燃料电池膜电极在干湿循环－化学衰减循环联合工况下的失效分析
Accelerated Stress Testing of Fuel Cell Membranes Subjected to Combined Mechanical／Chemical Stressors and Cerium Migration

Yeh－Hung Lai

Kenneth M． Rahmoeller

James H． Hurst

Ratandeep S． Kukreja

Mohammed Atwan

Andrew J． Maslyn

Craig S． Gittleman

Abstract

      A highly accelerated stress test （HAST） has been developed to generate local stressful conditions that are representative of those in automotive fuel cell stacks． Using a 50－cm2 cell cycled between 0．05 and 1．2 A／cm2 with a low inlet RH in the co－flow configuration， the HAST creates a distribution of combined mechanical／chemical stressors in the membrane with the maximum chemical stress occurring near the gas inlets and the maximum mechanical stress near the outlets． Conducting HASTs using a current distribution measurement tool and a shorting／crossover diagnostic method to track the state of health of a robust membrane containing both a mechanical support and a chemical stabilizing additive， the result shows that the membrane location with the most severe thinning coincides with that of the deepest membrane hydration cycling． Upon examination of the cerium redistribution patterns after the test， it was found that the severe humidity cycling generated by the HAST condition near the outlet region not only generated the highest membrane mechanical stress but also resulted in the strongest water flux， which may cause local depletion of the cerium added as chemical stabilizer． Further study is required to decouple the cerium migration effect from the possible mechanical／chemical synergistic degradation effect．

Figure 1． The stack－up schematic of the single fuel cell．

RTD：resistance temperature detector

Figure 2． （a） The layout of the current distribution circuit board， （b） the layout of the flowfield， （c） flowfield／CD board assembly， （d） the layout of the RTD temperature circuit board．

Figure 3． （a） An MEA overlaid on the flowfield． The GDL， electrode， and
membrane layers are of the same size． The active area is framed by a Kapton subgasket with a 50 cm2 opening． Note that portions of the membrane and electrode layers are outside the active area． （b） The current distribution grid． The four CD segments along the R1 row are the inlet segments while the four along R25 form the outlet segments．

The subgasket had a window frame of 2．73 cm × 18．30 cm to form an active area of 50 cm2．可以计算出分区电流块的大小6．825mm＊7．32mm。

Figure 4． The HFR ［ohm cm2］ distribution under 0．05 A／cm2 at various RHs and 80◦C．

Figure 5． The HFR－λ calibration data and best fit curves for the C3 column．

Figure 6． The crossover current vs． hydrogen concentrations for the two crossover measurement steps． The slopes are equal to the H2 crossover current at 100％ H2 while the intercepts are the shorting currents at 0％ H2．

Figure 7． The distributions of （a） shorting resistance ［ohm］， （b） diffusive H2 crossover current ［mA／cm2］ at 300／300 kPa （H2／N2）， （c） H2 crossover current ［mA／cm2］ at 300／200 kPa （H2／N2）， and （d） convective H2 crossover current ［mA／cm2］， which is the difference between （b） and （c）．

Figure 8． Snapshot of cell current， HFR， temperature， and voltage over
600 sec．

前7张图几乎没有什么需要说的，第8张图中的工况和缓解干湿循环的机械因素造成燃料电池电极质膜失效的五种方法（其五）燃料电池电解质膜在混合模式加速测试下的原位诊断和分区降解分析

中的工况并不完全相同。缓解干湿循环的机械因素造成燃料电池电极质膜失效的五种方法（其五）燃料电池电解质膜在混合模式加速测试下的原位诊断和分区降解分析文中的电流密度变化从0．08到0．8 A cm－2，而这片文章中电流密度变化从0．05到1．2 A cm－2，跨越的范围更大。流道设计也不一样。

而且电流密度的增加不是随时间线性状态的。

Figure 9． The distributions of current， HFR， temperature， and water content at （a） 0．05 and （b） 1．2 A／cm2， respectively at 80◦C nominal temperature．

Figure 10． （a） The snapshot of the membrane hydration and dehydration in the segments of Column C3 during current cycling at 80◦C． （b） The close－up of λ （solid lines） and current （dashed line） profiles from 790 to 930 sec．

因为是同流，入口相对偏干，出口偏湿。总体趋势电流密度入口偏小，出口偏大，但不是线性增加。

Figure 11． The H2 crossover current distribution ［mA／cm2］ of the 80◦C HAST at various test hours． （a） The progression of the diffusive crossover current change relative to the beginning－of－test values． （b） The corresponding convective crossover current distribution． Gas flow direction is from top to bottom．

