相关对比讨论见第二条推文

90度30％RH下氧浓度、膜厚度对燃料电池电解质膜降解速率的影响

Degradation of Polymer－ElectrolyteMembranes in Fuel Cells I． Experimental

T． Madden

D． Weiss

N． Cipollini

D． Condit

M． Gummalla

S． Burlatsky

V． Atrazhev

Abstract

In this work， chemical degradation isstudied using highly controlled measurements of the fluoride ion release fromsubscale cells in degrading environments using perfluorosulfonic－acid－basedmembrane electrode assemblies， primarily with cast， 25um／1mil thick membranes．Effects of key variables， such as oxygenconcentration， relative humidity （RH）， temperature， and membrane thickness onthe fluoride ion emission rate （FER） are described under open－circuit decayconditions． Some of the observed trends are expected or consistent withprevious observations， such as decreasing FER with decreasing temperature andincreasing RH． Other trends observed are not expected， such as a logarithmic decrease of FER with oxygenconcentration and increasing FER with increasing membrane thickness．Cross－sectional transmission electron microscopy analysis of decayed membranesindicates a surprisingly homogeneous distribution of small Pt particles （3 to 20nm in diameter）， presumably from dissolution and migration from the cathode．The experimental results are consistent with radical generation at these Ptparticles from crossover hydrogen and oxygen， subsequent radical migration， andpolymer attack． The response of the FER to new experimental conditions in thisstudy suggests that the attack can existat any plane within the membrane， not just the “Xo” plane of maximum Ptprecipitation．

Figure 1． BOL performance curve for the  cast 1 mil MEA with type 1 electrodes showing cell voltage data （uncorrected）  and corrected for ohmic resistance data （current interrupt method）．  Performance curves were measured with hydrogen and air at saturator dew  points of 65 and 62°C， respectively， and the cell temperature controlled to  65°C．

Two types of electrodes were used in this  study： type 1 electrodes were applied by a vendor， had higher performance，  and exhibited higher FER under equivalent conditions vs 2 electrodes． Type 2  electrodes were laboratory－made samples that included high－surface－area 50 wt  ％ Pt on Ketjen carbon support （supplied by Tanaka Kikinzoku Group） mixed with  Nafion 1100 equivalent weight solution （supplied by Solution Technologies）．

Figure 2． Hydrogen crossover currents  measured at 0．5 V with pure hydrogen／nitrogen （100％ RH） applied at the anode／cathode，  respectively， for the cast 1 mil MEA with type 1 electrodes at BOL． Balanced （0  psig） and unbalanced （5 psi） anode overpressure measurements are used to  characterize cell integrity at BOL．

Figure 3． Plot of cumulative fluoride emitted  （umol／cm2） vs time for various O 2 concentrations during a single－cell test using  the cast 1 mil MEA with type 1 electrodes at 90°C， 30％ RH， and open－circuit  conditions． The FER is obtained by least－squares fit of the data during the O  2 introduction．

Table I． Replicate values for the  fluoride emission rate measured for six unique cell builds at the following  conditions： 90°C， 30％ RH， 100％ O 2 （dry basis）， and open－circuit potential．  All samples employed type 1 electrodes．

Figure 4． Plot of FER （umol／cm 2 h） and  cathode－to－anode FER ratio vs O 2 concentration during a single－cell test  using the cast 1 mil MEA with type 1 electrodes at 90°C， 30％ RH， and  open－circuit conditions． The line is a least－squares logarithmic fit．

Figure 5． Plot of FER （umol／cm 2 h） vs RH  using the cast 1 mil MEA with type 1 electrodes at 90°C， 100％ cathode O 2 ，  and open－circuit conditions． The dotted line is a guide for the eyes．

Figure 6． Plot of ln FER （umol／cm 2 h） vs  temperature （at 55， 70， 80， and 90°C） using the cast 1 mil MEA with type 1  electrodes at 30％ RH， 100％ cathode O 2 ， and open－circuit conditions． The line  is a least－squares linear fit corresponding to ～70 kJ／mol．

Figure 7． Plot of FER （umol／cm 2 h） vs membrane  thickness using type 2 electrodes at 90°C， 30％ RH， 100％ cathode O 2 ， and  open－circuit conditions．

Figure 8． Plot of Pt particle locations  using two TEM imaging techniques． DF－STEM was better for determining particle  location but was not accurate for particle size． BF－TEM was more difficult  but yielded more accurate particle sizes．

Two different techniques were used to determine  the Pt particles： bright－field （BF） TEM and dark－field （DF） scanning TEM （STEM）

Conclusions

Experimental methods have been refined to  provide accurate determinations of FER over a wide range of conditions． Some  of the observed trends are expected， such as decreasing FER with decreasing  temperature and increasing RH．

Other trends observed were not expected，  such as a logarithmic decrease of FER with oxygen concentration and  increasing FER with increasing membrane thickness．

Cross－sectional TEM analysis of decayed  membranes indicates a surprisingly homogeneous distribution of small Pt  particles， presumably from dissolution and migration from the cathode． The  experimental results are consistent with radical generation and migration at  these Pt particles from crossover hydrogen and oxygen． The response of the  FER to new experimental conditions in this study suggests that the attack  throughout the volume of the membrane populated with Pt particles and not  just the “Xo” plane of maximum Pt precipitation．