Impurity diffusion along ceria grain boundary

Abstract

Cerium-based oxide (ceria) is among the most intensively investigated materials, in heterogeneous chemical catalysis as an active support and in electrochemical catalysis as a non-metallic oxygen-ion-conducting electrode for applications such as three-way catalysts (TWCs) and solid oxide fuel cells (SOFCs). In such devices, ceria is typically polycrystalline and is used in close contact with other components at high temperatures. Thus, cation impurities from other components can diffuse into the ceria, especially through grain boundaries, which are a fast cation transport path. In this study, I investigated how quickly metal impurities diffuse through the grain boundaries of acceptor-doped ceria and how one can control their diffusion kinetics. Furthermore, I monitored how the oxygen-ion transport properties and surface oxidation reactivity of ceria change according to the impurity diffusion phenomena. More specific research topics are as follows.

1. Investigate the diffusivity and solubility of impurities inside ceria grain boundaries. 2. Analyze the specific grain boundary conductivity of ceria with impurities. 3. Synthesize reactive metal nanoparticles on ceria surfaces through impurity diffusion.

For these purposes, I prepared dense polycrystalline thin films of acceptor-doped $CeO_2$ via pulsed laser deposition. The remarkably high grain boundary density of polycrystalline thin films with vertically-oriented, nanosized-columnar grains enabled accurate analysis of the diffusion kinetics and solubility of Ni impurities by means of time-of-flight secondary ion mass spectroscopy. Impurities diffuse unexpectedly quickly through ceria grain boundaries, although cation diffusion in a cubic fluorite lattice is known to be exceedingly slow at temperatures below $750^\circ$C. For example, in the case of a 0.5 % Sm-doped ceria thin film, heat treatment at $750^\circ$C only for 10 hours diffuses Ni to more than 1 micron, whereas it takes about ten million years for Ni to diffuse to the same depth in bulk lattice. Interestingly, the diffusion coefficients of Ni vary strongly with the choice and concentration of acceptor dopant, as well as with gas atmosphere, which provides a clue to the control parameters of the diffusion kinetics. Moreover, it was found that a considerable amount of impurities are dissolved in the grain boundaries. Above all, I concluded that the ceria grain boundary can act as a reservoir of impurities and also as a highway of their transport.

Next, I analyzed how the diffused impurities alter the in-plane oxygen-ion conductivity of Sm-doped $CeO_2$ grain boundaries. For this, both epitaxial and polycrystalline ceria thin films were grown via pulsed laser deposition, and the change of ionic transport property of one single grain boundary after impurity diffusion was investigated in terms of space charge theory. It was observed that when certain impurities are present inside ceria grain boundaries, the positively charged space charge potential of the ceria grain boundaries is noticeably reduced, thereby greatly improving the conduction characteristics. These observations suggest that the impurity diffusion phenomena through the ceria grain boundary can be utilized to improve the performance of solid oxide electrolytes.

Lastly, using impurity grain boundary diffusion, I developed a novel synthesis method for metal nanoparticles on the top surface of ceria. Appropriate reduction heat treatment out-diffuses relatively reducible and catalytically active cation impurities in the grain boundary of ceria and subsequently forms nano-sized metal particles on the ceria surface. Diverse types of metal nanoparticles were successfully synthesized by this method on pristine and Sm-doped $CeO_2$ surfaces. Interestingly, the synthesized nanoparticles became deeply embedded at grooved grain boundary sites, and so they possessed excellent sintering and coking resistance as well as redox stability. Furthermore, the ceria surfaces with synthesized metal nanoparticles showed improved reactivity toward H2 electro-oxidation and CO oxidation, suggesting that this method can be applied to various catalytic applications.