Structural and electrochemical properties of the perovskite oxide Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ

Structural and electrochemical properties of the perovskite oxide Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ

Available online at www.sciencedirect.com Solid State Ionics 179 (2008) 725 – 731 www.elsevier.com/locate/ssi Structural and electrochemical propert...

719KB Sizes 0 Downloads 11 Views

Available online at www.sciencedirect.com

Solid State Ionics 179 (2008) 725 – 731 www.elsevier.com/locate/ssi

Structural and electrochemical properties of the perovskite oxide Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ Shanwen Tao ⁎, John T.S. Irvine School of Chemistry, University of St Andrews, Fife KY16 9ST, Scotland, UK Received 27 November 2007; received in revised form 11 April 2008; accepted 27 April 2008

Abstract The perovskite oxide Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ was synthesised by a combustion method. Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ obtained at 1400 °C has been shown to have an orthorhombic structure with space group Pnma (62), a = 5.4388(1)Å, b = 7.6969(1)Å, c = 5.4584(1)Å, V = 228.50(1)Å3 according to X-ray diffraction. The material is redox stable and maintains its structure in a reducing atmosphere. After reducing in 5% H2 at 900 °C for 190 h, Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ still exhibits an orthorhombic structure. A lattice volume expansion of 0.61% was observed during the reduction, which may be attributed to reduction of Pr, Cr and Ni ions accompanying loss of lattice oxygen. TGA analysis and EDS indicate Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ shows enhanced resistance to nickel reduction. The conductivities of this material in air and 5% H2 were 27.4 and 1.37 S/ cm respectively at 900 °C. Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ exhibits semiconductor behaviour in both air and 5% H2. The anode polarisation resistance of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ reached 0.98 Ωcm2 at 900 °C in wet H2 but still not good enough as a good SOFC anode although it could be further improved by optimisation of microstructure. © 2008 Elsevier B.V. All rights reserved. Keywords: Electrical conductivity; Perovskite; Structure; Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ; Fuel cell

1. Introduction High temperature solid oxide fuel cells (SOFCs) promise high efficiencies in a range of fuels. Unlike for lower temperature alternative fuel cells, carbon monoxide is a fuel rather than a poison and so hydrocarbon fuels can be used directly, through internal reforming or even direct oxidation. This provides a key entry strategy for fuel cell technology into the current energy economy. Present development of SOFCs is mainly based on the yttria-stabilised zirconia (YSZ) electrolyte because it exhibits good thermal and chemical stability, high oxide-ion conductivity and mechanical strength at high temperature [1]. The most commonly used anode materials for zirconia-based SOFCs are Ni/YSZ cermets, which display excellent catalytic properties for fuel oxidation and good current collection but do exhibit disadvantages, such as low tolerance to ⁎ Corresponding author. Present address: School of Engineering & Physical Sciences, Heriot-Watt University, Edinburgh EH14 4AS, UK. Tel.: + 44 131 451 4299; fax: +44 131 451 3180. E-mail address: [email protected] (S. Tao). 0167-2738/$ - see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ssi.2008.04.027

sulphur [2] and carbon deposition [3] when using hydrocarbon fuels and poor redox cycling causing volume instability. In order to overcome the disadvantages of the traditional Ni-YSZ cermet anode, alternative oxides are being investigated as potential anodes for SOFCs. Materials with different structure have been investigated as the alternative anode materials for SOFCs which has been reviewed recently [4]. Among these materials, those with perovskite structure such as chromites and titanates are promising SOFC anode materials [5–13]. LaCrO3-based materials have been investigated as interconnect materials for SOFCs [1]; however, they are also potential anode materials for SOFCs due to their relatively good stability in both reducing and oxidising atmospheres at high temperatures [2,9,14,15]. No significant weight loss was observed when LaCrO3 was exposed to a reducing atmosphere (oxygen partial pressure pO2 = 10− 21 atm at 1000 °C) [16]. This indicates that chromium strongly retains its six-fold coordination. The introduction of other transition elements into the B-site of La1 − xSrxCr1 − yMyO3 (M = Mn, Fe, Co, Ni) has been shown to improve the catalytic properties for methane reforming [17]. Of the various dopants, nickel seems to be the most successful and

