干湿循环条件下膜电极屈曲变形造成催化层裂缝的原因分析与对策

Catalyst layer cracks by buckling deformation of membrane electrode assemblies under humidity cycles and mitigation methods

Tomoaki Uchiyama

Hideyuki Kumei

Toshihiko Yoshida

In this study， the formation of catalyst layer （CL） cracks in a membrane electrode assembly （MEA） under multiple humidity cycles is analyzed to propose solutions to mitigate the crack formation． The MEA is fabricated by the thermal transfer of catalyst layers （4 μm thickness） to both sides of the NR211 （25 μm thickness）． The buckling deformations are created by microscopic 300 μm diameter holes in the polyimide （PI） films． The microscopic holes model the clearance between the MEA and gas diffusion layers （GDLs）． The clearance height is adjusted by the film thickness． The MEA and PI film are sandwiched between the GDLs and exposed to humidity cycles in test cells． When the clearance height decreases， the CL cracks do not tend to develop after repeated humidity cycles． In the FEM analysis， swelling of the NR211 causes large plastic strains on the CL in areas that correspond to crack locations in the experiment． For the lower clearance heights， the plastic strain in the CL is reduced due to reduction of the in－plane swelling ratio of the NR211 as it swells． A reduction in in－plane swelling of the PEM by thoughtful structure design is directly effective in preventing CL cracks as well．

Fig． 1． Proposed mechanical degradation mechanism of membrane electrode assemblies under multiple humidity cycles．

Fig． 2． Diagram of a buckling test in the membrane electrode assemblies under multiple humidity cycles．

这个装配名义压强很小，不理解为什么用这么小。

仿真参数，可以作为对标参数。

Fig． 3． Approximated mechanical properties of （a） NR211 and （b） the catalyst layer （CL） under several relative humidities．

80度不同湿度下的拉伸测试结果，这种结果都极为宝贵。

Because the isolated CL was not analyzed by the tensile test， the parameters of the CL are determined from data for the NR211 and NR211－CL．

Fig． 4． Dimensional changes of the NR211 and the modeled PEM with relative humidity changes．

Swelling ratios of the NR211 are defined in the in－plane and
through－plane directions． In－plane swelling （εin） was found to be
12．7％， and through－plane swelling （εth） was 24．6％．

压缩率很小。

Fig． 6． FEM analysis procedures for simulating the NR211eCL buckling test．

Fig． 7． A SEM observation of the NR211eCL bulge deformation after 5 humidity cycles at 100 um clearance height． The deformation is captured from an oblique view．

Fig． 8． A SEM observation from an upper surface of the NR211eCL bulge after 1000humidity cycles at 100 um clearance height．

Fig． 9． SEM observations from an upper surface of the NR211eCL bulge （a） after 1000 cycles at 100 mm clearance height， （b） after 1000 cycles at 62．5 mm clearance height， （c） after 2000 cycles at 37．5 mm clearance height （CL cracks are circled） and （d） after 8500 cycles at 25．0 mm clearance height．

间隙越小干湿循环耐受性越好。

Fig． 10． Deformation behaviors of the NR211－CL at 100 um clearance height during a humidity cycle． The contour indicates the plastic strain in radial direction． The conditions were （a） initial state at 50 RH％， （b） swelled state at 100 RH％， （c） dry state at 5 RH％ and （d） return to initial state at 50 RH％．

间隙100um，极易产生塑性变形。

Fig． 11． Plastic strain and true stress behaviors at the central and topside portion of the CL versus relative humidity change at 100 mm clearance height． Radial and circumferential components are similar and are shown in this figure．

Fig． 12． Equivalent plastic strains at 100 RH％ （swelled state） for chosen in－plane swelling ratios of the NR211 （εin） and clearance heights （h）． （a） εin＝12．7％， h＝100 um， （b）εin＝12．7％， h＝62．5 um， （c） εin＝12．7％， h ＝37．5 um， （d） εin＝12．7％， h＝25．0 um， （e） εin＝12．7％， h＝10．0 um and （f） εin＝5．0％， h＝100．0 um．

Fig． 13． Equivalent plastic strain behaviors of the CL at the upper bending portion during a humidity cycle．

Fig． 14． Thicknesses at the central and topside portion of （a） NR211 and （b） CL at 100 RH％ （swelled state）．

这个结果出乎意料，可以和三维打印制作的流场夹具用于燃料电池中液态水分布的可视化测量对比，膜的溶胀大约是25％，也就是实际膜溶胀后厚度31．25微米。而仿真告诉我们这个值竟然与膜和气体扩散层之间的间隙、膜的面内膨胀率有关。

Fig． 15． Strain components at the central and topside portion of NR211 at 100 RH％（swelled state）．

Conclusion

In this paper， experiments and FEM simulations were performed
to address the issue of CL crack formation by bulge deformation
under multiple humidity cycles． The formation of cracks in the CL
was analyzed by performing buckling tests at several clearance
heights． From the data of repeated humidity cycles， the CL cracks
happen at the central and topside bulges， and the occurrence of
cracks was reduced at lower clearance height． From the FEM
analysis of the buckling tests， the higher plastic strain portion seen
for the swollen state of the NR211－CL corresponds to the cracked
CL portions in the experiment． The equivalent plastic strain in the
CL decreases with a decrease in the clearance height． The mechanism
that allows suppression of CL crack formation at lower
clearance heights could be explained by analyzing the strain components of the NR211． When the NR211 swells， in－plane strain
decreases with a decrease in the clearance height whereas throughplane strain increases． In the simulation of a modeled PEM with a low in－plane swelling ratio， the plastic strain in the CL was reduced．
To reduce the strain in the CL， lower clearance heights between
the MEA and GDL and a lower in－plane swelling ratio in the PEM are recommended． A flat MPL， a rigid GDL and a narrow channel for the gas flow fields are useful methods to obtain lower clearance height． Lower in－plane swelling of the PEM can be achieved by alterations to the PEM composition such as reinforcement． Throughout this work， the mechanisms of MEA degradation and solutions to suppress the degradation under repeated humidity cycles were indicated．

根据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（其一、其四）：燃料电池电解质膜的破裂与针孔形成 中的结论，也不知道膜的MD和TD方向膨胀对于这个现象的解决有没有差异。