扩散氢渗透系数和对流氢渗透系数趋势相当，都在出口出现。

Figure 12． The H2 crossover current distribution ［mA／cm2］ of the 90◦C HAST at various test hours． （a） The progression of the diffusive crossover current change relative to the beginning－of－test values． （b） The corresponding convective crossover current distribution． Gas flow direction is from top to bottom．

名义温度从80度增加到90度，出口处出现较大针孔的时间起始点从1213hr缩短到648hr，也就是寿命缩短为50％。

Figure 13． （a） CD segment grid； （b） hydration amplitude deltaλ ［H2O／SO3H］； （c） diffusive crossover current change ［mA／cm2］ after 1213 h； （d） cerium concentration XRF image of a fresh sample； （e） cerium concentration XRF image of the end－of－test sample over－laid on the flowfield； （f） end－of－test membrane thickness ［μm］； and （g） cross－sectional images within selected segments． Data are from the 80◦C HAST test．

回头看前文The nominally 12 μm thick membrane contains an ePTFE matrix to enhance its mechanical robustness．也就是说在最后的16％活性区域长度范围内电解质膜的厚度出现了大幅减薄，hydration amplitude deltaλ ［H2O／SO3H］达到13。

Figure 14． （a） CD segment grid； （b） hydration amplitude λ ［H2O／SO3H］； （c） diffusive crossover current change ［mA／cm2］ after 648 h； （d） cerium concentration XRF image of a fresh sample； （e） cerium concentration XRF image of the end－of－test sample over－laid on the flowfield； （f） end－of－test membrane thickness ［μm］； and （g） cross－sectional images within selected segments． Data are from the 90◦C HAST test．

90度下的电解质膜失效状态大致相当。

Figure 15． The cerium map of 90◦C HAST near the outlet， superposed with the outlines of active area， cathode channels region， and anode channels region．

Conclusions

A highly accelerated stress test （HAST） was developed to generate
local stressful conditions that are representative of those in a
power－generating fuel cell stack in automotive application． Using a
50 cm2 small scale cell cycled between 0．05 and 1．2 A／cm2 with a
constant inlet gas flows and low inlet RH operating in the co－flow
configuration， the HAST creates a smoothly changing distribution of
combined mechanical and chemical stressors in the membrane； with
the maximum chemical stress occurring near the gas inlets due to that part of the cell experiencing the driest conditions and the maximum mechanical stress occurring near the gas outlets due to a high swing in water content．

In this study， HAST experiments were conducted at 80 and 90◦C
to investigate the degradation of a robust membrane containing both
a mechanical support and a chemical stabilizing additive． Using a
current distribution measurement tool and a shorting／crossover diagnostic method based on variable anode H2 concentration to track the local shorting， diffusive H2 crossover， and convective H2 crossover currents， the state of health of the membranes under various combined stressors was monitored for the three degradation modes． While the 90◦C tests failed in less than half the time of the 80◦C test （744 and 1645 hours， respectively）

关注一下这里，作者认为的失效点时的渗氢电流大于134mA／cm2，而不是25mA／cm2。

， the degradation behavior and the eventual failure modes in the two tests are remarkably similar， which suggests
that the 90◦C HAST can be a useful test to accelerate the learning
cycles． In both cells， the membrane location with the most severe
thinning and crossovers coincides with that of the deepest membrane
hydration cycling and highest mechanical stress． However there is no
evidence of mechanical damage in the electrodes or membranes in
either MEA， even in the locations of membrane suffering from the
most chemical degradation and highest mechanical stress． This result
suggests that high mechanical stress might have created a synergistic nanoscale condition that accelerates chemical degradation even
in the absence of macroscopic defects． However， upon examination
of the cerium distribution patterns after the test， it is found that the
severe humidity cycling generated by the HAST condition near the
outlet region not only generated the highest membrane mechanical stress but also resulted in the strongest water flux， which may cause depletion of the cerium added as chemical stabilizer． Further study is required to decouple the cerium migration effect from the possible mechanical／chemical synergistic degradation effect．

Regardless of whether the membrane chemical degradation is
caused by the depletion of cerium or by the synergistic effect of
combined mechanical and chemical stressors， the discovery of the colocation of the maximum mechanical stress with the maximum cerium depletion by humidity cycling found in the HAST is significant as it poses a rather severe scenario for membrane durability in fuel cell
stacks operating under dynamic automotive conditions．

colocation这个词我还以为是杜撰的，作者的意思容易理解，但是用的词感觉不太对。