Page 183 - High Temperature Solid Oxide Fuel Cells Fundamentals, Design and Applications

160 High Temperature Solid Oxide Fuel Cells: Fundamentals, Design and Applications

spectra. The charge transfer process is the only strictly electrical component of
the system, and its time constant can be estimated. The resistive component
should not exceed 1 Q, and the reactive component, the interfacial capacitance,
should be around 10 pF/cm2 [21] implying a characteristic frequency of the
order of 10 kHz. The feature I (Figure 6.7) is therefore identified with the charge
transfer process. Other experimental work concurs in this identification [ 191,
where the variation of the capacitance with temperature and with gas
environment, particularly the partial pressures of hydrogen and steam indicate
some dependence on adsorption in the interface region. Diffusion and mass
transfer processes must therefore be associated with the slower features. A very
slow mechanism may in fact be an artefact, due to variation of the Nernst
potential in the vicinity of the active anode sites over the time of the modulation,
of the order of 1 s (1 Hz), as measured against a reference electrode exposed to a
constant gas composition, for example on the cathode side of the sample. This
concentration polarisation effect was suppressed using an anode-side reference
[18], clearly demonstrating that it is a function of the experimental system, not
specific to the electrode. In the same work the feature I1 could also be suppressed
if gas transport to the test anode was electrochemical, by direct contact to a
second identical device exchanging reagent and reaction product, rather than
gas-phase diffusion. Given that there was no gas transport out of the system, the
corresponding impedance spectral feature does not appear.

6.7 Behaviour of Anodes Under Current Loading
In the discussion of impedance spectroscopy so far, it has been assumed that the
modulation is applied to an anode under equilibrium conditions on each
electrode with no net d.c. current transfer. Further information can be obtained
by impedance analysis of the behaviour under load, the overpotential then being
an independent variable. In this way a further confirmation of the identification
of spectral features can be obtained, because there is a logarithmic relationship
between current and overpotential (Figure 6.4) in the Tafel region. As a
consequence the impedance spectroscopic feature corresponding to this reaction
rapidly shrinks with increasing overpotential, reducing both resistive and
reactive components and decreasing the apparent interface polarisation. At
the same time a further reactance, in the inductive or opposite sense to the
capacitative phenomena observed thus far, may emerge near the low-frequency
limit. High-frequency inductive effects are usually artifacts of the measurement
system, associated with self-inductance, whereas the low-frequency inductive
feature is generally interpreted as evidence of adsorbed reaction intermediates,
and may be associated with autocatalytic effects. This is illustrated in Figure 6.8.
Mass transport effects may also be influenced by overpotential, for example by
change of concentration gradients, increased occupation of surface sites
by reaction products, or the formation of reaction intermediates. They may also
be temperature influenced, for example due to thermal desorption, or even gas
viscosity changes with temperature. A clear example when the anode