Article pubs.acs.org/IECR
Asymmetric Thin Samarium Doped Cerium Oxide−Carbonate DualPhase Membrane for Carbon Dioxide Separation Bo Lu and Y. S. Lin* Chemical Engineering, School for Engineering of Matter, Transport and Energy, Arizona State University, Tempe, Arizona 85287, United States ABSTRACT: Samarium doped ceria (SDC)−carbonate dual-phase membranes are permselective to carbon dioxide. This paper reports a method to prepare thin SDC−carbonate membranes on an adequate base support to improve CO2 permeance of the membranes. It was found that macroporous support made of a physical mixture of SDC and 40 vol % bismuth−yttrium− samarium oxide (BYS) has the desired pore structure, ionic conductivity, carbonate nonwettability, and mechanical compatibility with the thin SDC−carbonate membrane layer. Asymmetric porous supports consisting of a thin, porous SDC top layer on an SDC−BYS base were prepared by the copressing method. The porous SDC top layer was filled with molten carbonate by the direct infiltration method. The final membranes consist of a 150 μm, hermetic SDC−carbonate layer on the macroporos SDC− BYS base support. The thin SDC−carbonate dual-phase membrane offers significantly improved CO2 permeance as compared to thick SDC−carbonate membranes. Both membrane thickness and the structure of the SDC phase affect CO2 permeance. The thin SDC−carbonate membranes exhibit CO2 flux of 1.33 × 10−3−6.55 × 10−3 mol/s/m2 at 550−700 °C, with steady state operation for at least 160 h. The CO2 permeation flux is related to upstream and downstream CO2 partial pressures by a power law, consistent with the theoretical model. supports have been prepared by several research groups,22−26 and their CO2 permeation properties are summarized in Table 1. Table 1 shows that all these ceramic−carbonate membranes are CO2 permselective and CO2 permeances increase with increasing oxygen ionic conductivity of the ceramic phase. In addition, the pore structure of both the carbonate phase and the ceramic phase, as characterized by the porosity to tortuosity ratio, has a strong effect on CO2 transport properties.27 Some of the ceramic−carbonates, such as La0.6Sr0.4Co0.8Fe0.2O3−δ− carbonate membrane, are not stable when operated in streams containing CO 2 . 2 6 Samarium doped ceria (SDC, Ce0.8Sm0.2O1.9)−carbonate membrane offers both high CO2 permeance24,26 and stability25 and therefore is the most attractive dual-phase membrane for practical applications. Based on the transport models describing CO2 permeation through the ceramic−carbonate dual-phase membranes,27−29 oxygen ionic transport is a controlling factor and reduction of membrane thickness would effectively improve membrane performance. Lu and Lin30 reported the concept of making a thin ceramic−carbonate dual-phase membrane to increase the CO2 flux. The membrane was prepared on a two-layer asymmetric support consisting of a thick, larger pore base and a thin, smaller pore oxygen ionic conducting or mixed conducting ceramic top layer. The base support has to be molten carbonate nonwettable to prevent carbonate infiltration into the base support and ensures formation of a thin, hermetic dual-phase membrane top layer on the porous support. The
1. INTRODUCTION High temperature CO2 permselective membranes offer potential for uses in various processes for CO2 separation, including CO2 capture for fossil fuel burning power plants.1,2 Progress has been made on making membranes specifically designed for CO2/N2 separation using polymers,3−8 zeolites,9−12 carbon,13,14 and silica.15,16 These membranes take advantage of molecular sieving properties and have been found to be particularly promising in CO2 separation at low temperature. However, their CO2 permeances and selectivities seriously descend when they are operated at moderately high temperature (>300 °C). For some applications, such as fuel gasification for hydrogen production which is operated in the high temperature regime (400−1100 °C)17 or separation in acid flue gas at high temperature (>350 °C), an efficient high temperature CO2 permselective membrane (>400 °C) is highly desirable. A new concept of a dual-phase (metal−carbonate) membrane permselective to CO2 at high temperature was first reported by Lin and co-workers.18 The membranes are CO2 permselective at high temperature due to the transport of carbonate ion (and oxygen ion) in the liquid carbonate phase and electron in the metal phase. The metal−carbonate dualphase membranes were synthesized with stainless steel18 and silver19 as the metal phase. However, these metal−carbonate dual-phase membranes either suffer a stability problem due to reaction between the metal phase and molten carbonates at high temperature18 or require removal of oxygen from the permeate products.18,19 To improve the membrane stability and allow membrane permeation CO2 without O2, Lin and coworkers20,21 prepared ceramic−carbonate dual-phase membranes by replacing the metal support with ionic conducting or mixed conducting metal oxide ceramics. Dual-phase membranes using various mixed or oxygen ionic conducting porous © 2014 American Chemical Society
Received: Revised: Accepted: Published: 13459
May 23, 2014 July 1, 2014 July 11, 2014 July 11, 2014 dx.doi.org/10.1021/ie502094j | Ind. Eng. Chem. Res. 2014, 53, 13459−13466
Industrial & Engineering Chemistry Research
Article
Table 1. CO2 Permeation Properties for Ceramic−Carbonate Dual-Phase Membranes ceram phase
membr thickness (μm)
temp (°C)
σv (S/cm)
CO2 flux (10−3 mol/s/m2)
stability (h)
selectivity (CO2/N2)
La0.6Sr0.4Co0.8Fe0.2O3−δ CGO Y0.16Zr0.84O2−δ Bi1.5Y0.3Sm0.2O3−δ Ce0.8Sm0.2O1.9a Ce0.8Sm0.2O1.9b Ce0.8Sm0.2O1.9c
375 200−400 200−400 ∼50 1500 1500 ∼150
900 850 750 650 700 950 900
0.173 0.126 0.029 0.137 0.21 0.155 0.11
2.43 1.51 1.02 0.55 3.35 6.40 11.6
∼80 N/A N/A ∼80 N/A >30 days N/A
225 0.2−4 0.2−4 N/A N/A >1000 >1000
a Prepared from a “coprecipitation” and “sacrificial-template” synthesis. bPrepared from uniaxially pressing and sintering method. cPrepared from centrifugal method, tubular membrane.
