Durability of Pt Catalysts Supported on Graphitized Carbon-Black during Gas-Exchange Start-Up Operation Similar to That Used for Fuel Cell Vehicles

.

2][3] Nevertheless, the widespread commercialization of PEFCs has been impeded because of the large amount of the platinum catalyst required, which must be minimized to reduce the system cost. 4The minimization can be achieved, for example, by the improvement of utilization of Pt nanoparticles dispersed on carbon black supports (Pt/CB) with high surface area used as the cathode catalyst for PEFCs as well as the improvement of the durability.During the start-up and shutdown (SU/SD) cycles of the PEFCs, air and H 2 coexist transiently in the anode until the replacement of the former with the latter (or vice versa) is completed.Reiser  et al. showed that this situation causes the cathode potential to climb to more than 1.5 V due to the so-called "reverse current mechanism", 5 which significantly accelerates the degradation of the catalyst due to the oxidation of the CB and the agglomeration or dissolution of the Pt nanoparticles. 6][9] The high dispersion of the Pt nanoparticle catalyst particles on the GCB support prepared by this method provides enhanced oxygen reduction reaction (ORR) activity.It was also found that the durability of the n-Pt/GCB catalyst was improved during a potential cycling evaluation that simulated the variation of the cathode potential during the SU/SD of the PEFCs.These studies have demonstrated important roles of the support carbon and the state of the Pt particles on the supports to lead to high activity and durability for PEFC electrocatalysts.
In order to clarify the mechanism of the performance degradation under the actual SU/SD conditions of PEFCs, Kreitmeier et al. investigated the carbon support corrosion by measuring local CO 2 formation with online mass spectrometry. 10Lamibrac et al. measured the internal current in the cathode during the SU/SD of PEFCs by using a segmented cell. 11Durst et al. also used a segmented cell to correlate the local performance loss of the catalyst layer (CL) with its degradation. 12However, these measurements have included the effects of both the SU and the SD on both the cell performance and the catalyst degradation, so that the individual effects of each are still unclear.
The three unique points of this research are summarized as follows.First, while in our previous research, it was found that a well dispersed Pt catalyst, n-Pt/GCB, showed higher durability than commercial Pt/GCB under a potential cycling evaluation that simulates SU/SD, in the present research, we demonstrated that n-Pt/GCB has high durability under actual FCV SU conditions, which involve gas exchange.The changes of the electrochemical active surface area (ECA), cell performance, particle size distribution of Pt and the degradation of the carbon supports were investigated by electrochemical measurements, transmission electron microscopy (TEM), scanning TEM (STEM), and micro-Raman spectroscopy.Second, we applied glancing incidence X-ray diffraction (GIXD) 13 for the analysis of Pt degradation in the CL and membrane.This is the first application of GIXD in the PEFC field.GIXD is a simple, non-destructive analytical technique and is a useful way to analyze Pt degradation quickly.Finally, while it has been well recognized that the cathode CL in the gas outlet region degrades severely under the SU condition caused by the "reverse current mechanism," we found that the catalyst degrades not only in the outlet region but also in the inlet region during SU.We associated this degradation in the inlet region with the potential excursions caused by the ORR during the SU by using a single cell with two RHEs.

