Article pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
The Fluorite-Like Phase Nd5Mo3O16±δ in the MoO3−Nd2O3 System: Synthesis, Crystal Structure, and Conducting Properties Jordi Jacas Biendicho,*,† Helen Y. Playford,‡ Seikh M. H. Rahman,§ Stefan T. Norberg,§ Sten G. Eriksson,§ and Stephen Hull‡ †
Catalonia Institute for Energy Research, Jardins de les Dones de Negre 1, 08930 Sant Adrià del Besos, Spain The ISIS Facility, STFC Rutherford Appleton Laboratory, Chilton, Didcot, Oxfordshire OX11 0QX, United Kingdom § Department of Chemistry and Chemical Engineering, Chalmers University of Technology, SE-412 96 Gothenburg, Sweden ‡
S Supporting Information *
ABSTRACT: This paper describes a study of the system MoO3− Nd2O3 using a combination of X-ray powder diffraction (XRD), neutron powder diffraction (NPD), thermogravimetric analysis (TGA), and ac impedance spectroscopy (IS). A phase-pure material is observed at a composition of 45.5 mol % Nd2O3, which corresponds to an ideal stoichiometry of Nd5Mo3O16.5. XRD and NPD show that the crystal structure is a superstructure of the fluorite arrangement, with long-range ordering of the two cation species leading to a doubled unit cell parameter. The sample is found to be significantly oxygen deficient, i.e. Nd5Mo3O15.63(4), when it is prepared by a solid-state reaction at 1473 K in air. TGA measurements indicate that the sample loses only minimal mass on heating to 1273 K in O2. IS studies of the mean conductivity under different atmospheres show that the sample is a mixed conductor between ambient temperature and 873 K, with a dominant electronic component at higher temperatures, as demonstrated by measurements under inert atmosphere. NPD measurements indicate that the anion vacancies are preferentially located on the O2 sites, while studies of the temperature dependence performed under an O2 atmosphere to 1273 K show significantly anisotropic thermal parameters of the anions. Together with analysis of the total neutron scattering data, this supports a model of oxygen ions hopping between O2 positions, with a vacancy, rather than interstitial, mechanism for the anion diffusion.
1. INTRODUCTION Solid oxide fuel cells (SOFCs) are currently the subject of considerable research worldwide, as a potentially highly efficient and environmentally benign technology to generate electrical power. Their high temperature of operation (typically over 1000 K) allows the direct use of hydrocarbon fuels, offering a potential route to exploit the lower emissions associated with fuel cells, but without the need for an extensive supply network for pure hydrogen. However, there are a number of practical difficulties to be overcome, including relatively long start-up times, gas sealing of the device, and poisoning by sulfur impurities in the fuel supply. For an introduction to SOFCs and a more detailed discussion of these issues, see refs 1 and 2. Oxide ceramics based on molybdenum(VI) oxide, MoO3, have been shown to be relatively resistant to sulfur impurities,3 and several possess high values of oxide ion conductivity at elevated temperatures required to fulfill the role of a solid electrode or electrolyte within SOFCs.4,5 These include La2Mo2O9,6 which has been shown to possess an ionic conductivity at elevated temperatures comparable to that the currently favored electrolyte material, the anion-deficient, fluorite-structured ZrO2 doped with Y2O3.7 An alternative stoichiometry, R2MoO6, has been reported to occur across the © XXXX American Chemical Society
majority of the series of R = rare earth, Y, and La systems, with the crystal structure formed being dependent on the size of the R cation and, in some cases, the synthesis method.8 For the largest trivalent cation, R = La, a tetragonal structure (labeled γ) in space group I41/acd has been determined for La2MoO69 and, in the case of the smaller rare earths, compounds R2MoO6 with R = Tb, Dy, Ho, Er, Tm, Yb, and Y have been shown to adopt a monoclinic (α) structure in space group C2/c.8,9 A cubic phase labeled (β) has been observed for the compound Ce2MoO68,10 and appears to be an ordered superstructure of the fluorite arrangement, since the XRD pattern show extra reflections that are not indexed in Fm3̅m symmetry. Indeed, the structures of the three α, β, and γ phases are all believed to be related to the cubic fluorite structure.9 Compounds with a closely related stoichiometry, R5Mo3O16.5, with R = La, Pr, Nd, Sm, and Gd, were shown to possess high conductivities by Tsai et al. in 1989.11 In the case of Nd 5 Mo 3 O 16.5 , the system was more recently “rediscovered” by Voronkova et al.12 (without citing the earlier work) and shown to achieve values of conductivity close to Received: March 19, 2018
