Cation Diffusion and Segregation at the Interface between Samarium

Subscriber access provided by READING UNIV

Article

Cation diffusion and segregation at the interface between samarium-doped ceria and LSCF or LSFCu cathodes investigated with X-ray microspectroscopy Francesco Giannici, Giovanna Canu, Alessandro Chiara, Marianna Gambino, Chiara Aliotta, Alessandro Longo, Vincenzo Buscaglia, and Antonino Martorana ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b13377 • Publication Date (Web): 01 Dec 2017 Downloaded from http://pubs.acs.org on December 3, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Cation diffusion and segregation at the interface between samarium-doped ceria and LSCF or LSFCu cathodes investigated with X-ray microspectroscopy

Francesco Giannici$,*, Giovanna Canu§, Alessandro Chiara$, Marianna Gambino$, Chiara Aliotta$, Alessandro Longo£,%, Vincenzo Buscaglia§, Antonino Martorana$

$

Dipartimento di Fisica e Chimica, Università di Palermo, Viale delle Scienze, I-90128 Palermo,

Italy §

Institute of Condensed Matter Chemistry and Technologies for Energy (CNR-ICMATE),

Consiglio Nazionale delle Ricerche, via De Marini 6, I-16149 Genova, Italy £

Istituto per lo Studio dei Materiali Nanostrutturati (CNR-ISMN), Consiglio Nazionale delle

Ricerche, via La Malfa 153, I-90146 Palermo, Italy %

Dutch-Belgian Beamline (DUBBLE), ESRF – The European Synchrotron, CS40220, F-38043

Grenoble, France.

*

corresponding author. Email: [email protected]; phone +39(0)9123897927; address

Dipartimento di Fisica e Chimica, Università di Palermo, Viale delle Scienze, I-90128 Palermo, Italy

Keywords: SOFC, cathode, ceria, electrolyte, X-ray microspectroscopy, XANES, compatibility

Abstract The chemical compatibility between electrolytes and electrodes is an extremely important aspect governing the overall impedance of solid-oxide cells. Since these devices work at elevated temperatures, they are especially prone to cation interdiffusion between the cell components, possibly resulting in secondary insulating phases. In this work, we applied X-ray microspectroscopy ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 27

to study the interface between a samarium-doped ceria electrolyte and lanthanum ferrite cathodes (La0.9Sr0.1Fe0.85Cu0.15O3 – LSCF; La0.4Sr0.6Fe0.8Co0.2O3 - LSFCu), at a submicrometric level. This technique allows to combine the information about the diffusion profiles of cations on the scale of several micrometers, together with the chemical information coming from space-resolved X-ray absorption spectroscopy. In SDC-LSCF bilayers, we find that the prolonged thermal treatments at 1150 °C bring about the segregation of samarium and iron in micrometer-sized perovskite domains. In both SDC-LSCF and SDC-LSFCu bilayers, cerium diffuses into the cathode perovskite lattice Asite as a reduced Ce3+ cation, while La3+ is easily incorporated in the ceria lattice, reaching 30 at.% in the ceria layer in contact with LSFCu.

1. Introduction Solid-oxide fuel cells (SOFC) converting hydrogen and hydrocarbons from renewable sources to electricity with high efficiency, have been at the forefront of applied materials chemistry research for the last decades.1 In particular, most recent efforts have been devoted to lowering the operating temperature of the devices to the so-called intermediate temperature (IT) range, around 600 °C. Efficient IT-SOFC systems with good performance in the intermediate temperature range have been acknowledged as “a promise for the 21st century”:2 while the availability of commercial SOFC systems is still lower than what envisaged in the past, the fundamental research activity has yielded much knowledge in recent years, and a remarkable variety as what concerns the formulation of electrode and electrolyte materials, and their interaction.3 The chemical and mechanical compatibility of the different components of a SOFC is a key point in the stability of the device, since they are held in intimate contact at high temperatures during operation. Furthermore, the high temperature required for the cell fabrication may also lead to unwanted chemical reactions between the components. The fundamental mechanisms behind the long-term durability of SOFC systems have been reviewed by Yokokawa et al.4 A well-known example is the reaction between yttria-stabilized zirconia (YSZ) and lanthanum-containing ACS Paragon Plus Environment

