF in the similar conditions.¹⁴ Fig. 8 shows the temperature dependence of the total conductivity of the SFN in air, O₂, N₂ and CO₂ atmospheres. The total conductivity can be assumed to be electronic conductivity since the ionic conductivity of SFN is much lower than that of electronic conductivity. The conductivities of SFN are exhibiting p-type semi-conductivity behavior and electronic conductivities are in the same order of magnitude to those of BSCF in corresponding atmosphere,⁶⁰ e.g. 22.6 and 16.9 S cm⁻¹ in O₂ and air, respectively, at 850 °C. Most importantly, only a slight reduction of the electronic conductivities of the SFN is observed in CO₂ when compared to those in N₂ atmosphere. Since the SFN layer is ∼45 nm in thickness according to the TEM observation, and both D_chem and k_chem are only one order of magnitude lower, as well as the electronic conductivities of SFN are similar to those of BSCF, the coated SFN layer will not influence the oxygen permeation flux significantly, although the 1 mm SFN membrane shows relatively low oxygen permeation flux. The slight difference of the oxygen permeation flux between BSCF and SFN/BSCF may be mainly due to the relatively low surface exchange coefficients of the SFN materials when compared to the values of BSCF.

Fig. 7 ECR curves of SFN at various temperatures after a sudden change in the oxygen partial pressure from 0.21 to 0.1 atm.

Fig. 8 Temperature dependence of the electronic conductivities of SFN in different atmospheres.

The protective effect of the SFN layer in the CO₂-containing atmosphere was then investigated with a 1 mm SFN membrane first. As can be seen from Fig. 9a, the oxygen permeation fluxes of SFN under the 10% CO₂-containing atmosphere are only slightly lower than those of SFN in the pure He sweep gas, and the flux can be maintained at a stable level during the investigated time range at 850 °C. Furthermore, the oxygen permeation flux can be recovered to the original value when the sweep gas was switched to the pure He. These results suggest that SFN may be a good candidate for membranes operating in CO₂ containing atmosphere, which has also been reported by Yi et al. very recently.³⁶ Since SFN membrane itself can be steadily operated in CO₂ containing atmosphere and the flux is relatively low, the SFN-coated BSCF membrane was further investigated both with He and 10% CO₂ + 90% He as the sweep gas. As shown in Fig. 9b, when switching the sweep gas from He to 10% CO₂ + 90% He mixture, the oxygen flux decreases slightly to ∼2.25 mL cm⁻² min⁻¹, and then remains steady at this permeability in a short-term measurement at 850 °C. This indicates that the SFN layer can enhance the stability of BSCF material under the CO₂ containing atmosphere dramatically with slight influence on the oxygen permeation. In the case of the initial slight reduction of the oxygen flux of SFN, due to the competitive adsorption, when switching the sweep gas to 10% CO₂ + 90% He mixture, the possible reason may be that CO₂ can also be partially adsorbed at the active sites for the oxygen reduction on the surface of SFN perovskite oxides, which blocks the oxygen surface exchange reaction in some exposed surface area. Similarly, the oxygen flux of SFN/BSCF was rapidly retrieved to the original value when the sweep gas was switched back to pure He. Indeed, it has already been observed that oxygen exchange kinetics and oxygen permeation flux of CO₂-treated BSCF material itself can also be easily recovered after exposing to CO₂-free He or air at relatively high temperatures.^14,61 To exclude the regenerative properties of the BSCF membrane itself, the BSCF membrane without SFN protective layer was also investigated under the sweep gas of 10% CO₂ + 90% He mixture for comparison, and severe degradation of the oxygen permeation flux of BSCF material occurred. As shown in Fig. 9c, the oxygen permeation flux of BSCF immediately declines down to as low as ∼0.4 mL cm⁻² min⁻¹ at 850 °C when 10% CO₂ + 90% He mixture is used as sweeping gas, which accounts for less than 20% of the original flux. This also indirectly suggests that the tolerance to the CO₂ could be enhanced by coating a protective layer. As expected, though the oxygen permeation flux of BSCF can also be recovered by switching to the pure He, the flux in CO₂ containing sweeping gas is obviously lower than that of the SFN/BSCF membrane. Based on these results, the protective effect of the nanostructured SFN layer in the SFN/BSCF membrane can be confirmed.