726

S. Tao, J.T.S. Irvine / Solid State Ionics 179 (2008) 725–731

the lowest polarisation resistances have been reported for 10% Ni-substituted lanthanum chromite [18]. Certainly nickel oxides would not be stable in fuel atmospheres, and although the nickel may be stabilised by the lattice in higher oxidation state, there will always be the suspicion that the activity of nickel doped perovskites is due to surface evolution of nickel metal and hence questions about long term stability. It has been reported that a composite anode of 5% Ni with a 50/50 mixture of La0.8Sr0.2Cr0.8Mn0.2O3 and Ce0.9Gd0.1O1.95 was used for SOFCs with different fuels and it was found that a small amount of Ni provides a substantial electrocatalytic effect [19]. It is difficult to introduce the oxygen vacancies that are required for oxide-ion conduction into the LaCrO3 lattice. When the B-sites are doped by other multivalent transition elements that do tolerate reduced oxygen co-ordination, such as Mn, Fe, Co, Ni and Cu, oxygen vacancies may be generated at the B-site dopants in a reducing atmosphere at high temperature. A significant degree of B-site dopant is required to generate enough oxygen vacancies in order to achieve high oxide-ion conductivity. We have reported that introduction of a large amount of transition elements such as Mn and Fe into the Bsites of LaCrO3 can significantly improve the anode performance and catalytic property but the material is not redox stable at a high doping level of Ni at B-site [20–22]. Therefore we tried to introduce a small amount of nickel (10%) at the B-site in order to improve the catalytic properties. On the other hand, introduction of multivalent lanthanide at the A-site could be beneficial to the redox stability, conductivity and catalytic property. Praseodymium is a well known multivalent lanthanide prefers 4+/3+ mixed oxidation state in oxides. It was reported that the conductivity of Pr0.65Sr0.3MnO3 − δ is higher than its Laanalogue La0.65Sr0.3MnO3 − δ [23]. Similarly, replacing lanthanum by praseodymium in LnCrO3-based oxide could improve the conductivity. The redox nature of Prn+ ions at the A-site may help to stabilise the transition element such as Nin+ ions at the B-site in perovskite lattice because Prn+ could be partially reduced as well. In this report, the structure and electrical properties of the perovskites oxide Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ with a multivalent lanthanide at the A-sites have been investigated. 2. Experimental Materials were prepared by a combustion method which employed the corresponding metal nitrates, citric acid and ethylene glycol as precursors. Pr6O11 (Alfa, 99.9%) and SrCO3 (Aldrich, 99.9%) were dissolved into dilute nitric acid to form nitrates then appropriate amounts of Cr(NO3)3·9H2O (Fisher, Acros, 99%) and Ni(NO3)2·6H2O (Strem Chemicals, 99.9+%) was added to form a mixed nitrate solution of the required stoichiometry. Citric acid and ethylene glycol were added and the solution refluxed at 80 °C for 2 h. After further concentrating the mixture at higher temperatures, a gel formed, combustion of which yielded a brown sponge-like material. Firing this above 1100 °C with intermediate grindings resulted in the formation of the sought perovskite related phase. The material was pre-fired at 1100 °C for 4 h then pressed into