Experimental
Preparation of CCMs.-Commercial 30 wt% Pt-loaded GCB catalyst supplied by Tanaka Kikinzoku Kogyo K. K. (c-Pt/GCB) and 30 wt% Pt-loaded GCB catalyst prepared in house by the nanocapsule method (n-Pt/GCB) 9 were used for the cathodes of the CCMs.A commercial Nafion membrane (NRE 212, DuPont, 50 μm thickness) was used as the polymer electrolyte membrane (PEM).As the anode, commercial 50 wt% Pt catalyst supported on carbon black supplied by Tanaka Kikinzoku Kogyo K. K. (c-Pt/CB) was used for all experiments.The CCMs were prepared in the same manner as reported in our previous work. 8Briefly, the catalyst pastes were prepared from the each Pt-loaded carbon catalysts with Nafion binder (ion exchange capacity 0.9 meq g −1 , DE521, E. I. Du Pont de Nemours & Co., Inc.) and the mass ratio of Nafion binder (dry basis) to carbon black (Nafion binder/carbon) was adjusted to 0.7.The catalyst paste was directly sprayed onto the PEM to prepare the CCM by use of a pulseswirl-spray apparatus (PSS, Nordson Co.), and then dried at 60 • C in an electric oven.The Pt loading on both cathode and anode sides was 0.5 ± 0.1 mg -Pt cm −2 , and the active geometric area of the electrode was 29.2 cm 2 .The CCM was pressed at 140 • C and 1.0 MPa for 3 min.The CCM was sandwiched between two gas diffusion layers (GDLs, 25BCH, SGL Carbon Group Co., Ltd.) and was then assembled into a single serpentine pattern cell (Japan Automotive Research Institute (JARI) standard cell) consisting of two carbon separator plates.

Procedure of CCMs evaluation for durability and electrochemical properties.
-The durability evaluation was performed according to a gas exchange protocol simulating air start (AS) conditions, as shown in Fig. 1 and Table I.The initial ECA of the Pt catalyst in the cathode before the durability evaluation was evaluated by use of cyclic voltammetry (CV) at 40 • C and 65 • C with 100% RH N 2 in the cathode and 100% RH H 2 in the anode with a potentiostat (HZ-5000 Automatic Polarization System, Hokuto Denko Co.).The anode was used as both reference electrode and counter electrode for the CV.At 65 • C, the cathode potential was swept at 20 mV s −1 from 0.07 V to 1.0 V, and at 40 • C, it was swept at 50 mV s −1 from 0.05 V to 1.15 V.The value of ECA was determined from the hydrogen adsorption charge, referred to Q H • = 0.21 mC cm −2 , the value used conventionally for polycrystalline platinum. 14,15The current-voltage (I-E) curves were measured galvanostatically supplied to the anode and the cathode by use of an electronic load (PLZ-664WA, Kikusui Electronics Co.) operated in the constant current mode (5 min acquisition at each current), controlled by a measurement system (FCE-1, Panasonic Production Engineering Co.) at 65 • C with 100% RH H2 and air under ambient pressure (1 atm) with gas utilizations of 70% and 40%, respectively.The durability evaluation of the c-Pt/GCB and n-Pt/GCB catalysts was performed by repetitive cycles in which the gases supplied to the anode / cathode were switched in the following sequence: (1) dry air/ humidified (100% RH) air, (2) humidified H 2 / humidified (100% RH) air, and (3) dry N 2 / N 2 , at a cell temperature of 45 • C. The supply gas was purged to the cell with a holding period of 90 s for air / air and H 2 / air and a holding period of 60 s for N 2 / N 2 during each cycle.The gas flow rates used during the AS cycles are shown in Table I.After every 200 cycles, the ECA values were estimated from CV, and the I-E curves were obtained under the same cell operating conditions as those used initially.
Raman spectroscopy.-Thestructural changes of the GCB support after the AS cycling were evaluated from the change of Raman spectra recorded by a confocal micro-Raman instrument (LabRAM HR-800 spectrometer, HORIBA Jobin Yvon, Ltd.).The Raman spectra were obtained by excitation with the radiation from a He-Ne laser operated at approximately 2 mW.The Raman spectroscopic measurements were performed with a 100 × magnification objective, a 100-μm confocal aperture for the sample illumination and collection of the scattered light, a 600-line mm −1 grating, and a Peltier-cooled CCD camera as a detector.The instrument control and spectral analysis were performed with the software programs Labspec 5 (HORIBA Jobin Yvon, Ltd.) and GRAMS/AI 8.0 (Thermo Fisher Scientific, Inc.).In order to easily compare the results for each sample, the Raman spectra of the Pt catalysts illustrated in the figures were normalized by the intensity of the peak at ca. 1580 cm −1 , which corresponds to the spectrum for a stable graphitic lattice, as discussed below.Curve fitting for the determination of spectral parameters was performed with the software program GRAMS/AI 8.0.

