Synthesis of Co-Doped LaP3O9 by Precipitation in Phosphoric Acid Solutions

Several lanthanum phosphates exhibit proton conductivity when doped with alkaline earth metals such as Sr. Especially the electrical conductivity of Sr-doped lanthanum polyphosphate (Sr-doped LaP3O9) is larger than those of other Sr-doped lanthanum phosphates [1-4], and it does not depend on the partial pressure of H2O in contrast to other lanthanum phosphates[1-3]. Therefore LaP3O9 is one of the promising candidates for the electrolytes in intermediate-temperature fuel cells. However the conductivity of Sr-doped LaP3O9 is still on the order of 10 S cm at 500 °C[3], while ~10 S cm is required for practical use. Therefore it is necessary to improve the conductivity of doped LaP3O9. The proton conductivity is expected to increase with the concentration of protons, and protons are introduced to compensate the charge difference between La and Sr. Hence increasing Sr-doping level might be a strategy to improve the conductivity. Amezawa et al. investigated the electrical conductivity of 1 – 5 mol% Srdoped LaP3O9, and reported that the conductivity increased with Sr-doping level, but the Sr-doping level in LaP3O9 prepared by solid-state reaction method was limited to only 5 mol%. However Hatada recently reported that they prepared 14 mol% Sr-doped LaP3O9 by precipitation in phosphoric acid solutions at around 230 °C[5]. It was also reported that other divalent cations, Ca or Ba doped LaP3O9 also have protonic conductivity[6]. Therefore we tried to synthesize highly Ca or Ba doped LaP3O9 by solution synthesis. As the result of our previous work[7], the highest doping levels of Ca and Ba were about 8 mol% and 3 mol% respectively. These were smaller than that of Sr. The lower Ca and Ba doping levels might be due to the difference of the ionic radii. As the ionic radii of La, Ca, Sr and Ba in eightfold coordination are 1.16, 1.12, 1.26 and 1.42 A respectively[8], Ba is expected to generate larger lattice volume change than that generated by Sr. Therefore one of the ideas to increase the doping level is co-doping with smaller and larger dopants to suppress the lattice volume change. In this study, we investigated the possibility of doping LaP3O9 with multiple elements simultaneously by synthesis in phosphoric acid solutions. When initial composition in phosphoric acid solution was set at La : M : P = 0.6 : 0.4 : 15 in molar ratio (M = Ca, Sr, Ba), La1-xMxP3O9 was precipitated at 230 °C. Thus the initial concentrations of each dopant were set at 0.4 in molar ratio in all experiments. Figure 1 shows the lattice volume of divalent-cations-doped LaP3O9 determined by Rietveld analysis, and Figure 2 shows the doping levels determined by Inductively Coupled Plasma Atomic Emission Spectrometry (ICPAES). As seen in Figure 1, the lattice volume change was proportional to the doping levels in the case of singleelement doping, and the slopes of the dashed lines exhibit the lattice volume change per 1 mol% of the doping levels. That of Ba-doped LaP3O9 was larger than that of Sr-doped, and that of Ca-doped was negative because Ca is smaller than La. Therefore the lattice volume change is positively correlated with the difference of the ionic radii between La and dopant ions. The lattice volume changes of co-doped samples with a total doping level of 1 mol% were between those of single-element-doped samples, and were considered to be weighted-averaged. However, as seen in Figure 2, the doping levels of each individual element were not enhanced by multi-element doping. Thus, at present, we do not consider that the lattice volume change restrict the doping level, though it depends on the doping level. However in the case of Ca, Sr and Ba-doped sample, the doping levels of each dopant were almost as large as those in single-element-doped samples, and the total doping level reached 25 mol%. Since this value is twice larger than the maximum of single-element doping, it is very interesting to study the conductivity.