| Proton exchange membrane（PEM）fuel cell,with high efficiency and low pollution,is one of the most promising electrochemical conversion devices,attracting considerable attention for many applications.As the key component of PEM fuel cell,catalyst layer（CL）provides the sites for electrochemical reactions.CL contains the carbon-supported catalyst particles and polymer electrolyte,of which the conformation is complex.Moreover,multiple transport processes occur simultaneously,including the reactant gas transport,water transport,proton transport,and electronic conduction,all of which significantly affect the CL performance.Therefore,exploring the transport properties in CL is significant for advanced CL fabrication.Current experimental technologies can’t capture the transport phenomena in the nanometer-sized CL clearly.By contrast,the molecular dynamics（MD）simulation is a powerful tool in exploring the transport in molecular-scale structures.This dissertation explores the transport phenomena in molecular-scale structures of CL mainly using MD simulation.Besides,machine learning and experimental tests are also involved.Firstly,the simulated density and transport properties,including the water and hydronium diffusivities in bulk hydrated perfluorosulfonic acid（PFSA）ionomer,the O2 diffusivity,solubility,and permeability in bulk hydrated PFSA,the O2 diffusivity and solubility in pure water,and the O2 transport resistance in the ionomer film on a Pt surface,are well validated with the previous experimental and simulated results.After model validation,the MD simulation is used to explore the effects of PFSA side chain length on the structure,oxygen transport,and thermal conductivity of bulk hydrated PFSA.The simulation result shows that the structures of hydrated PFSA strongly affect the O2 diffusivity and solubility.The small cavities in the poorly hydrated PFSA mainly result in the low O2 diffusivity and high O2 solubility.With increasing water content,the number of small cavities in the PFSA with shorter side chains decreases more rapidly,leading to a higher O2 diffusivity and lower O2 solubility at high water content.The thermal conductivity of the bulk PFSA with shorter side chains is higher.The oxygen transport in the ionomer film on planar Pt surface is also investigated.The ionomer film can be divided into three regions.The dense ultrathin sublayer with a tight arrangement of PFSA chains in the ionomer-Pt interface has a density～1.5-2 times higher than that in the bulk-like ionomer.The ionomer-Pt interface plays a dominant role in the O2 transport resistance due to its dense structure.O2 mainly permeates via the water sites in the ionomer-Pt interface and thus a lower resistance is present at higher water content.In the bulk-like ionomer region,the O2 permeation routes are different at different water contents.Two ways including adding ionic liquid（IL）and modifying ionomer-substrate interaction strength are proposed to enhance the oxygen transport across the ionomer film.Adding the IL significantly alters the ultrathin sublayer structure by inhibiting the tight arrangement of PFSA chains.The IL addition provides a larger free space for O2 dissolution in the ionomer ultrathin sublayer.Consequently,the O2 density in the ultrathin sublayer is improved by an order of magnitude and the O2 transport flux across the ionomer film is increased by up to 8 times due to IL molecules’presence.This dissertation finds that the medium ionomer-substrate interaction strength benefits the O2 transport across the ionomer film.Coupling MD simulation and machine learning is applied to optimize the interaction strengths between the substrate and the hydrophobic and hydrophilic parts of ionomer film.The data-driven surrogate model based on support vector machine（SVM）predicts similar results as the MD simulation.The genetic algorithm（GA）is further employed to obtain the optimal interaction strength,which increases the O2 transport flux by 6 times compared with the highest interaction strength.The ionomer film structure,O2 density distribution,transport fluxes,and permeation routes are investigated for carbon-supported polyhedral Pt particles.The results show that a dense ultrathin sublayer with a tight arrangement of PFSA is present on the Pt facets（namely region A）.In the ionomer near the Pt edges and corners（namely region B）,the structure is less dense due to the weaker Pt attraction,resulting in a higher O2 density than that in region A.O2 fluxes show that 90%of O2 molecules reach the Pt cube and tetrahedron particles via their upper corner and edge regions.In the vicinity of Pt particles,O2 permeation routes are inferred to penetrating region B to the Pt upper corners or edges instead of region A to the Pt facets.Finally,the morphology of water domain in the ionomer film on Pt particles and the effect mechanism of water content on electrochemical surface area（ECSA）are explored by MD simulation and experimental tests.The MD results show that the morphology of water domain in ionomer film is isolated water clusters at low water content,resulting in the poor ECSA due to the lack of proton transport paths.The proton transport paths are quickly formed due to the morphology transition of water domain from isolated water clusters to water channel network with increasing water content,thereby leading to the rapid increase of ECSA.The slight increase of ECSA at high water content mainly results from the increased contact area between water domain and Pt particle. |