A
DOI: 10.1021/acs.inorgchem.8b00734 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Table 1. Compositions in the System (Nd2O3)x(MoO3)1−x Prepared by Solid-State Reactions in Air, Including the Composition Determined by XRF Methods and the Weight Percent of the Secondary Phase As Obtained by Rietveld Refinement of the NPD Data nominal (wt %)
XRF (wt %)
composition x in (Nd2O3)1−x(MoO3)x
Nd
Mo
Nd
Mo
3/7 = 0.429 11/25 = 0.440 5/11 = 0.455 7/15 = 0.467 9/19 = 0.474 12/25 = 0.480
69.28 70.26 71.47 72.45 73.01 73.51
30.72 29.74 28.53 27.55 26.99 26.49
68.6 69.8 70.8 71.8 72.1 72.3
31.4 30.2 29.2 28.2 27.9 27.7
10−2 Ω−1 cm−1 at 1073 K. This has led to significant interest in the structural and conducting properties of Nd5Mo3O16.513−16 and, to a lesser extent, its Pr analogue,15,17,18 including studies of the effects of cation doping onto the Nd3+ 19−22 and Mo6+ 23 sites. The crystal structure of Nd5Mo3O16.5 can be described as a supercell of the cubic fluorite structure with the lattice parameter a ≈ 2afluorite, space group Pn3n̅ , long-range ordering of the Nd3+ and Mo6+ over the cation sites, and significant displacements of the anions away from their locations within an ideal fluorite lattice.13,15 However, on the basis of the formal cation valences, the metal to oxygen ratio is 1:2.0625, implying anion excess with respect to the conventional 1:2 ratio for stoichiometric fluorite-structured compounds. To reflect this, the chemical formula is better written as Nd5Mo3O16+δ, where 0 < δ ≤ 0.5 corresponds to an anion-excess fluorite structure, with full Mo6+ oxidation at δ = 0.5. At δ = 0, the compound has the ideal fluorite stoichiometry, with reduction of some Mo6+ to Mo5+, while δ < 0 describes an anion-deficient structure in which all the Mo6+ cations are reduced to Mo5+ at δ = −1.0. Within the literature, there are reports of both anionexcess11,12,15 and anion-deficient13 cases. In this paper, we have prepared several compositions in the (Nd2O3)x(MoO3)1−x system by solid-state reactions to establish the stability range of the fluorite-like phase. X-ray powder diffraction, neutron powder diffraction, thermogravimetric analysis, and impedance spectroscopy measurements have been performed under different atmospheres and as a function of temperature, to investigate the structure and conducting properties of the material.
phase analysis (wt %) using Rietveld refinement of NPD data 2afluorite; 85(1) 2afluorite; 94(1)
Nd2Mo3O12; 15(1) Nd2Mo3O12; 6(1) 2afluorite; 100 2afluorite; 88(1) Nd2MoO6; 12(1) 2afluorite; 74(1) Nd2MoO6; 26(1) 2afluorite; 56(1) Nd2MoO6; 44(1)
2.3. Neutron Powder Diffraction (NPD). Neutron powder diffraction (NPD) data were collected at the Polaris diffractometer at the ISIS neutron source, STFC Rutherford Appleton Laboratory, U.K.,25 with high-temperature studies performed using a standard furnace constructed with a vanadium heater element and heat shields. Approximately 10 g of sample, in the form of pellets of 6 mm diameter, were placed in a flow-through quartz gas cell26 and heated from 300 to 1272 K in O2. Typical data collection times were 60 min, with a 30 min stabilization time prior to each measurement. The pellets did not show any significant change in pleochroism after in situ measurement. The sample temperature was measured using a type K thermocouple located ∼5 mm above the sample, while Rietveld refinement of the NPD data used the program GSAS.24 The Bragg peak profile was described using function type 3 in GSAS (a convolution of a pseudoVoigt peak shape with two back-to-back exponential functions) and the background fitted using function type 2 (a cosine Fourier series). To complement the time-averaged structural models provided by Rietveld analysis of the XRD and NPD data, analysis of the total neutron scattering data (i.e. including both the Bragg and diffuse scattering components) was performed using scattering data collected at ambient temperature, normalized using the GUDRUN software27 and modeled using the RMCProfile code.28 The latter adjusts the positions of atoms within a large configuration (5 × 5 × 5 unit cells) to simultaneously fit the total scattering function, S(Q), the total pair distribution function, G(r) (obtained by Fourier transform of S(Q)), and the Bragg profile, while subject to a number of chemical constraints. For further details of this approach, see refs 29−32. 2.4. Thermogravimetric Analysis (TGA). Thermogravimetric analysis (TGA) studies were performed using a Netzsch STA 409 PC instrument. Measurements were carried out on dried (after heating to 1273 K) powder samples under two sequences from ambient temperature to 1273 K, followed by a cooling step to 473 K. Samples were measured under flowing dry O2 (H2O
