1 Introduction
ABO3 perovskite oxides are widely used as oxygen electrode materials in both solid oxide fuel cells and solid oxide electrolysis cells, owing to their excellent electrical conductivity, electrochemical catalytic performance, as well as low cost [1-3]. Typically, rare earth and alkaline earth elements are used at the A-site and reducible transition metals such as Mn, Fe, Co, Ni are used at the B-site [4]. ABO3 oxides with Sr2+ substitution in the A-site, including La1-xSrxMnO3−δ (LSM), La1-xSrxCo1-yFeyO3−δ (LSCF), La1-xSrxFeO3−δ (LSF), La1-xSrxCoO3−δ, Sm1-xSrxCoO3−δ, and Ba1-xSrxCo1-yFeyO3−δ, possess much better O2− transport ability and lower polarization resistance for oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) [5-11]. However, such Sr-doped electrodes suffer from Sr surface segregation (SSS), which seriously decreases the ORR and OER activities [12-15]. Therefore, it is urgent to investigate the relationship between Sr substitution and the stability of the ABO3 oxides for developing chemically stable Sr-substituted electrodes with high catalytic activities.
To date, numerous reports have been published on the elemental surface segregation phenomena, including some reviews on the materials, mechanisms, and methods. Understanding the correlation between the surface structure and stability of the electrodes has been emphasized [16-17], which is critical for the rational design of more efficient perovskites and perovskite-related electrode materials for intermediate-temperature solid oxide fuel cells [18]. In case of the La1–xCaxFe0.8Ni0.2O3–δ system, the content of the A-site Ca atoms determines whether the material is monophasic or not. [19]. Low-energy ion scattering spectroscopy has been used to study the outer atomic surfaces of a series of perovskite materials. The results indicate that the segregated A-site cations dominate the outer surfaces while the B cations are enriched below the surface [20]. Total reflection X-ray fluorescence, X-ray photoelectron spectroscopy, and static secondary ion mass spectrometry have also been used to study the Sr enrichment on the surface [21-23]. According to previous reports, many conditions influence Sr segregation. In particular, the Sr concentration on the surface increases with elevating temperature and decreasing partial pressure of oxygen [21, 22], and increased SrO is formed on the surface in humidified air [24]. Furthermore, polarization is another important factor that significantly influences the formation of insulating SrO on the cathode surface or its removal from the surface. Anodic polarization induces the formation of SrO while cathodic polarization removes SrO [23]. Moreover, SSS phenomenon is also associated with Cr contamination of the surface of cathodes under the operating conditions, and the segregated SrO also enhances the deposition of Cr species in the presence of a Fe-Cr alloy metallic interconnect. Additionally, Mn2+ and SrO in LSM and LSCF are considered as the nucleation centers of Cr deposition as well [25-28]. Recently, SSS has also been intensively explored via the first principles studies. Further, surface charge minimization and strain relaxation have been considered as the two possible thermodynamic driving forces during SSS [14]. In particular, tensile strain and size mismatch between the cations at the A-site are demonstrated to be the fundamental factors leading to surface enrichment in selected elements (Ca, Sr, Ba) doped LnMnO3 (Ln = La, Sm) at elevated temperatures [29]. Based on these findings, the suppression of SSS in LSCF is achieved by the infiltration of Sm-doped CeO2, BaO, La0.8Sr0.2MnO3−δ, La0.1Sr0.9Co0.5Mn0.5O3−δ, PrSrCoMnO6−δ, and Pr0.75Sr0.25MnO3−δ [30-34].
In view of the crystal structure, Sr stability in perovskite oxides with Sr substitution is closely related to the fine structures around the Sr atoms. However, little effort has been focused on establishing the relationship between the local geometric structure of the perovskite electrodes and Sr segregation. In this work, we utilized X-ray absorption spectroscopy (XAS), an excellent method to probe the local atomistic and electronic structure [35-37], to establish correlation between the Sr stability and local geometric structures around the Sr atoms in LSM and LSF. Scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), and X-ray diffraction (XRD) were also employed to identify the segregated compounds formed on the surfaces of LSM and LSF. Further, the effect of Cr deposition on Sr segregation was also been taken into consideration in the research.
