Could you explain the significance of your article to the non-specialist?

The solid oxide fuel cell (SOFC) constitutes a promising technology for meeting future energy needs over a range of scales, from MW power generation to power supply (in the form of micro-SOFC) for laptops and handheld electronics. SOFCs generally use yttria-stabilised zirconia (YSZ) as the oxygen ion conducting electrolyte. One recent approach to achieving successful application of SOFC technology has been to examine whether the oxygen ion conductivity of polycrystalline YSZ can be improved by reducing the grain size to the nanometre scale. Unfortunately, answers to this question have been contradictory: some report a large enhancement in total conductivity; others see little if no effect. By following in our measurements the motion of oxygen ions alone, we show that grain boundaries hinder oxygen ion transport; consequently we can say that, where observed, increased conductivity has a different origin (e.g. protons).

What has motivated you to conduct this work?

The rationale behind reducing the grain size is the belief that diffusion along grain boundaries is faster than in the grains, as is commonly found in metals. Thus, reducing the grain size, and thereby increasing the volume fraction of grain boundaries, should result in enhanced oxygen ion transport. In ionic solids, however, this simple picture can be complicated by the presence of space-charge regions at the grain boundaries. In order to reach unambiguous conclusions regarding the various possible transport processes (in the grain, in the space-charge, in the grain boundary core), one requires nano-grain-size samples that are free of pores and microcracks, since such microstructural imperfections also act as fast diffusion paths. Such samples have up to now been unobtainable with conventional methods of preparation. Using Spark Plasma Sintering (SPS) we have recently managed to produce macroscopic, 100% dense samples of nano-YSZ (UC Davis) and perform oxygen diffusion measurements on them (RWTH Aachen).

Where do you see this work developing in the future?

Our work suggests that strain or dissolved water in the form of protons may explain the enhanced total conductivity of nano-YSZ. In particular we have started to examine the incorporation and migration of protons in nano-YSZ. Another direction for future research concerns alternative SOFC electrolytes. We believe that our results for nano-YSZ are transferable to the current alternative oxygen ion conducting electrolytes, doped ceria and doped lanthanum gallate. It remains to be seen if there are ceramic electrolytes for which nano-grained samples show enhanced oxygen ion transport.

Are there any particular challenges facing future research in this area?

In our simplified model of possible diffusion processes in a polycrystalline oxide with space-charge regions at the grain boundaries, one can consider four separate diffusion processes: within a grain, along the grain boundary core, and parallel and perpendicular to the space-charge regions. With current models one can only differentiate, as we have done, between a grain diffusivity and an effective interface diffusivity that combines the three other diffusivities. One outstanding challenge is to develop appropriate models that would allow the extraction from experimental data of all four diffusion coefficients.