In situ exsolution of Ni particles on the PrBaMn2O5 SOFC electrode material monitored by high temperature neutron powder diffraction under hydrogen

Mona Bahout *a, Praveen B. Managutti a, Vincent Dorcet a, Annie Le Gal La Salle b, Serge Paofai a and Thomas C. Hansen c
aUniv Rennes, CNRS, Institut des Sciences Chimiques de Rennes-UMR6226-ScanMAT-UMS2001, 263 Avenue du Général Leclerc, 35042 Rennes, France. E-mail: mona.bahout@univ-rennes1.fr
bInstitut des Matériaux Jean Rouxel (IMN), CNRS UMR 6502, Université de Nantes, 2 rue de la Houssinière, B.P. 32229 Nantes Cedex 3, France
cInstitut Laue-Langevin, 1 avenue des Martyrs CS 20156, 38042 Grenoble Cedex 9, France

Received 14th September 2019 , Accepted 21st December 2019

First published on 23rd December 2019


NiO has been incorporated into the Pr0.5Ba0.5MnO3−δ perovskite to produce, upon heating under a hydrogen atmosphere, in situ exsolved Ni-catalysts supported on the PrBaMn2O5 anode material. Transmission electron microscopy (TEM) and neutron powder diffraction (NPD) showed that the initial composition obtained by annealing in air at 950 °C consists of two perovskite phases: orthorhombic Pr0.65Ba0.35Mn0.975Ni0.025O3 (S.G. Ibmm, ∼75 wt%) and 2H-hexagonal BaMnO3−δ (S.G. P63/mcm, ∼25 wt%). On heating the two-phase sample under wet hydrogen, MnO particles exsolve at T ∼ 500 °C meanwhile the orthorhombic phase transforms to tetragonal (S.G. I4/mcm) then to cubic (S.G. Pm[3 with combining macron]m) at T ∼ 665 °C. When the temperature approaches 900 °C, the emergence of Ni metal particles was detected in the neutron diffraction patterns meanwhile the two perovskite phases start to transform into a Ni-free layered double perovskite, PrBaMn2O5. In situ real time observation of the structural changes under hydrogen atmosphere provided evidence of the simultaneity of Ni exsolution and phase transformation within our timescale resolution. From quantitative Rietveld analysis, the fraction of exsolved nickel represents the whole amount of Ni introduced in the synthesis. Impedance spectroscopy measurements in a 5% H2/Ar atmosphere show promising electrochemical performance for the Ni-exsolved layered perovskite electrode with a polarization resistance of 0.4 Ω cm2 at 800 °C (0.135 Ω cm2 at 850 °C) without any optimization.


1. Introduction

Solid oxide fuel cells (SOFCs) which transform chemical energy into electrical energy by an electrochemical conversion process with high efficiency are considered among the cleanest promising technologies to generate electricity.1–3 Conventional SOFCs employ Ni-based cermet anodes which display high electronic conductivity, excellent electrocatalytic activity for fuel oxidation and good compatibility with zirconia- or ceria-based electrolytes but suffer from serious drawbacks such as large volume change on redox cycling and Ni coarsening during operation, as well as carbon build-up (coking) and sulfur contamination from hydrocarbon fuels.4–6 Several materials have been investigated as alternatives to Ni-based cermet anodes to overcome coking and impurity poisoning issues. Among them, A-site layered double perovskite (LDP) manganites, LnBaMn2O5+δ (Ln = Pr, Nd) have received particular attention owing to their mixed ionic and electronic conductivity (MIEC), redox stability, superior resistance to coking and sulfur poisoning and good mechanical compatibility with common electrolytes. However the (electro)catalytic activity of these materials for fuel oxidation is insufficient for practical applications without addition of catalysts.7–11 Recently, exsolution of metal nanoparticles from a host oxide lattice has emerged as a promising catalyst design to improve the electrode performance and stability in replacement to the infiltration and impregnation technique.12–16 Potential drawbacks of these latter methods include non-uniform distribution of the nanoparticles, reduced porosity of the electrode and complicated processing procedures17 as well as coarsening of the nanoparticles at elevated temperature.18 In the exsolution process, (electro)catalytic elements are introduced in the crystal lattice during the synthesis under oxidizing conditions forming a solid solution and precipitate (exsolve) on the surface of the oxide phase upon heating the sample in hydrogen at T ≥ 800 °C. For instance, Pd-, Pt- and Rh-exsolved LaFeO3 perovskites are well-known catalysts for automotive emission control.19 Unlike the conventional impregnation method, in situ exsolution delivers thermally stable and uniformly dispersed nanoparticles (NPs).20–26 In particular, the A-site deficiency in perovskites giving the stoichiometry A1−αBO3−δ has been shown to promote exsolution of the reducible B cations.27–30 Indeed, A1−αBO3−δ perovskite manganites are used to minimize A-site cation segregation31,32 which is detrimental to exsolution.29

A few years ago, we reported the first real time in situ high temperature neutron powder diffraction (NPD) of the layered double perovskite (LDP) manganite, NdBaMn2O5+δ, under a hydrogen atmosphere.8,9 In the work reported here, we target the synthesis of a Ni-doped A-site deficient LDP manganite, (LnBa)1−x/2Mn2−xNixO5+δ with Ln = Pr and x = 0.05. In addition to being a good model system, this composition is representative of a range of oxides that have been shown to be active for metal and metal alloy exsolution (Mn, Co, Ni, Fe, and FeCo) and suitable as electrodes in solid oxide electrochemical cells.33,34