2. EXPERIMENTAL SECTION 2.1. Membrane Synthesis. Powders of Bi1.5Y0.3Sm0.2O3−δ (BYS) and Ce0.8Sm0.2O1.9 (SDC) were prepared by the citrate method.23,25 Stoichiometric amounts of the corresponding metal nitrates, i.e., Bi(NO3)3·5H2O (98%, Sigma-Aldrich), Y(NO3)3·6H2O (99.9%, Sigma-Aldrich), Ce(NO3)3·6H2O (99.5%, Alfa-Aesar), and Sm(NO3)3·6H2O (99.9%, SigmaAldrich), were fully dissolved in a dilute nitric acid solution (10 vol % concentrated HNO3), followed by addition of anhydrous citric acid (99.5%, Alfa Aesar). The solutions were heated to 90−110 °C forming a sticky gel-like solid, and a fast and uniform self-ignition of the organics was performed at 400 °C. After self-ignition, both powders were further calcined at 900 °C for 6 h. Macroporous disks of BYS, SDC, or SDC/BYS were prepared by placing about 3.5 g of powder of BYS, SDC, or a mixture of SDC/BYS, with the addition of about 5 wt % graphite (Alfa Aesar, APS 7−11 μm), in a stainless steel mold (22 mm in diameter), and pressing at 200 MPa for 5 min. The green disks of BYS, SDC, or SDC/BYS were then respectively sintered at 750, 1100, and 950 °C for 24 h. Addition of graphite powders will ensure that the support base contains macropores. To make a support with the structure shown in Figure 1, about 0.2−0.3 g of SDC powders ball-milled for 72 h (to reduce particle size) were packed into the stainless steel mold and pressed under a hydraulic pressure of 5 MPa for 10 s. Then about 3.5 g of BYS powders without ball-milling, with an addition of 5 wt % graphite, was added on top of the SDC layer with a pressure of 150 MPa for 5 min, as shown in steps 1−4 in Figure 2. The green body was removed from the mold and sintered at 950 °C for 20 h with a ramping rate of 1 °C/min to obtain BYS supported SDC. Asymmetric porous supports consisting of thick, macroporous SDC/BYS support sandwiched by two thin SDC layers (to balance sintering stress) were prepared by a similar method
CO2 permselectivity is guaranteed by the thin and dense dualphase top layer only. A thin dual-phase membrane using yttria stabilized zirconia (YSZ) as the ceramic phase was successfully synthesized.30 The 10 μm thick membrane achieved a CO2 flux of 3.89 × 10−3 mol/s/m2 at 650 °C that was 10 times larger than that for thick YSZ−carbonate membrane.22 As discussed above, SDC−carbonate membranes exhibit high CO2 permeability and stability for CO2 separation applications. Therefore, it is important to study methods to prepare thin SDC−carbonate membranes on a proper support base. The basic requirements for such base support are that30 (1) it should be porous providing minimum mass transport resistance, (2) it should be oxygen ionic conducting so a support−carbonate plug, if formed in part of the support, would not provide much transport resistance for CO2, (3) it should not wet carbonate to avoid infiltration of carbonate into the base support, and (4) it should be compatible with the porous SDC top layer. Following the work of Lu and Lin,30 macroporous, oxygen ionic conducting, carbonate nonwettable Bi1.5Y0.3Sm0.2O3−δ (BYS) would be selected as the base, coated with a porous SDC layer to be infiltrated with carbonate, as shown in Figure 1. However, as will be discussed in this paper,
Figure 1. Schematic configuration of thin SDC−carbonate dual-phase membrane.
this concept does not work for preparing supported thin SDC− carbonate membranes. Thus, the present work examined the use of a macroporous SDC/BYS base support for synthesis of thin SDC−carbonate membrane and studied the properties of thin SDC−carbonate membranes on these supports. The outcome of this study would provide guidance on the synthesis of thin ceramic−carbonate dual-phase membranes by other methods, such as thin SDC−carbonate in tubular shape fabricated by a centrifugal slip-casting method recently reported by Lin and co-workers.26 The objective of this paper is to present a rational design of the structure of the base support and supported thin SDC− carbonate membranes and properties of the thin membranes prepared in such a design by a simple copressing method. High temperature gas permeation properties of the membrane are studied as well.