pellets and further fired at 1400 °C for 40 h. Samples fired at 1400 °C were used for XRD and SEM analysis. XRD analysis of powders reacted at different temperatures were carried out on a Stoe Stadi-P diffractometer to determine phase purity and measure crystal parameters. Structure refinement was performed by the Rietveld method using the program General Structure Analysis System (GSAS) [24]. Thermal analysis of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ was carried out on a Rheometric Scientific TG 1000 M Plus, and a TA Instruments SDT2960 from room temperature to 900 °C (5 °C/min), holding at 900 °C for 120 min then cooling down to 30 °C (5 °C/min) under flowing 5% H2 in argon at a rate of 20 ml/min. SEM observations were carried out on a 5600 scanning electron microscope integrated with a Mica Element Analysis System (Oxford Instrument). The d.c. conductivity was measured by a conventional fourterminal method using a Keithley 220 Programmable Current Source to control current and a Schlumberger Solartron 7150 Digital Multimeter to measure the voltage. The Pr0.7Sr0.3Cr0.9 Ni0.1O3 − δ samples were mounted with four Pt wire electrodes to measure the d.c. conductivity dependence upon pO2 in a slowly varying atmosphere, which was monitored by a zirconia oxygen sensor. The conductivity was measured by the four-terminal d.c. method in air and 5% H2. A.c. impedance spectroscopy was carried out using a Schlumberger Solartron 1255 Frequency Response Analyser coupled to a 1287 Electrochemical Interface and controlled by Z-plot electrochemical impedance software. A three electrode arrangement has been applied to measure the anode polarisation which is the same as described in detail elsewhere [21]. The impedance spectra of the electrochemical cell were recorded at open circuit voltage (OCV) with a 20 mV a.c. signal amplitude over the frequency range 105–0.01 Hz. 3. Results and discussion 3.1. Structure of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in air and after reduction After firing the reaction precursors for Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ at 1400 °C for 40 h, the XRD pattern of the Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ powders may bepindexed with an orthorhombic lattice parameter of p ffiffiffi ffiffiffi 2ap  2ap  2ap , where ap ≈ 4 Å, and is the lattice parameter of the cubic primitive perovskite. The ionic sizes for Cr3+ and Ni2+ are 0.615 Å and 0.69 Å respectively for six coordination. The ionic sizes for the possible species Cr4+ are 0.55 Å, for Ni3+ are 0.56 (LS) and 0.60 Å (HS) [25]. Anyway, the ionic sizes of chromium and nickel are quite close therefore ordering is unlikely. The most common space group for unordered perovskite is Pnma(62), corresponding to the a−b+a− tilting system using Glazer's symbol [26,27]. The structure of La0.75Sr0.25CrO3 at room temperature is ¯c(167) by neutron diffraction rhombohedral with space group R3 [28]. Replacement of lanthanum at the A-sites by praseodymium and introduction of nickel at the B-sites change the Goldschmidt tolerance factor which is related to the symmetry of perovskite oxides. The Goldschmidt tolerance factor is defined as [29], ðrA þ rO Þ t ¼ pffiffiffi 2ð r B þ r O Þ

ð1Þ

S. Tao, J.T.S. Irvine / Solid State Ionics 179 (2008) 725–731

where rA, rB and rO are the ionic radii of the constituent at Asites, B-sites and oxygen ions. Ullmann et al. [30] investigated the relationship between the Goldschmidt tolerance factor and symmetry in perovskite oxides. After examining 54 various perovskites, they find that the borders from monoclinic → orthorhombic → cubic → and hexagonal are at t ≈ 0.89, 1.01, 1.07 respectively. The rhombohedral structure fits into the cubic range. The praseodymium ionic size is smaller than lanthanum under the same environment due to lanthanide contraction although the exact ionic size of Pr3+ at twelve coordination is unavailable. Therefore it is expected that Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ is likely to have a lower t value than La0.75Sr0.25CrO3. From this point of view, it is logical for Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ to exhibit an orthorhombic structure. Therefore, the space group Pnma (62) was chosen for Rietveld refinement using the GSAS program suite. The two A-sites are shared by both praseodymium and strontium and B-sites by chromium and nickel (Table 1). During the refinement, the atom position and the thermal factor of the shared sites were constrained to be equal. Refinement of the oxygen site occupancy was not very conclusive due to the insensitivity of powder X-ray diffraction to small variations in oxygen content in the presence of higher atomic number elements although the refinements were consistent with full occupation of these sites. The oxygen occupancy at the 4c and 8d sites was fixed at unity during the refinement. After the refinement, reasonable thermal factors, R-values and good pattern fit were obtained indicating that it is a reasonable model. The observed, calculated and difference profiles for the refinement of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ are shown in Fig. 1a. The final refined structure data are given in Table 1. Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ obtained at 1400 °C has been shown to have an orthorhombic structure with space group Pnma (62), a = 5.4388(1)Å, b = 7.6969(1)Å, c = 5.4584(1)Å, V = 228.50(1)Å3 according to X-ray diffraction. To examine the stability of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ under SOFC anode condition, the powder obtained at 1400 °C was further reduced in 5% H2 at 900 °C for 190 h. XRD analysis of reduced sample indicates that the symmetry of the perovskite is unchanged and no nickel was observed. A good fit was achieved using the same model Pnma (62) indicating that the structure remained during the reduction. The observed, calculated and difference profiles for the refinement of reduced Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ are shown in Fig. 1b. Table 2 lists the Table 1 Structure parameters of La0.7Sr0.3Cr0.9Ni0.1O3 − δ sample prepared in air at 1400 °C

