Modelling CO poisoning and O2 bleeding in a PEM fuel cell anode

Abstract

SUMMARYFuel gas containing carbon monoxide severely degrades the performance of a polymer electrolyte membrane (PEM) fuel cell. However, CO poisoning can be mitigated by introducing oxygen into the fuel (oxygen bleeding). A mathematical PEM fuel cell model is developed that simulates both CO poisoning and oxygen bleeding, and obtains excellent agreement with published, experimental data. Modelling efforts indicate that CO adsorption and desorption follow a Temkin model. Increasing operating pressure or temperature mitigates CO poisoning, while use of reformate fuel increases the severity of the poisoning effect. Although oxygen bleeding mitigates CO poisoning, an unrecoverable performance loss exists at high current densities due to competition for reaction sites between hydrogen adsorption and the heterogeneous catalysis of CO.

A transient PEMFC model with CO poisoning and mitigation by O2 bleeding and Ru-containing catalyst

“…This model assumes linearly bonded CO and H 2 on Pt, with temperature, potential and coverage dependent rate constants. Kinetics for O 2 bleeding are taken from [28] and assume that the gas-phase oxidation of CO and H 2 is negligible. Those on Pt-Ru are an extension of the model in [36], which neglects the adsorption of CO on Ru and the H 2 kinetics entirely, and are based on the bifunctional mechanism proposed by Watanabe and Motoo [39].…”

“…Equations for the evolution of the site coverages θ i , i ∈ {O, H, CO, M}, are as follows (using notation similar to that in [28]) [45] for CO adsorption and desorption were proposed by Springer et al [25] (see references therein) based on analysis of the dependence of the critical current density on θ CO . The same approach is adopted here, including the dependence of the forward step on θ CO , with symmetry factor β and interaction parameter r. The Langmuir model assumes that coverage does not exceed a monolayer, that there is no interaction between adsorbed molecules and that the apparent standard free energy of adsorption is coverage-independent.…”

“…Modelling of membrane degradation is hampered by uncertainty in the reactions driving the structural changes [24]. On the other hand, due to the work of Springer et al [25], Baschuk and Li [26][27][28] and Bhatia and Wang [29], good models for the kinetic mechanism of CO poisoning of the anode have been developed. This issue is of paramount importance as long as reformate gas is intended as the main source of H 2 , whether on-or off-board a vehicle, until carbon-free pathways are developed, [30].…”

“…Baschuk and Li [27] placed the same mechanism inside a one-dimensional, steady-state, single-phase MEA framework, later extending their work to include the effects of O 2 bleeding using a more simplified MEA model [28]. A three-dimensional steady-state, single-phase phase model can be found in [33].…”

“…A similar model, which includes the diffusion of gas, can be found in [34]. A different approach can be found in [35], in which there is assumed to be a competition between linearly bonded and bridge-bonded adsorption, in contrast to the purely linear-bonded models in [25][26][27][28][29]33,34]. This model neglects the transport problem, is steady-state, employs temperature-independent rate constants and simplifies the CO adsorption kinetics.…”