A-site deficiency has not been investigated in LDP manganites so far. In our targeted composition, the A-site deficiency is correlated to the Ni-content which have been chosen such that the LDP composition approaches stoichiometry upon total Ni exsolution under hydrogen atmosphere, according to the reaction:

(LnBa)1−x/2Mn2−xNixO6−δ → (1 − x/2) LnBaMn2O5 + xNi

A modest Ni2+ doping levels (x = 0.05) has been used due to low solubility of Ni in the host lattice. We used high temperature in situ neutron powder diffraction (NPD) under wet hydrogen (5% H2/He) and Rietveld phase analysis to monitor Ni exsolution and investigate the underlying structural evolution. To date the characterization of the exsolution process has been performed mostly under ex situ conditions; by X-ray diffraction (XRD), scanning and transmission electron microscopy (SEM/TEM), atomic force microscopy and X-ray photoelectron spectrometry (XPS).35,36 An in situ study reports on atomic-scale visualization of Co exsolution on the PrBaMn1.8Co0.2O5 double perovskite using an environmental transmission electron microscope.34 For the whole perovskite family, in situ characterization of the exsolution phenomena by neutron diffraction techniques is unavailable. This work aims to respond to this lack of information by a detailed high temperature neutron powder diffraction study of nickel exsolution in double perovskite manganites.30,37,38

2. Experimental

Polycrystalline powder of (PrBa)1−x/2Mn2−xNixO6−δ with x = 0.05 was prepared using a citrate–nitrate sol–gel route. Stoichiometric amounts of nitrate metal precursors, Pr(NO3)3·6H2O (Aldrich, 99.9%), Ba(NO3)2 (Aldrich, 99%), Mn(NO3)2·4H2O (Aldrich, 98%), and Ni(NO3)2·6H2O (Aldrich, 98.5%), were dissolved in distilled water under continuous stirring and heating (T ∼ 40 °C) and then citric acid (CA) (Acros Organics, 99%) and ethylene glycol (EG) (Sigma-Aldrich) were introduced. The molar ratio of CA/EG/metal ions is 3[thin space (1/6-em)]:[thin space (1/6-em)]1.5[thin space (1/6-em)]:[thin space (1/6-em)]1. The pH value was adjusted to ∼8 with ammonia (NH4OH 28 vol%) to avoid precipitation of the cations. The temperature was increased to T ∼ 180 °C and the solution was kept stirring and heating until a resin containing homogeneously distributed cations forms and autoignition occurred. The product was calcined at T ∼ 600 °C overnight to decompose most of the organic components. The resulting ash-like precursor was ground in an agate mortar, pressed into pellets (2 mm thick, 13 mm diameter), heated at T ∼ 950 °C for 24 h in a muffle furnace and then cooled to room temperature (RT) at a rate of ∼ 4 °C min−1. This sample, hereafter called “as-prepared PBMN”, was used for the NPD and DSC (differential scanning calorimetry) experiments. In order to examine the behavior of the as-prepared PBMN in a hydrogen atmosphere, thermal gravimetric analysis (TGA) and DSC analysis were simultaneously carried out on a Netzsch (STA 449F3) instrument. The sample was heated under dry 5% H2/N2 (40 mL min−1) from 20 to 1000 °C and subsequently cooled (heating/cooling rate of 20 °C min−1). TEM was performed using a JEOL JEM-2100 LaB6 transmission electron microscope operating at 200 kV and equipped with an Oxford silicon drift detector (SSD) X-MaxN 80T for energy dispersive X-ray scattering (EDS) measurements. The samples were crushed in dry ethanol and a drop of the suspension was deposited on a carbon-coated film (copper grid).

Temperature-dependent neutron powder diffraction experiments were carried out in the temperature range of 20–900 °C on the two-axis D20 neutron diffractometer at the high-flux reactor of the Institut Laue-Langevin (ILL, Grenoble, France).39 The set-up is the same as that described previously.40 A take-off angle of 90° from the (115) plane of a germanium monochromator was chosen, giving a wavelength of λ ∼ 1.54 Å and a resolution of Δd/d ∼ 2.9 × 10−3, while retaining a high flux on the sample (∼1.6 × 107 n cm2 s−1). A sample of ∼4.5 g was loaded at the centre of a quartz tube (8 mm inner diameter). Two K-type thermocouples were placed in the quartz tube; one a few mm above the sample to monitor the temperature and another below the sample to adjust the temperature. The temperature was programmed to increase linearly at 10 °C min−1 from 20 to 300 °C and at 2 °C min−1 from 300 °C to 900 °C and subsequent cooling. The data were summed, and the counters were reset to zero every 600 s, thus giving a temperature resolution of 50 and 20 °C, respectively. Besides the data collected on ramping, a few isothermal datasets were recorded at relevant temperatures; 300 °C (6 × 10 min), 800 °C (4 × 30 min) and 900 °C (4 × 30 min) as well as at the beginning and at the end of the reducing cycle. The temperature profile used under a wet 5% H2/He flow for the as-prepared PBMN–PBMN sample is shown in Fig. S.I. 1.

The diffraction patterns were analyzed by the Rietveld method41,42 using the program FullProf and its graphic interface WinPLOTR.43 The background was modelled using linear interpolation between ∼60 points and two asymmetry parameters were refined below 2θ = 55°. The unit cell parameters and zero-point shift were refined. The peak profile was modeled using a Thompson-Cox-Hastings pseudo-Voigt profile function.44 The atomic coordinates, the oxygen occupancy and the isotropic displacement parameters (Biso) of all the atoms were refined.