2

Page 3 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


perovskites used as electrode, which form an insulating layer of La2Zr2O7 at the interface between the two materials. Even when the electrolyte and electrodes do not form a secondary phase, their interface is still one of the key aspects that determine the overall performance of the SOFC, influencing not only the kinetics of electrochemical reactions at the gas-solid phase boundaries, but also the electrolyte and electrode bulk conductivity.5 In any case, the possible interdiffusion of cations affects the conduction properties of the materials, and it also modifies the thermal expansion of the various layers, therefore affecting the mechanical stability of the interfaces.4 Recently, we presented the application of X-ray microspectroscopy using synchrotron radiation to the study of SOFC materials, investigating the interface between the proton-conducting electrolyte Ca-doped LaNbO4, and (La,Sr)MnO3 cathodes after prolonged thermal annealing. The X-ray fluorescence maps provide information about the cation distribution, while space-resolved X-ray absorption spectroscopy allows the mapping of chemical and coordination state of cations.6 In this work, we apply X-ray microspectroscopy to a system of broader interest, involving an oxygen ion-conducting electrolyte, samarium-doped ceria (SDC), in contact with a lanthanum strontium cobalt ferrite (LSCF) cathode. Trivalent-doped ceria is an ubiquitous ceramic material, with appealing properties in oxygen ion conduction and catalysis.7,8 In SOFC research, it is routinely used as a protective layer between the state-of-the-art YSZ electrolytes and lanthanumcontaining cathodes, in order to reduce interface reactivity.9 It has also been extensively studied as an electrolyte in the IT range,10 due to its lower activation energy and higher ionic conductivity compared to YSZ, and in composite cathodes.11-13 Among ceria-based electrolyte materials, SDC is one of the best performing, due to high purely ionic conductivity and even distribution of samarium in the ceria matrix.14,15 LSCF is a typical SOFC cathode, having attractive ionic and electronic conductivity, and it has been extensively studied in the past in combination with zirconia or ceriabased electrolytes.16 Alongside LSCF, we also investigated the compatibility between SDC and the relatively less studied La0.9Sr0.1Fe0.85Cu0.15O3 (LSFCu): this latter compound has been recently proposed as a ACS Paragon Plus Environment

3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 27

cobalt-free cathode for IT-SOFC, displaying interesting performance as a cathode either alone or in combination with Gd-doped ceria.17-19 In particular, classical chemical compatibility tests with powder mixtures of La0.6Sr0.4Fe0.8Cu0.2O3 and SDC showed chemical compatibility up to a temperature of 1100 °C for 10 h.17

2. Experimental Ce0.8Sm0.2O2-x (SDC) and La0.9Sr0.1Fe0.85Cu0.15O3-x (LSFCu) powders were prepared by solution combustion

synthesis.

Ce(NO3)3·6H2O,

Sm(NO3)3·6H2O,

La(NO3)3·6H2O,

Sr(NO3)2,

Cu(NO3)2·2.5H2O and Fe(NO3)3·9H2O (Sigma–Aldrich) were dissolved in water in a stainless steel beaker, adding citric acid as combustion fuel, and NH3 to set pH = 6. The combustion reactions were carried out in stoichiometric condition, setting a fuel-to-metals cation molar ratio equal to 2. NH4NO3 was added to reaction mixture as oxidant additive.20 The solution was stirred at 80 °C until the gel formation, then the temperature was further increased until combustion. The resulting powders were eventually fired in air at 1000 °C for 5 hours. La0.6Sr0.4Fe0.8Co0.2O3-x (LSCF) powders (Marion Technologies) were used as received. The SDC powders were isostatically pressed (1500 bar) and sintered at 1550 °C for 4 hours, obtaining a cylindrical pellet with relative density of 96%. The resulting pellet was cut in disks of about 10 mm diameter and 1 mm thickness. These were then polished, surrounded by either LSCF or LSFCu powder, and uniaxially pressed at 7 tons in a 1-inch die. After pressing, the bilayers were annealed at 1150 °C for either 12 or 72 hours. The samples were then embedded in resin, cut to expose the interface, and the cross sections were mechanically polished down to 1 µm. Phase purity of the starting electrode materials powders and of the SDC ceramic was ensured by X-ray diffraction. Scanning X-ray microscopy measurements were carried out at the SXM-II end station of the ID21 beamline of ESRF (Grenoble, France).21 The incident beam was monochromatized with a doublecrystal silicon monochromator and focused down to a beam size of 850 x 400 nm2 (H x V) using ACS Paragon Plus Environment

4

Page 5 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Kirkpatrick-Baez mirrors, with a flux between 1010 and 2·1010 ph/s. The beam energy was scanned in the vicinity of the Ce L3-edge (5.72 keV) or the Fe K-edge (7.11 keV), and microXRF and microXAS spectra were collected in fluorescence mode using a SDD detector. The sample was mounted with the interface parallel to the horizontal plane, to exploit the smaller beam size in the vertical direction, and X-ray scanning microscopy was performed by moving the sample in 2 dimensions with piezoelectric motors with respect to the fixed X-ray beam. Data reduction and analysis was performed with PyMca.22 All maps were corrected for the incident beam intensity and for detector dead-time, and analyzed using a self-consistent XRF fitting taking into account self-absorption and matrix effects. We recently demonstrated the accuracy of microXRF data acquired with this setup for the elemental quantification of cations in dense LaMnO3/LaNbO4 matrices.6

3. Results and discussion The high temperature thermal treatments (either 12 or 72 hours) are carried out to simulate the conditions of prolonged operation, and also the thermal treatment for cell preparation, with the aim to investigate the resulting cation diffusion. The microXRF maps provide information on the distribution of cations across the interface after thermal treatments. The samples were investigated at the Ce L3-edge and at the Fe K-edge: at each edge, the elements emitting an X-ray fluorescence signal at lower energy lower or equal than the incident beam are detected. For the samples under investigation, these are respectively (Sr, La, Ce) for the Ce L3-edge, and (Sr, La, Ce, Sm, Fe) for the Fe K-edge. When not specified otherwise, the concentration profiles are calculated from the concentration maps by averaging over the whole map width, perpendicularly to the electrode/electrolyte boundary. When considering the concentration profiles after diffusion, it is not possible to fix a definite boundary that separates the electrode from the electrolyte because: a) the concentration profiles are asymmetric; b) the diffusion lengths vary for the different cations. Thus, the following practical ACS Paragon Plus Environment

