High Temperature Oxide Ion Conduction Path for Pb1-XLaxWO4+x/2

Abstract

Introduction Partially substituting lanthanum ions for the Scheelite-type structured PbWO4, oxide ion interstitials are formed as Pb1-xLaxWO4+x/2 and high oxide ion conduction appears at high temperatures [1]. While most of the oxide ion conductors including zirconias, cerias, or LSGM employ the oxide ion vacancy for the high-oxide ion conduction, the present system utilize the slight amount of oxide ion interstitials. Since there is no sufficient space for the oxide ion to migrate through only the interstitial sites, oxide ion diffusion should be made using regular and interstitial sites. However, unlike the vacancy diffusion mechanism, actual diffusion path is hardly estimated without direct experiment such as neutron diffraction. We have previously measured the neutron diffraction on Pb1-xLaxWO4+x/2(x = 0.1 and 0.2) at room temperature to evaluate the diffusion path from the interatomic distances and direction of anisotropic temperature factors of regular oxide ions [2]. Nevertheless, the estimated model only support the 2 D diffusion path, and oxide ion distribution at high temperature would be different from that at room temperature.We therefore carried out the neutron diffraction experiments on Pb0.8La0.2WO4.1in the temperature range from 200°C to 800°C and estimated the high temperature conduction path of oxide ions by means of maximum entropy method (MEM). Experimental Pb0.8La0.2WO4.1 sample has been prepared by solid state reaction method. Stoichiometric mixtures of PbO, H2WO4 and La2O3 were at first calcined at 800°C in air for 10 h, finely ground, pressed into pellets under a hydrostatic pressure of 100 MPa, and finally sintered at 900°C in air for 10 h. The base material of PbWO4 was also prepared for comparison. Crystalline phase of sintered sample was identified by X-ray diffraction measurement. About 10 g of the sample pellets were placed in vanadium holder through a thin quartz tube and the neutron powder diffraction experiments were performed in the temperature range between 200°C and 800°C by using SHRPD diffractometer in J-PARC and the Rietveld refinements were carried out by using Z-Rietveld code, and the oxide ion conduction path at high temperatures was estimated with maximum entropy method using Z-MEM. Results and Discussions Fig.1 represent the refined lattice parameters, a and c, of Pb1-xLaxWO4+x/2 (x = 0 and 0.2), both of which grow larger with increasing temperature. On the other hand, the interatomic distances between W and O, d W-O, gradually decreases with temperature in x = 0.2, while it grows larger for x = 0 as shown in Fig. 2. This indicates that the regular oxide ion is distorted by the diffusing oxide ions through the interstitial sites at high temperature. As for the oxide ion interstitials, two possible sites are deduced for the present high-temperature diffraction data in x = 0.2. The fractional coordinates of Oi(1) and Oi(2) are (0 0 0.208) and (0.667 0.333 0.238), respectively, which are represented in Fig. 3. The former corresponds to the previously reported one obtained from the room temperature neutron diffraction data. On the other hand, the latter was found by the present high-temperature diffraction experiments. Therefore, a new diffusion route through Oi(2) site is expected at high temperature in comparison with the previous route estimated at room-temperature. Fig.4 shows the nuclear density diagram of Pb0.8La0.2WO4.1 at 800°C. It is indicated that oxide ions migrate through Oi(2) as well as Oi(1) in Pb0.8La0.2WO4.1. It is indicated that actual diffusion path through Oi(1) (black arrow) is rather complex in comparison with that estimated from previous room temperature data (red arrow). In addition, additional new diffusion path is found through both Oi(1) and Oi(2) in the present high temperature experiment. Using these two typical conduction paths, oxide ion can migrate to achieve 3 D diffusion network. On the other hand, PbWO4doesn’t have such diffusion path even at high temperature.