Solid state ionic conductors based on Lu-doped δ-Bi2O3

Abstract

Regarding limited fossil energy resources and hence the increasing demands for new highly efficient and environmentally friendly energy conversion devices, the widespread use of solid oxide fuel cells (SOFCs) might become a keystone in near future. An SOFC is consisted of dense electrolyte which is sandwiched between two porous electrodes. Since the electrolyte is the most important part of an SOFC, oxide ion conductors applicable in SOFCs became the hot topic of modern research. The main requirement is to find a stable dense electrolyte material with increased conductivity at intermediate temperature. Two commercial electrolytes, yttria stabilized zirconia (YSZ) and gadolinia doped ceria (GDC), are far to be ideal. The use of YSZ requires operating temperatures above 1000 °C while devices based on GDC are not efficient enough due to GDC lower conductivity at intermediate temperatures. The best candidate for an electrolyte in intermediate temperature SOFCs is undoubtedly fluorite structured bismuth oxide, i.e. δ-Bi2O3, being the fastest known ionic conductor. However, this material is unstable upon heating/cooling but the use of dopant, such as Tm, led to the impressive stability and high conductivity at intermediate temperatures [1]. Our findings indicated that Lu-doped δ-Bi2O3 could show even better performances since Lu is smaller and more rigid dopant than Tm. In this study, the possibility to stabilize -Bi2O3 in Bi2O3–Lu2O3 system was investigated. Two starting mixtures of α-Bi2O3 and Lu2O3 with the following compositions (Bi1–xLux)2O3, x = 0.20 and 0.25, were dry homogenized in an agate mortar, heat treated at 750 °C for 3 h and then slowly cooled. The obtain powders were characterized by XRD and DTA techniques. Based on these results, the targeted cubic single-phase δ-Bi2O3 was successfully obtained within both systems. The unit cell parameter of obtained Lu-doped -Bi2O3 decreases as the dopant content increases, as expected since Lu3+ is smaller cation than Bi3+. According to cyclic DTA curves, no phase transitions (25 – 980 °C) were observed for both phases, (Bi0.8Lu0.2)2O3 and (Bi0.75Lu0.25)2O3, indicating that these -Bi2O3 phases are stable. Afterwards, half of obtained quantity of each powder was mechanochemically treated (in planetary ball mill Retsch PM-100) in order to decrease the crystallite size. Namely, before performing EIS measurements the powders undergo to pressing and then sintering to obtain dense ceramic pellets. This is also important for their future application in SOFC since the density of almost 100% is mandatory in order to avoid the direct contact of air (oxygen) and fuel (hydrogen). The density of sintered pellets obtained from both untreated and mechanochemically treated powders will be compared. Using EIS technique, the ionic conductivity will be measured for the samples having the highest density. We expect that these materials will exhibit the conductivities which are higher than those obtained for Tm-doped -Bi2O3 (0.1 – 0.4 S cm–1 at 550 – 800 °C). Such stability and extraordinary conductivity would open the possibility for application of (Bi0.8Lu0.2)2O3 and (Bi0.75Lu0.25)2O3, which could result in the significant enhancement of electrochemical performance of intermediate temperature SOFCs but also in their good stability over long time service.