5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 27

approach is used: for each concentration profile, a midpoint is found on the x-axis, defined as the position where the amount of cation diffused from the first phase equals the amount diffused into the second phase (for mass conservation). For different concentration profiles, different midpoints are found by using the above procedure: the spread of these points, spanning a few microns, is then loosely defined as “the interface”. The origin of the x-axis in all the concentration plots displayed in the following is placed in the middle of this interface region.

3.1. LSCF-SDC In general, doped ceria does not show the marked reactivity displayed by zirconia towards lanthanum perovskite cathodes. Interdiffusion between ceria and lanthanum strontium cobaltites (LSC) or ferrites (LSF) may still take place, but it has been studied only by a few groups, and a somewhat erratic evidence has been obtained due to the variety of compounds and techniques used.23-27 In the total fluorescence maps of the LSCF-SDC samples, shown in figures 1 and 2, the interface between SDC and LSCF looks sharp after both 12h and 72h, without evident formation of secondary phases: the diffusion of cations is however more pronounced after 72h. In both cases, the concentration profiles of cerium and lanthanum are complementary. In the case of lanthanum diffusing towards SDC, this means that it most likely replaces cerium in the fluorite lattice of SDC, resulting in a (Sm, La) co-doped ceria: according to the literature, in fact, La3+ solubility in ceria is very high, up to 50 at%.28

ACS Paragon Plus Environment

6

Page 7 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1 – LSCF-SDC_12h at the Ce L3-edge. Left: heatmap of the total fluorescence intensity and color scale; right: concentration profiles of cerium (black), lanthanum (red) and strontium (blue).

Figure 2 – LSCF-SDC_72h at the Ce L3-edge. Left: heatmap of the total fluorescence intensity; right: concentration profiles of cerium (black), lanthanum (red) and strontium (blue).

For cerium diffusion into LSCF, there are different possibilities: cerium could either substitute lanthanum as Ce3+ in the A-site of the perovskite cathode, or be placed in the perovskite B-site as a Ce4+ cation. We found no experimental evidence of secondary phases containing cerium on the LSCF side. In an XRD study involving SDC as a protective layer between LSF and YSZ, Martínez-


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 27

Amesti et al. put forward that even if no secondary phases were formed, Ce is incorporated in the A-site of the LSF perovskite.26 We find that the concentration of cations is perturbed inside LSCF at considerable distances from the interface: considering both Figures 1 and 2, it can be seen that the concentration of lanthanum in LSCF is appreciably lower than the stoichiometric amount around 10 µm from the interface. At the same time, cerium diffuses inside LSCF for a correspondingly long distance. Another interesting feature of the LSCF-SDC couple is that the migration of cerium and lanthanum is not significantly affected by the duration of the thermal treatment, i.e. most of the cation displacement already occurs after the first 12 hours of annealing. The main difference between samples treated for 12 and 72 hours lies in the shape of the concentration profiles, which are smooth after 12 h, but present a distinct change in slope in LSCF at 1 µm near the interface after 72 h. This layer between -1 µm and 2 µm in Figure 2 (72 h) represents a region in which cation diffusion proceeds with a different speed, and therefore possibly with different mechanisms, with respect to both the SDC and LSCF layer: this could then be tentatively identified as a reaction zone. Previous studies investigated the migration of strontium in various electrolyte/electrode couples. It was generally reported that Sr2+ is very mobile compared to La3+,27,29 and it may segregate out of LSCF in different ways: e.g. at the surface of LSCF electrodes after heating in air,30 under cathodic polarization,31 or at the LSCF/SDC interface.24 While Baumann et al. attributed the high activity towards oxygen reduction to the surface segregation,31 other authors claimed that strontium segregation is instead detrimental for the cell performance.24 In the annealed LSCF/SDC samples, we do not observe any peculiar segregation of Sr2+ at the micrometric level in LSCF or SDC. Strontium diffuses towards SDC in a similar fashion as lanthanum, although to a higher extent, and it is found in appreciable amount around -5 µm, and can still be detected at -10 µm (Figure 2). Although not widely studied, strontium doping in ceria was investigated recently;32 strontium and cerium may also form the SrCeO3 perovskite (tolerance factor = 0.79). The latter possibility is however not corroborated by the microXANES spectra taken at the ACS Paragon Plus Environment

8

Page 9 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Ce L3-edge, which are shown in Figures 3 and 4. All spectra present a similar overall shape, typical of CeO2 and tetravalent cerium compounds. The XANES spectrum of pure SrCeO3 presents shallower peaks and a more pronounced shoulder around 5726 eV,33 features which are not found in the microXANES spectra taken near the interface. The substantially unmodified features in the microXANES spectra indicate that cerium is maintained mainly in a Ce4+ state in the investigated points. The typical XANES feature of reduced Ce3+ is a sharp XANES peak at 5726 eV, which is not visible in any of the spectra: however, the slight shift of the first main peak towards lower energy hints at the presence of a small fraction of Ce3+ in points 6 and 7 of Figure 4 (in the cathode region), where Ce3+ may be incorporated in the A-site of the LSCF perovskite.

