Enhanced Sintering of Yttrium-Doped Barium Zirconate by Addition of ZnO Peter Babilo and Sossina M. Haile w Department of Materials Science, California Institute of Technology, Pasadena, California 91125 The inﬂuence of transition metal oxides additives, especially zinc oxide, on the densiﬁcation and electrical properties of doped barium zirconate have been examined. With the use of zinc ox- ide as a sintering aid, BaZr 0.85 Y 0.15 O 3–d was readily sintered to above 93% of theoretical density at 13001C. Scanning electron microscopic investigations showed Zn accumulation in the in- tergranular regions. Thermogravimetric analysis of the material under ﬂowing CO 2 showed ZnO-modiﬁed barium zirconate to exhibit excellent chemical stability. The conductivity, as meas- ured by A.C. impedance spectroscopy under H 2 O saturated ni- trogen, was slightly lower than that of unmodiﬁed barium zirconate. Electromotive force measurements under fuel cell conditions revealed the total ionic transport number to be B0.9 at 6001C. The combination of electrical and chemical properties and good sinterability render ZnO-modiﬁed barium zirconate an excellent candidate for reduced temperature solid oxide fuel cell applications. I. Introduction M ANY alkaline earth perovskites exhibit high proton con- ductivities at moderately elevated temperatures (4001– 7001C). These materials have attracted attention for their potential applications as electrolytes in fuel cells and chemical sensors. 1–5 Examples with well-documented proton conductivi- ties include barium cerate (BaCeO 3 ), 2,6–8 strontium cerate 9 and the perovskite-derivative BCN. 10–12 Many of these compounds suffer from chemical instability under CO 2 containing atmos- pheres, readily forming alkaline earth carbonates according to Eq. (1) 13,14 BaCeO 3 þ CO 2 ! BaCO 3 þ CeO 2 (1) This reactivity causes severe degradation of the electrolyte and precludes applications in fuel cells based on hydrocarbon fuels. In contrast, barium zirconate exhibits excellent stability under CO 2 , 15,16 rendering it highly attractive for applications in ag- gressive environments. While initial investigations suggested that barium zirconate did not share the high proton conductiv- ity of other members of the alkaline earth perovskite family, 15,17 it has been more recently recognized that doped BaZrO 3 , in fact, exhibits higher bulk conductivity than doped BaCeO 3 . 18–20 The early mis-interpretation of the behavior of doped barium zir- conate originates from the highly refractory nature of this ma- terial, which results in samples with small grain sizes and high total grain-boundary area. As a consequence, the resistive grain boundaries produce a material with an overall low conductivity, and, in the absence of low temperature A.C. impedance meas- urements (oB2001C), the individual bulk and grain-boundary contributions to conductivity were not resolved in the early literature. The refractory nature of doped barium zirconate leads to significant challenges to its implementation in fuel cells and oth- er devices. First, it is difﬁcult to process to a high density (493%), as is required of fuel cell electrolyte membranes. Typ- ically, extreme conditions, such as high temperature (17001– 18001C), long sintering times (24 h), and nanometer-sized particles are needed to prepare fully densiﬁed pellets. 13,19,21,22 Not only are these conditions costly to implement, they are in- compatible with most potential electrode materials and thereby preclude the fabrication of co-sintered structures. Second, it is the total resistance of the electrolyte, not only that of the bulk or grain interiors, that dictates electrochemical performance. Thus, strategies for improving total grain-boundary conductivity are required. Third, high temperature processing can be anticipated to induce barium oxide evaporation and thereby decrease con- ductivity, as has been observed in BaCeO 3 . 13 In addition, occa- sional abnormal grain growth 19 results in inhomogeneous properties, both electrical and mechanical, which are highly undesirable. In the present work, we demonstrate that ZnO is an effective sintering aid for BaZr 0.85 Y 0.15 O 3–d (BYZ), enhancing both dens- iﬁcation and uniform grain growth. Yttrium has been selected as the dopant because of the high proton conductivity this species imparts, presumably as a result of its good ionic radius match to Zr. 15 Of potential sintering aids, a previous study of Al 2 O 3 , MgO, and Y 2 O 3 on barium zirconate densiﬁcation showed a marginal improvement in sintering behavior with yttria. This additive yielded samples of 91%–92% dense at 16001C as com- pared with B90% density for pristine samples processed under identical conditions. 23 While no detailed mechanism for the slight enhancement in density was presented, it was speculated that yttria limits grain growth. z In this work, an initial screening of all transition elements in the series Sc to Zn, as discussed be- low, showed NiO, CuO, and ZnO to be the most effective ad- ditives for enhancing barium zirconate densiﬁcation. Of these, zinc oxide was selected as most suitable and accordingly sub- jected to further investigation and system optimization. II. Experimental Procedure (1) Sample Preparation Crystalline powders of BYZ were synthesized by a glycine-nitrate combustion synthesis process. 24 Starting materials were high pu- rity Ba(NO 3 ) 2 (Alfa Aesar, 99.95% purity), Y(NO 3 ) 3  6H 2 O (Alfa Aesar, 99.9% purity), and ZrO(NO 3 ) 2  xH 2 O (Alfa Aesar, Ward Hill, MA, 99.9% purity), where x 5 2.3, as determined by thermogravimetric analysis (TGA). The appropriate molar ratios of nitrates and glycine (NH 2 CH 2 COOH) were mixed in a min- imum volume of deionized water to obtain a transparent solu- tion. A glycine to nitrate ratio of 1:3 was used. The aqueous solution was dehydrated on a hot plate at a temperature of 1501C J ournal J. Am. Ceram. Soc., 88 [9] 2362–2368 (2005) DOI: 10.1111/j.1551-2916.2005.00449.x r 2005 The American Ceramic Society 2362 J. Drennan—contributing editor w Author to whom correspondence should be addressed. e-mail: smhaile@caltech.edu Manuscript No. 20302. Received July 26, 2004; approved March 14, 2005. Funding for this work was provided by the Department of Energy, Ofﬁce of Energy Efﬁciency and Renewable Energy, and by the Army Research Ofﬁce, Chemical Science Division, through subcontract with NexTech, Inc. Additional support has been provided by the National Science Foundation, through the Caltech Center for the Science and Engineering of Materials. z Although it was suggested that yttria additive limits grain growth in barium zirconate, from an examination of the micrographs presented, this does not appear to be the case.