166 Prakt. Met. Sonderband 46 (2014) 
migration has been reported in the literature for high-temperature superconductors and 
ferro-electrics [5-7]. 
The surface composition of a silver-coated sample has been characterized after testing 
by means of depth profiles of the first 500 nm of the sample surface using X-ray 
photoelectron spectroscopy combined with Ar-sputtering. No traces of silver could be 
identified from XPS-spectra within the first 500 nm of the surface region (Fig. 3). Estimating 
a XPS-detection limit of 0.1 at.% it appears that Ag completely evaporates during the high- 
temperature interval and does not interfere with the oxygen diffusion process by way of 
penetration into the material through grains or along grain boundaries. This suggests a 
rather poor solubility or diffusivity of silver in the nickelate compound or may be simply a 
consequence of the small thickness of the deposited Ag-layer which does not offer enough 
reservoir for large-scale contamination of the bulk. 
The source of the rather large carbon content is unclear but may be caused by 
hydrocarbon impurities in the measurement gases accumulating over time on the sample 
surface. Small amounts of bismuth were detected up to a depth of about 300 nm, originating 
from the gold paste used for attaching electrical contacts to the sample. 
89 — eee 
— 0 1s woe BY 4f 
70 —La3d52 Cs 
- Ni LM2 
+ 60 
ToT 
50: 
40 - 
BO a a 
AA at 
4 20- 
© 1 — 
10 ~ Tol 
LE a a a i 
0 100 200 300 400 500 
sputter depth / nm 
Fig. 3: XPS-depth profiles within the sample surface of initially Ag-coated La2NiOa+s after a 
complete temperature cycle between 600 and 850°C at 0.01 bar pO2. Dashed lines mark Fig. 4: S 
nominal compositions. between 
layer. Th 
3.2 Surface morphology of Ag-films 
The stability and morphology of silver metallization has been investigated after short-term Visual ins! 
annealing at 500, 600, 700 and 800°C by an independent study. Small test specimens of reflective : 
La2NiO4+5 were polished on one side and coated with a 200 nm Ag-layer. Each sample was lackluster 
annealed for 24 h at the respective temperature in a flowing Ar/O2-mixture at 0.01 bar pO. In Fig. 4. | 
Large temperature ramps of £10 K/min were used for heating and cooling in order to a homoge 
minimize the impact of heat ramping on the film morphology. After ther 
homogenr