(131e) Modeling Charged-Defect Transport in Proton-Conducting Ceramic Membranes

Certain doped perovskite ceramics, such as yttrium- and/or
cerium-doped barium zirconates (BCZY), can transport protons through their bulk.  Thus, they are potentially valuable for
application as separation membranes and in membrane reactors.   The most natural thought is that
these materials are hydrogen-selective permeable membranes.  However, depending upon the gas
compositions across the membrane, and possibly electrode polarization, the
membranes can effectively transport H₂, H₂O, and O₂.  

Figure 1 illustrates a shell and tube configuration in whicn
a thin (order 10 microns) BCZY membrane is supported on a porous-ceramic
tube.  At elevated temperatures around
600 ûC and above, the BCYZ materials are mixed ionic-electronic conductors
(MIEC).  As such, these materials
enable multicomponent defect transport of protons, oxygen vacancies, electrons
and electron holes.  The subject of
this presentation is the development of a model that can predict the radial
electrochemical defect transport through the membrane.

The model is based upon formulating and solving
defect-conservation equations where the defect fluxes are represented via the
Nernst-Planck equations.   At the membrane surfaces the defect
reactions are assumed to be in equilibrium with the gas phase, thus
establishing boundary conditions for the transport equations.   Internal electric-potential
gradients are formed within the membrane to maintain electroneutrality.  Thus the defect fluxes depend upon both
concentration gradients and the local electric fields.   The membrane performance depends
upon physical parameters including defect mobilities and equilibrium constants
for the gas-surface defect reactions.

The steady-state system of governing equations forms an
ordinary-differential-equation boundary-value problem (BVP) that is solved
computationally using finite-volume discretization.  The mathematical structure of the BVP is
somewhat unusual because of the requirement to determine the internal electric
fields.