Fig. 9 Oxygen permeation fluxes of SFN (a), SFN/BSCF (b) and BSCF (c) membranes under different sweeping gas flow.

In order to identify the phase stability of the SFN/BSCF membrane after oxygen permeation test, the SFN/BSCF disk sample, which was taken from the previously tested membrane (Fig. 9b), was analysed by XRD. As shown in Fig. 10, the characteristic peaks from both SFN and BSCF samples are preserved after the test, which is quite similar to the fresh SFN/BSCF sample (Fig. 1) before the stability test. It should be pointed out that new diffraction peaks appeared at 38.138°, 44.279°, 64.451° and 77.476° are due to the silver glue used for the membrane sealing.

Fig. 10 XRD pattern of SFN/BSCF membrane after the stability test.

In summary, PLD seems to be a feasible technique to prepare nanostructured layer on top of the membrane with the aim of achieving both high oxygen flux and CO₂ tolerance. SFN-coated BSCF dual-layer membrane exhibits several advantages compared to other state-of-the-art membranes. First, the oxygen permeation flux of the dual-layer membrane is only slightly lower than that of the BSCF membrane. Second, the SFN layer can prevent BSCF from CO₂ poisoning and the steady-state oxygen permeation flux demonstrated in CO₂-containing atmosphere is much higher than that observed in other configurations.^27,32,62

4. Conclusions

Nanostructured SFN has been deposited onto the BSCF membrane surface by using a PLD method in order to prevent the performance degradation of the membrane in CO₂-containing atmosphere and to obtain high oxygen permeability simultaneously. It was revealed by XRD and TEM that perovskite phase structure of the SFN layer was formed, and the interface between the SFN layer and BSCF membrane attached well. The SFN layer coated on the BSCF membrane showed a strong protection to the BSCF membrane in the CO₂-containing atmosphere. The oxygen permeation could be stably maintained at ∼2.25 mL cm⁻² min⁻¹ at 850 °C in 10% CO₂ + 90% He mixture sweeping gas, while BSCF membrane degraded quickly to lower than 0.4 mL cm⁻² min⁻¹ at 850 °C. The relatively high oxygen permeation flux after coating with a SFN layer is attributed to the nanostructured SFN layer prepared by PLD. The CO₂ tolerance and high oxygen permeability of this configuration makes it a promising candidate as OTM for oxyfuel process and membrane reactors, as well as electrodes in solid oxide fuel cells.

Acknowledgements

This work was partially supported by the “National Science Foundation for Distinguished Young Scholars of China” under contract no. 51025209. F.C. gratefully acknowledges HKUST for providing start-up funds and the Research Grants Council of Hong Kong for support through the projects DAG12EG07-12, DAG12EG06, and ECS 639713.