Figure 3 – LSCF-SDC_12h at the Ce L3-edge. Left: concentration map of cerium (blue); right: Ce L3-edge microXANES spectra measured at different points shown in the left panel.


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 27

Figure 4 – LSCF-SDC_72h at the Ce L3-edge. Left: concentration map of cerium (blue); right: Ce L3-edge microXANES spectra measured at different points shown in the left panel.

In LSCF/GDC diffusion couples, Uhlenbruck et al. reported a Gd3+ enrichment in the LSCF layer in contact with GDC, which was proposed to be detrimental for the chemical stability and conductivity of the cathode.25 We find that samarium diffuses out of SDC in a different way than cerium, as can be seen in the concentration profiles of Figures 5 and 6. In particular, samarium depletion in SDC is evident already at -2 µm, while cerium concentration remains at its maximum level before decreasing at the interface. A slight cerium enrichment can also be seen in SDC right before the interface. On the LSCF side, iron and strontium maintain their bulk concentration very close to the interface (about 3 µm), while lanthanum concentration decreases to about 80% of its bulk value, being substituted by samarium and cerium.


10

Page 11 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5 – LSCF-SDC_12h at the Fe K-edge. Left: heatmap of the total fluorescence intensity; right: concentration profiles of cerium (black), lanthanum (red), strontium (blue), samarium (orange) and iron (green).

Figure 6 – LSCF-SDC_12h at the Fe K-edge. Left: heatmap of the total fluorescence intensity; right: concentration profiles of cerium (black), lanthanum (red), strontium (blue), samarium (orange) and iron (green). This is a portion of the map shown in Figure 5, measured at higher resolution.


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 27

As shown in Figure 7, the Fe K-edge microXANES spectra of LSCF-SDC_12h do not show significant differences in the various points investigated. This suggests that iron maintains its pristine oxidation state (around +3.4) and coordination in the perovskite B-site, and it is not incorporated in the fluorite lattice of SDC. According to stoichiometry, the average oxidation state of the LSCF B-site is only determined by the concentration of strontium in the A-site, since in air the concentration of oxygen vacancies is virtually zero.34 In such conditions, it was reported that the oxidation state of cobalt and iron is the same.35 At high temperature and low oxygen partial pressure, cobalt may undergo reduction, compensated by the creation of oxygen vacancies, while iron maintains its valence. In LSCF-SDC_12h, the relative concentration of Sr in the ferrite phase remains the same in the different regions of the sample, and so the Fe microXANES spectra are unchanged.

Figure 7 – LSCF-SDC_12h at the Fe K-edge. Left: Concentration map of Fe; right: Fe K-edge microXANES spectra measured at different points shown in the left panel.

The LSCF-SDC_72h sample is best discussed by considering the elemental maps, shown in Figures 8 and 9, rather than the profiles, which are plotted in Figure 10. These show the concentration of cerium, iron and samarium: as a consequence of iron diffusion, in the SDC region iron and samarium cluster together in small regions about 3-7 µm wide, placed at around 20 µm far from the ACS Paragon Plus Environment

12

Page 13 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


interface, and in these regions cerium is correspondingly depleted. It is also worth noting that in such regions neither lanthanum nor strontium are detected. The concentration of iron in ceria outside such Sm-rich clusters is overall negligible, hinting at a substantial stabilization of iron in a newly formed secondary phase. Since the samarium and iron concentrations are roughly equal in the (Sm, Fe)-rich zones, the most likely product of the iron diffusion into SDC is the SmFeO3 perovskite (tolerance factor = 0.92). Samarium ferrites are primarily oxygen ion conductors tested as gas sensors and SOFC anodes, with much lower p-type electronic conductivity.36,37 This hypothesis is corroborated by the analysis of the microXANES spectra of iron in the LSCF and SDC regions of LSCF-SDC_72h (see Figure 11): the near-edge features are roughly the same in the various points investigated, indicating that iron keeps its coordination environment (perovskite Bsite). Upon further inspection, however, interesting correlations can be drawn between edge features and distance from the interface.


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 27

Figure 8 – LSCF-SDC_72h at the Fe K-edge. Clockwise from top left: heatmap of the total fluorescence intensity, and concentration maps of cerium (green), samarium (blue), iron (red), strontium (cyan) and lanthanum (pink).

Figure 9 – LSCF-SDC_72h at the Fe K-edge. Clockwise from top left: heatmap of the total fluorescence intensity and concentration maps of cerium (green), samarium (blue), iron (red), strontium (cyan) and lanthanum (pink).

The Fe XANES pre-edge peak around 7115 eV corresponds to a transition between 1s and 3d states of iron, and its intensity is proportional to the oxidation state. The main edge peak at 7130 eV is due to a 1s 4p transition, and its intensity decreases with the oxidation state instead.38,39 In order to evaluate the oxidation state of iron in the various points from the microXANES spectra, we calculated the ratio between the intensity of these two features (plotted in the Supporting Information).


