Gadolinium-doped ceria (GDC) and yttria-stabilised zirconia (YSZ) are very well known materials, used as components for solid-oxide fuel cells (SOFC). They both are very good electrolytes, and they have been vastly used in this sense. The main difference between GDC and YSZ when used as electrolytes is that GDC shows higher oxygen conduction at lower temperatures, leading to the possibility of developing intermediate temperatures SOFC (IT-SOFC). But their versatility allow us to use them also as anode materials, like YSZ, which in combination of nickel particles, is used to catalyse oxygen reduction. In addition, YSZ combined with nickel (Ni-YSZ) is also used as anode in SOFC, in combination with GDC as electrolyte.

In any case, the main role of both is related to oxygen migration, i.e. ionic conductivity. In GDC and YSZ oxygen migration undergoes via a vacancy hopping mechanism. Vacancy mobility depends on several aspects, like the association energy with respect to the dopants, dopants concentration, or temperature, whose effects are very well studied. Grain boundaries (GBs) are also known to have an important impact in conductivity, despite that little is known about their influence. From an experimental point of view, results suggests that the presence of these GBs
offers faster migration channels. On the other hand, dopants tend to segregate to GBs, thus trapping oxygen vacancies and reducing the ionic conductivity of the materials. To clearly understand the role of GBs in the oxygen conductivity, it is necessary to study the system at the atomic level. One of the main issues, however, is that the model must include all the chemical and structural complexity inherent to GBs, which force us to generate very large models.

In this work, and based on interatomic potential and molecular dynamic techniques, we present a strategy based on the work of Sayle et al. (2007) referred as Amorphisation and Recrystallization (A&R). With this methodology, consisting on melting and recrystallizing the system at high pressures, we are able to obtain accurate models that allow us to describe realistic GBs including dislocations and translocations. After generating different models, we have performed a series of molecular dynamic simulations in order to see the effect of the temperature in the structure,
ionic mobility and also oxygen diffusion and therefore its conductivity. Results seem to indicate that indeed, the presence of the grain boundaries are beneficial for the enhancement of the oxygen diffusion in the system.