2 Experimental Section
2.1. Synthesis of LSM and LSF materials
Powders of LSM and LSF were prepared by the conventional sol-gel method. For the synthesis of LSM, stoichiometric amounts of La(NO3)3•6H2O (Strem Chemicals), Sr(NO3)2 (Alfa Aesar), and Mn(NO3)2 (Alfa Aesar) were mixed and dissolved in distilled water to obtain a precursor solution. After that, citric acid (Alfa Aesar) was added as the complexing agent into the precursor solution. Then, Ethylenediaminetetraacetic acid (EDTA, J&K Chemical) was dissolved in ammonium hydroxide on a hot plate at 80 °C. The molar ratio of citric acid:EDTA:total metal ions was fixed at 1.2:1.5:1. The pH of the solution was adjusted to 7 using ammonium hydroxide and nitric acid solutions. The two prepared solutions were mixed and slowly evaporated at 80 °C under stirring over 5 h to obtain a gel. Then, the gel was decomposed at 150 °C in an oven for 24 h. Finally, the powder was ground in a mortar and calcined at 1000 °C for 4 h in air. The synthesis of LSF was carried out following the same procedure using Fe(NO3)3•9H2O (Strem Chemicals) instead of Mn(NO3)2.
2.2. XRD and XAS characterization
XRD patterns were collected in 0.02° steps over the 2θ range of 20° to 90° to verify the phase purity of raw powders using a Bruker D8 ADVANCE X-ray diffractometer with Cu Kα radiation. The K-edge X-ray absorption spectra of Sr, Mn, and Fe atoms in LSM and LSF were obtained using the BL14W1 beam line at the Shanghai Synchrotron Radiation Facility (SSRF), operated at 3.5 GeV with the injection current of 200 mA. A Si(111) double-crystal monochromator was used for energy selection. All the spectra were collected in the transmission mode. The powder to be characterized was pressed into a self-supporting disk and then sealed with Kapton membrane before subjecting to XAS measurements. XAS data analyses were carried out using the software package of Ifeffit [38]. The relative structural parameters such as coordination numbers (N), bond distance (R), the Debye-Waller factor (σ2), and inner potential shift (ΔE0) were calculated from the XAS data.
2.3. Preparation of LSM and LSF electrodes on GDC substrates
Gadolinium-doped ceria (GDC, Fuelcell materials) was pressed into disks and sintered at 1600 °C for 4 h in air to form the electrolyte pellets. The disks were approximately 20 mm in diameter. LSM and LSF slurries were obtained by mixing the powders and special inks and then screen-printed to obtain the GDC electrolyte, followed by sintering at 1100 °C for 2 h. The effective area of the electrode was fixed at 0.5 cm2.
2.4. Heat treatment of the electrodes and post-treatment analysis
All the electrodes were sintered at 900 °C for 20 h in a furnace; a commercial Fe-Cr alloy bulk was placed on top of each electrode as the source of Cr species. The alloy (RA446, 23–27 wt.% Cr, 1.5 wt.% Mn, 1 wt.% Si, 0.2 wt.% C, 0.12 wt.% N, and remaining Fe) was machined into cuboids (12 mm × 12 mm × 4 mm), with channels (1.2 mm × 1.2 mm) cut on one side. The rib of the Fe-Cr cuboid was in touch with the electrode, while the channel region of the alloy was not. Finally, SEM and EDS were performed to characterize the morphologies of the electrodes using a LEO1530VP FE-SEM (ZEISS, Germany) equipped with an Oxford energy- dispersive X-ray spectrometer. Since the surfaces of the samples were not flat enough, we selected three points in each region for EDS analysis and used the mean value of the elemental content to calculate the Sr/La ratio. Further, XRD was used to characterize the composition of the as-prepared and heat-treated porous electrodes.