14

Page 15 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In general, the pre-edge peak intensity in LSCF-SDC_72h is lower than in LSCF-SDC_12h, even in the deep LSCF region far from the interface: this is a further evidence of the incipient formation of a Sm- and Fe-rich perovskite phase in which the Fe oxidation state is closer to +3. The relative intensity of the pre-edge peak decreases to about two-thirds closer to the LSCF/SDC interface, which indicates a further lowering of the iron oxidation state. A possible explanation is that after interdiffusion with SDC, LSCF demixes in different perovskite phases. The driving force for such demixing is the ultimate segregation of Sm3+ and Fe3+ as shown in the elemental maps. In such regions, a perovskite phase enriched in Sm and Fe has a B-site whose valence is close to +3.

Figure 10 – LSCF-SDC_72h at the Fe K-edge. Left: heatmap of the total fluorescence intensity; right: concentration profiles of cerium (black), lanthanum (red), strontium (blue), samarium (orange) and iron (green).


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 27

Figure 11 – LSCF-SDC_72h at the Fe K-edge. Left: Concentration map of Fe. Right: Fe K-edge microXANES spectra measured at different points shown in the left panel.

3.2. LSFCu-SDC Reactivity of LSFCu electrode material with SDC electrolyte is very pronounced: the sample LSFCu-SDC_72h shows a dramatic diffusion of lanthanum, and a corresponding depletion of samarium and cerium, over a distance of tens of microns, in the SDC region near the interface, as shown in the concentration profiles of Figures 12 and 13. The inverse correlation of lanthanum on one hand, and samarium and cerium on the other, is evident from the elemental maps shown in Figure 14. To find an undisturbed portion of bulk SDC, one has to move about 50 µm away from the interface. The concentration profiles of La and Ce show a similar peculiar shape, which is composed of two different regions: going from bulk LSFCu to SDC, a steep descent marks the boundary of a reaction zone (about 5 µm wide); after that, on the SDC side, the concentration profile changes into a gentle slope resembling the Gaussian-shaped profile for limited source diffusion. Strontium behaves very differently from lanthanum, as its concentration profile only shows a simple diffusion in the reaction zone of about 5 µm.


16

Page 17 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 12 – LSFCu-SDC_72h at the Ce L3-edge. Left: heatmap of the total fluorescence intensity; right: concentration profiles of cerium (black), lanthanum (red) and strontium (blue).

Figure 13 – LSFCu-SDC_72h at the Fe K-edge. Left: heatmap of the total fluorescence intensity; right: concentration profiles of cerium (black), lanthanum (red), strontium (blue), samarium (orange) and iron (green).


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 27

Figure 14 – LSFCu-SDC_72h at the Fe K-edge. Clockwise from top left: heatmap of the total fluorescence intensity and concentration maps of cerium (green), samarium (blue), iron (red), strontium (cyan) and lanthanum (pink).

The peculiar concentration profile of La3+ may be explained by hypothesizing the following steps for its diffusion into SDC: 1) the formation of an interfacial layer of fixed composition; 2) the steady Fickian diffusion from the interfacial layer towards SDC. The concentration profile of La3+ resembles that observed between La0.9Sr0.1Ga0.8Mg0.2O3 and Ce0.8Gd0.2O2, suggesting that a similar mechanism of fast La3+ insertion into ceria occurs in these cases, but not in LSCF/SDC.40 Apparently, the diffusion of iron and strontium through the interfacial layer is much less favored, as they are not found in significant concentration in SDC. The same can be said for diffusion into LSFCu, as both the samarium and cerium profiles fall sharply a few microns after the boundary. The microXANES spectra at the Ce L3-edge show that the chemical state of cerium is modified between SDC and the interfacial reaction zone. The relevant spectra are reported in Figure 15. The


18

Page 19 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


XANES spectrum of CeO2 consists of three edge features: a shoulder at 5726 eV, and two main peaks at 5730 and 5737 eV: these are labeled A, B and C, respectively.14 As mentioned in the previous section, the XANES spectrum of trivalent cerium compounds features only a single intense peak at 5726 eV.41 Therefore, the intensity of the A peak can be used as a qualitative indication of the concentration of Ce3+. The A/B intensity ratio grows significantly when moving from SDC towards LSFCu (see Supporting Information): the A/C ratio shows the same behavior, since both the B and C peaks are due to the Ce4+ character. The presence of reduced Ce3+ in LSFCu indicates that Ce3+ substitutes La3+ in the perovskite A-site: this substitution arguably proceeds faster in LSFCu/SDC bilayer compared with LSCF/SDC, and it could be the driving force for the large La3+ diffusion towards SDC. Also in this case, as it was observed for the LSCF-SDC couple, the microXANES spectra at the Fe K-edge (shown in Figure 16) indicate a substantial unmodified oxidation state and coordination environment of iron in the different points investigated. The XANES edge and pre-edge features, compared with the La1-xSrxFeO3 series,38,39 are compatible with an oxidation state that is very close to +3.1: this also implies that the relative negative charge of Sr2+ in the A-site is compensated both by iron and copper, sharing the same oxidation state. However, the presence of some Cu3+ cannot be probed directly with the present data. The Fe pre-edge peak shows a slight but abrupt increase in the interface region (see Supporting Information), which means that the oxidation state of iron also increases. This is not correlated to an increment in the Sr concentration, so its chemical origin is currently unclear. Differently from the case of SDC-LSCF bilayers, however, no areas enriched in Sm and Fe are observed.