Figure 2. Schematic drawing of preparation of thin SDC layers on SDC/BYS supports by copressing method. 13460
dx.doi.org/10.1021/ie502094j | Ind. Eng. Chem. Res. 2014, 53, 13459−13466
Industrial & Engineering Chemistry Research
Article
increments, starting from 550 °C and upward to 700 °C. Once the desired temperature was reached, the system was kept at that temperature for 1 h before data collection. Membrane long-term stability was tested by measuring CO2 permeation at 700 °C for over 100 h. The gas composition in the sweep gas was measured by a HP 5890 Series II gas chromatograph with a TCD detector with an Alltech Hayesep DB 100/120 column of 30 ft × 1/8 in. × 0.85 in. stainless steel. The amount of N2 detected in the sweep gas during permeation experiments allowed determination of CO2 leak through the seal. Corrected CO2 flux was made by subtracting the CO2 associated with the leak. Gas leakage through the membrane or the seal is assumed to follow the Knudsen diffusion mechanism. Leaking nitrogen flux was around 3−5% of the total amount of CO2 measured. For each condition, we measured two to three data in around 30 min intervals. The errors of the data are within 10%.
shown in Figure 2. After formation of SDC and SDC/BYS layers in the mold (step 4 in Figure 2), 0.2−0.3 g of ball-milled SDC powders were again added on the top and all three layers were uniaxially pressed under a pressure of 200 MPa for 5 min. Asymmetric porous supports with a three-layer sandwich structure were achieved. The green body was removed from the mold, and sintered at 950 °C for 20 h with heating and cooling rates of 1 °C/min. One side of the SDC layer was then polished off by a SiC paper (Struers, No. 800) to obtain the support with the structure shown in Figure 1. All the disks prepared are about 20 mm in diameter and about 2 mm in thickness. Thin dual-phase membrane was obtained via a direct infiltration of molten carbonates into the top SDC layer of asymmetric porous support.18 Lithium (Li), sodium (Na), and potassium (K) carbonates (Fisher Scientific, Li2CO3, 99.2%; Na2CO3, 99.9%; K2CO3, 99.8%) were mixed in 42.5/32.5/25 mol % ratio and heated to 550 °C in a box furnace. Before infiltration, the asymmetric support was preheated in the furnace for 30 min above the molten carbonate and then was dipped, with the SDC layer facing down, into the molten carbonate. The support was held in contact with molten carbonate for 20 min to allow the molten carbonate to infiltrate the thin porous SDC top layer via capillary action. A continuous carbonate layer covered the surface of the SDC support, and was removed by the “sponge” soaking method reported by Lu and Lin.30 Finally, membrane was cooled to room temperature at a rate of 1 °C/min. As-synthesized thin SDC−carbonate dual-phase membrane is denoted as thin SDC−MC. 2.2. Membrane Characterization and High Temperature CO2 Permeation Measurements. Helium permeances of SDC/BYS supports were measured by the unsteady state permeation method. The porosity of the ceramic layer is determined mainly by the particle size and sintering conditions. In this work, the porosity of the base supports was measured by the liquid nitrogen infiltration method.31 The porosity of the top layers was estimated from the data of the support of the top layer material and particle size measured by the liquid nitrogen method. The sizes of the supports before and after sintering were measured by a caliper. The molten carbonate wettability of SDC/BYS supports was checked by observing molten carbonate liquid drop on the supports at 600 °C. Phase structures of powders, porous supports, and dual-phase membranes were determined by X-ray diffraction (XRD) (Bruker AXS, D8 Focus Diffractometer, Cu Kα). The morphologies of the membranes were examined by scanning electron microscopy (SEM, Phillips, FEI XL-30). High temperature CO2 permeation tests were carried out in a homemade permeation setup as described in our previous publications.23,30 The thin dual-phase membrane was sealed with graphite seals in the middle of a high temperature cell. CO2 (50 mL/min, STP) and N2 (50 mL/min, STP) were introduced to the feed side, and He (100 mL/min, STP) was introduced to the downstream as sweep gas during the heating and permeation measurements. Before the system was heated, inert gases of N2 (50 mL/min, STP) to the feed side and He (50 mL/min, STP) to the sweep side were introduced for 20 min to eliminate oxygen in the membrane cell and protect the graphite seals from decomposition. The system was heated at a rate of 1 °C/min to 550 °C. After allowing the system to remain at 550 °C for 2 h reaching steady state, data collection began. Measurements were taken in 25 °C (1 °C/min ramp)
3. RESULTS AND DISCUSSION 3.1. Membrane Synthesis. Asymmetric porous supports containing one thin SDC layer and a thick BYS support were prepared by copressing a thin SDC layer directly on pure BYS base support followed by sintering. However, the support was seriously warped after sintering at 950 °C, as shown in Figure 3.