To check the catalytic effect of the exsolved nanoparticles on H2 oxidation, electrochemical impedance spectroscopy (EIS) measurements were performed on symmetric cells under a 5% H2/Ar flow in the temperature range of 850–650 °C. Two compositions were studied: Ni-exsolved-PBMN and PrBaMn2O5 (PBM) for comparison. The inks were prepared by mixing 60 wt% of the powder sample with 40 wt% of the α-terpineol (99% Acros Organics)/ethyl cellulose (Aldrich) (95/5 w/w) mixture. Dense 8YSZ (8% yttria-stabilised zirconia TOSOH) discs were used as the electrolyte. Due to the reactivity of the electrode materials with the 8YSZ electrolyte, a Ce0.9Gd0.1O1.95 (CGO)10,45 buffer layer (500 nm) was deposited by physical vapor deposition on both sides of the electrolyte. The electrode ink was deposited on both sides of the electrolyte followed by sintering in air at 1100 °C for 3 h. The current collectors consist of gold grid discs (A = 0.95 cm2) connected to the electrodes and linked to the external current and voltage circuits. The discs were placed into the open-flange setup™ provided by the Swiss company Fiaxell as described previously.46 It contains an oven and an Inconel 600 & 601 support in order to maintain the cell in the furnace. EIS was performed in potentiostatic mode using a VersaSTAT device and associated VersaStudio software in the 0.1–10[thin space (1/6-em)]000 Hz frequency range under open circuit voltage (OCV) conditions with an AC signal amplitude of 20 mV or 10 mV; the amplitude value was optimized for each measurement in order to get the best signal to noise ratio without loss of the transfer function linearity.

The cells were heated in air at 2 °C min−1 up to T ∼ 600 °C and then fed with a 5% H2/Ar gas mixture at a flow rate of 200 mL min−1 while the temperature was increased to 850 °C. Before starting the measurements, the samples were stabilized at T ∼ 850 °C until no variation of the impedance spectra was observed (24 h). Data were collected on cooling from 850 to 600 °C at 50 °C intervals with a stabilization time of 4 h at each temperature. The impedance diagrams were analysed using the ZView® software (D. Johnson, ZView: A software program for IES analysis, Version 2.8, Scribner associates, INC, Southern Pines, NC, 2002).

Preliminary analysis of the product

The crystalline structure of as-prepared PBMN examined by powder X-ray diffraction (XRD) is consistent with the existence of cubic and hexagonal perovskite structures without any secondary phases. Electron diffraction analysis of the cubic perovskite crystallite shows weak superstructure reflections (Fig. 1a and b, see arrows) related to octahedral tilting consistent with a (a√2 × a√2 × 2a) supercell and the Ibmm space group, reported for Pr0.7Ba0.3MnO3.47Fig. 1c displays a typical selected area electron diffraction (SAED) pattern of a hexagonal perovskite crystallite oriented along the [[1 with combining macron]10] zone axis. The structure is mainly two-layer hexagonal (2H) with the lattice parameters, a ≈ 5.7 Å and c ≈ 4.8 Å (space group P63/mmc) consistent with the XRD data. The presence of streaks along the c* direction can be attributed to disordered stacking faults as described by Parras et al.32 TEM-EDS analysis shows the orthorhombic and hexagonal perovskite compositions to be close to Pr0.65(3)Ba0.35(3)Mn0.97(1)Ni0.03(1)O3 and BaMnO3, respectively (Fig. S.I. 2).
image file: c9ta10159d-f1.tif
Fig. 1 SAED of as-prepared PBMN: zone axis pattern (a) [001] and (b) [0[1 with combining macron]1] of an orthorhombic Ibmm perovskite crystallite and (c) [[1 with combining macron]10] of a hexagonal perovskite crystallite.

The as-prepared PBMN sample was heated under 5% H2/Ar at T ∼ 900 °C for 18 h to produce the Ni-exsolved layered perovskite manganite. TEM-EDS analysis of few tens of “particles” showed the existence of two phases: a Ni-free PrBaMn2Ox phase (Fig. S.I. 3) and Ni particles/clusters of ∼80 nm diameter (Fig. 2). The analyses carried out on a large number of particles indicate the absence of nickel in the layered perovskite meaning that Ni has been completely exsolved from the perovskite structure.


image file: c9ta10159d-f2.tif
Fig. 2 Bright field TEM image of reduced PBMN. The grey square area is enlarged highlighting a Ni exsolved particle.

The results of thermal analysis in Fig. S.I. 4 indicate two transitions on the DSC curve on heating; at T ∼ 450 and ∼ 870 °C and, one transition on cooling; at T ∼ 525 °C. The TGA curve shows a rapid mass decrease between 400 °C and 500 °C on heating, consistent with the DSC signal at T ∼ 450 °C and subsequent smoother mass loss up to 800 °C due to diffusion phenomena. The temperature range of the first mass variation is consistent with that reported for Pr0.5Ba0.5MnO3−δ assigned to the phase transformation to double perovskite.11 However, as shown in the neutron section, the sample has not transformed yet to a layered perovskite in this temperature range. Indeed, the mass variation of ∼ 3 wt% involved cannot be attributed to oxygen loss only. If so, it would correspond to a perovskite with 2.55 oxygen/f.u. which is inconsistent with the oxygen content refined from NPD at T ∼ 500 °C. Consequently, the first mass loss may include dehydration or removal of adsorbed water. When the temperature exceeds 800 °C, a second weight loss, concomitant with the presence of the small exothermic peak at T ∼ 870 °C on the DSC curve, proceeds and completes with the increase in temperature. It can be speculated that this weight loss of ∼0.7 wt% corresponds to the removal of oxygen atoms bonding Ni2+ and their reduction to Ni metal meanwhile the layered perovskite is being constructed, as demonstrated by in situ NPD.