TEM and STEM measurements.-The morphology of CLs and
Pt dispersion were characterized by TEM (H-9500, Hitachi High-Technologies Co.; acceleration voltage 300 kV) and by STEM (HD-2700, Hitachi High-Technologies Co.; acceleration voltage 200 kV), respectively.The average Pt particle size and standard deviation for Pt-loaded carbon catalysts were obtained from the image analysis of at least 200 randomly selected Pt particles in the TEM and STEM images for the initial condition and after the AS cycling, respectively.To analyze the distribution of the Pt particle degradation along the through-plane direction of the CL, ultrathin slices of the CCM were prepared with a focused ion beam system (FIB, FB-2200, Hitachi High-Technologies Co.).
GIXD measurements.-ThePt particle size as function of depth along the through-plane direction of the CLs was evaluated by use of the glancing incidence X-ray diffraction (GIXD) method with an X-ray diffractometer (XRD, Smart-Lab, Rigaku Co.; Cu Kα, 40 kV, 20 mA) equipped with a confocal mirror (beam diameter 30 μm).The glancing angle (θ) was varied from 0.2 • to 5.0 • with steps of 0.2 ).The X-ray diffraction intensity of each scanning angle (2θ) was measured with a fixed exposure time of 3 s at each step.In order to detect the θ value corresponding to the interface between the CL and PEM, the GIXD was performed on a MEA, separately prepared in the same manner as that described above but with an additional thin layer of sputtered gold (Au) between them.

Measurements of the electrode potential during the AS cycling.-
The electrode potentials of the c-Pt/GCB cathode during the AS cycling were measured by use of a JARI cell with two reversible hydrogen electrodes (RHE).Those two RHEs, which can observe the potentials of the gas inlet and outlet regions, were individually constructed on the top and bottom of the anode-side bipolar plate (Fig. 2a).These RHEs were constructed from gas diffusion electrodes using c-Pt/CB loaded on GDLs (0.5 ± 0.1 mg-Pt cm −2 ; 4 mm diameter).
Fig. 2b shows the cell voltage and the cathode potential changes in both the inlet and outlet region during SU conditions (AS cycle).The cathode potential in the outlet region increased to around 1.6 V when hydrogen was introduced to the cell.This phenomenon has been explained by the reverse current mechanism. 5On the other hand, the cathode potential in the inlet region temporarily increased and then decreased.We consider that the potential increase in the inlet region was caused by the reverse current, because the H 2 -air front gradually moved from inlet region to the outlet region, and the immediate decrease was caused by the simultaneous ORR associated with the reverse current in the outlet region.