19


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 27

Figure 15 – LSFCu-SDC_72h at the Ce L3-edge. Left: concentration map of cerium (blue); right: Ce L3-edge microXANES spectra measured at different points shown in the left panel.

Figure 16 – LSFCu-SDC_72h at the Fe K-edge. Left: concentration map of iron (blue); right: Fe Kedge microXANES spectra measured at different points shown in the left panel.

4. Conclusions Scanning X-ray microscopy with synchrotron radiation is a powerful technique, combining elemental microanalysis (microXRF) with spatially-resolved information about the oxidation state and chemical environment of a given element (microXANES, microEXAFS).


20

Page 21 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


We have applied this technique to evaluate the chemical compatibility of important SOFC materials, i.e. two different lanthanum ferrites cathodes in contact with SDC electrolyte, after prolonged annealing at high temperature. Our results demonstrate that SDC shows some interdiffusion with the commonly used LSCF cathode, and extensive reaction with LSFCu, recently proposed as a cobalt-free cathode. In both cases, microXANES spectra show that cerium can be incorporated in the A-site of the perovskite phase, while iron cannot enter the fluorite lattice of SDC. Most interestingly, in LSCF-SDC bilayers we find a peculiar segregation of Fe and Sm in the SDC layer after prolonged thermal annealing, most likely due to the formation of a SmFeO3 phase. LSFCu, recently proposed as a cobalt-free alternative to LSCF and other electrode materials, shows a marked reactivity with SDC, involving a major diffusion of La3+, but it does not feature micrometric segregation of cations as in the case of LSCF/SDC. We can conclude, in general, that the two cathode materials examined here are not completely stable when annealed in contact with the SDC electrolyte for an extended period of time. While the temperature used in this work is higher than the operating temperature of an IT-SOFC device, the observed effects may be brought about by the thermal treatments during fabrication, and long term operation. These results underline the importance of microprobe techniques for the evaluation of materials compatibility. In the search for a stable interface between SOFC electrolyte and electrodes, X-ray microspectroscopy studies can be successfully used to determine the origin of performance degradation in various electrolyte/electrode couples of SOFC devices.

Acknowledgements We acknowledge funding through MIUR projects Futuro in Ricerca “INnovative Ceramic and hYbrid materials for proton conducting fuel cells at Intermediate Temperature: design, characterization and device assembly (INCYPIT)”, PON R&C 2007-1013 “Tecnologie ad alta Efficienza per la Sostenibilità Energetica ed ambientale On-board (TESEO)”, PRIN2010 “Celle a ACS Paragon Plus Environment

21


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 27

combustibile ad ossido solido operanti a temperatura intermedia alimentate con biocombustibili (BIOITSOFC)”. We acknowledge the ESRF for provision of beamtime, and we thank the staff of beamline ID21 for assistance during the measurements. We thank Dr. Wout De Nolf (ESRF) for useful discussions and assistance during data analysis.

Supporting Information. Spatial resolution of the X-ray microbeam. Comparison of the concentration profiles of LSCF-SDC_12h and LSCF-SDC_72h. Ce L3-edge and Fe K-edge microXANES data analysis. Concentration maps of Ce and La measured at the Ce L3-edge in LSFCu-SDC_72h.

References (1) Mahato, N.; Banerjee, A.; Gupta, A.; Omar, S.; Balani, K. Progress in material selection for solid oxide fuel cell technology: A review. Prog. Mater. Sci. 2015, 72, 141–337. (2) Huijsmans, J. P. P.; Van Berkel, F. P. F.; Christie, G. Intermediate temperature SOFC – a promise for the 21st century. J. Power Sources 1998, 71, 107–110. (3) Kilner, J. A.; Burriel, M. Materials for Intermediate-Temperature Solid-Oxide Fuel Cells. Annu.

Rev. Mater. Res. 2014, 44, 365–393. (4) Yokokawa, H.; Tu, H.; Iwanschitz, B.; Mai, A. Fundamental mechanisms limiting solid oxide fuel cell durability. J. Power Sources 2008, 182, 400–412. (5) Liu, M.; Wu, Z. Significance of interfaces in solid-state cells with porous electrodes of mixed ionic-electronic conductors. Solid State Ionics 1998, 107, 105–110. (6) Giannici, F.; Canu, G.; Gambino, M.; Longo, A.; Salomé, M.; Viviani, M.; Martorana, A. Electrode–Electrolyte Compatibility in Solid-Oxide Fuel Cells: Investigation of the LSM–LNC Interface with X-Ray Microspectroscopy. Chem. Mater. 2015, 27, 2763–2766. (7) Inaba, H.; Tagawa, H. Ceria-based solid electrolytes. Solid State Ionics 1996, 83, 1–16.