Figure 3. Morphology of thin SDC top layer prepared on pure BYS support before and after sintering (BYS is facing up).
BYS and SDC have different sinterabilities as reflected by the different melting points of 1100 °C for BYS32 and 2200 °C for SDC.24 For example, BYS support can maintain a desirable porosity after sintering at temperatures to 750 °C30 while SDC support could remain porous up to 1400 °C.25 This means that, at the same sintering temperature, the BYS layer would shrink more than the SDC layer. The difference in the shrinkage rate causes the warp toward the BYS side of the two-layer base support, as shown in Figure 3. To overcome the issue, SDC powder was mixed with BYS powder to prepare the SDC/BYS base support. In this way, addition of SDC in the base BYS support helps elevate the allowable sintering temperature of the base support and ensures better sintering comparability with the top thin SDC layer. SDC/BYS base supports with SDC of 20, 40, and 60 vol %, denoted as SDC20BYS80, SDC40BYS60, and SDC60BYS20, were prepared. The characteristics of the BYS/SDC supports are summarized in Table 2. Pure BYS and SDC supports are also included for comparison. Figure 4 shows helium permeances and porosities of SDC/ BYS supports with different mixing ratios. SDC60BYS40 exhibits the highest permeance and porosity, and it could provide minimum gas transport resistance through the membrane. As shown in Table 2, the morphology change (disk diameter) of SDC60BYS40 after sintering is almost same as that of pure SDC support. Addition of SDC powders in BYS 13461
dx.doi.org/10.1021/ie502094j | Ind. Eng. Chem. Res. 2014, 53, 13459−13466
Industrial & Engineering Chemistry Research
Article
Table 2. Characteristics of SDC/BYS Supports support
diam of green disk (mm)
diam after sintering (mm)
He permeance (10−6 mol/s/m2/Pa)
porosity (%)
carbonate wettability
pure BYS SDC20BYS80 SDC40BYS60 SDC60BYS40 pure SDC
22 22 22 22 22
18 18.9 20.0 21.2 21.3
0.001 0.15 1.48 8.81 8.55
5 28 40 46 45
nonwettable nonwettable nonwettable nonwettable wettable
Figure 4. Porosity and helium permeance dependence on volume fraction of SDC powders in SDC/BYS support.
improves sintering compatibility between the base support and the top layer. Figure 5 shows XRD patterns of BYS, SDC, and Figure 6. Examination of molten carbonates wettability on the various BYS, SDC60BYS40, and SDC supports: (a) dry carbonate powders on top of the supports at room temperature and (b) same supports heated and staying at 600 °C.
support. The BYS is carbonate nonwettable30 and SDC is carbonate wettable.25 Melting carbonate droplets are also observed on SDC60BYS40 support, exhibiting its carbonate nonwettability. Although molten carbonate is wettable to SDC, the existence of nonwettable BYS powder apparently modifies the surface property of the support. The support pores become carbonate nonwettable and avoid being filled by molten carbonate during infiltration. High permeance, carbonate nonwettable SDC60BYS40 was chosen as the base support for thin dual-phase membrane. Crack-free and flat asymmetric porous supports consisting of SDC60BYS40 sandwiched by two macroporous SDC layers were obtained after sintering. Helium permeance of the asymmetric support is around 10−15% lower than that of the base support due to resistance offered by the two thin SDC layers on both sides of the base support. Figure 7a shows the SEM image of surface morphology of macroporous SDC membrane on SDC60BYS40 base support. The membrane is homogeneous and of good quality. The thin SDC layer on the base support exhibits a morphology similar to that of thick SDC membrane.25 Figure 7b presents a cross section of thin SDC membrane on the support. The thickness of the top layer is around 150 μm. Compared to the warping SDC layer shown in Figure 3, the support is flat and not warped. Addition of SDC in the BYS base support successfully improves the sintering
Figure 5. XRD patterns of powders of BYS and SDC and composite SDC60BYS40 support after sintering at 950 °C.