727

Fig. 1. X-ray diffraction profiles of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ obtained at 1400 °C (a) and after further reducing in 5% H2 at 900 °C for 190 h (b).

final refined structure data. After reducing in 5% H2 at 900 °C for 190 h, Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ still exhibits an orthorhombic structure with space group Pnma (62), a = 5.4532(2)Å, b = 7.7130(2)Å, c = 5.4660(1)Å, V = 229.90(1)Å3 . A lattice volume expansion of 0.61% was observed during the reduction, which may be attributed to the loss of lattice oxygen. The reduction of Pr, Cr and/or Ni ions due to loss of lattice oxygen may lead to increased ionic size resulting in lattice expansion. The loss of lattice oxygen itself may lead to the formation of oxygen vacancies which may also cause lattice expansion. The thermal factors of the metal ions increase in the reduced samples (Tables 1 and 2).

Table 2 Structure parameters of La0.7Sr0.3Cr0.9Ni0.1O3 − δ sample prepared in air at 1400 °C then reduced in 5% H2 at 900 °C for 190 h

Atom Site Occupancy x

Y

Z

Uiso (Å2)

Atom Site Occupancy x

Y

Pr Sr Cr Ni O(1) O(2)

0.25 0.25 0 0 0.25 0.0209(29)

0.0027(9) 0.0027(9) 0.5 0.5 − 0.0610(37) − 0.2932(28)

0.0118(5) 0.0118(5) 0.0115(8) 0.0115(8) 0.0124(28) 0.0124(28)

Pr Sr Cr Ni O(1) O(2)

0.25 0.0003(29) 0.25 0.0003(29) 0 0.5 0 0.5 0.25 0.039(7) 0.0316(31) − 0.3194(35)

4c 4c 4b 4b 4c 8d

0.7 0.3 0.9 0.1 1 1

0.0197(4) 0.0197(4) 0 0 0.4875(37) 0.3000(31)

Note. Space group Pnma (62); a = 5.4388(1)Å, b = 7.6969(1)Å, c = 5.4584(1)Å, V = 228.50(1)Å3, Z = 4. Rwp = 4.81%, Rp = 3.49%, χ2red = 1.823. From room temperature XRD data.

4c 4c 4b 4b 4c 8d

0.7 0.3 0.9 0.1 1 1

0.0202(6) 0.0202(6) 0 0 0.491(4) 0.303(4)

Z

Uiso (Å2) 0.0242(7) 0.0242(7) 0.0269(11) 0.0269(11) 0.011(4) 0.011(4)

Note. Space group Pnma (62); a = 5.4532(2)Å, b = 7.7130(2)Å, c = 5.4660(1)Å, V = 229.90(1)Å3, Z = 4. Rwp = 4.50%, Rp = 3.28%, χ2red = 1.781. From room temperature XRD data.

