The Relationship between the Active Pt Fraction in ... - ACS Publications

Article pubs.acs.org/JPCC

The Relationship between the Active Pt Fraction in a PEFC Pt/C Catalyst and the ECSA and Mass Activity during Start-Up/Shut-Down Degradation by in Situ Time-Resolved XAFS Technique Kotaro Higashi,† Gabor Samjeské,† Shinobu Takao,† Takuma Kaneko,† Oki Sekizawa,† Tomoya Uruga,†,§ and Yasuhiro Iwasawa*,† †

Innovation Research Center for Fuel Cells, The University of Electro-Communications, Chofugaoka, Chofu, Tokyo 182-8585, Japan Japan Synchrotron Radiation Research Institute, SPring-8, Sayo, Hyogo 679-5198, Japan

§

S Supporting Information *

ABSTRACT: Transient-response kinetics of the transformations (six elementary steps) of the Pt valence, coordination number of Pt−Pt bonds, and coordination number of Pt−O bonds of a Pt/C cathode catalyst in a polymer electrolyte fuel cell (PEFC) under cyclic voltage operations (0.4 → 1.4 → 0.4 VRHE) during anode−gas exchange (AGEX) treatments (startup/shut-down) has been studied by in situ time-resolved quick X-ray absorption fine structure (QXAFS, 100 ms/spectrum). The transient-response analysis identified the existence and fractions of three different kinds (active, less active, and inactive) of Pt nanoparticles in the Pt/C cathode. The active Pt nanoparticles degraded to less active and inactive Pt nanoparticles by the AGEX cycles. The degradation probability and mechanism were clarified by the transient-response kinetics. The electrochemical surface area (ECSA) and mass activity (MA) of the Pt/C cathode catalyst also decreased with increasing AGEX cycles. It was found that the change in the sum of the fractions of the active and less active Pt nanoparticles correlates with the change in the ECSA and MA during the AGEX treatments. The in situ time-resolved QXAFS analysis provides direct information on the dynamic behavior of the Pt/C catalyst relevant to the electrochemical performance and property under the operando conditions for thorough understanding of the degradation process toward PEFC improvement.

1. INTRODUCTION For widely spread commercialization of polymer electrolyte fuel cell (PEFC) vehicles, it is indispensable to further improve the oxygen reduction reaction (ORR) activity and, particularly, the long-term durability of Pt/C cathode catalysts in the membrane electrode assembly (MEA) of PEFC, which enables the remarkable reduction of the Pt amount and cost of PEFC stacking.1−18 The limited durability of MEAs in PEFC vehicles under dynamic operation conditions especially during start-up/ shut-down (“on/off”) cycles of PEFCs is a major problem to solve in the development of highly durable next-generation PEFCs.19−23 Measured or theoretically predicted cathode potentials vary over a range of about 1.2 to 1.75 V, resulting in the conditions facilitating carbon corrosion and PEFC performance loss.19,22 A powerful tool to investigate the catalyst changes under operando conditions is in situ time-resolved quick X-ray absorption fine structure (QXAFS).8,9,24−51 Tada et al. were the first to report the elementary steps and kinetic rate constants for the ORR at the Pt/C cathode under PEFC operating conditions by means of the in situ time-resolved XAFS technique.26 The transient-response kinetic behavior of Pt nanoparticles in the Pt/C cathode catalyst layer during the © XXXX American Chemical Society

course of the degradation has been demonstrated as a key issue in order to understand and predict the degradation of Pt/C cathodes in PEFCs.52 Recently, repeated quick anode−gas exchange (AGEX) cycles were applied to Pt/C cathode catalysts in a PEFC MEA to simulate start-up/shut-down conditions and to examine a possible countermeasure for their effects on Pt/C catalyst degradation by in situ time-resolved QXAFS technique.19 The rate constants for the transformations of the Pt valence, coordination number (CN) of Pt−Pt bonds, and CN of Pt−O bonds of the cathode Pt nanoparticles under the transient-response voltage operations 0.4 → 1.0 VRHE/1.4 → 0.4 VRHE (vs RHE) decreased with increasing AGEX cycles. We also showed the decreases of the electrochemically active surface area (ECSA), power density maximum performance, and mass activity of MEA Pt/C by the AGEX cycles.19 End-oflifetime (EOL) analysis after the AGEX by spatially resolved nano-XAFS and TEM/STEM-EDS revealed that both Pt2+ ion Received: July 23, 2017 Revised: September 17, 2017 Published: September 19, 2017 A

DOI: 10.1021/acs.jpcc.7b07264 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 1. (a) Repeated AGEX experiments for MEA in a QXAFS cell by a four-way valve; (b,c) changes in cathode, anode, and cell voltages with sharp spikes in every AGEX (typical 50 cycles) (H2/air and air/air).

different sorts of active Pt nanoparticles responsible for the performance of the Pt/C cathode catalyst in the degraded MEAs by in situ time-resolved QXAFS at 100 ms time resolution. The in situ time-resolved QXAFS experiments were performed at three different areas of an MEA, top (close to gas inlet), middle, and bottom (close to the gas outlet), to examine spatial differences of Pt/C cathode deterioration. It was found that active Pt nanoparticles in the aging Pt/C cathode were transformed to less active and inactive Pt nanoparticles by repeated AGEX (start-up/shut-down) cycles, and the change in the sum of the fractions of the active and less active Pt nanoparticles correlated with the change in the ECSA and mass activity of the MEA Pt/C cathode catalyst during the AGEX treatments. The finding provides a new insight into understanding the degradation of MEA Pt/C cathode catalysts.