3. Neutron diffraction

Fig. 3 shows the Rietveld analysis of the neutron powder diffraction data collected at T ∼ 60 °C prior to reduction. The as-prepared PBMN sample consists of two perovskites: an orthorhombic phase (space group Ibmm #74, a = 5.5079(3) Å, b = 5.4893(2) Å, c = 7.7539(4) Å) and a two-layer BaMnO3−δ hexagonal (2H) phase (space group P63/mmc #194, a = 5.6869(2) Å, c = 4.7916(3) Å). No NiO secondary phase (space group Fm[3 with combining macron]m, a = 4.176 Å) was detected.
image file: c9ta10159d-f3.tif
Fig. 3 Observed, calculated, and difference patterns for as-prepared PBMN at T ∼ 60 °C. Tick marks refer to (upper) (Pr,Ba)MnO3−δ (S.G. Ibmm) and (lower) 2H-BaMnO3−δ (S.G. P63/mmc) perovskites. Peaks unindexed in cubic symmetry are labelled (*).

The ratio of the Pr and Ba cations at the A site of the orthorhombic and hexagonal perovskites could not be refined due to small difference between the scattering lengths of Ba (5.07 fm) and Pr (4.58 fm).48 The compositions of the orthorhombic and hexagonal phases were therefore fixed at Pr0.65Ba0.35Mn0.975Ni0.025O3 and BaMnO3, respectively, on the basis of the EDS analysis. The A-site deficiency targeted in our synthesis was shown not to affect the quality of the fit and was neglected in the refinement. The refined oxygen occupancy at the O1 (4e) and O2 (8g) sites of the orthorhombic phase converged to 0.93(7) and 1.05(5), respectively, resulting in an oxygen composition close to stoichiometry. The refinement of the Ni content in the hexagonal perovskite converged to a negative value meaning that this phase is Ni-free, in agreement with the EDS analysis. The oxygen occupancy at the 6h site converged to 0.81(5) resulting in 2.4(1) oxygen/formula unit (f.u.). The two-phase model produced a good fit (χ2 ∼ 12, Rwp ∼ 8.0% and Rp ∼ 9.7%) with a weight ratio for the orthorhombic and hexagonal phases of ∼77(4) and ∼23(1) wt%, respectively. The refined structural parameters are given in Table S.I. 1. We should mention that the X-ray diffraction patterns reported for analogous Co-, Ni- or Fe-doped manganites were shown to consist of a mixture of hexagonal and cubic (rather than orthorhombic) phases.37

The Pr- and Ba-content of the orthorhombic and hexagonal phases within our as-prepared PBMN sample is consistent with the Goldschmidt tolerance factor (t) which evaluates the crystal structure distortion of ABO3 perovskites. This factor is estimated from the following equation: image file: c9ta10159d-t1.tif, where rA, rB and rO are the average ionic radii at the A site, B site and oxygen site, respectively.

Since the ionic radius of Pr3+ is smaller than that of the Ba2+ one (rPr3+ = 1.18 Å, rBa2+ = 1.61 Å),49 it is expected that the orthorhombic perovskite contains larger amount of Pr3+ cations than Ba2+ while the reverse prevails for hexagonal perovskites. Indeed, various Pr-rich manganites such as Pr0.65Ba0.35MnO3 and Pr0.7Ba0.3MnO3 crystallize in the orthorhombic Ibmm S.G. while the hexagonal perovskite manganites reported so far contain only Ba at the A site.48 Other space groups were reported for Pr1−xBaxMnO3−δ perovskites annealed in air, such as Pbnm for the compositions with x = 0.1,47 0.2, 0.3[thin space (1/6-em)]50 and 0.33[thin space (1/6-em)]51 or tetragonal symmetry for Pr0.65Ba0.35MnO3 at low temperature (S.G. I4/mcm at T ∼ 210 K, S.G. I4/m at T ∼ 210 and ∼5 K).50

Heating of the sample

The two-phase sample was heated to T ∼ 300 °C at a fast rate (10 °C min−1) since no particular changes were expected in this temperature range according to the TGA-DSC analysis and, at a lower rate (2 °C min−1) from T ∼ 300 °C to T ∼ 900 °C and subsequent cooling. When the temperature approached 400 °C, additional peaks ascribed to MnO (S.G. Fm[3 with combining macron]m) were observed at 2θ ∼ 34.7 and 70° in the NPD pattern. Moreover, at T ≥ 400 °C the oxygen occupancy at the O1 site in the body-centered phase starts decreasing while the amount of MnO increases. The Rietveld plot for the data collected at T ∼ 500 °C is shown in Fig. S.I. 5. No significant changes were detected on heating from T ∼ 500 to 665 °C besides the decrease of the orthorhombic distortion of the Ibmm phase whose a and b lattice parameters become closer, as illustrated in Fig. 4. The percentages of the body-centered and hexagonal phases remain unchanged as well as their oxygen content, respectively close to 2.7 and 2.5 atoms/formula unit (f.u.).
image file: c9ta10159d-f4.tif
Fig. 4 Evolution of the (red) a and (black) b lattice parameters of (Pr,Ba)MnO3−δ perovskite (S.G. Ibmm) on heating under wet 5% H2/He.