SDC60BYS40 supports. Both BYS and SDC are of cubic fluorite structure. SDC60BYS40 shows a XRD pattern that combines the characteristic patterns of pure BYS and SDC, and no impurities are observed. This indicates that BYS and SDC in the support are present as separate phases and the powder mixtures are chemically stable after high temperature sintering. To examine molten carbonate wettability to the supports, Figure 6 show porous support disks holding carbonate solid powder at room temperature (Figure 6a) and after heat treatment in air at 600 °C (Figure 6b). Compared with Figure 6a, clear liquid drops of melting carbonates can be observed on top of BYS support and no residuals are found on the SDC 13462
dx.doi.org/10.1021/ie502094j | Ind. Eng. Chem. Res. 2014, 53, 13459−13466
Industrial & Engineering Chemistry Research
Article
reported.30 Figure 7c shows the morphology of the cross section of the top SDC layer after carbonate infiltration. The SDC−carbonate layer, of 150 μm thickness, looks dense showing complete filling of the SDC pores by molten carbonate. Thin membrane exhibits a 4 order of magnitude reduction in helium permeance compared to the support before carbonate infiltration. This membrane is referred to as “d-SDC/ SDC60BYS40” membrane for CO2 permeation study. 3.2. High Temperature CO2 Permeation Test. High temperature permeation tests were performed with SDC dualphase membrane under different conditions. Table 3 Table 3. Experimental Parameters of CO2 Permeation Tests for 150 μm Thick SDC−Carbonate Dual-Phase Membrane (d-SDC/SDC60BYS40) test param
temp (°C)
P′CO2 (atm)
temper long-term stab P′CO2
550−700 700 700
0.5 0.5 0.1−0.9
FHe (mL/min) 100 100 100
P″CO2
700
0.5
75−175
summarizes experimental parameters for the tests. Figure 8a shows the temperature dependence on CO 2 flux. CO 2 permeation flux increases from 1.33 × 10−3 to 6.55 × 10−3 mol/s/m2 when the temperature changes from 550 to 700 °C. Figure 8b exhibits CO2 permeation long-term stability of thin SDC−MC membrane at 700 °C. The membrane was exposed
Figure 7. SEM images of (a) surface, (b) cross section of thin SDC layer about 150 μm thick copressed on SDC60BYS40 support, and (c) cross section of the SDC layer after molten carbonate infiltration.
comparability of two layers. No cracks are observed between the top SDC layer and SDC60BYS40 base support. The thin dual-phase membrane was prepared by directly infiltrating the molten carbonate into the asymmetric SDC60BYS40 support. The carbonate residual layer formed after infiltration was removed by the method previously
Figure 8. CO2 permeation flux through d-SDC/SDC60BYS40 support as a function of (a) temperature and (b) permeation time. 13463
dx.doi.org/10.1021/ie502094j | Ind. Eng. Chem. Res. 2014, 53, 13459−13466
Industrial & Engineering Chemistry Research
Article
in the SDC phase, can be described by the following equation:25
to CO2:N2 (50:50) feed with a total pressure of 1 atm. The membrane exhibits a stable CO2 flux about 6.55 × 10−3 mol/s/ m2 for 160 h. The test was discontinued due to system leakage from decomposition of graphite seals above 650 °C. Longer stability is expected with the proper sealing materials. Figure 9 shows the pressure effect on CO2 permeation through the SDC/SDC60BYS40 dual-phase membrane. The
JCO = 2
RT 1 ⎛⎜ ε ⎞⎟ σi[P′CO2 n − P″CO2 n ] 4nF 2 L ⎝ τ ⎠
(1)
where R is the ideal gas constant, n is a constant specific to the ceramic phase, T is the system temperature, F is Faraday’s constant, L is the membrane thickness, ε and τ are the SDC volume fraction and tortuosity for oxygen ion conductivity, and σi is the oxygen ionic conductivity in the SDC phase at PO2 = 1 atm. As shown by eq 1, the CO2 permeation flux increases with increasing upstream CO2 partial pressure, P′CO2, and decreasing downstream CO2 partial pressure, P″CO2 The downstream CO2 partial pressure is equal to the downstream total pressure times the ratio of CO2 permeation flow to the total flow rate in the permeate side (CO2 permeation flow rate plus sweep gas flow rate). Therefore, eq 1 explains the experiment results of CO2 partial pressure dependence of CO2 permeation flux shown in Figure 9. Figure 10 plots the CO2 flux (combining data at different feed side CO2 partial pressures and sweep side helium flow
Figure 9. CO2 permeation flux through d-SDC/SDC60BYS40 support as a function of (a) feed side upstream CO2 partial pressure, P′CO2, and (b) permeate side downstream CO2 partial pressure, P″CO2.
