| With the commercialization of proton exchange membrane fuel cells（PEMFC）worldwide,there is an urgent need to address challenges regarding their cost,performance,and durability.Catalysts account for over 50%of the total cost of PEMFC,and they have a crucial impact on the performance and durability of PEMFC.The electrochemical kinetics of methanol oxidation reaction（MOR）in anode catalytic layer is several orders of magnitude slower than that of hydrogen oxidation reaction,which makes the performance of direct methanol fuel cell（DMFC）lower than that of hydrogen fuel cell.At present,compared to the supported catalytic layer,the self-supporting ultra-thin catalytic layer exhibits many unique structural advantages.The ultra-thin catalytic layer can maximize the mass transfer efficiency of substances passing through the catalytic layer.Research has found that controlling the composition,microstructure design,and structure-activity relationship of catalysts is beneficial for improving the utilization rate of platinum（Pt）and catalytic performance of nanoscale catalysts.The microstructure parameters of the catalytic layer surface,porosity and twist,play a crucial role in the material transport ability.In addition,battery temperature,ionomer content,proton/ion conductivity,material flow rate,and material concentration all affect the properties of the catalyst layer,thereby affecting the proton exchange membrane fuel cell.The catalytic layer was studied using multi-scale methods such as first principles,numerical simulation,and machine learning.The research content is as follows:（1）Surface engineering plays a crucial role in improving the performance of electrocatalysts for fuel cells and energy applications.In this chapter,GCO and GOH are used as descriptors,and density functional theory（DFT）is used to explore the influence of Pt Cu catalysts with Cu vacancy defects on the adsorption behavior and electronic structure changes of CO and OH.DFT calculations show that Cu vacancy defects can reduce the adsorption capacity for CO and exhibit good adsorption capacity for OH,indicating that the surface vacancy structure is energetically advantageous for the MOR process.In order to better understand the influence of Cu vacancy defects on MOR,the full path calculation and path selective transition state analysis of MOR were carried out from the thermodynamic and kinetic aspects.The calculation results indicate that the Pt Cu alloy catalyst can reduce the toxicity of CO and choose a non CO pathway,while the Pt Cu alloy rich in Cu vacancy defects has the lowest potential barrier to generate formic acid,and the formic acid pathway is the most favorable.Through this study,we have deeply recognized the importance of defect engineering in improving catalytic performance.（2）Using a combination of numerical simulation and machine learning to evaluate the impact of different catalyst layer parameters on the performance of DMFC.Develop a suitable DMFC model through numerical simulation.Introduce different parameter ranges of the catalytic layer into the mathematical model for calculation,establish a database,and solve the prediction of power density as a regression problem.Using seven machine learning algorithms to predict the database,it is found that tree integration methods（Random Forest Regression（RFR）,Gradient Boosting Regression with XGBoost（XGB）and Extra Trees Regression（ETR））work well.Then,based on the three optimized ensemble algorithms,the feature importance ranking was performed on 19 feature values,with the cathode specific surface area,anode specific surface area,and anode catalytic layer thickness having the highest feature importance score.Then,the sequential model-based algorithm configuration program obtains the expected improvement（EI）values for various parameter combinations.The combination samples with the top ten EI values were brought into the simulation model for calculation,and the polarization curve obtained was compared with the polarization curve of the maximum power density in the original database.It was found that there was a 50%probability that it exceeded the original data.Artificial intelligence can provide direction for experimental design and improve the efficiency and quality of experimental work.（3）Using numerical simulation methods,investigate the effects of material transport and electrochemical reactions of dominant porous structures with different porosity in ultra-thin catalytic layers on PEMFC cathodes.Using the principle of immiscible two-phase separation of fluids,dominant porous structures with different porosity were obtained,and the relationship between porosity,tortuosity,and permeability of different structures was calculated.It was found that tortuosity slightly decreased with the increase of actual porosity,while permeability increased with the increase of porosity.Due to the slow oxygen reduction reaction kinetics of Pt based catalysts,activation losses are caused,with severe cathodic activation losses.Therefore,dominant nano porous structures with different porosity were used as PEMFC cathode catalytic layers for multi physical field simulation.The research results showed that the higher the porosity and power density,the better the battery performance.This chapter provides some insights into the impact of catalyst layer microstructure on PEMFC through research. |