At T ≥ 510 °C, the multi-phase refinement using the tetragonal space group I4/mcm gave similar goodness-of-fit indicators to the orthorhombic Ibmm one and therefore the former was used to model the body-centered phase.

When the temperature reached 665 °C, small peaks arising from a cubic perovskite phase (S.G. Pm[3 with combining macron]m) whose composition was the same as the tetragonal one were detected. Therefore, sequential refinements accounting for the coexistence of three perovskite phases (tetragonal, cubic and hexagonal) and MnO were undertaken up to T ∼ 800 °C. To avoid instability in the multiphase refinements, the overall isotropic temperature factor Bov, for each phase was refined instead of Biso of the individual atoms. Fig. 5 shows the evolution of the phase fractions of the tetragonal and cubic perovskites in the temperature range of 665–800 °C.


image file: c9ta10159d-f5.tif
Fig. 5 Evolution of the weight fractions on heating as-prepared PBMN between 665 and 800 °C: (red) body-centered perovskite and (black) cubic perovskite. The hexagonal phase remains at 15 wt%. The insert shows the weight fraction evolution of MnO.

After 1 h of isothermal heating at ∼800 °C, the tetragonal perovskite has completely transformed into a cubic one. During this phase transition, the amount of the hexagonal phase remains constant (∼15 wt%) as well as its oxygen content (2.5 atoms/f.u.) while the quantity of MnO has increased to ∼1.9 wt%. The profile of two representative patterns collected during this transition is shown in Fig. S.I. 6 and S.I. 7.

After 2 h of heating at 800 °C, the temperature was increased at 2 °C min−1 to 900 °C at which the sample was held for 2 h while four datasets were collected at 30 min intervals. Fig. 6 shows relevant patterns recorded on heating from 800 to 900 °C. An additional peak at 2θ ∼ 44° arising from Ni metal is detected at T = 870 °C and its intensity increases on heating. At the same time, the phase fraction of the hexagonal perovskite decreases as revealed by the decrease of the peaks at 2θ ∼ 27, 38, and 41° (labelled * in Fig. 6) and a layered perovskite phase emerges (Fig. S.I. 8). Its construction completed within 60 min heating at 900 °C (grey pattern, Fig. 6), Table S.I. 2 lists the corresponding refined structural parameters.


image file: c9ta10159d-f6.tif
Fig. 6 In situ patterns collected on heating as a function of temperature/time. Ni exsolution and formation of the layered perovskite occur simultaneously as revealed by the increase of the Ni peak and decrease of the peaks of the hexagonal perovskite labelled (*). The (101) and (103) peaks from the layered perovskite are indicated in the last pattern at 900 °C.

Table S.I. 3 lists the weight fractions of the different phases in the 800–900 °C temperature range. Both Fig. 6 and Table S.I. 3 highlight that Ni exsolution and layered perovskite formation occur simultaneously within the timescale of our experiment and completed within 60 min at 900 °C. Consequently, Ni exsolution does not proceed by intracrystalline diffusion of the Ni2+ ions from the layered perovskite backbone, as reported previously.34,37 The unmixing of Ni ions with Mn at the B-site of the layered perovskite is consistent with the lack of layered double perovskite nickelates LnBaNi2O5+δ (Ln = lanthanide) or nickel-substituted manganites, LnBaM2−xNixO5+δ. However, layered double perovskites with Ni ions sharing the B-site with either Co or Fe have been reported.52

The onset temperature of the layered perovskite construction is concomitant with Ni particle formation in agreement with the mass loss and exothermic peak detected at T ∼ 870 °C in the TGA-DSC curves (Fig. S.I. 4). This temperature is much higher than that reported at T ∼ 400 °C.11 According to our results, the DSC peak at T ∼ 400 °C corresponds to the onset of MnO exsolution from as-prepared PBMN.

Fig. 7 summarizes the main structural changes undergone by as-prepared PBMN heated in a wet 5% H2/He atmosphere.


image file: c9ta10159d-f7.tif
Fig. 7 In situ NPD data highlighting the structural transformations of PBMN heated under wet 5% H2/He. Transition from orthorhombic (Ibmm) to tetragoanl (I4/mcm) perovskite proceeds at T ∼ 500 °C, to cubic (Pm[3 with combining macron]m) at T ∼ 800 °C and finally to layered perovskite (P4/mmm) at T ∼ 900 °C.

Cooling from 900 °C to room temperature

The sample was cooled under wet hydrogen at a rate of 2 °C min−1 and the structure of the double perovskite was sequentially refined in the space group P4/mmm. The space group changes to P4/nmm at T ∼ 550 °C due to Mn3+ and Mn2+ charge ordering in agreement with previous results.9Fig. 8 shows the Rietveld fit in the end of the reducing cycle, at T ∼ 60 °C (lowest temperature reached on cooling) and Table S.I.4 lists the corresponding structural parameters. From the line broadening of the (111) Ni peak at 2θ ∼ 44°, the mean crystallite size of the exsolved Ni particles was estimated at ∼ 60 nm using the Scherer equation,53 and is in good agreement with TEM observation (Fig. 2).
image file: c9ta10159d-f8.tif
Fig. 8 Rietveld profile at T ∼ 60 °C at the end of the heating/cooling cycle: (upper) PrBaMnO5, (middle) Ni and (bottom) MnO. A few peaks accounting for charge ordering and indexed in S.G. P4/nmm are labelled (+). The “ordered” Mn3+O5/Mn2+O5 polyhedra are drawn in pink/purple in the insert.

