Systematic and predictive trends to chromium poisoning in solid oxide fuel cell cathodes

Abstract

High performance computing simulations of Cr-poisoning are used to develop systematic trends for overpotential driven degradation modes in microstructurally resolved fuel cell cathodes. Oxygen reduction and chromium oxide deposition at triple phase boundaries (tpbs), and species transport, are numerically computed within 19 microstructures (103 µm3 domains) over a range of operating conditions. Three primary input parameters drive degradation: tpb density of the microstructure ρtpb, normalized charge-transfer current density for Cr-poisoning i∗, and galvanostatic current density j. Simulations converged over large fractions of the potential operating time, ranging from thousands to over a hundred thousand hours. For all simulations, normalized overpotential η∗ versus time t curves can be modeled using a progressive, asymptotic function η∗=ma(ta−t) with two fitting parameters: the initial slope m and the asymptote a. A third output parameter tl was defined as the time when a specific fraction of overpotential range was lost. Several methods to determine each output parameter are presented. More importantly, using the large amount of data generated from over 95 time-dependent simulation series (over 29,000 individual simulations), simple correlations are made between the input and output parameters that provide predictive capability of operational lifetime tl (and m and a), without needing further high-performance computing.