Figure 10. CO2 permeation flux through SDC−carbonate dual-phase membrane versus P′CO2n − P″CO2n.
tests were performed at 700 °C with a total feed pressure of 1 atm. Figure 9a shows the effect of feed side CO2 partial pressure (P′CO2) on the CO2 permeation flux. The sweep helium flow rate was fixed at 100 mL/min. P′CO2 was adjusted by changing the feed side CO2 flow rate from 10 to 90 mL/min. The CO2 permeation flux increases from 2.45 × 10−3 to 9.17 × 10−3 mol/s/m2 with upstream CO2 partial pressure from 0.1 to 0.9 atm. Figure 9b shows the effect of the sweep side CO2 partial pressure (P″CO2) on CO2 permeation. The feed side CO2 partial pressure was maintained at 0.5 atm. Different P″CO2 values were obtained by changing the sweep side helium flow rate from 75 to 175 mL/min. The CO2 permeation flux decreases with an increase of downstream CO2 partial pressure. A smaller P″CO2 (higher helium flow rate) leads to a larger CO2 flux. The CO2 permeation flux through SDC−carbonate membranes, for which the carbonate ionic conductivity is about 2 orders of magnitude larger than the oxygen ionic conductivity
rates) versus P′CO2n − P″CO2n, and n is equal to 0.5 at 700 °C.25 The value of n is related to the defect equilibrium between the oxygen vacancy concentration in SDC and the CO2 partial pressure, as discussed previously.25 The plot gives a straight line going through the origin of the coordinate and CO2 flux proportionally increases with P′CO2n − P″CO2n, confirming the CO2 pressure dependence of the CO2 permeation flux through the SDC−carbonate membrane. The slope of the corresponding straight line in Figure 10 is 0.011 mol/s/m2/atm0.5. In a previous study,25 for thick (1.5 mm) SDC−carbonate dualphase membrane tested at 700 °C, the slope for the line of CO2 flux versus pressure difference is 0.002 mol/s/m2/atm0.5. The slope for CO2 permeation data for the thin dual-phase membrane is about 5.5 times that of the thick membrane, but the thickness of the latter (1.5 mm) is 10 times that of the former (0.15 mm). As shown by eq 1 for the same material, the slope is proportional to (1/L)(ε/τ). Not only the membrane thickness, L, but also the SDC structure, ε/τ, affects the CO2 13464
dx.doi.org/10.1021/ie502094j | Ind. Eng. Chem. Res. 2014, 53, 13459−13466
Industrial & Engineering Chemistry Research
Article
SDC65MC35 dual-phase membrane reported by Huang and co-workers24 has a higher CO2 permeation flux than SDC−MC membrane25 of the same thickness. The SDC support of the SDC65MC35 membrane was prepared by “template-sacrificial” synthesis,24 while that of SDC−MC was prepared by powder press-and-sintering.25 It is possible that SDC support prepared by the “template-sacrificial” method has a larger ε/τ value than the SDC support of the SDC−MC membrane. The thin dSDC/SDC60BYS40 dual-phase membrane has a larger CO2 permeation flux than SDC65MC35 with the support having a porosity similar to that of SDC65MC35. This shows that reducing membrane thickness is still an effective way to improve CO2 permeation flux.
permeation. The SDC/SDC60BYS40 support was prepared at a sintering temperature of 950 °C, while the thick SDC support was prepared at 1100 °C.25 The value of ε/τ for porous ceramics increases with increasing sintering temperature. This is because increasing temperature leads to coarsening of sintered particles and growth of the pores. The pore growth gives a larger porosity, ε, and the particle coarsening results in a smaller tortuosity, τ. Thus, the value of ε/τ increases with increasing sintering temperature.27 Therefore, the value of ε/τ for thick SDC−carbonate membrane is larger than that for the thinner SDC. This explains the less-than-expected increase in CO2 permeation flux for the thinner membrane.
4. CONCLUSIONS Thin samarium doped ceria (SDC)−carbonate dual-phase membrane can be prepared on macroporous SDC/BYS base support with desired ionic conductivity, carbonate nonwettability, mechanical stability, and compatibility with the top layer. The final membranes consist of a 150 μm, hermetic SDC−carbonate layer on the macroporous 60% SDC−40% BYS base support. The thin SDC−carbonate dual phase membrane offers significantly improved CO2 permeance as compared to thick SDC−carbonate membranes. Both membrane thickness and the structure of the SDC phase affect CO2 permeance. CO2 permeation through the membranes exhibits increases with temperature (550−700 °C). The thin SDC− carbonate membrane has a CO2 permeation flux proportional to P′CO2n − P″CO2n (n = 0.5 at 700 °C). Reducing the membrane thickness effectively improves the CO2 permeation flux. However, the structure of the SDC support also has a strong effect on CO2 permeation. The dual-phase membrane is stable for longer than 160 h.
Figure 11. Comparison of CO2 permeation fluxes for four different SDC−carbonate membranes of different thicknesses and microstructures: thin 0.15 mm d-SDC/SDC60BYS40 disk membrane (in this work), 1.5 mm SDC65MC35 disk membrane,24 thin 0.15 mm SDC−MC tubular membrane,26 and 1.5 mm SDC−MC disk membrane.25 Data in solid lines are reported or measured values, while points in dashed lines are estimated from reported activation energies.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected].