Electrochemical measurements

To evaluate the catalytic activity towards H2 oxidation, the electrochemical performance of PBMN symmetric cells was studied by electrochemical impedance spectroscopy under a 5% H2/Ar flow in the temperature range of 850–650 °C and compared to that of PrBaMn2O5 (PBM). The impedance diagrams can be fitted with the equivalent circuit consisting of two resistance (R)-constant phase element (CPE) parallel circuits connected to an inductance (L) and a series resistance (Rs). The inductance, L, primarily arises from the electrical wires and ranges between 0.5 × 10−6 H and 1.0 × 10−6 H for the present system, in agreement with the literature data using a similar set-up.54 The series resistance, Rs, is mainly correlated with the ohmic losses originating from the electrolyte. A conductivity of ∼ 2 × 10−2 S cm−1 was calculated from the fitted Rs value of the data obtained at 800 °C and is in good agreement with the literature.55Fig. 9a presents the impedance results obtained for the two samples at 800 °C. In order to emphasize the anodic part of the impedance response, the diagrams are presented after subtraction of the L and Rs contributions and consist of two elementary contributions.
image file: c9ta10159d-f9.tif
Fig. 9 Electrochemical performance of symmetric cells with PBMN and PBM electrodes in 5% H2/Ar: (a) Nyquist impedance diagrams at 800 °C at OCV and (b) Arrhenius plot.

For PBM, the capacitances C1 and C2 associated with these two processes are 0.95 mF cm−2 and 3.6 mF cm−2, showing that the polarization resistance, Rp, image file: c9ta10159d-t2.tif is only due to mass transfer, such as adsorption of H2, dissociation of H2, and charge transfer-diffusion in the electrode.56 In contrast, for PBMN, the polarization resistance Rp is lower, with C1 and C2 capacitances of 2.5 μF cm−2 and 17 mF cm−2, respectively, suggesting that in this case, only the low frequency contribution is related to mass transfer. The decrease of the mass transfer resistance for PBMN reveals the promoted effect of Ni-exsolution on the catalytic activity for the oxidation of H2 and the possibility of using PBMN as a hydrogen electrode material for SOFCs. The thermal variation of the polarization resistance, Rp, is shown in Fig. 9b. The Rp values of PBM are always larger than those of PBMN and decrease with increasing temperature with the same activation energy of 1.38 eV. At T = 850 °C, Rp values for PBM and PBMN correspond to 0.43 and 0.135 Ω cm2, respectively.

The performance of the PBMN anode is similar to that of PBM electrodes doped with much larger amounts of transition metals37 and to composite electrodes such as PBM-YSZ (Rp = 0.13 Ω cm2 at 850 °C).57 With further optimization of PBMN electrode composition and thickness, interfaces, microstructural and architectural design, the EIS performance can be largely improved as demonstrated for La0.75Sr0.25Cr0.5Mn0.5O3−δ (LSCM) whose polarization resistance decreased considerably when composite (LSCM-YSZ/CGO) or graded electrodes were used.58,59

4. Conclusion

The present work demonstrates that the introduction of a small amount of nickel into the (PrBa)0.5MnO3 perovskite results in a mixture of orthorhombic Pr0.67Ba0.33Mn0.975Ni0.025O3 and hexagonal BaMnO3−δ perovskite phases after annealing in air at 950 °C. On heating the two-phase sample under wet hydrogen, NPD reveals exsolution of MnO at T ≥ 500 °C and structural phase transitions from Ibmm to I4/mcm then to Pm[3 with combining macron]m symmetry. In the 800–900 °C temperature range, further oxygen loss resulting in ∼ 2.5 oxygen atoms/f.u. for both hexagonal and cubic perovskite phases triggers the construction of a double perovskite. During this crystal reconstruction, the Ni2+ ions initially present in the A-site disordered perovskite are pulled outside the structure and reduce to metal (nano)particles. The fraction of the exsolved Ni evaluated at ∼ 0.06 wt% from Rietveld phase analysis corresponds to the whole amount of nickel introduced in the synthesis and differs from the value of ∼ 58% previously reported.37 These findings may provide new insights into the exsolution mechanism and implications for the regeneration of the nanocatalysts. The Ni-exsolved layered perovskite (PBMN) presents better electrochemical performance in a hydrogen atmosphere than the Ni-free material (PBM) which can be improved by optimizing the electrode architecture (composition, thickness, and sintering temperature).

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the PhD grants provided to P. M. from the MESR (Ministry of Higher Education, of Research and Innovation). We thank the ILL for the beam time allowed and Alain Daramsy for his technical help at ILL. We are grateful to P. Briois (Université de Technologie de Belfort-Montbéliard) for doped ceria deposition by PVD on the YSZ electrolytes and for helpful discussions.

