Jessica Hermet

Molecular dynamics simulations of oxygen diffusion in GdBaCo2O5.5

The mechanisms of oxygen diffusion in GdBaCo2O5.5 compound are investigated by molecular dynamics simulations. The results confirm that diffusion is mainly bidimensional with oxygen moving in the (a,b) plane while diffusion along the c axis is much more difficult. Between 1000 and 1600 K, the activation energy for diffusion is about 0.6 eV, close to experimental values. Going deeper inside the oxygen diffusion mechanism, we see that this diffusion occurs mainly in the cobalt planes while most of the oxygen vacancies are kept in the Gd planes. Analysis of oxygen motions show that Gd planes can be seen as source-sink for the oxygen vacancies rather than as fast pathways.

Oxygen diffusion mechanism in the mixed ion-electron conductor NdBaCo2O5+x

Double perovskite cobaltites were recently presented as promising cathode materials for solid oxide fuel cells. While an atomistic mechanism was proposed for oxygen diffusion in this family of materials, no direct experimental proof has been presented so far. We report here the first study that directly compares experimental and theoretical diffusion pathways of oxygen in an oxide, namely in the double cobaltite compound, NdBaCo2O5+x. Model-free experimental nuclear density maps are obtained from the maximum entropy method combined with Rietveld refinement against high resolution neutron diffraction data collected at 1173 K. They are then compared to theoretical maps resulting from classical molecular dynamics calculations. The analysis of 3D maps of atomic densities allows identifying unambiguously the pathways and the mechanisms involved in the oxide ion diffusion. It is shown that oxygen diffusion occurs along a complex trajectory between Nd- and Co-containing a,b planes. The study also reveals that Ba-containing planes act as a barrier for oxygen diffusion. The diffusion mechanism is also supported through the oxygen sites occupancy analysis that confirms the increase of oxygen vacancies in the cobalt-planes on heating. The use of such combined experimental and theoretical analysis should be considered as a very powerful approach for materials design.

How dopant size influences the protonic energy landscape in BaSn1−xMxO3−x/2 (M = Ga, Sc, In, Y, Gd, La)

The energy landscape of the protonic defect is investigated in acceptor-doped barium stannate using density-functional calculations. Several trivalent dopants are studied (Ga, Sc, In, Y, Gd, La), covering a wide range of ionic radii. All the dopants are found attractive with respect to the proton, with (negative) association energies varying from −0.40 to −0.07 eV. A radius rc1 is defined to separate the “small” dopants that induce tensile stress from the “large” ones that induce compressive stress in the host matrix (rc1 ≈ 0.72 A). The protonic energy surface exhibits a non-trivial evolution with ionic radius of the dopant: for low dopant radii, the most stable protonic site is the oxygen first-neighbor of the dopant, while for high dopant radii, the most stable position is obtained when the proton is bonded to an oxygen second-neighbor of the dopant. This evolution of the protonic energy surface with dopant ionic radius is smooth and the transition takes place between In and Y, i.e. for a critical radius rc2 between 0.80 and 0.90 A (rc2 > rc1 significantly). The dopant–proton association energy exhibits a minimal value ≈−0.07 eV (weakest attraction) at this transition, i.e. in the case of yttrium, for which the first-neighbor and second-neighbor positions are almost degenerate. Other dopants, smaller or larger, are more attractive to protons. The present study gives useful information about the modification of the trapping effect according to the dopant ionic size.

Oxide ion and proton transport in Gd-doped barium cerate: a combined first-principles and kinetic Monte Carlo study

Oxide ion and proton transport properties in the fuel cell electrolyte Gd-doped BaCeO3 are investigated by first-principles density-functional calculations and kinetic Monte Carlo simulations. Behind the apparent complexity of the energy landscape (related to the low-symmetry of the system), general tendencies governing the energy barriers can be extracted. In particular, the set of barriers is tested with respect to the Bell–Evans–Polanyi (BEP) principle that relates the activation energy Ea of a series of similar chemical reactions to their reaction enthalpies ΔH. This rule is found poorly satisfied in the case of oxide ion migration, but much better satisfied for proton hopping mechanisms. Protonic reorientations, by contrast, do not obey the BEP rule. Kinetic Monte Carlo simulations give insight into the macroscopic transport properties. We observe that dopants act as traps for oxygen vacancies and slow down their motion, whereas their effect on the protonic diffusion coefficient is more complex. This is related to the fact that a two-state picture roughly applies to the oxygen vacancy energy landscape, while it does not apply to the protonic one. Oxide ion migration exhibits strong anisotropy, the motion of the oxygen vacancies being favored in the equatorial planes, while protonic diffusion is found more isotropic.