3 Results and Discussion
3.1. XRD characterization
The structure of the LSM and LSF powders was characterized by XRD (Fig. 1). XRD data indicates that both samples are of a single phase with the hexagonal perovskite structure belonging to the R-3c space group. The diffraction angles shift to smaller values as the B-site element changes from Mn to Fe, owing to the different ionic radii of these two metals. The refined lattice parameters are shown in Table 1. A smaller volume and smaller lattice parameters are observed for LSM as compared to those of LSF. The average crystalline sizes of LSM and LSF calculated from the full width at half-maximum of the (110) peak using the Scherrer equation are 21.5 and 35.0 nm, respectively.
| Materials | Space group | Volume (Å3) | Lattice parameters (Å) | ||
|---|---|---|---|---|---|
| a | b | c | |||
| LSM | R-3c | 350.49 | 5.502(3) | 5.502(3) | 13.367(1) |
| LSF | R-3c | 354.38 | 5.521(4) | 5.521(4) | 13.422(5) |
3.2. X-ray absorption measurements
Since the XRD data reveals the long-range structural order, it cannot provide information on the local environment of a specific element. Therefore, to identify the local structure around the Sr atoms, we collected the K-edge X-ray absorption spectra of Sr, Mn, and Fe atoms for LSM and LSF (Fig. 2). X-ray absorption near edge structure (XANES) and Fourier transform extended X-ray absorption fine structure (FT-EXAFS) are presented to suggest the oxidation state and the local structural details, respectively.
To quantitatively discuss the differences between the local structure around the Sr, Mn, and Fe atoms, curve fitting was performed. For fitting, the following equations [39] were used:
where, χ is the oscillation in k space, N is the coordination number, f (k, π) is the back-scattering amplitude, σ2 is the Debye–Waller (DW) factor, R is the inter-atomic distance, λ is the mean free path, and ϕ is the phase shift of the photoelectrons. f (k, π), λ and ϕ can be calculated using the FEFF software.
From Sr K-edge XANES spectra in Fig. 2a, we find that Sr atoms in LSM and LSF have overlapped pre-edges and absorption edges. Further, no binding energy shifts can be observed. These results indicate that the Sr atoms in both LSM and LSF have the same oxidation state and other similar features. In Fig. 2b, the peaks emerging at 2.85 Å for LSM and 2.88 Å for LSF can be assigned to the contributions of Sr–O bond in each sample. The structural parameters obtained by fitting the K-edge EXAFS data are listed in Table 2. The average Sr–O bond length in LSM is shorter than that in LSF, indicating that the Coulomb interaction of Sr and O in LSM is stronger than that in LSF. The DW factors of Sr–O and Sr–Mn in LSM are smaller than those of the corresponding bonds in LSF, which demonstrates that LSM has lower local structural disorder. Fig. 2c shows the Mn and Fe K-edge EXAFS spectra and the fitted curves. The average Mn–O bond length is 1.94 ± 0.02 Å, while the interatomic distance between Fe and O is 1.96 ± 0.01 Å, implying slight lattice expansion. The DW factor for Mn–O is smaller than that for Fe–O. All these results suggest that LSM is more stable than LSF.
| Sample | Path | N | R (Å) | σ2(×10-3 Å2) | ΔE0(eV) |
|---|---|---|---|---|---|
| La0.6Sr0.4MnO3−δ | Sr-O1 | 3 | 2.55 ± 0.02 | 7.0 ± 3.1 | −10.0 |
| Sr-O2 | 6 | 2.81 ± 0.02 | 6.0 ± 3.6 | 8.0 | |
| Sr-O3 | 3 | 2.94 ± 0.03 | 2.7 ± 4.5 | −3.5 | |
| Sr-Mn | 8 | 3.34 ± 0.01 | 7.2 ± 0.4 | −5.0 | |
| Mn-O | 6 | 1.94+/−0.02 | 2.5+/−2.3 | −0.9 | |
| La0.6Sr0.4FeO3−δ | Sr-O1 | 3 | 2.55 ± 0.02 | 8.0 ± 3.0 | −10.0 |
| Sr-O2 | 6 | 2.84 ± 0.02 | 11.6 ± 5.4 | 8.0 | |
| Sr-O3 | 3 | 2.99 ± 0.03 | 6.6 ± 6.5 | −3.5 | |
| Sr-Fe | 8 | 3.36 ± 0.01 | 9.3 ± 0.5 | −5.0 | |
| Fe-O | 6 | 1.96+/−0.01 | 3.8+/−1.9 | −3.9 |
A simplified slab model with R-3c structure for ABO3 (B = Mn/Fe, A = Sr) is built as shown in Fig. 2d for further discussion of the local structural changes geometrically. The hexahedron model consists of one Sr atom, three O atoms from different Sr–O shells, and one B atom. Compared with the perfect cubic perovskite structure, the SrBO3 model is distorted. The distance between Sr and different O atoms varies, while the bond length of B–O bond is the same, in accordance with the EXAFS results. As discussed above, the Fe–O bond is longer than the Mn–O bond. Therefore, when the Mn atom is replaced with Fe, the hexahedron structure will be expanded, resulting in an increase in the average Sr–O bond length. Meanwhile, the substitution of the transition metal elements may lead to more severe structural distortion of the octahedron, which also contributes to the Sr–O bond length. To sum up, the difference between the Sr–O bond lengths in different shells is due to the combined effects of structural expansion and structural distortion. Therefore, we obtain varied bond lengths of 2nd and 3rd Sr–O bonds, while no obvious change is found for the length of the 1st Sr–O bond.