Notes and references

1. K. Zhang, J. Sunarso, Z. Shao, W. Zhou, C. Sun, S. Wang and S. Liu, RSC Adv., 2011, 1, 1661–1676.
2. X. Dong, W. Jin, N. Xu and K. Li, Chem. Commun., 2011, 47, 10886–10902.
3. J. Sunarso, S. Baumann, J. Serra, W. Meulenberg, S. Liu, Y. Lin and J. Diniz da Costa, J. Membr. Sci., 2008, 320, 13–41.
4. H. J. Bouwmeester and A. J. Burggraaf, The CRC handbook of solid state electrochemistry, 1997, p. 481.
5. V. Kharton, A. Yaremchenko, A. Kovalevsky, A. Viskup, E. Naumovich and P. Kerko, J. Membr. Sci., 1999, 163, 307–317.
6. M. Balaguer, C. Solís and J. M. Serra, Chem. Mater., 2011, 23, 2333–2343.
7. E. Pfaff and M. Zwick, Mechanical Properties and Performance of Engineering Ceramics and Composites III, Ceram. Eng. Sci. Proc., 2007, 28, 23–31.
8. D. A. Lashof and D. R. Ahuja, Nature, 1990, 344, 529–531.
9. H. J. Bouwmeester, Catal. Today, 2003, 82, 141–150.
10. A. Thursfield and I. S. Metcalfe, J. Mater. Chem., 2004, 14, 2475–2485.
11. H. Luo, Y. Wei, H. Jiang, W. Yuan, Y. Lv, J. Caro and H. Wang, J. Membr. Sci., 2010, 350, 154–160.
12. W. Yang, H. Wang, X. Zhu and L. Lin, Top. Catal., 2005, 35, 155–167.
13. H. Jiang, H. Wang, F. Liang, S. Werth, T. Schiestel and J. Caro, Angew. Chem., Int. Ed., 2009, 48, 2983–2986.
14. D. Chen and Z. Shao, Int. J. Hydrogen Energy, 2011, 36, 6948–6956.
15. J. Yi, M. Schroeder, T. Weirich and J. Mayer, Chem. Mater., 2010, 22, 6246–6253.
16. X. Tan, N. Liu, B. Meng, J. Sunarso, K. Zhang and S. Liu, J. Membr. Sci., 2012, 389, 216–222.
17. S. Engels, T. Markus, M. Modigell and L. Singheiser, J. Membr. Sci., 2011, 370, 58–69.
18. A. Yan, M. Cheng, Y. Dong, W. Yang, V. Maragou, S. Song and P. Tsiakaras, Appl. Catal., B, 2006, 66, 64–71.
19. Y. Chen, F. Wang, D. Chen, F. Dong, H. J. Park, C. Kwak and Z. Shao, J. Power Sources, 2012, 210, 146–153.
20. M. Arnold, Q. Xu, F. D. Tichelaar and A. Feldhoff, Chem. Mater., 2009, 21, 635–640.
21. Z. Shao, H. Dong, G. Xiong, Y. Cong and W. Yang, J. Membr. Sci., 2001, 183, 181–192.
22. Z. Shao, W. Yang, Y. Cong, H. Dong, J. Tong and G. Xiong, J. Membr. Sci., 2000, 172, 177–188.
23. T. Chen, H. Zhao, Z. Xie, J. Wang, Y. Lu and N. Xu, J. Power Sources, 2013, 223, 289–292.
24. X. Zhu, M. Li, H. Liu, T. Zhang, Y. Cong and W. Yang, J. Membr. Sci., 2012, 394, 120–130.
25. H. Luo, H. Jiang, T. Klande, Z. Cao, F. Liang, H. Wang and J. r. Caro, Chem. Mater., 2012, 24, 2148–2154.
26. X. Zhu, H. Liu, Y. Cong and W. Yang, Chem. Commun., 2012, 48, 251–253.
27. H. Luo, K. Efimov, H. Jiang, A. Feldhoff, H. Wang and J. Caro, Angew. Chem., Int. Ed., 2011, 50, 759–763.
28. Z. Wang, W. Sun, Z. Zhu, T. Liu and W. Liu, ACS Appl. Mater. Interfaces, 2013, 5, 11038–11043.
29. J. Xue, Q. Liao, Y. Wei, Z. Li and H. Wang, J. Membr. Sci., 2013, 443, 124–130.
30. K. Zhang, L. Liu, J. Sunarso, H. Yu, V. Pareek and S. Liu, Energy Fuels, 2013, 28, 349–355.
