No abstract

A modeling study on a polymer electrolyte membrane fuel cell by means of non-equilibrium thermodynamics is presented. The developed model considers a one-dimensional cell in steady-state operation. The temperature, concentration and electric potential profiles are calculated for every domain of the cell. While the gas diffusion and the catalyst layers are calculated with established classical modeling approaches, the transport processes in the membrane are calculated with the phenomenological equations as dictated by the non-equilibrium thermodynamics. This approach is especially instructive for the membrane as the coupled transport mechanisms are dominant. The needed phenomenological coefficients are approximated on the base of conventional transport coefficients. Knowing the fluxes and their intrinsic corresponding forces, the local entropy production rate is calculated. Accordingly, the different loss mechanisms can be detected and quantified, which is important for cell and stack optimization.

No abstract

Metal oxides with oxygen vacancies are widely used in electrochemical processes at high temperature due to their ionic conductivity. These processes are strongly influenced by the electrostatic potential of the ions because it is closely related to the electrochemical potential. We calculate the partial molecular Coulomb internal energy for different compositions of yttria-stabilized zirconia (YSZ) with molecular dynamics (MD) at different temperatures and zero pressure. On the basis of thermodynamic considerations, we assume that these quantities correspond to the electrostatic potential of ZrO 2 and Y 2 O 3. We also calculate the mean electrostatic potential of the ions and develop a mixing rule between this potential and the electrostatic potential of the molecules. With this mixing rule and following the thermodynamic framework proposed in this study, one can calculate the Coulomb contribution of other thermodynamic properties like the entropy or the Nernst–Planck diffusivities for YSZ-like metal oxides. Furthermore, the methods proposed here can be extended for other electrolyte mixtures.

Despite the high advances of classical molecular simulation to study bulk phases, classical force fields (FFs) to describe interactions at interfaces are rarely available in the literature. In this study, FFs to describe fluid | solid interfaces are developed by matching forces and energies from ab initio simulation and by using a newly developed genetic algorithm (GA). The interfacial FFs are parameterized to be combined with existing classical bulk FFs. Our procedure is tested on the methanol (CH 3 OH) | ZnO interface. The results for the forces, energies, and some structural adsorption properties calculated using an own parameterized interfacial FF are comparable with results from ab initio and experimental data. With this, we illustrate the potential of the proposed procedure to yield accurate models for interfacial systems to be combined with available bulk FFs.