728

S. Tao, J.T.S. Irvine / Solid State Ionics 179 (2008) 725–731

Table 3 Selected bond length of La0.7Sr0.3Cr0.9Ni0.1O3 − δ obtained in air and after reduction in 5% H2 at 900 °C for 190 h (from room temperature XRD data) Bond

Length in sample before reduction (Å) Length in sample after reduction (Å)

Pr/Sr–O(1)

2.417(1) 2.568(1) 2.915(1) 3.0527(1) 2.418(1) × 2 2.559(1) × 2 2.836(1) × 2 3.144(1) × 2 1.954(1) × 2 1.941(1) × 2 1.990(1) × 2

Pr/Sr–O(2)

Cr/Ni–O(1) Cr/Ni–O(2)

2.524(1) 2.575(1) 2.895(1) 2.952(1) 2.285(1) × 2 2.572(1) × 2 2.876(1) × 2 3.297(1) × 2 1.941(1) × 2 1.940(1) × 2 2.064(1) × 2

There are three redox elements in Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ therefore it is difficult to predict which element tends to loss oxygen first. In (La0.75Sr0.25)Cr0.5Mn0.5O3 − δ, it was observed that oxygen atoms between Mn–O–Mn bond lose on reduction based on XAFS study [31]. It is assumed that nickel is easier to be reduced than chromium. However, it is difficult to know that, between Pr and Ni, which is easier to be reduced in a perovskite lattice. The bond lengths of M–O (M = Pr, Ni) of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ before and after reduction are listed in Table 3. The BO6 octahedra become more distorted on reduction. However, XRD is not sensitive for oxygen for a delicate analysis. Neutron diffraction is required to better determine the position and occupancy of oxygen vacancies in the reduced Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ.

there is insignificant weight change apart from some buoyancy drift (Fig. 2a). This indicates kinetic stability of Pr0.7Sr0.3Cr0.9 Ni0.1O3 − δ. On reoxidation of the sample which has been prereduced in 5% H2/Ar at 900 °C for 190 h, 0.3 wt.% or ~0.04 oxygen atoms per formula unit are regained starting from ~500 °C (Fig. 2b) due to reoxidation. The initial weight loss before 500 °C is due to the loss of adsorbed water and gases. In comparison, the weight loss of its lanthanum analogue La0.7Sr0.3Cr0.95Ni0.05O3 − δ is more than 0.4 wt.% between room temperature and 900 °C on initial reduction (Fig. 2c). The smaller oxygen loss of Pr0.7Sr0.3 Cr0.9Ni0.1O3 − δ (Fig. 2a) and its much slower kinetics indicates significant stabilisation. To observe any possible microstructure change during reduction, scanning electron microscopy was applied. The SEM pictures of the powders before and after the reduction are shown in Fig. 3. The sample is quite homogeneous with secondary particle size of about 5 μm (Fig. 3a). The morphology of the sample did not change significantly after reducing the sample in 5% H2 at 900 °C for 190 h (Fig. 3b). EDS mapping of the elements of the reduced Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ indicates that Pr, Sr, Cr

3.2. Thermal analysis and SEM observation of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ Fig. 2 compares the thermogravimetric programmed reduction of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ performed in 5% H2/95%Ar (a) with the reoxidation of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ that had been previously reduced at 900 °C for 190 h (b) and the thermogravimetric reduction of the La0.7Sr0.3Cr0.95Ni0.05O3 − δ (c). On initial reduction

Fig. 2. TGA analysis of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in 5% H2 (a); Pr0.7Sr0.3Cr0.9 Ni0.1O3 − δ pre-reduced at 900 °C for 190 h then weight change in air (b); weight change of La0.7Sr0.3Cr0.95Ni0.05O3 − δ in 5% H2 (c).

Fig. 3. The SEM pictures of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ obtained at 1400 °C (a) and after ageing at 900 °C in 5% H2 for 190 h (b).