3.3. Morphology and analysis of the components
We then employed SEM to study the morphology of the surface of pristine LSM and sintered LSM (LSM was sintered in the presence of Fe-Cr alloy interconnect for 20 h) (Fig. 3). Upon comparing the surface of pristine and sintered LSM, it is observed that limited agglomerated particles are present on both the planar surfaces and the particle size increases slightly during 20 h of sintering treatment. None of the boundaries of the rib region and channel region are observed on the surface of either of the LSM samples.
However, the surface of LSF after heat treatment is distinguishable from that of the pristine LSF. As shown in Fig. 4a, several isolated grown grains are mixed in the LSF coating, owing to the sintering of particles during the stage of elevated temperature. The microstructures of sintered LSF coating are shown in Fig. 4b, 4c, and 4d. The morphologies of the LSF electrode under the rib and channel regions of the interconnect are apparently different (Fig. 4b). The rib region of the electrode surface (Fig. 4d) is covered by a dense film consisting of larger particles owing to the severe deposition of Cr species from the Fe-Cr interconnect. Even in the channel region shown in Fig. 4c, much more agglomerated grains are found in comparison with that found in the fresh LSF electrode.
From Fig. 5, the SEM-EDS results further reveal that the molar ratio of Sr/La on the LSM surface increases from 0.52 to 0.64 after sintering in the presence Fe-Cr for 20 h. In contrast, the molar ratio of Sr/La on the LSF surface increases much more from 0.57 to 1.01 for the channel region and 8.19 for the rib region. This result indicates that Fe accelerates SSS, especially in the presence of Cr species.
The compounds formed on the LSM and LSF electrodes surfaces were characterized by XRD. As shown in Fig. 6, the XRD pattern of sintered LSM sample is similar to that of the pristine one, indicating that no impurity phase appears on the surface of the LSM electrode or that the content of the impurity is below the limit of detection. However, the characteristic peak of SrCrO4 is observed in the XRD pattern of sintered LSF sample, indicating that Sr segregation occurs on the surface of LSF and further leads to a dense film observed in Fig. 4d. Since the formation of the Sr-Cr-O oxide particles is associated with the amount of segregated Sr atoms on the surface of the electrodes, SrCrO4 can be regarded as an indicator of the surface enrichment with Sr.
4 Conclusion
In this study, synchrotron-based X-ray spectroscopy was used to investigate the local structure of Sr atoms in LSM and LSF. The bond length of Sr–O in LSF is longer by 0.03 Å than that in LSM, and the Coulomb interaction between Sr and O in LSF is weaker than that in LSM, indicating that the Sr–O bond in LSF can be broken easily. After 20-h heat treatment in the presence of a Fe-Cr metallic interconnect, SrCrO4 composite was found on the surface of LSF, while no Cr species were evidenced on the surface of LSM. Moreover, contact with Cr species results in much more segregation of Sr on the surface of the LSF electrode.