Evaluations of Pt catalyst durability by electrochemical
measurements.-Fig.3a shows cathode ECA changes of the cells using c-Pt/GCB and n-Pt/GCB during the AS cycles.The ECA values were calculated from the H adsorption peaks of the CV for c-Pt/GCB and n-Pt/GCB at every 200th cycle up to 1000 cycles.The ECA values of both catalysts decreased to approximately half the initial values at 1000 cycles.However, the ECA value of n-Pt/GCB at the 1000th cycle was equal to that of c-Pt/GCB at about the 700th cycle, because the initial ECA value of n-Pt/GCB was higher, and the degradation rates were nearly the same.
Fig. 3b shows the changes in the I-E curves of the cells using the c-Pt/GCB and n-Pt/GCB cathode catalysts before and after the AS cycling; the cell performances for both catalysts decreased.The performance of the n-Pt/GCB cell was superior to that of c-Pt/GCB.In the region of current density higher than 0.5 A cm −1 , the performance degradation of the c-Pt/GCB cell was clearly more severe than that of the n-Pt/GCB cell.In this current density region, the oxygen mass transport strongly influences the cell performance. 16The ohmic resistances of the cells using both catalysts exhibited nearly the same values after the durability cycles compared with the initial values.
STEM examination of the catalysts: initial and after the durability evaluation.-Fig.4 shows the dark-field STEM images of both (a) c-Pt/GCB catalyst and (b) n-Pt/GCB catalyst in the initial state and after the AS durability tests, located in both the inlet and outlet regions.In the initial images, the Pt particles had been more uniformly dispersed on the GCB in the n-Pt/GCB catalyst than in the c-Pt/GCB catalyst.The mean Pt diameters for c-Pt/GCB and n-Pt/GCB were 3.3 nm and 2.4 nm, respectively (Table II).The images of both the inlet and outlet regions show the cross-section of the CL close to the PEM.In order to remove the effect of FIB-induced ion beam damage, the mean Pt particle-diameters were calculated from the Pt particles that were not directly on the top-surface of the CL slice.The results are shown in Table II, together with those of the center, and close to the GDL side.The numbers in parentheses indicate the distances in μm from the PEM surface.The mean Pt diameter for c-Pt/GCB in the outlet region was not obtainable, because the thickness of the CL decreased (Fig. 5) and the Pt particles were severely agglomerated (Fig. 4).The distances from the PEM surface in the outlet region were significantly smaller than those in the inlet region, because the overall

Table II. Pt particle diameters (nm) of c-Pt/GCB and n-Pt/GCB in the initial state and both inlet and outlet regions at several depth positions after the AS durability evaluation obtained by STEM images (Fig. 4). Values in parentheses show distances from the interface between the PEM and the CL (μm).
After thickness of the CL decreased substantially due to the severe carbon corrosion that occurred in the outlet region.In both regions, the Pt particle sizes of both catalysts increased compared with the initial values.In the inlet region, the Pt particle sizes of both c-Pt/GCB and n-Pt/GCB catalysts at the PEM side were smaller than those at the GDL side.In contrast, in the outlet region, the Pt particle size for n-Pt/GCB at the PEM side was larger than those at the center and at the GDL side.The values observed after the AS cycles for n-Pt/GCB became smaller than those for c-Pt/GCB.These results indicate that the uniform distribution of Pt particles in the n-Pt/GCB catalyst, which can be observed in the initial image of Fig. 4b, prevented both the increase of the Pt particle size in the inlet region and the severe aggregation of Pt in the outlet region, compared with the case of c-Pt/GCB.In the next section, we discuss the degradation behavior in the depth direction of each region in detail.6a shows the relationship between the incident angle of the X-ray and the crystallite size of Pt calculated by the Scherer equation.The crystallite sizes in the inlet and outlet region of both catalysts after AS cycling were larger than the initial values.In the inlet region, the crystallite sizes of both catalysts after cycling increased with increasing the X-ray incident angle.In the outlet region, these crystallite sizes decreased in the incident angle range from 0.2 • to 2.0 • and then increased in the range larger than 2.0 • .From ancillary measurements in which Au particles were deposited at the interface between the CL and PEM, the incident angle of about 2.0 • was established as a marker for this boundary, with smaller angles corresponding to the CL itself.In the CL region, i.e., incident angle smaller than 2.0 • , the Pt crystallite sizes for the c-Pt/GCB cathode in the inlet region increased to 4.2-5.0nm from the initial value of 2.6 nm after cycling, and those in the outlet region increased to 8.6 nm -12 nm.In the case of the n-Pt/GCB cathode, the Pt crystallite sizes in the inlet region increased to 4.0 nm-4.2 nm from the initial value of 2.5 nm, and those in the outlet region increased to 5.4 nm-9.5 nm.The Pt crystallite sizes of both CLs in the inlet region were smaller than the Pt particle sizes obtained by the STEM measurements (Fig. 4 and Table II).These results indicate that severe aggregation of the Pt particles had occurred, together with Pt re-deposition, during the AS cycles, because the present STEM images of the Pt particles were not able to discriminate between large single crystallites and aggregated Pt particles.The Pt crystallite sizes after the cycling of the n-Pt/GCB cathode were also smaller than those for the c-Pt/GCB cathode, similar to the results from the STEM images (Fig. 4 and Table II).These results also indicate that the uniform distribution of the Pt particles in the n-Pt/GCB catalyst prevented both the severe aggregation of Pt, particularly in the outlet region, compared with the case of c-Pt/GCB.