22

Page 23 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(8) Vinodkumar, T.; Rao, B. G.; Reddy, B. M. Influence of isovalent and aliovalent dopants on the reactivity of cerium oxide for catalytic applications. Catal. Today 2015, 253, 57–64. (9) Mai, A.; Haanappel, V. A. C.; Uhlenbruck, S.; Tietz, F.; Stöver, D. Ferrite-Based perovskites as cathode materials for anode-Supported solid oxide fuel cells. Solid State Ionics 2005, 176, 1341– 1350. (10) Steele, B. C. H. Appraisal of Ce1−yGdyO2−y/2 electrolytes for IT-SOFC operation at 500°C.

Solid State Ionics 2000, 129, 95–110. (11) Leng, Y.; Chan, S.; Liu, Q. Development of LSCF–GDC composite cathodes for lowTemperature solid oxide fuel cells with thin film GDC electrolyte. Int. J. Hydrogen Energ. 2008, 33, 3808–3817. (12) Perry Murray, E.; Barnett, S. A. (La,Sr)MnO3–(Ce,Gd)O2−x composite cathodes for solid oxide fuel cells. Solid State Ionics 2001, 143, 265–273. (13) Kharton, V. V.; Marques, F. M. B.; Atkinson, A. Transport properties of solid oxide electrolyte ceramics: a brief review. Solid State Ionics 2004, 174, 135–149. (14) Giannici, F.; Gregori, G.; Aliotta, C.; Longo, A.; Maier, J.; Martorana, A. Structure and Oxide Ion Conductivity: Local Order, Defect Interactions and Grain Boundary Effects in Acceptor-Doped Ceria. Chem. Mater. 2014, 26, 5994–6006. (15) Artini, C.; Pani, M.; Carnasciali, M. M.; Plaisier, J. R.; Costa, G. A. Lu-, Sm-, and Gd-Doped Ceria: A Comparative Approach to Their Structural Properties. Inorg. Chem. 2016, 55, 10567– 10579. (16) Perry Murray, E.; Sever, M. J.; Barnett, S. A. Electrochemical performance of (La,Sr)(Co,Fe)O3–(Ce,Gd)O3 composite cathodes. Solid State Ionics 2002, 148, 27–34. (17) Zhou, Q.; Xu, L.; Guo, Y.; Jia, D.; Li, Y.; Wei, W. La0.6Sr0.4Fe0.8Cu0.2O3−δ perovskite oxide as cathode for IT-SOFC. Int. J. Hydrogen Energ. 2012, 37, 11963–11968.


23


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 27

(18) Zurlo, F.; Di Bartolomeo, E.; D'Epifanio, A.; Felice, V.; Natali Sora, I.; Tortora, L.; Licoccia, S. La0.8Sr0.2Fe0.8Cu0.2O3−δ as “cobalt-free” cathode for La0.8Sr0.2Ga0.8Mg0.2O3−δ electrolyte. J. Power

Sources 2014, 271, 187–194. (19) Zurlo, F.; Natali Sora, I.; Felice, V.; Luisetto, I.; D'Ottavi, C.; Licoccia, S.; Di Bartolomeo, E. Copper-Doped lanthanum ferrites for symmetric SOFCs. Acta Materialia 2016, 112, 77–83. (20) Aliotta, C.; Liotta, L. F.; La Parola, V.; Martorana, A.; Muccillo, E. N. S.; Muccillo, R.; Deganello, F. Ceria-based electrolytes prepared by solution combustion synthesis: The role of fuel on the materials properties. Appl. Catal. B 2016, 197, 14–22. (21) Salomé, M.; Cotte, M.; Baker, R.; Barrett, R.; Benseny-Cases, N.; Berruyer, G.; Bugnazet, D.; Castillo-Michel, H.; Cornu, C.; Fayard, B.; Gagliardini, E.; Hino, R.; Morse, J.; Papillon, E.; Pouyet, E.; Rivard, C.; Solé, V. A.; Susini, J.; Veronesi, G. The ID21 Scanning X-Ray Microscope at ESRF. J. Phys. Conf. Series 2013, 425, 182004. (22) Solé, V. A.; Papillon, E.; Cotte, M.; Walter, P.; Susini, J. A multiplatform code for the analysis of energy-Dispersive X-Ray fluorescence spectra. Spectrochim. Acta B 2007, 62, 63–68. (23) Jordan, N.; Assenmacher, W.; Uhlenbruck, S.; Haanappel, V.; Buchkremer, H.; Stöver, D.; Mader, W. Ce0.8Gd0.2O2−δ protecting layers manufactured by physical vapor deposition for ITSOFC. Solid State Ionics 2008, 179, 919–923. (24) Simner, S. P.; Bonnett, J. F.; Canfield, N. L.; Meinhardt, K. D.; Sprenkle, V. L.; Stevenson, J. W. Optimized Lanthanum Ferrite-Based Cathodes for Anode-Supported SOFCs. Electrochem.