S. Tao, J.T.S. Irvine / Solid State Ionics 179 (2008) 725–731

and Ni were homogeneously distributed in the material (not displayed). No significant nickel evolution was observed which is consistent with the observation of XRD. This observation is different from previous reports in which nickel evolution was found from 10% Ni-doped lanthanum chromites in fuel condition [32,33]. The stability of nickel in the Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in a reducing atmosphere could be improved by the replacement of lanthanum by praseodymium because Prn+ ions at the A-sites might be reduced as well which might play a role as a buffer. The stability of nickel in a perovskite is related to the structure, components, temperature and pO2. High temperature and extreme reducing atmosphere may destroy the perovskite lattice causing nickel evolution. With the introduction of multivalent lanthanide, such as praseodymium, it is possible to introduce a small amount of nickel or less stable first row transition element into the B-sites of LnCrO3-based material to get redox stable perovskite oxides.

729

Fig. 5. The total conductivity of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in air and 5% H2 at different temperatures.

3.3. Conductivity of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ The conductivity of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ was measured by the 4-terminal d.c. method. The conductivity of Pr0.7Sr0.3Cr0.9 Ni0.1O3 − δ is higher in air than that in 5% H2 indicating p-type conduction at low pO2. The total conductivities of Pr0.7Sr0.3Cr0.9 Ni0.1O3 − δ in air and 5% H2 were measured as 27.4 and 1.37S/cm respectively at 900 °C (Fig. 4). The conductivity of La0.7Sr0.3 Cr0.95Ni0.05O3 − δ is about 3S/cm in air at 900 °C [34] which is lower than that of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ under the same condition although the dopping level of nickel is slightly different. At 150 °C, the total conductivity is 7.67S/cm in air but only 2.03 × 10− 3 S/cm in 5% H2. The apparent conduction activation energy of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in air is 0.13± 0.01 eV between 92–900 °C; in 5% H2 0.38 ± 0.02 eV between 120–570 °C and 0.46 ± 0.02 eV between 570–900 °C indicating it is a semiconductor in both atmospheres. The slope change at around 570 °C on the conductivity curve in 5% H2 (Fig. 5) might be attributed to a phase transition which is common in perovskite oxides [34]. The material is a p-type conductor at low oxygen partial pressure. As shown in Fig. 6, at pO2 values below 10− 15 atm, the conductivity decreases with decreasing pO2. The logσ vs.

Fig. 4. Isothermal conductivity vs. pO2 diagram for Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ starting at 900 °C.

logpO2 relation of p-type electronic conductivity at low pO2 gives a slope 1/4. The defect chemistry in this material would be quite complicated since multivalent elements are involved in both A- and B-sites. The high electronic conductivity of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in air will involve the 3d orbitals of the ions at the B-sites. The charge compensation for introduction of strontium at the A-sites may cause the charge change from Cr3+ to Cr4+ and/or Ni2+ to Ni3+. In this case the concentration of electronic holes increase therefore the p-type conductivity increases. The conductivity of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in air is about one order of magnitude higher than its lanthanum analogue La0.7Sr0.3Cr0.95Ni0.05O3 − δ [34]. The more nickel at the B-site, the more Cr4+ ions at the B-sites for charge compensation in air if the formation of oxygen vacancies is not favoured. Therefore the conductivity of Pr0.7Sr0.3Cr0.9Ni0.1 O3 − δ should be higher due to more nickel ions in the lattice. The cell volume of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in air is 228.50(1)Å3 which

Fig. 6. The anode polarisation of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ anode in different wet atmospheres (3% H2O).