GIXD measurements of the catalysts: initial and after the durability evaluation.-Fig.
In contrast, in the region corresponding to incident angles larger than 2.0 • , these values were nearly the same and increased with increasing X-ray incident angle θ (Fig. 6a).Fig. 6b shows the cross section of dark-field STEM images of the nearby interface between  the cathode CL and the PEM of both CCMs.These Pt particle sizes increased with increasing distance from the cathode CL.The trend of increasing Pt crystallite sizes for both catalysts at the larger θ-values provides information on the Pt particles not only in the CL but also in the PEM, because of the XRD from deeper positions within the PEM.In the PEM region, it is considered that dissolved Pt in ionic form was deposited as the metal by hydrogen permeating from the anode, and the size of the Pt crystallites depended on the hydrogen concentration, which would be higher closer to the anode.

Raman spectroscopy measurements of the catalyst: initial condition and after the durability evaluation.
-To clarify the carbon corrosion phenomena during the AS cycling, Raman spectroscopic analysis was carried out according to procedures described in our previous work. 8As an indicator of the carbon corrosion, the ratios of A D1-band / A G-band are shown in Fig. 7.The values of the A D1-band / A G-band ratio in the inlet region for both catalysts after the AS cycling were nearly the same compared with those for the initial condition.These results show that the carbon corrosion in the inlet region was negligible.In contrast, the A D1-band / A G-band ratios in the outlet regions of both c-Pt/GCB and n-Pt/GCB were ca.1.6 times and ca.1.3 times higher, respectively, after the AS cycling than the initial values.These results show that the carbon corrosion that was caused by the reverse current and the cathode potential change observed in Fig. 2b predom-inated in the outlet region, in agreement with the results of the CL thickness changes (Table II).The A D1-band / A G-band ratio for n-Pt/GCB in the outlet region after the AS cycling was lower than that for c-Pt/GCB.This result shows that the carbon support degradation for n-Pt/GCB was milder than that for c-Pt/GCB, also in agreement with the CL thickness change results (Table II).In our previous research, 8 in the case of c-Pt/GCB, Pt particles were attached densely at defect sites on the GCB support.In the case of n-Pt/GCB, Pt particles were dispersed not only at defect sites, which are easily corroded, but also at stable sites of the graphitized layer on the GCB support.We consider that the carbon corrosion is accelerated by catalytic oxidation with Pt oxide.Thus, the degradation of n-Pt/GCB with well dispersed Pt particles on the GCB support was milder than that on c-Pt/GCB, due to most Pt particles existing on stable sites.
The carbon corrosion accelerates both the destruction of the CL and the detachment and aggregation of Pt particles, as well as leading to the degradation of cell performance, particularly in the high current density region.Therefore, we consider that the cell performance using the n-Pt/GCB cathode was superior to that using c-Pt/GCB in the high current density region after the durability cycling (Fig. 3b), because the carbon corrosion was suppressed by the highly uniform Pt distribution of n-Pt/GCB. 7,8vestigation of the Pt degradation in the inlet region.-According to the above results, we found that degradation of the Pt catalyst occurred not only in the outlet region but also in the inlet region.We considered two possible scenarios of the Pt degradation in the inlet region.The first is that the interim electrochemical measurements carried out during the durability evaluation could have affected the Pt dissolution, because the ORR performance in the outlet region decreased during the durability evaluation, and the load (current density) in the inlet region would have increased in order to compensate. 17The second is that the ORR in the inlet region (shown in Fig. 2) carried out during the durability evaluation could have accelerated the Pt dissolution and re-deposition during the potential cycling.