Factors governing oxygen reduction in solid oxide fuel cell cathodes
. Chemical Rev. 104, 4791-4843 (2004). doi: 10.1021/cr020724oIntermediate temperature solid oxide fuel cells
. Chem. Soc. Rev. 37, 1568-1578 (2008). doi: 10.1039/B612060CElectrode materials and reaction mechanisms in solid oxide fuel cells: a brief review
. J. Solid State Electr. 15, 1007-1040 (2011). doi: 10.1007/s10008-011-1341-8Correlation between thermal expansion and oxide ion transport in mixed conducting perovskite-type oxides for SOFC cathodes
. Solid State Ionics 138, 79-90 (2000). doi: 10.1016/S0167-2738(00)00770-0Sm0.5Sr0.5CoO3 cathodes for low-temperature SOFCs
. Solid State Ionics 149, 11-19 (2002). doi: 10.1016/S0167-2738(02)00131-5A high-performance cathode for the next generation of solid-oxide fuel cells
. Nature 431, 170-173 (2004). doi: 10.1038/nature02863High-performance lanthanum-ferrite-based cathode for SOFC
. Solid State Ionics 176, 457-62 (2005). doi: 10.1016/j.ssi.2004.09.007Electrochemical performance and degradation of (La0.6Sr0.4)0.99CoO3 − δ as porous SOFC-cathode
. Solid State Ionics 179, 1422-1426 (2008). doi: 10.1016/j.ssi.2007.11.010Development of lanthanum strontium manganite perovskite cathode materials of solid oxide fuel cells: a review
. J. Mater. Sci. 43, 6799-6833 (2008). doi: 10.1007/s10853-008-2966-6Enhanced oxygen diffusion in epitaxial lanthanum-strontium-cobaltite thin film cathodes for micro solid oxide fuel cells
. Energy Environ. Sci. 6, 116-20 (2013). doi: 10.1039/C2EE21647GLa0.8Sr0.2FeO3−δ as fuel electrode for solid oxide reversible cells using LaGaO3-based oxide electrolyte
. J. Phys. Chem. C 120,16110-16117 (2016).doi: 10.1021/acs.jpcc.5b12755Issues on development of (La,Sr)MnO3 cathode for solid oxide fuel cells
. J. Power Sources 124, 390-402 (2003). doi: 10.1016/S0378-7753(03)00814-0Origin of Sr segregation at La1-xSrxMnO3surfaces
. Phys. Rev. B 83, 155437 (2011). doi: 10.1103/PhysRevB.83.155437Suppression of Sr surface segregation in La1-xSrxCo1-yFeyO3: a first principles study
. Phy. Chem. Chem. Phys. 15, 489-496 (2013). doi: 10.1039/C2CP43148CSr and Zr diffusion in LSCF/10GDC/8YSZ triplets for solid oxide fuel cells (SOFCs)
. J. Power Sources 258, 281-289 (2014). doi: 10.1016/j.jpowsour.2014.02.046Improved chemical and electrochemical stability of perovskite oxides with less reducible cations at the surface
. Nat. Mater. 15, 1010-1016 (2016).doi: 10.1038/nmat4659Controlling cation segregation in perovskite-based electrodes for high electro-catalytic activity and durability
. Chem. Soc. Rev. 46, 6345-6378 (2017). doi: 10.1039/C7CS00120GMaterials for intermediate-temperature solid-oxide fuel cells
. Ann. Rev. Mater. Res. 44, 365-393 (2014). doi: 10.1146/annurev-matsci-070813-113426The formation of performance enhancing pseudo‐composites in the highly active La1–xCaxFe0.8Ni0.2O3 system for IT‐SOFC application
. Adv. Funct. Mater. 23, 5131-5139 (2013). doi: 10.1002/adfm.201300481Surface termination and subsurface restructuring of perovskite-based solid oxide electrode materials
. Energy Environ. Sci. 7, 3593-3599 (2014). doi: 10.1039/C4EE01497AInvestigation of chemisorbed oxygen, surface segregation and effect of post-treatments on La0.8Sr0.2MnO3 powder and screen-printed layers for solid oxide fuel cell cathodes