730

S. Tao, J.T.S. Irvine / Solid State Ionics 179 (2008) 725–731

is smaller than 232.80 Å3 for La0.7Sr0.3Cr0.95Ni0.05O3 − δ. The small lattice means shorter A–O and B–O bonding which would facilitate charge transfer through the M–O bonds if some multivalent lanthanide such as praseodymium is involved. On the other hand, short M–O bond also implies high binding energy which may hinder the charge transfer as well. The Pr3+ and Pr4+ ions at the A-site might have contributions to the high conductivity of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in air which is consistent with a previous report in manganite oxides [23] although other effects such as lattice contract could be dominant. 3.4. Anode performance of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ in wet H2 and CH4 The impedance spectra of the pure Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ anode in wet 5% H2/Ar, wet H2 and wet CH4 at open circuit voltage (OCV) are shown in Fig. 6 (3 electrode test configuration). The anode polarisation resistance in wet H2 was about 1.49 Ωcm2 at 850 °C but improved to 0.98 Ωcm2 at 900 °C. The anode polarisation resistances in wet 5% H2 and wet CH4 were 2.39 and 3.84 Ωcm2 respectively at 900 °C. Carbon deposition was not observed after the experiment when CH4 was used as the fuel. A large response at low frequency for all atmospheres was observed. It is believed that the low frequency response is related to noncharge processes, mainly gas diffusion and surface exchange [10,35–37]. This is related to the microstructure of the anode, the current collector and the gas diffusion therefore could be improved. In addition, the anode performance may be further optimised by integrating some YSZ with the anode to improve the anode/electrolyte interface which is demonstrated to be successful in the LSCM anode [21]. 4. Conclusions Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ obtained at 1400 °C has been shown to have an orthorhombic structure with space group Pnma (62), a = 5.4388(1)Å, b = 7.6969(1)Å, c = 5.4584(1)Å, V = 228.50(1)Å3 according to X-ray diffraction. The material is redox stable and maintains its structure in a reducing atmosphere. After reducing in 5% H2 at 900 °C for 190 h, Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ exhibits the same structure. A lattice volume expansion of 0.61% was observed during the reduction, which may be attributed to reduction of Pr, Cr and Ni ions accompanying with the loss of oxygen. TGA analysis indicates that Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ is very resistant to nickel reduction especially at short time scale. The morphology of this material does not significantly change on reduction according to SEM observation. Element mapping indicates nickel is homogenously distributed in the material and nickel evolution is not detected. The conductivities of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ is about one magnitude higher than its La-analogue which mean multivalent lanthanide could be better for A-sites for mixed conducting perovskite oxides. The enhanced conductivity with in Pr-containing perovskite has been observed by other group as well [23]. The decrease of d.c. conductivity of Pr0.7Sr0.3Cr0.9 Ni0.1O3 − δ at low pO2 indicates p-type electronic conduction.