To investigate these effects, three durability protocols were evaluated with the same CCM construction using the c-Pt/GCB catalyst.In the first, we excluded the interim measurements (CV and I-E) from the AS cycles in order to avoid their influence.In the second and third protocols, potential cycling profiles of 0.8 V to 1.2 V and 0.3 V to 1.2 V, respectively, were evaluated in order to define the effect of the lower potential limit in the inlet region that would be caused by the AS cycles.These potential cycling evaluations were applied for 1000 cycles with 100% RH air (360 ml min −1 flow rate) at the cathode and 100% RH H 2 (38 ml min −1 flow rate) at the anode, as shown in Fig. 8. Fig. 9a shows CVs at 40 • C before and after the three evaluation protocols described above and also an AS evaluation with the interim measurements (AS with CV and I-E) included for comparison.The AS evaluation that was carried out without the interim measurements is shown as "AS without CV and I-E."The two types of potential cycles with different lower potential limits are shown as "PC 0.8-1.2V" and "PC 0.3-1.2V." Fig. 9b shows the ECAs calculated from the H adsorption peaks of the CVs shown in Fig. 9a.In Fig. 9b, the ECA measured after AS without interim measurements was around two times higher than that for the AS with interim measurements.This result shows that the interim measurements enhanced the Pt particle degradation. 17The ECA value obtained after the evaluation for PC 0.8-1.2V was nearly the same as that for the AS without the interim measurement, and the ECA value obtained after the evaluation for PC 0.3-1.2V was lower than those of either the AS without interim measurements or PC 0.8-1.2V.9][20] the lower potential limit of the potential cycling enhances the Pt degradation.These results suggest that the degradation of the Pt particles in the inlet region was caused by potential cycling due to a local ORR effect associated with the reverse current in the outlet region.Fig. 10 shows the relationship between the incident X-ray angle and the Pt crystallite size in the inlet region of the cathode with the c-Pt/GCB catalyst before and after the AS evaluation with/without the interim measurements.The Pt crystallites examined after the AS without interim measurements also increased in size.In the incident angle range from 2.0 • to 5.0 • , which is the PEM region, the increase of the Pt crystallite size after the AS evaluation without the interim measurements was smaller than that for the AS with the interim measurements.These results indicate that the AS accelerated the Pt dissolution in the inlet region, and the severe growth of the Pt crystallites was caused by the interim measurements.
From the above results, we conclude that the degradation of the Pt particles in the inlet region can be ascribed to the potential cycling caused by the ORR in the inlet region during the AS durability evaluation.In addition, we conclude that the severe degradation of the Pt catalyst that resulted from the interim measurements, which is similar to the load cycle operation in the actual FCV, occurs not only in the outlet region but also in the inlet region.