References

  1. E. D. Wachsman, C. A. Marlowe and K. T. Lee, Energy Environ. Sci., 2012, 5, 5498 RSC .
  2. R. J. Gorte, S. Park, J. M. Vohs and C. Wang, Adv. Mater., 2000, 12, 1465 CrossRef CAS .
  3. N. P. Brandon, S. Skinner and B. C. H. Steele, Annu. Rev. Mater. Res., 2003, 33, 183 CrossRef CAS .
  4. W. Wang, C. Su, Y. Wu, R. Ran and Z. Shao, Chem. Rev., 2013, 113, 8104 CrossRef CAS PubMed .
  5. M. Liu, M. E. Lynch, K. Blinn, F. M. Alamgir and Y. Choi, Mater. Today, 2011, 14, 534 CrossRef CAS .
  6. S. McIntosh and R. J. Gorte, Chem. Rev., 2004, 104, 4845 CrossRef CAS PubMed .
  7. S. Choi, S. Sengodan, S. Park, Y. W. Ju, J. Kim, J. Hyodo, H. Y. Jeong, T. Ishihara, J. Shin and G. Kim, J. Mater. Chem. A, 2016, 4, 1747 RSC .
  8. F. Tonus, M. Bahout, V. Dorcet, R. K. Sharma, E. Djurado, S. Paofai, R. I. Smith and S. J. Skinner, J. Mater. Chem. A, 2017, 5, 11078 RSC .
  9. F. Tonus, M. Bahout, V. Dorcet, G. H. Gauthier, S. Paofai, R. I. Smith and S. J. Skinner, J. Mater. Chem. A, 2016, 4, 11635 RSC .
  10. O. L. Pineda, Z. L. Moreno, P. Roussel, K. Świerczek and G. H. Gauthier, Solid State Ionics, 2016, 288, 61 CrossRef CAS .
  11. S. Sengodan, S. Choi, A. Jun, T. H. Shin, Y.-W. Ju, H. Y. Jeong, J. Shin, J. T. S. Irvine and G. Kim, Nat. Mater., 2015, 14, 205 CrossRef CAS PubMed .
  12. W. H. Kan, A. J. Samson and V. Thangadurai, J. Mater. Chem. A, 2016, 4, 17913 RSC .
  13. L. Thommy, O. Joubert, J. Hamon and M.-T. Caldes, Int. J. Hydrogen Energy, 2016, 41, 14207 CrossRef CAS .
  14. B. Hua, M. Li, Y.-F. Sun, J.-H. Li and J.-L. Luo, ChemSusChem, 2017, 10, 3333 CrossRef CAS PubMed .
  15. S. P. Jiang, J. Mater. Sci. Eng. A, 2006, 418, 199 CrossRef .
  16. T. Ishihara, J. Korean Ceram. Soc., 2016, 53, 469 CrossRef CAS .
  17. X. Lou, Z. Liu, S. Wang, Y. Xiu, C. P. Wong and M. Liu, J. Power Sources, 2010, 195, 419 CrossRef CAS .
  18. Y. Gong, D. Palacio, X. Song, R. L. Patel, X. Liang, X. Zhao, J. B. Goodenough and K. Huang, Nano Lett., 2013, 13, 4340 CrossRef CAS .
  19. H. Tanaka, M. Uenishi, M. Taniguchi, I. Tan, K. Narita, M. Kimura, K. Kaneko, Y. Nishihata and J. i. Mizuki, Catal. Today, 2006, 117, 321 CrossRef CAS .
  20. J.-h. Myung, D. Neagu, D. N. Miller and J. T. S. Irvine, Nature, 2016, 537, 528 CrossRef CAS .
  21. G. Yang, W. Zhou, M. Liu and Z. Shao, ACS Appl. Mater. Interfaces, 2016, 8, 35308 CrossRef CAS .
  22. Z. Du, H. Zhao, S. Yi, Q. Xia, Y. Gong, Y. Zhang, X. Cheng, Y. Li, L. Gu and K. Świerczek, ACS Nano, 2016, 10, 8660 CrossRef CAS PubMed .
  23. S. Cui, J. Li, X.-W. Zhou, G. Wang, J.-L. Luo, K. T. Chuang, L. Qiao and Y. Bai, J. Mater. Chem. A, 2013, 1, 9689–9696 RSC .
  24. S. Liu, Q. Liu and J.-L. Luo, ACS Catal., 2016, 6, 6219 CrossRef CAS .
  25. T. Jardiel, M. T. Caldes, F. Moser, J. Hamon, G. Gauthier and O. Joubert, Solid State Ionics, 2010, 181, 894 CrossRef CAS .
  26. T. Delahaye, T. Jardiel, O. Joubert, R. Laucournet, G. Gauthier and M. T. Caldes, Solid State Ionics, 2011, 184, 39 CrossRef CAS .
  27. Y.-F. Sun, J.-H. Li, M.-N. Wang, B. Hua, J. Li and J.-L. Luo, J. Mater. Chem. A, 2015, 3, 14625 RSC .
  28. Y.-F. Sun, J.-H. Li, L. Cui, B. Hua, S.-H. Cui, J. Li and J.-L. Luo, Nanoscale, 2015, 7, 11173 RSC .
  29. D. Neagu, G. Tsekouras, D. N. Miller, H. Ménard and J. T. S. Irvine, Nat. Chem., 2013, 5, 916 CrossRef CAS PubMed .
  30. Y. Gao, D. Chen, M. Saccoccio, Z. Lu and F. Ciucci, Nano Energy, 2016, 27, 499 CrossRef CAS .
  31. M. Ellouze, W. Boujelben, A. Cheikhrouhou, H. Fuess and R. Madar, Solid State Commun., 2002, 124, 125 CrossRef CAS .