31. K. Zhang, Y. Zou, C. Su, Z. Shao, L. Liu, S. Wang and S. Liu, J. Membr. Sci., 2013, 427, 168–175.
32. K. Zhang, Z. Shao, C. Li and S. Liu, Energy Environ. Sci., 2012, 5, 5257–5264.
33. Y. Wei, O. Ravkina, T. Klande, H. Wang and A. Feldhoff, J. Membr. Sci., 2013, 429, 147–154.
34. Q. Zeng, Y.-b. Zuo, C.-g. Fan and C.-s. Chen, J. Membr. Sci., 2009, 335, 140–144.
35. Y. Chen, Q. Liao, Y. Wei, Z. Li and H. Wang, Ind. Eng. Chem. Res., 2013, 52, 8571–8578.
36. J. Yi, M. Schroeder and M. Martin, Chem. Mater., 2013, 25, 815–817.
37. L. Bi, S. Zhang, S. Fang, Z. Tao, R. Peng and W. Liu, Electrochem. Commun., 2008, 10, 1598–1601.
38. W. Chen, C. Chen and L. Winnubst, Solid State Ionics, 2011, 196, 30–33.
39. H. Luo, B. Tian, Y. Wei, H. Wang, H. Jiang and J. Caro, AIChE J., 2009, 56, 604–610.
40. F. Wang, T. Nakamura, K. Yashiro, J. Mizusaki and K. Amezawa, Solid State Ionics, 2014 DOI:10.1016/j.ssi.2014.01.045.
41. K. Zhang, R. Ran, L. Ge, Z. Shao, W. Jin and N. Xu, J. Membr. Sci., 2008, 323, 436–443.
42. F. Wang, D. Chen and Z. Shao, Electrochim. Acta, 2013, 103, 23–31.
43. D. Ding, X. Li, S. Y. Lai, K. Gerdes and M. Liu, Energy Environ. Sci., 2014, 7, 552–575.
44. D. Chen, G. Yang, F. Ciucci, M. O. Tadé and Z. Shao, J. Mater. Chem. A, 2014, 2, 1284–1293.
45. M. Balaguer, J. García-Fayos, C. Solís and J. M. Serra, Chem. Mater., 2013, 25, 4986–4993.
46. I. García-Torregrosa, M. P. Lobera, C. Solís, P. Atienzar and J. M. Serra, Adv. Energy Mater., 2011, 1, 618–625.
47. W. Zhou, F. Liang, Z. Shao and Z. Zhu, Sci. Rep., 2012, 2, 327.
48. Y. Liu, X. Zhu, M. Li, H. Liu, Y. Cong and W. Yang, Angew. Chem., Int. Ed., 2013, 52, 3232–3236.
49. Y. Gong, D. Palacio, X. Song, R. L. Patel, X. Liang, X. Zhao, J. B. Goodenough and K. Huang, Nano Lett., 2013, 13, 4340–4345.
50. M. Risch, K. A. Stoerzinger, S. Maruyama, W. T. Hong, I. Takeuchi and Y. Shao-Horn, J. Am. Chem. Soc., 2014, 136, 5229–5232.
51. B. He, K. Zhang, Y. Ling, J. Xu and L. Zhao, J. Membr. Sci., 2014, 464, 55–60.
52. D. Chen, G. Yang, Z. Shao and F. Ciucci, Electrochem. Commun., 2013, 35, 131–134.
53. F. Ciucci, Solid State Ionics, 2013, 239, 28–40.
54. D. Chen, F. Wang and Z. Shao, Int. J. Hydrogen Energy, 2012, 37, 11946–11954.
55. D. Chen, R. Ran, K. Zhang, J. Wang and Z. Shao, J. Power Sources, 2009, 188, 96–105.
56. M. Arnold, H. Wang and A. Feldhoff, J. Membr. Sci., 2007, 293, 44–52.
57. L. Ge, R. Ran, K. Zhang, S. Liu and Z. Shao, J. Membr. Sci., 2008, 318, 182–190.
58. K. Zhang, L. Ge, R. Ran, Z. Shao and S. Liu, Acta Mater., 2008, 56, 4876–4889.
59. K. Nomura, Y. Ujihira, T. Hayakawa and K. Takehira, Appl. Catal., A, 1996, 137, 25–36.
60. Z. Chen, R. Ran, W. Zhou, Z. Shao and S. Liu, Electrochim. Acta, 2007, 52, 7343–7351.
61. M. Arnold, H. Wang and A. Feldhoff, J. Membr. Sci., 2007, 293, 44–52.
62. J. Xue, Q. Liao, Y. Wei, Z. Li and H. Wang, J. Membr. Sci., 2013, 443, 124–130.

---