Conclusions
The first key point of the present research is that, while we previously found that a well dispersed Pt catalyst, n-Pt/GCB, showed higher durability than commercial Pt/GCB under a potential cycling evaluation that simulates SU/SD, we have now demonstrated that n-Pt/GCB has high durability under actual FCV SU conditions, which involve gas exchange.The spatial distribution of the degradation of the carbon-supported Pt was investigated to evaluate in more detail the durability of the fuel cell under SU-like conditions, in which the anode gas was successively cycled between air, hydrogen, and nitrogen.Both the ECA values and the I-E performance of the cathode using the n-Pt/GCB catalyst, which was prepared by the nanocapsule method, were maintained at higher levels than that using c-Pt/GCB after 1000 durability cycles.The second key point is that we applied GIXD, which is a simple, quick, non-destructive technique with which to analyze Pt degradation in the CL and membrane, as the first application of GIXD in the PEFC field.Both STEM and GIXD measurements indicated that the uniform distribution of Pt particles in the n-Pt/GCB catalyst largely prevented their severe aggregation, particularly in the outlet region, compared with the case of c-Pt/GCB, with its nonuniform Pt particle distribution.Raman spectroscopy measurements indicated that the carbon corrosion was also suppressed by the uniform Pt distribution of n-Pt/GCB.STEM and GIXD measurements also showed that the degradation of the Pt catalyst occurred not only in the outlet region but also the inlet region.The final key point is that the degradation of Pt particles in the inlet region can be ascribed to both the interim measurements carried out after every 200th cycle and the potential excursions caused by the ORR in the inlet region during the AS evaluation.Thus, while it has been well recognized that the cathode CL in the gas outlet region degrades severely under the SU condition, we found that the catalyst degrades not only in the outlet region but also in the inlet region during SU.

Figure 1 .
Figure 1.Experimental procedures for the AS durability evaluation.

Figure 2 .
Figure 2. (a) Schematic of the anode separator with RHE at the inlet region, RHE inlet and outlet region, RHE outlet and (b) changes of the cell and cathode potential during the AS cycling.

Figure 3 .
Figure 3. (a) ECA changes of c-Pt/GCB and n-Pt/GCB at 65 • C with 100% RH N 2 in the cathode and 100% RH H 2 in the anode measured during the AS cycling and (b) I-E curves of c-Pt/GCB and n-Pt/GCB at 65 • C with 100% RH H 2 in the anode and air in the cathode before and after the entire AS-SU durability evaluation.

Figure 4 .
Figure 4. Dark-field STEM images of (a) c-Pt/GCB and (b) n-Pt/GCB of the initial state and both the inlet and outlet regions after the AS cycling.Pt particle sizes were calculated from the Pt particles that were not directly on the surface of the FIB slice of the CL within the white frames shown in the images of the inlet region for c-Pt/GCB and both inlet and outlet regions for n-Pt/GCB.

Figure 5 .
Figure 5. Dark-field STEM images of (a) inlet region and (b) outlet region of c-Pt/GCB after the AS cycling.

Figure 6 .
Figure 6.(a) Relationship between the incident X-ray angle and the crystallite size of c-Pt/GCB and n-Pt/GCB before and after the AS durability evaluation in the inlet and outlet regions and (b) dark-field STEM images of MEA crosssections of c-Pt/GCB and n-Pt/GCB after the AS durability evaluation in the inlet and outlet region.The CL thicknesses for c-Pt/GCB and c-Pt/GCB became small due to the ion beam used to prepare the ultrathin slices of the PEMs.

Figure 7 .
Figure 7. Raman spectra for (a) c-Pt/GCB and (b) n-Pt/GCB and (c) A D1-band / A G-band ratios for c-Pt/GCB and n-Pt/GCB before and after the AS durability evaluation in the inlet and outlet regions.

Figure 8 .
Figure 8.(a) Experimental procedures for the durability evaluations of the AS without interim measurements and potential cycling (PC) and (b) cathode potential changes during the PC.

Figure 9 .
Figure 9. (a) CVs of c-Pt/GCB at 40 • C with 100% RH N 2 in the cathode and 100% RH H 2 in the anode before and after the AS durability evaluation both with and without interim measurements (CV and I-E) and potential cycling (PC) with lower limit potential 0.8 V (PC 0.8-1.2V) and 0.3 V (PC 0.3-1.2V); (b) ECAs of c-Pt/GCB calculated from the CVs (a).

Figure 10 .
Figure 10.Relationship between the incident X-ray angle and the crystallite sizes of c-Pt/GCB before and after the durability evaluation of AS with and without interim measurements (CV and I-E).