The d.c. conductivity of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ at low pO2 exhibits a pO21/4 dependence. The anode polarisation resistance of Pr0.7Sr0.3Cr0.9Ni0.1O3 − δ reached 0.98 Ωcm2 at 900 °C in wet H2 but still not good enough as a good SOFC anode but may be further improved by tailoring microstructure and integrating some electrolyte materials into the anode to improve the anode/ electrolyte interface. Acknowledgements We would like to thank the EPSRC, European SOFC-600 and EaStCHEM for funding. Tao thanks EaStCHEM for a fellowship and Irvine thanks EPSRC for a senior fellowship. References [1] N.Q. Minh, J. Am. Ceram. Soc. 76 (1993) 563. [2] Y. Matsuzaki, I. Yasuda, Solid State Ionics 132 (2000) 261. [3] B.C.H. Steele, I. Kelly, H. Middleton, R. Rudkin, Solid State Ionics 28–30 (1988) 1547. [4] S.W. Tao, J.T.S. Irvine, Chem. Rec. 4 (2004) 83. [5] G. Pudmich, B.A. Boukamp, M. Gonzalez-Cuenca, W. Jungen, W. Zipprich, F. Tietz, Solid State Ionics 135 (2000) 433. [6] O.A. Marina, L.R. Pederson, in: J. Huijsmans (Ed.), Proc. 5th European Solid Oxide Fuel Cell Forum, 2002, p. 481, (European SOFC Forum, Switzerland ). [7] O.A. Marina, N.L. Canfield, J.W. Stevenson, Solid State Ionics 149 (2002) 21. [8] J. Canales-Vázquez, S.W. Tao, J.T.S. Irvine, Solid State Ionics 159 (2003) 159. [9] S. Primdahl, J.R. Hansen, L. Grahl-Madsen, P.H. Larsen, J. Electrochem. Soc. 148 (2001) A74. [10] P. Holtappels, J. Bradley, J.T.S. Irvine, A. Kaiser, M. Mogensen, J. Electrochem. Soc. 148 (2001) A923. [11] J.C. Ruiz-Morales, J. Canales-Vazquez, C. Savaniu, D. Marrero-Lopez, W.Z. Zhou, J.T.S. Irvine, Nature 439 (2006) 568. [12] Y.H. Huang, R.I. Dass, Z.L. Xing, J.B. Goodenough, Science 312 (2006) 254. [13] B.D. Madsen, W. Kobsiriphat, Y. Wang, L.D. Marks, S.A. Barnett, ECS Trans. 7 (2007) 1339. [14] H. Yokokawa, N. Sakai, T. Kawada, M. Dokiya, Solid State Ionics 52 (1992) 43. [15] P. Vernoux, M. Guillodo, J. Fouletier, A. Hammou, Solid State Ionics 135 (2000) 425. [16] T. Nakamura, G. Petzow, L.J. Gauckler, Mater. Res. Bull. 14 (1979) 649. [17] J. Sfeir, P.A. Buffat, P. Möckli, N. Xanthopoulos, R. Vasquez, H.J. Mathieu, J. Van Herle, K.R. Thampi, J. Catal. 202 (2001) 229. [18] J. Sfeir, J. Van Herle, R. Vasquez, in: J. Huijsmans (Ed.), Proc. 5th European Solid Oxide Fuel Cell Forum, 2002, p. 570, (European SOFC Forum, Switzerland). [19] J. Liu, B.D. Madsen, Z.Q. Ji, S.A. Barnett, Electrochem. Solid State Lett. 5 (2002) A122. [20] S.W. Tao, J.T.S. Irvine, Nat. Mater. 2 (2003) 320. [21] S.W. Tao, JT.S. Irvine, J. Electrochem. Soc. 151 (2004) A252. [22] S.W. Tao, J.T.S. Irvine, Chem. Mater. 16 (2004) 4116. [23] H. Ullmann, N. Trofimenko, F. Tietz, D. Stover, A. Ahmad-Khanlou, Solid State Ionics 138 (2000) 79. [24] A.C. Larson, R.B. Von Dreele, GSAS-Generalised Crystal Structure Analysis System, Los Alamos National Laboratory Report No. LA-UR86-748, 1994. [25] R.D. Shannon, Acta Crystallogr. A32 (1976) 751. [26] M.W. Lufaso, P.M. Woodward, Acta Crystallogr. B57 (2001) 725. [27] A.M. Glazer, Acta Crystallogr. B28 (1972) 3384. [28] K. Tezuka, Y. Hinatsu, A. Nakamura, T. Inami, Y. Shimojo, Y. Morii, J. Solid State Chem. 141 (1998) 404. [29] V.M. Goldschmidt, Naturwissenschaften 14 (1926) 477.

S. Tao, J.T.S. Irvine / Solid State Ionics 179 (2008) 725–731 [30] H. Ullmann, N. Trofimenko, J. Alloys Compd. 316 (2001) 153. [31] S.W. Tao, J.T.S. Irvine, S.M. Plint, J. Phys. Chem., B 110 (2006) 21771. [32] A.-L. Sauvet, J.T.S. Irvine, in: J. Huijsmans (Ed.), Proc. 5th European Solid Oxide Fuel Cell Forum, 2002, p. 490, (European SOFC Forum, Switzerland). [33] A.L. Sauvet, J.T.S. Irvine, Solid State Ionics 167 (2004) 1.

731

[34] S.W. Tao, J.T.S. Irvine, Chem. Mater. 18 (2006) 5453. [35] S. Primdahl, M. Mogensen, J. Electrochem. Soc. 146 (1999) 2827. [36] S.B. Adler, J.A. Lane, B.C.H. Steele, J. Electrochem. Soc. 143 (1996) 3554. [37] S.W. Tao, J.T.S. Irvine, J. Electrochem. Soc. 151 (2004) A497.