Page 162 - High Temperature Solid Oxide Fuel Cells Fundamentals, Design and Applications

Cathodes 139

tends to be tetravalent. This 'enthalpy effect' leads to a driving force to form
LaCrOs and AMn03 (A = Ca, Sr) [62]. Thus, there are two different driving forces
which promote interdiffusion in opposite directions.
Since the A-site-deficient lanthanum manganite shows better compatibility
with YSZ, its compatibility with oxide interconnect is of particular interest.
Nishiyama et al. [62] found the following interesting behaviour of the A-site-
deficient manganite at the interface with (La,Ca)Cr03 (LCC) that had excess CaO
to enhance its air sinterability:
(i) With porous manganite cathode, the elemental distribution across the
cathode/interconnection interface suggests the following replacement
reaction because of the above second driving force:

LaMn03 (in LSM) + CaCrO3 (in LCC) = LaCrO3 (in LCC)
(15)
+ CaMnO3 (in LCC)

Here, LaCrOs is formed as a dense layer next to the original LCC while
CaMn03 is formed as a porous layer next to the LaCr03 layer. This
reaction is triggered by the presence of calcium oxychromates in the
original LCC which can be squeezed out of grain boundaries under
oxygen potential gradient.
(ii) With dense manganite cathode, CaO instead of calcium oxychromate
initiates the precipitation of manganese oxide at the interface, suggesting
that oxygen potential distribution plays an important role in determining
the mass transfer. Inside the dense manganite, the oxygen potential is
lowered due to low oxide ion diffusivity.

5.4.2 Compatibility of Cathodes with Metajlic fnterconnects
Even though many alloys have been investigated as metallic interconnects,
almost all form chromia as a protective oxide scale [63]. The main issues
associated with the use of such chromia-forming alloys are chemical reactions at
the interface with the cathode (and also with anode) material and chromium
poisoning of the cathode. Taniguchi et al. [59] found that degradation by
chromium poisoning occurs more severely at lower temperatures (see Figure
5.12) and that degradation measured in terms of the cathode life time is
proportional to the logarithmic oxygen activity derived from the overpotential
values, q, as follows:

t(degradation) 0: Alogao(= 2qF/2.303RT) (16)

Their observations on distribution of chromium in the electrolyte/cathode
vicinity (as shown in Figure 5.13) indicated that although the average Cr
content in the cathode layer increases with increasing temperature, this
quantity is not directly related to cathode degradation. Instead, concentrated
chromium deposition on the three-phase boundaries can be directly related u7ith
the cathode lifetime.