Figure 11 and Table 4 compare CO2 permeation fluxes of four SDC−carbonate dual-phase membranes of different thicknesses, with the SDC supports prepared by different methods. As mentioned above, not only the membrane thickness but also the SDC structure plays an important role in affecting the CO2 permeation flux. The disk and tubular thin dual-phase membranes have the same thickness but different CO2 permeation fluxes because the two supports were prepared by different methods (copressing for disk in this work and centrifugal casting for tube26), resulting in different SDC structures. The CO2 permeation flux of thin d-SDC/ SDC60BYS40 membrane is about 5 times that of the thick SDC−MC membrane due to the effects of both membrane thickness and SDC microstructure as explained above.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
The authors acknowledge the support of National Science Foundation (CBET-0828146) and the U.S. Department of Energy (DE-FE000470) on the work reported here. The authors are grateful for the assistance of the School for Matter, Transport, and Energy lab manager Fred Pena. We also acknowledge the support and use of facilities in the LeRoy Eyring Center for Solid State Sciences (LE-CSSS) at Arizona State University.
Table 4. CO2 Permeation Properties of SDC Based Dual-Phase Membranes membr (shape)
membr thickness (mm)
temp range (°C)
thin SDC−MC (disk) SDC65MC35 (disk) asym SDC−MC (tube) SDC−MC (disk)
∼0.15 1.5 ∼0.15 1.5
550−700 550−700 700−900 700−950
a
CO2 flux (10−3 mol/s/m2) εa (%) 1.38−6.55 0.55−3.35 3.14−11.6 1.3−6.40
35 37 35 36
synth method
ref
copressing coprecipitation and sacrificial template centrifugal casting uniaxially pressing
this work 24 26 25
ε, porosity. 13465
dx.doi.org/10.1021/ie502094j | Ind. Eng. Chem. Res. 2014, 53, 13459−13466
Industrial & Engineering Chemistry Research
■
Article
Inorganic Membranes; Bredesen, R., Rader, H. Eds.; SINTEF: Oslo, 2006; p 678. (21) Anderson, M.; Lin, Y. S. Carbonate−ceramic dual-phase membrane for carbon dioxide separation. J. Membr. Sci. 2010, 357, 122−129. (22) Wade, J. L.; Lee, C.; West, A. C.; Lackner, K. S. Composite electrolyte membranes for high temperature CO2 separation. J. Membr. Sci. 2011, 369, 20−29. (23) Rui, Z.; Anderson, M.; Li, Y.; Lin, Y. S. Ionic conducting ceramic and carbonated dual phase membranes for carbon dioxide separation. J. Membr. Sci. 2012, 417−418, 174−182. (24) Zhang, L.; Xu, N.; Li, X.; Wang, S.; Huang, K.; Harris, W. H.; Chiu, W. K. S. High CO2 permeation flux enabled by highly interconnected three-dimensional ionic channels in selective CO2 separation membranes. Energy Environ. Sci. 2012, 5, 8310−8317. (25) Norton, T.; Lu, B.; Lin, Y. S. Carbon dioxide permeation properties and stability of samarium doped ceria-carbonate dual-phase membrane. J. Membr. Sci. 2014, 467, 244−252. (26) Dong, X. L.; Ortiz-Landeros, J.; Lin, Y. S. Asymmetric tubular ceramic-carbonate dual phase membrane for high temperature CO2 separation. Chem. Commun. 2013, 49, 9654−9656. (27) Ortiz-Landeros, J.; Norton, T.; Lin, Y. S. Effects of support pore structure on carbon dioxide permeation of ceramic-carbonate dualphase membranes. Chem. Eng. Sci. 2013, 104, 891−898. (28) Wade, J. L.; Lackner, K. S.; West, A. C. Transport model for a high temperature mixed conducting CO2 separation membrane. Solid State Ionics 2007, 178, 1530−1540. (29) Rui, Z.; Anderson, M.; Lin, Y. S.; Li, Y. D. Modeling and analysis of carbon dioxide permeation through ceramic-carbonate dual-phase membranes. J. Membr. Sci. 2009, 345, 110−118. (30) Lu, B.; Lin, Y. S. Synthesis and characterization of thin ceramiccarbonate dual-phase membranes for carbon dioxide separation. J. Membr. Sci. 2013, 444, 402−411. (31) Harry, K. G.; Johnson, A. A non-destructive technique for measuring ceramic porosity using liquid nitrogen. J. Archaeol. Sci. 2004, 31, 1567−1575. (32) Zeng, Y.; Lin, Y. S. Synthesis and properties of copper and samarium doped yttria-bismuth oxide powders and membranes. J. Mater. Sci. 2011, 36, 1271−1276.