. Appl. Surf. Sci. 253, 4641-4648 (2007). doi: 10.1016/j.apsusc.2006.10.019In situ characterization of strontium surface segregation in epitaxial La0.7Sr0.3MnO3 thin films as a function of oxygen partial pressure
. Appl. Phys. Lett. 93:151904 (2008). doi: 10.1063/1.2987731In situ study of activation and de-activation of LSM fuel cell cathodes – Electrochemistry and surface analysis of thin-film electrodes
. J. Catal. 294, 79-88 (2012). doi: 10.1016/j.jcat.2012.07.010Stability of strontium-doped lanthanum manganite cathode in humidified air
. J. Power Sources 248, 196-204 (2014). doi: 10.1016/j.jpowsour.2013.08.098Chromium deposition and poisoning in dry and humidified air at (La0.8Sr0.2)0.9MnO3+δ cathodes of solid oxide fuel cells
. Int. J. Hydrogen Energy 35, 2477-2485 (2010). doi: 10.1016/j.ijhydene.2009.12.185Chromium deposition and poisoning of cathodes of solid oxide fuel cells – A review
. Int. J. Hydrogen Energy 39, 505-531 (2014). doi: 10.1016/j.ijhydene.2013.10.042Chromium deposition and poisoning at La0.6Sr0.4Co0.2Fe0.8O3oxygen electrodes of solid oxide electrolysis cells
. Phys. Chem. Chem. Phys. 17, 1601-1609 (2015). doi: 10.1039/c4cp05110fFormation of chromium containing impurities in (La,Sr)MnO3 solidoxidefuelcell cathodes under stack operating conditions and its effect on performance
. Ceram. Int. 42, 9467-9474 (2016). doi: 10.1016/j.ceramint.2016.03.010Cation size mismatch and charge interactions drive dopant segregation at the surfaces of manganite perovskites
. J. Am. Chem. Soc. 13, 7909-7925 (2013). doi: 10.1021/ja3125349La0.6Sr0.4Co0.2Fe0.8O3−δ cathodes infiltrated with samarium-doped cerium oxide for solid oxide fuel cells
. J. Power Sources 195, 4704-4708 (2010). doi: 10.1016/j.jpowsour.2010.02.049Effect of La0.1Sr0.9Co0.5Mn0.5O3−δ protective coating layer on the performance of La0.6Sr0.4Co0.8Fe0.2O3−δ solid oxide fuel cell cathode
. J. Power Sources 197, 12-19 (2012). doi: 10.1016/j.jpowsour.2011.09.017Efficient electro-catalysts for enhancing surface activity and stability of SOFC cathodes
. Adv. Energy Mater. 3, 1149-1154 (2013). doi: 10.1002/aenm.201200984Development of La0.6Sr0.4Co0.2Fe0.8O3−δcathode with an improved stability via La0.8Sr0.2MnO3-film impregnation
. Int. J. Hydrogen Energ. 38, 5375-5382 (2013). doi: 10.1016/j.ijhydene.2013.02.091Highly chromium contaminant tolerant BaO infiltrated La0.6Sr0.4Co0.2Fe0.8O3-[small delta] cathodes for solid oxide fuel cells
. Phys. Chem. Chem. Phys. 17, 4870-4874 (2015). doi: 10.1039/c4cp04172kThe Fe K-edge X-ray absorption characteristics of La1−xSrxFeO3−δ prepared by solid state reaction
. Mater. Res. Bull. 44, 1397-1404 (2009). doi: 10.1016/j.materresbull.2008.11.026X-ray absorption investigation of the valence state and electronic structure of La1−xCaxCoO3−δ in comparison with La1−xSrxCoO3−δ and La1−xSrxFeO3−δ
. J. Solid State Chem. 184, 3163-3171 (2011). doi: 10.1016/j.jssc.2011.09.027Rational SOFC material design: new advances and tools
. Materials Today14, 534-546 (2011). doi: 10.1016/S1369-7021(11)70279-6IFEFFIT: interactive XAFS analysis and FEFF fitting
. J. Synchrotron Radiat. 8, 322-324 (2001). doi: 10.1107/S0909049500016964Local structural analysis for oxide ion transport in La0.6Sr0.4FeO3−δ cathodes
. J. Mater. Chem., 21, 14013-14019 (2011). doi: 10.1039/C1JM11358E