  32. M. Parras, J. M. González-Calbet, J. Alonso and M. Vallet-Regí, J. Solid State Chem., 1994, 113, 78 CrossRef CAS .
  33. S. Sengodan, Y.-W. Ju, O. Kwon, A. Jun, H. Y. Jeong, T. Ishihara, J. Shin and G. Kim, ACS Sustainable Chem. Eng., 2017, 5, 9207 CrossRef CAS .
  34. Y.-F. Sun, Y.-Q. Zhang, J. Chen, J.-H. Li, Y.-T. Zhu, Y.-M. Zeng, B. S. Amirkhiz, J. Li, B. Hua and J.-L. Luo, Nano Lett., 2016, 16, 5303 CrossRef CAS PubMed .
  35. B. D. Madsen, W. Kobsiriphat, Y. Wang, L. D. Marks and S. A. Barnett, J. Power Sources, 2007, 166, 64 CrossRef CAS .
  36. C. Arrive, T. Delahaye, O. Joubert and G. Gauthier, J. Power Sources, 2013, 223, 341 CrossRef CAS .
  37. O. Kwon, S. Sengodan, K. Kim, G. Kim, H. Y. Jeong, J. Shin, Y.-W. Ju, J. W. Han and G. Kim, Nat. Commun., 2017, 8, 15967 CrossRef CAS PubMed .
  38. D. Neagu, T.-S. Oh, D. N. Miller, H. Ménard, S. M. Bukhari, S. R. Gamble, R. J. Gorte, J. M. Vohs and J. T. S. Irvine, Nat. Commun., 2015, 6, 8120 CrossRef PubMed .
  39. T. C. Hansen, P. F. Henry, H. E. Fischer, J. Torregrossa and P. Convert, Meas. Sci. Technol., 2008, 19, 034001 CrossRef .
  40. F. Tonus, M. Bahout, P. D. Battle, T. Hansen, P. F. Henry and T. Roisnel, J. Mater. Chem., 2010, 20, 4103 RSC .
  41. H. M. Rietveld, J. Appl. Crystallogr., 1969, 2, 65 CrossRef CAS .
  42. L. B. McCusker, R. B. Von Dreele, D. E. Cox, D. Louër and P. Scardi, J. Appl. Crystallogr., 1999, 32, 36 CrossRef CAS .
  43. J. Rodríguez-Carvajal, Phys. B, 1993, 192, 55 CrossRef .
  44. J.-F. Berar and G. Baldinozzi, J. Appl. Crystallogr., 1993, 26, 128 CrossRef .
  45. S. Joo, O. Kwon, K. Kim, S. Kim, H. Kim, J. Shin, H. Y. Jeong, S. Sengodan, J. W. Han and G. Kim, Nat. Commun., 2019, 10, 697 CrossRef PubMed .
  46. F. Ricoul, A. Subrenat, O. Joubert and A. Le Gal La Salle, J. Solid State Electrochem., 2018, 22, 2789 CrossRef CAS .
  47. S. V. Trukhanov, V. A. Khomchenko, L. S. Lobanovski, M. V. Bushinsky, D. V. Karpinsky, V. V. Fedotova, I. O. Troyanchuk, A. V. Trukhanov, S. G. Stepin, R. Szymczak, C. E. Botez and A. Adair, J. Exp. Theor. Phys., 2006, 103, 398 CrossRef CAS .
  48. V. F. Sears, Neutron News, 1992, 3, 26–37 CrossRef .
  49. R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751 CrossRef .
  50. Z. Jirak, E. Pollert, A. F. Andersen, J. C. Grenier and P. Hagenmuller, Eur. J. Solid State Inorg. Chem., 1990, 27, 421 CAS .
  51. S. Hcini, S. Zemni, A. Triki, H. Rahmouni and M. Boudard, J. Alloys Compd., 2011, 509, 1394 CrossRef CAS .
  52. J. H. Kim and A. Manthiram, Electrochim. Acta, 2009, 54, 7551 CrossRef CAS .
  53. A. L. Patterson, Phys. Rev., 1939, 56, 978 CrossRef CAS .
  54. D. Marrero-López, J. Peña-Martínez, J. C. Ruiz-Morales, M. Gabás, P. Núñez, M. A. G. Aranda and J. R. Ramos-Barrado, Solid State Ionics, 2010, 180, 1672 CrossRef .
  55. F. Ricoul, A. Subrenat, O. Joubert and A. Le Gal La Salle, Int. J. Hydrogen Energy, 2017, 42, 21215 CrossRef CAS .
  56. Q. X. Fu, F. Tietz and D. Stöver, J. Electrochem. Soc., 2006, 153, D74 CrossRef CAS .
  57. Y.-F. Sun, Y.-Q. Zhang, B. Hua, Y. Behnamian, J. Li, S.-H. Cui, J.-H. Li and J.-L. Luo, J. Power Sources, 2016, 301, 237 CrossRef CAS .
  58. I. Jung, D. Lee, S. O. Lee, D. Kim, J. Kim, S.-H. Hyun and J. Moon, Ceram. Int., 2013, 39, 9753 CrossRef CAS .
  59. S. He, H. Dai, G. Cai, H. Chen and L. Guo, Electrochim. Acta, 2015, 152, 155 CrossRef CAS .

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ta10159d

This journal is © The Royal Society of Chemistry 2020