(a) Time dependence of the overall conductivity measured parallel and perpendicular to the interface at 500 K during the heat treatment. Starting from a given data point the sample was heated to 780 K and then cooled down to 500 K to obtain the next data point. The heating/cooling cycle consisted of a stepwise transition by 50 K and a respective waiting period of 1 h. (b) Arrhenius plots for the transient and stationary conductivities in both directions. The conductivity combination of isolated layers for the two directions is shown for comparison.
The average conductivities measured parallel to the interface as a function of inverse interfacial spacings at for sample . Nonstationary, stationary data, and the conductivity combination of isolated layers are shown for comparison.
Derived defect concentration profiles in sample at 500 K (a) for the stationary data after annealing and (b) for the transient data before annealing. For the parallel conductivity of , a (modified) Mott–Schottky model applies (Ref. 19), while for the perpendicular conductivity of , a Gouy–Chapman model has been used (Ref. 20).
Cross-section TEM images for (a) the nonannealed and (b) the annealed heterolayers. The insets give the diffraction pattern of the respective heterolayer.
Schematic representation of the ion transfer at the interface for (a) before contact, (b) equilibrium contact (after annealing), and (c) freshly contacted (before annealing) with the dashed area representing the unstable microstructural defects at the interface.
Article metrics loading...
Full text loading...