Oxygen Permeation Through Composite Oxide-Ion and Electronic Conductors

Oxygen permeation through composites consisting of four well-known oxide-ion conductors and a noble metal, Pd or Ag, is reported. The oxides were Zr 0.9Y0.1O1.95 (YSZ), (Bi1.75Y0.25O3)0.95(CeO2)0.05 (BYC5), Ce0.8Sm0.2O1.9 (SSC), and La0.8Sr0.2Ga0.83Mg0.17O2.815 (LSGM). The results show that (BYC5 + Ag) yields the highest oxygen permeation flux, but the composite deteriorates with time. The composites (SSC + Pd), (LSGM + Pd), and (YSZ + Pd) give stable, but relatively lower oxygen per meation flux in the order of (SSC + Pd) > (LSGM + Pd) > (YSZ + Pd). The composite microstructures indicate that (BYC5 + Ag) has the best percolating network for both oxide-ion and electronic pathways while (SSC + Pd) has the longest triple-phase boundary lengths with the smallest grains, which is beneficial to the surface oxygen exchange. It is shown that the microstructure of the composites, which strongly influences the competition between surface reaction and bulk diffusion, is technically as important as the oxide-ion conductivity. The activation energy appears to be related more to the morphology of the metallic phase than to that of the oxide phase. These results suggest that (SSC + Pd) is a promising composite mixed conductor for applications requiring oxygen separation. © 1999 The Electrochemical Society. S1099-0062(98)11-052-0. All rights reserved. Manuscript submitted November 12, 1998; revised manuscript received January 5, 1999. Available electronically May 20, 1999. Electrochemical synthesis of syngas via ceramic membranes has been considered an economic and efficient alternative to transform natural gas into liquid fuel. The central component of this electrosynthesis is the oxygen-permeable ceramic membrane that is solely an oxide-ion and electronic conductor at high temperatures. The use of mixed conductors with both high oxide-ion and electronic conductivity assures a high oxygen permeation flux and a fast surface exchange rate. These membranes may be categorized into (i) homogenous mixed oxide-ion and electronic conductors (MIECs) and (ii) composite oxide-ion electrolytes and metals. Most of the Cocontaining single-phase perovskite oxides, 1-3 for example, are good MIECs that belong to the first category since both oxide ions and electrons inside these perovskites are mobile at elevated temperatures. A mixed oxide-ion and electronic conduction is also possible in a composite of an oxide-ion electrolyte and a noble metal if each component has a percolation pathway through the composite, a concept patented by Mazanec et al. 4,5 These composites function the same way as the MIECs. An immediate example is the oxide-ion conductor Y 2 O 3 -stablized ZrO 2 plus the noble metal Pd6-9 that has been shown to have an appreciable oxygen permeation rate at elevated temperatures. The relative advantages of each type of mixed conductor depend on the choice of materials and their preparations. Homogenous mixed conductors produce higher oxygen permeation flux, but those previously studied suffer degradation with time 2,3 due to either oxygen ordering or compositional alteration at the oxygen-lean side. Composite mixed conductors, on the other hand, can yield a stable oxygen permeation flux, but with a relatively lower oxygen permeation flux. From the commercial point of view, stability of the oxygen permeation is the most important criterion. Therefore, composite mixed conductors have been considered more attractive for industrial applications even though they are relatively costly. In this paper, we report measurements of oxygen permeation flux on composite mixed conductors consisting of the oxide-ion conductors Zr 0.9 Y 0.1 O 1.95 (YSZ), (Bi 1.75 Y 0.25 O 3 ) 0.95 (CeO 2 ) 0.05 (BYC5), Ce 0.8 Sm 0.2 O 1.9 (SSC), and La 0.8 Sr 0.2 Ga 0.83 Mg 0.17 O 2.815 (LSGM), and a noble metal, either palladium or silver. The oxygen permeation fluxes of the above composites were measured with gas chromatography. The compositions and microstructures of the prepared composites were examined by X-ray diffraction (XRD) and scanning electron microscopy (SEM), respectively. Experimental Preparation of composites.—All the composites investigated in this study were made by solid-state reaction. The YSZ powders were purchased from the Tosoh Company, and all other oxide powders were made in our laboratory. The starting BYC5 powder was made by solid-state reaction after calcining at 700°C, while the starting SSC and LSGM powders were the products of decomposition at intermediate temperatures of polymer precursors made by a wetchemical method. The details of the procedures are described in Ref. 10-12. The starting oxides had been thoroughly mixed with a required amount of PdO or silver powder. The mixtures were then pressed into 2.54 cm diam pellets before sintering at high temperatures. The sintering temperature for each composite is listed in Table I. At these temperatures PdO is reduced to elemental Pd. X-ray diffraction and SEM.—The component phases and the reactivity between the oxides and Pd or Ag were examined with a powder X-ray diffractometer PW 1740. The powder diffraction was conducted from 10 to 80° with a Cu Kα target and Ni filter. The microstructure and elemental analysis of the as-prepared composites after polishing were observed with a JEOL 35C scanning electron microscope equipped with a Kevex energy dispersed spectrum system. In order to distinguish clearly the oxide and metal phases, a backscattered electron image was chosen. Measurement of oxygen permeation.—The sintered composite pellets (~ 1.75 cm in diam) were ground on both sides with a diamond