燃料电池气体扩散层微孔层中孔孔径对水管理的改善：以40度过饱和加湿条件运行为例［设计34］
Improving water management in fuel cellsthrough microporous layer modifications Fast operando tomographic imaging ofliquid water
Yasutaka NagaiJens EllerTatsuya HatanakaSatoshi YamaguchiSatoru KatoAkihiko KatoFederica MaroneHong XuFelix N． Büchi
Abstract
Polymer electrolyte fuel cells （PEFCs） havebeen actively developed for a wide range of power generation applications． Atthe high power densities required for automotive applications， sophisticatedwater management is vital to further improve cell performance． In particular，the gas diffusion layer （GDL）， which includes a microporous layer （MPL）， playsa crucial role in optimizing both water drainage and gas transport between thecatalyst layer and gas channels． The present work studied the effect of cathode MPL porosity on the water distribution in theGDL． The results show that cells in which MPL materials have larger， micron－sized pores exhibit betterperformance． Advanced 4D operando X－ray imaging （3D structure plus time）was employed to analyse the water content in GDLs， and demonstrated that thesuperior performance of cells with large MPL pores is due to the efficientformation of water pathways． These pathways are based on water clustersallowing percolation in the through－plane direction from the bottom to the topof the GDL． Such pathways decrease the liquid water level in the entire cathodeGDL． Especially， large MPL pores mergenumerous small liquid water pathways in the catalyst layer and stabilize themmorphologically， thus creating primary pathways for effective water drainage．
Fig． 1． a） Components of the operando cellemployed for XTM and b） the XTM cell， complete with head socket and heated gasfeed tube， and unheated exhaust gas lines at the cell top． XTM reconstructions ofthe operando fuel cell： c） horizontal slice through the cell and d） in－planeslice through the cathode GDL； in c） and d） the regions of interest for the evaluationof channel （red dashed line）， rib （blue） and repetition unit （green） areindicated．
XTM ：X－ray tomographicmicroscopy
Table 1 Operating conditions forelectrochemical performance and XTM experiment
Fig． 2． Cross－sectional slice images of theGDLs obtained from micro XTM at 0．325 μm／voxel： a） GDL－M1 and b） GDL－M2．In－plane slice images of the MPL region obtained from micro XTM at 0．65μm／voxel： c） GDL－M1 and d） GDL－M2． In－plane slice images of the MPL regionobtained from nano XTM at 70 nm／voxel： e） GDL－M1 and f） GDL－M2．
Fig． 3． a） Pore size distribution in theMPL region obtained from micro （0．65 μm／voxel） and nano （70 nm／voxel） XTMmeasurements． GDL－M1（open symbols）， GDLM2 （solid symbols）． b） Percolation pathsof GDL－M1 and GDL－M2 in the MPL region obtained from micro XTM （0．65 μm／voxel）．c） Variations in water pressure during water permeability tests and d） liquidwater distributions of GDL－M1 and GDL－M2 during the water injection at 100 kPa，obtained from micro XTM at 0．65 μm／voxel．
Fig． 4． Current－voltage curves for Cell－M1and Cell－M2 （solid symbols， left yaxis）， iRcorrected curves （open symbols） andhigh－frequency resistance values （solid symbols， right y－axis）． Testconditions： cell temperature 40 ᵒC， dry gas velocities 7．6 m s－1 for air and5．3 m s－1 for H2， relative humidity105％ for air and of 100％ for H2， gaspressure 100 kPa．
Fig． 5． a） Diagram showing the differentwater cluster types in the cathode GDL domain． Temporal variations in theliquid water saturation in the cathode GDL under constant current operation： b）Cell－M1 at 0．5 A cm－2， c） Cell－M2 at 0．5 A／cm－2， d） Cell－M1 at 1．0 A cm－2 ande） Cell－M2 at 1．0 A cm－2． Conditions： cell temperature 40 C，dry gas velocities 7．6 m s－1 for air and 5．3 m s－1 for H2， relative humidity105％ for air and 100％ for H2， gas pressure 100 kPa．
Fig． 6． Temporal variations in the liquidwater saturation in the rib and channel areas of the cathode GDL under constantcurrent operation at 1．0 A cm－2： a） Cell－M1 channel， b） Cell－M2 channel， c）Cell－M1 rib and d） Cell－M2 rib．
Fig． 7． 3D renderings of the cathode GDLand channel domains with liquid water， viewing towards cell outlet at 1．0 Acm－2 operation after 300s： a） Cell－M1 and b） Cell－M2 view of GDL solid andwater in the channel． c） Cell－M1 and d） Cell－M2 renderings of the differentwater cluster types， the classification is the same as in Figure 5a； 2Dprojections of the liquid water volume fractions in the cathode GDL in the through－planee） Cell－M1 and f） Cell－M2． Black dashed lines indicate the edges of the flowfield ribs／ channels．
Fig． 8． Schematic representation of waterdrainage and gas diffusion in the cathode GDL： a） Cell－M1 and b） Cell－M2．
Conclusion
the MPL pore structure in PEFC watermanagement， and proposed a new structural configuration leading to improvedcell performance． The 3D structures of water clusters in the cathode GDL wereassessed within a cell without large pores in the cathode MPL （Cell－M1） and acell having large micron－sized pores in the cathode MPL （Cell－M2） underoperating conditions． The aim of these trials was to obtain an understanding ofthe liquid water pathways and oxygen gas diffusivity in the GDL． Cell－M2 showedsuperior performance because the efficient formation of water pathways in thisunit reduced the amount of liquid water in the entire cathode GDL． The dataalso suggest that the presence of large pores in the MPL facilitates themerging of many liquid water pathways generated in the catalyst layer and soforms stable primary pathways that readily connect water clusters in thesubstrate． As a result， water is efficiently drained． Due to this effectivedrainage function in the GDL， oxygen readily diffuses into the substrate（passing through small pores in the MPL） to reach the catalyst layer withlittle resistance． A number of previous studies have predicted this MPL watermanagement mechanism based on 2D radiography， simulated ex situ experiments andother techniques， and our results verify the validity of this hypotheticalprocess． Also in keeping with prior suggestions， the present results clearlyshow that improvements in the MPL structure， involving both larger pores for liquid water transport and smaller pores for gasdiffusion， effectively enhance water management in the entire GDL， with subsequentimprovement of the cell performance． Depending on the pore size， density and distribution state of the large pores in theMPL region， the effect on water transport which contributes to the cellperformance is considered to be different． We plan to systematically investigatethe effect of large pores of MPLs on the water management in order to developbetter GDLs．
This research demonstrates a possible water management mechanism at anintermediate temperature of 40 C under high humidity conditions． However，actual automotive FCs are operated at various temperatures with a wide range ofother parameters， including current density， gas velocity and humidity． Inparticular， the cell temperature is known to significantly affect cell performanceas a result of variations in water management properties． When a fuel cell isoperated at higher temperatures （e．g．， 80 C）， the water vapour generated at thecatalyst layer will migrate through the MPL and GDL． However， if a largetemperature gradient is present across the GDL， this vapour condenses and formsliquid water clusters within the substrate， leading to a more complicated watertransport mechanism through the GDL that cannot be described by a singlemechanism． Thus， further investigation using this operando time－resolved XTMtechnique is required in order to understand the different water managementmechanisms associated with various operating conditions． We expect that thisreport will assist in such investigations， leading to the development ofnext－generation FCVs．
