J. E. H. Sansom

Solid state 29Si NMR studies of apatite-type oxide ion conductors

Apatite-type silicates have been attracting considerable interest as a new class of oxide ion conductor, whose conduction is mediated by interstitial oxide ions. We report here the first 29Si solid state NMR studies of these materials with a systematic investigation of thirteen compositions. Our results indicate a correlation between the silicon environment and the observed conductivity. Specifically, samples which show poor conductivity demonstrate a single NMR resonance, whereas fast ion conducting compositions show more complex NMR spectra. For the oxygen excess samples La9M(SiO4)6O2.5 (M = Ca, Sr, Ba) two peaks are observed at chemical shifts of ≈−77.5 and −80.5 ppm, with the second peak correlated with a silicate group adjacent to an interstitial oxygen site. On Ti doping to give La9M(SiO4)6−x(TiO4)xO2.5 (x = 1,2) the second peak disappears, which is consistent with the “trapping” of interstitial oxygens by Ti and the consequent lowering in oxide ion conductivity. The samples La9.33(SiO4)6O2 and La9.67(SiO4)6O2.5 show a further third weak peak at a chemical shift (≈−85.0 ppm) consistent with the presence of some [Si2O7]6− units in these samples, due to condensation of two [SiO4]4− units. The effect of such condensation of [SiO4]4− units will be the creation of additional interstitial oxide ion defects, i.e. 2 [SiO4]4− → [Si2O7]6− + Oint2−. Overall, the results further highlight the importance of the [SiO4]4− substructure in these materials, and additionally suggest that 29Si NMR could potentially be used to screen apatite silicate materials for oxide ion conductivity

Neutron diffraction and atomistic simulation studies of Mg doped apatite-type oxide ion conductors

In this paper, detailed studies of the effect of Mg doping in the apatite-type oxide ion conductor La9.33Si6O26 are reported. Mg is confirmed as an ambisite dopant, capable of substituting for both La and Si, depending on the starting composition. A large enhancement in the conductivity is observed for Si site substitution, with a reduction for substitution on the La site. Neutron powder diffraction studies show that in agreement with cation size expectations, an enlargement of the unit cell is observed on Mg substitution for Si, with a corresponding increase in the size of the tetrahedral sites. For Mg substitution on the La site, a contraction of the unit cell is observed, and the neutron diffraction results indicate that there is preferential occupancy of Mg on the La2 (1/3, 2/3, approximately 0.5) site. Atomistic simulation studies show significant local structural changes affecting the oxide ion channels in both cases. Mg doping on the Si site leads to a local expansion of the channels, while doping on the La site results in a large displacement of the silicate O4 site, such that it encroaches the oxide ion channels. The observed differences in conductivities are discussed with respect to these observations.

Doping strategies to optimise the oxide ion conductivity in apatite-type ionic conductors

The apatite-type phases, La(9.33+x)(Si/Ge)(6)O(26+3x/2), have recently been attracting considerable interest as potential electrolytes for solid oxide fuel cells. In this paper we report results from a range of doping studies in the Si based systems, aimed at determining the key features required for the optimisation of the conductivities. Systems examined have included alkaline earth doping on the rare earth site, and P, B, Ga, V doping on the Si site. By suitable doping strategies, factors such as the level of cation vacancies and oxygen excess have been investigated. The results show that the oxide ion conductivities of these apatite systems are maximised by the incorporation of either oxygen excess or cation vacancies, with the former producing the best oxide ion conductors. In terms of samples containing cation vacancies, conductivities are enhanced by doping lower valent ions, Ga, B, on the Si site. The presence of higher valent ions on these sites, e.g. P, appears to inhibit the incorporation of excess oxygen within the channels, and so limits the maximum conductivity that can be obtained. Overall the results suggest that the tetrahedral sites play a key role in the conduction properties of these materials, supporting recent modelling studies, which have suggested that these tetrahedra aid in the motion of the oxide ions down the conduction channels by co-operative displacements.

Structural studies of apatite-type oxide ion conductors doped with cobalt

A series of Co doped lanthanum silicate apatite-type phases, La9.83Si4.5Co1.5O26, La9.66Si5CoO26, La10Si5CoO26.5 and La8BaCoSi6O26, have been synthesised, and neutron diffraction, EXAFS and XANES used to investigate their structures in detail. All compositions were shown to possess the hexagonal apatite structure, and the results confirmed that cobalt can be doped onto both the La and Si sites within the structure depending on the starting composition. The Co doping is shown to cause considerable local distortions within the apatite structure. In the case of Si site doping two compositions showed anisotropic peak broadening, which has been attributed to incommensurate ordering of oxygen within the apatite channels.

Synthesis and characterisation of the perovskite-related cuprate phases YSr2Cu2MO7+y (M = Co, Fe) for potential use as solid oxide fuel cell cathode materials

In this paper we report the synthesis and characterisation of the perovskite cuprate phases YSr2Cu2MO7+y (M = Co, Fe) in order to examine their potential for use as cathode materials in solid oxide fuel cells (SOFCs). Both samples showed conductivities of ≈10 S cm−1 at 900 °C and were also shown to be stable at this temperature in N2. For YSr2Cu2FeO7+x, semiconducting behaviour was observed up to ≈550 °C, with a decrease in conductivity at higher temperatures, attributed to oxygen loss reducing the charge carrier concentration. In the case of YSr2Cu2CoO7+y, semiconducting behaviour was observed over the range of temperatures studied, although a small but significant steep increase in conductivity was observed above 800 °C. High temperature X-ray diffraction studies of this particular phase showed that this increase in conductivity coincided with an orthorhombic–tetragonal structural transition, accompanied by a significant reduction in cell volume. In addition to measurements in air, conductivities were also measured with varying p(O2) (0.2–10−5 atm) at 900 °C, and these data showed significant hysteresis between measurements on reducing and re-oxidising, suggesting poor oxide ion transport, poor oxygen surface exchange kinetics, or significant structural changes on varying p(O2). Chemical compatibility studies of these phases with SOFC electrolytes at temperatures between 900 and 1000 °C showed reaction in all cases. In the case of CeO2 based electrolytes, the reaction led to the formation of the “fluorite-block” phases, (Y/Ce)2Sr2Cu3−xMxO9+y (M = Co, Fe), and samples of these were subsequently prepared and the conductivities measured. Similar hysteresis between conductivity measurements on reducing and re-oxidising were also observed for these samples.

Synthesis and electrical characterisation of the apatite-type oxide ion conductors Nd9.33+xSi6-yGayO26+z

Apatite-type oxides have been attracting interest as a new class of oxide ion conductors. In this paper we examine the effect of Ga doping on the conductivity of the apatite silicate system, Nd9.33+xSi6O26+3x/2 and compare the results to those reported for similar doping studies in La9.33+xSi6O26+3x/2. The highest conductivities are observed for samples containing oxygen excess, which is in agreement with previous reports that interstitial oxide ions are important for high oxide ion conduction in these materials. For oxygen stoichiometric materials, i.e. Nd9.33+x/3Si6-xGaxO26, the Ga doping results in a significant increase in activation energy and a consequent lowering of the low temperature conductivity. This is contrary to results previously reported for the La containing analogues, which showed an enhancement of conductivity on Ga doping up to x=1.5. Possible explanations for the differences between the two systems are discussed.