1484
(Invited) Modeling the Chemically Induced Stress in Lscf Oxygen-Separation Membranes

Wednesday, 1 June 2016: 10:00
Aqua 305 (Hilton San Diego Bayfront)
B. Euser, H. Zhu, J. Berger (Colorado School of Mines), and R. J. Kee (Mechanical Engineering, Colorado School of Mines)
Strontium- and iron-doped lanthanum cobaltite perovskites (La1-xSrxCo1-yFeyO3-δ) are used for a number of electrochemical applications, including solid oxide fuel cells (SOFCs), oxygen separation membranes, chemical sensors, and catalysts.  This paper focuses specifically on La0.6 Sr0.4Co0.8Fe0.2O3-δ, LSCF6482.  The material is a mixed ionic and electronic conductor (MIEC), with good oxide-ion conduction.  The electronic conduction is via small polarons associated with reduced and oxidized states of cobalt.  An important aspect of the present study is to understand the coupled effects of defect transport and stress, especially in oxygen-separation membranes.  Varying the environmental oxygen pressure on the membrane surfaces induces spatially transient responses of the charge-carrying-defect concentrations, electrostatic potential, and hydrostatic stress.  The stress is the result of atomic-scale strain and distortion within the crystal lattice structure.  Sufficiently high stress, especially tensile stress, can cause material degradation or failure. 

 A Nernst–Planck–Poisson (NPP) model is extended to account for the coupled effects of defect transport and chemically induced stress.  In addition to the effects of concentration and electrostatic-potential gradients, the Nernst–Planck fluxes include the influence of hydrostatic-stress gradients.  The stress model is based on the elastic response of a thin plate, including the effects of oxide-ion vacancy concentrations.  The material properties for LSCF6482, including mechanical properties, thermodynamic parameters for the defect chemistry, and charged-defect diffusivities, are obtained by interpreting previously published experimental data.  The model is formulated as a system of partial differential equations and solved computationally, predicting transient profiles of defect concentrations, electrostatic potential, and hydrostatic stress.