REFERENCES
(1) Armin, J. A. R.; Ebner, D. State-of-the-art adsorption and membrane separation process for carbon dioxide production from carbon dioxide emitting industries. Sep. Sci. Technol. 2009, 44, 1273− 1421. (2) Anderson, M.; Wang, H. B.; Lin, Y. S. Inorganic membranes for carbon dioxide and nitrogen separation. Rev. Chem. Eng. 2012, 28, 101−121. (3) Kim, S.; Ida, J.; Guliants, V.; Lin, Y. S. Functionalized mesoporous silica membrane for the separation of carbon dioxide. Int. J. Environ. Technol. Manage. 2004, 4, 21−31. (4) Poshusta, J.; Tuan, A. V.; Pape, E.; Noble, R.; Falconer, J. Separation of light gas mixtures using SAPO-34 membranes. AIChE J. 2000, 46, 779−789. (5) Koros, W. J.; Mahajan, R. Pushing the limits on possibilities for large scale gas separation: which strategies? J. Membr. Sci. 2000, 175, 181−196. (6) Ostwal, M.; Singh, R. P.; Dec, S. F.; Lusk, M. T.; Way, J. D. 3Aminopropyltriethoxysilane functionalized inorganic membranes for high temperature CO2/N2 separation. J. Membr. Sci. 2011, 369, 139− 147. (7) Hu, T.; Dong, G. X.; Li, H. Y.; Chen, V. Improved CO2 separation performance with additives of PEG and PEG−PDMS copolymer in poly (2,6-dimethyl-1,4-phenyleneoxide) membranes. J. Membr. Sci. 2013, 432, 13−24. (8) Bai, H.; Ho, W. S. W. New carbon dioxide-selective membranes based on sulfonated polybenzimidazole (SPBI) copolymer matrix for fuel cell applications. Ind. Eng. Chem. Res. 2009, 48, 2344−2354. (9) White, J. C.; Dutta, P. K.; Shqau, K.; Verweij, H. Synthesis of ultrathin zeolite Y membranes and their application for separation of carbon dioxide and nitrogen gases. Langmuir 2010, 26, 10287−10293. (10) Nicolas, C. H.; Sublet, J.; Schuurman, Y.; Pera-Titus, M. Role of adsorption and diffusion path ways on the CO2/N2 separation performance of nanocomposite B-MFI-alumina membranes. Chem. Eng. Sci. 2011, 66, 6057−6068. (11) Bernal, M. P.; Coronas, J.; Menendez, M.; Santamaria, J. Separation of CO2/N2 mixtures using MFI-type zeolite membranes. AIChE J. 2004, 50, 127−135. (12) Ping, E. W.; Zhou, R. F.; Funke, H. H.; Falconer, J. L.; Noble, R. D. Seeded-gel synthesis of SAPO-34 single channel and monolith membranes for CO2/CH4 separations. J. Membr. Sci. 2012, 415−416, 770−775. (13) He, X.; Lie, J. A.; Sheridan, E.; Hagg, M. B. Preparation and characterization of hollow fiber carbon membranes from cellulose acetate precursors. Ind. Eng. Chem. Res. 2011, 50, 2080−2087. (14) Kai, T.; Kazama, S.; Fujioka, Y. Development of cesiumincorporated carbon membranes for CO2 separation under humid conditions. J. Membr. Sci. 2009, 342, 14−21. (15) Sakamoto, Y.; Nagata, K.; Yogo, K.; Yamada, K. Preparation and CO2 separation properties of amine-modified mesoporous silica membranes. Microporous Mesoporous Mater. 2007, 101, 303−311. (16) Moon, J. H.; Park, Y. J.; Kim, M. B.; Hyun, S. H.; Lee, C. H. Permeation and separation of a carbon dioxide/nitrogen mixture in a methyltriethoxysilane templating silica/α-alumina composite membrane. J. Membr. Sci. 2005, 250, 195−205. (17) Rui, Z. B.; Ji, H. B.; Lin, Y. S. Modeling and analysis of ceramiccarbonate dual-phase membrane reactor for carbon dioxide reforming with methane. Int. J. Hydrogen Energy 2011, 36, 8292−8300. (18) Chung, S. J.; Park, J. H.; Li, D.; Ida, J. I.; Kumakiri, I.; Lin, Y. S. Dual-phase metal-carbonate membrane for high-temperature carbon dioxide separation. Ind. Eng. Chem. Res. 2005, 44, 7999−8006. (19) Xu, N.; Li, X.; Franks, M. A.; Zhao, H.; Huang, K. Silver-molten carbonate composite as a new high-flux membrane for electrochemical separation of CO2 from flue gas. J. Membr. Sci. 2012, 401−402, 190− 194. (20) Anderson, M.; Lin, Y. S. Synthesis and Characterization of Carbonate-Ceramic Dual-Phase Membranes for Carbon Dioxide Separation. In Proceedings of the 9th International Conference on 13466
dx.doi.org/10.1021/ie502094j | Ind. Eng. Chem. Res. 2014, 53, 13459−13466