Solid-State Lett. 2002, 5, A173–A175. (25) Uhlenbruck, S.; Moskalewicz, T.; Jordan, N.; Penkalla, H.-J.; Buchkremer, H. Element interdiffusion at electrolyte–cathode interfaces in ceramic high-temperature fuel cells. Solid State

Ionics 2009, 180, 418–423. (26) Martínez-Amesti, A.; Larrañaga, A.; Rodríguez-Martínez, L. M.; Aguayo, A. T.; Pizarro, J. L.; Nó, M. L.; Laresgoiti, A.; Arriortua, M. I. Reactivity between La(Sr)FeO3 cathode, doped CeO2


24

Page 25 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


interlayer and yttria-stabilized zirconia electrolyte for solid oxide fuel cell applications. J. Power

Sources 2008, 185, 401–410. (27) Sakai, N.; Kishimoto, H.; Yamaji, K.; Horita, T.; Brito, M. E.; Yokokawa, H. Interface Stability of Perovskite Cathodes and Rare-Earth Doped Ceria Interlayer in SOFCs. J. Electrochem.

Soc. 2007, 154, B1331–B1337. (28) Aliotta, C.; Liotta, L. F.; La Parola, V.; Martorana, A.; Muccillo, E. N. S.; Muccillo, R.; Deganello, F. Ceria-Based electrolytes prepared by solution combustion synthesis: The role of fuel on the materials properties. Appl. Catal. B 2016, 197, 14–22. (29) Lein, H.; Wiik, K.; Grande, T. Kinetic demixing and decomposition of oxygen permeable membranes. Solid State Ionics 2006, 177, 1587–1590. (30) Oh, D.; Gostovic, D.; Wachsman, E. D. Mechanism of La0.6Sr0.4Co0.2Fe0.8O3 cathode degradation. J. Mater. Res. 2012, 27, 1992–1999. (31) Baumann, F. S.; Fleig, J.; Konuma, M.; Starke, U.; Habermeier, H.-U.; Maier, J. Strong Performance Improvement of La0.6Sr0.4Co0.2Fe0.8O3−δ SOFC Cathodes by Electrochemical Activation. J. Electrochem. Soc. 2005, 152, A2074–A2079. (32) Zheng, Y.; He, S.; Ge, L.; Zhou, M.; Chen, H.; Guo, L. Effect of Sr on Sm-Doped ceria electrolyte. Int. J. Hydrogen Energ. 2011, 36, 5128–5135. (33) Reynolds, E.; Blanchard, P. E. R.; Zhou, Q.; Kennedy, B. J.; Zhang, Z.; Jang, L.-Y. Structural and spectroscopic studies of La2Ce2O7: Disordered fluorite versus pyrochlore structure. Phys. Rev.

B 2012, 85, 132101. (34) Mantzavinos, D.; Hartley, A.; Metcalfe, I.S.; Sahibzada, M. Oxygen stoichiometries in La1−xSrxCo1−yFeyO3−δ perovskites at reduced oxygen partial pressures. Solid State Ionics 2000, 134, 103-109. (35) Itoh, T.; Shirasaki, S.; Ofuchi, H.; Hirayama, S.; Honma, T.; Nakayama, M. Oxygen partial pressure dependence of in situ X-ray absorption spectroscopy at the Co and Fe K edges for (La0.6Sr0.4)(Co0.2Fe0.8)O3−δ. Solid State Commun. 2012, 152, 278–283. ACS Paragon Plus Environment

25


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 27

(36) Aono, H.; Sato, M.; Traversa, E.; Sakamoto, M.; Sadaoka, Y. Design of Ceramic Materials for Chemical Sensors: Effect of SmFeO3 Processing on Surface and Electrical Properties. J. Am.

Ceram. Soc. 2004, 84, 341–347. (37) Bukhari, S. M.; Penwell, W. D.; Giorgi, J. B. Doped Samarium Ferrite Perovskites as Carbon and Sulfur Resistant Anodes for Low Temperature Solid Oxide Fuel Cells. ECS Trans. 2013, 57, 1507–1515. (38) Haas, O.; Vogt, U.; Soltmann, C.; Braun, A.; Yoon, W.-S.; Yang, X.; Graule, T. The Fe KEdge X-Ray absorption characteristics of La1−xSrxFeO3−δ prepared by solid state reaction. Mater.

Res. Bull. 2009, 44, 1397–1404. (39) Blasco, J.; Aznar, B.; García, J.; Subías, G.; Herrero-Martín, J.; Stankiewicz, J. Charge disproportionation in La1−xSrxFeO3 probed by diffraction and spectroscopic experiments. Phys. Rev.

B 2008, 77, 054107. (40) Hrovat, M.; Ahmad-Khanlou, A.; Samardzija, Z.; Holc, J. Interactions between lanthanum gallate based solid electrolyte and ceria. Mater. Res. Bull. 1999, 34, 2027-2034. (41) Zhang, J.; Wu, Z.; Liu, T.; Hu, T.; Wu, Z.; Ju, X. XANES study on the valence transitions in cerium oxide nanoparticles. J. Synchrotron Rad. 2001, 8, 531–532.


26

Page 27 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC graphics


27

Cation Diffusion and Segregation at the Interface between Samarium

Recommend Documents