164 Prakt. Met. Sonderband 46 (2014)
2. EXPERIMENTAL
Oxygen transport parameters were determined by means of the conductivity relaxation
technique [3]. With this method, the pO2 of the atmosphere around the sample is abruptly
changed and the transient of electrical conductivity of the sample is recorded. Nonlinear
least squares fitting of the appropriate diffusion model to the data allows the extraction of
kinetic parameters, i.e. the chemical surface exchange coefficient kchem and chemical
diffusion coefficient Dchem of Oxygen. Oxygen partial pressures were switched between
0.10.15 bar and 0.010.015 bar. Depending on whether the pO: is adjusted to higher or
lower values the relaxation measurement is referred to as oxidation or reduction experiment,
respectively.
Thin layers of silver have been deposited on the sample surface by sputtering Ag-foil
targets (99.99% purity) in an Ar-plasma using a Baltek MED 20 sputtering device. The
chemical composition of sample surfaces was quantified by X-ray photoelectron
spectroscopy (XPS). XPS-measurements were performed at room temperature in ultrahigh
vacuum at a base pressure of 2x10'° mbar utilizing a Thermo MultiLab 2000 spectrometer
equipped with an alpha 110 hemispherical analyzer from Thermo Electron. Surface
morphologies of Ag-coated test specimens after annealing were investigated by use of a
scanning electron microscope (SEM) Zeiss EVO 50, Germany, with LaBes-cathode at an Fig.
acceleration voltage of 15 kV. LazNiO
from a
3. RESULTS AND DISCUSSION
Conductivity relaxation experiments were conducted on a bar-shaped sample of La2NiO4+5
between 600 and 850°C at oxygen partial pressures of 0.1 and 0.01 bar, which corresponds Ba
to the expected operating conditions of an intermediate temperature SOFC-cathode.
Without silver coating, the surface exchange of oxygen was the rate-determining step in the
two-stage oxygen exchange process and only surface exchange coefficients could be
obtained from the experiments. After Ag-deposition the surface exchange kinetics was ew]
significantly enhanced at temperatures below ~700°C due to the high catalytic activity of
silver for the oxygen redox reaction (Fig. 1) [4]. Each measurement series was started at
600°C and a complete temperature cycle with a maximum temperature of 850°C was =
performed with relaxation measurements taken in steps of 25°C. i
A pronounced temperature hysteresis was observed due to the removal of silver at B55
temperatures above ~700°C via volatile Ag-species. Results of kchem during the cooling run
were found to be in good agreement with data obtained from independent relaxation
measurements performed on a thin Ag-free sample, suggesting the complete removal of the Fig. 2: A
silver layer during the high-temperature treatment (Fig. 1).
From Ag-coated samples of LazNiOa4+5 oxygen diffusion coefficients could be obtained
with improved accuracy due to the increased contribution of the diffusion process to the
overall oxygen exchange process. Results obtained for Dchem and kchem at two different 3.1 Surfa
oxygen partial pressures are given in Fig. 2. Diffusion coefficients were determined during
the heating run of each measurement series while surface exchange data are results Visual ins
obtained from the cooling period after the silver layer has evaporated. All data show good residuals
linearity within the investigated temperature range in Arrhenius representation. Activation on the sur
energies of both kinetic parameters have been calculated from the slopes of linear as to cher
regression lines and amount to 50-60 kJmol! for Dchem and 130-140 kJmol™' for kchem. A elevated
comparison of the diagrams in Fig. 2 shows that Dchem is practically not affected by changes material.
in pO2 while kchem increases slightly for higher oxygen partial pressures.