dissolution and isolated Pt nanoparticle detachment from carbon support occur around the boundary of the cathode layer with the Nafion electrolyte membrane as well as the microcracks and voids produced in the degraded cathode layer due to carbon corrosion.53−55 These results suggest the formation of different sorts of Pt nanoparticles with different performances heterogeneously distributed in the cathode layer as a consequence of the harsh reaction conditions caused by the reverse current decay mechanism (RCD) during the start-up/ shut-down phase. The Pt nanoparticle sizes in the cathode layer remained almost unchanged during the AGEX cycles and, hence, the heterogeneous and decreasing Pt nanoparticle performances cannot be ascribed to Pt nanoparticle growth. The heterogeneous and decreasing Pt nanoparticle performances cannot be ascribed to Pt utilization fraction change either because the first-order rate constants for the above transformation processes determined by the in situ time-resolved QXAFS data are independent of the amount of active Pt species.37,52 In other words, if the Pt/C cathode has only Pt sites with single activity and the active Pt amount decreases in the degradation process, the first-order rate constant should remain constant during the course of the degradation. If the rate constant decreases from that for the active aging Pt/C sample (no AGEX) by increasing AGEX cycles, then the formation and fraction of degraded Pt nanoparticles with smaller rate constants and without any activity during the AGEX cycles may be the key issues of the AGEX degradation to be explored in relation to the electrochemical surface area and mass activity. In the present study we have examined the structural kinetics of the transformations (six elementary steps) of the Pt valence, CN (Pt−Pt), and CN (Pt−O) of MEA Pt/C in the transientresponse processes under the voltage operations 0.4 → 1.4 VRHE and 1.4 → 0.4 VRHE after aging (no AGEX), 100 AGEX cycles, 200 AGEX cycles, and 300 AGEX cycles to determine the rate constants for the transformations and the fractions of

2. EXPERIMENTAL SECTION 2.1. MEA Pt/C Cathode Catalyst. Pt/C (TEC10E50E, Tanaka Kikinzoku Kogyo) was used as a cathode catalyst for PEFC (0.6 mg of Pt cm−2; Pt: 46.1 wt %). The surface area for the Pt/C was estimated to be 385 m2/g, and the mean pore diameter was estimated to be 9.6 nm.36 The MEAs (3 × 3.3 cm2) used in this study were provided by EIWA FC Development Co, Ltd. and are described in the Supporting Information (SI). An MEA was mounted into a homemade XAFS cell with cutouts in the “top”, “middle”, and “bottom” positions (Figure S1). Ru/KB (Tanaka Kikinzoku Kogyo, TEC30Ru(ONLY)E; Ru: 30 wt %; 0.6 mg of Ru/cm2) was used as an anode catalyst to avoid interference against XAFS measurements of the Pt/C cathode. 2.2. Electrochemical Measurements and AGEX Treatments. The cathode was connected as the working electrode and the anode (hydrogen-fed) served as the combined counter and reference electrode (Figure S1). All potentials are referenced to this pseudohydrogen reference electrode B


Article


transmission mode at the BL36XU station in SPring-8 using ion chambers (I0: Ar 15%/N2 85%; It: N2 100%) for incident and transmitted X-rays, respectively, and using a Si(111) double-crystal monochromator while measuring the current/ charge of PEFC during the potential operating processes, as shown in Figure S1.19,35−37 The cell voltage was changed from the open-circuit voltage (OCV) to 0.4 VRHE, where the voltage was kept for 40 s, followed by the rapid voltage jump from 0.4 VRHE to 1.4 VRHE (at time = zero for the transient voltage up process), and this voltage was kept for 40 s, and then reversely, the cell voltage was jumped rapidly from 1.4 VRHE to 0.4 VRHE (at time = zero for the transient voltage down process) as shown in Figure S2. The transient-responses of the MEA cathode catalyst under the voltage cycling operations were measured by QXAFS at a time resolution of 100 ms for 30 s from 10 s before each voltage jump (Figure S2). QXANES and QEXAFS spectra were analyzed in the similar way to the previous reports,8,9,36,37,42−46 using the Larch code containing the IFEFFIT Package ver.2 (Athena and Artemis).24,56 The white line peaks of normalized QXANES spectra were analyzed by curve fitting with a Lorentzian function and an arctangent function. Error ranges of the QXANES curve fitting were estimated as 95% confidence intervals. Background subtraction in the QEXAFS analysis was performed using Autobk.57 The extracted k2-weighted QEXAFS oscillations were Fourier-transformed to R-space over k = 30−120 nm−1, and the curve fittings were performed in the R-space (0.14−0.30 nm), which covers the Pt−O and Pt−Pt contribution because the length of the Pt−O bonds with surface and subsurface oxygen atoms on Pt nanoparticles in MEA Pt/C cathode catalysts under the present conditions is ∼0.200 nm. We conducted the curve-fitting analysis systematically in the identical k and R ranges through all time-resolved XAFS data. The fitting parameters for each shell were coordination number (CN), interatomic distance (R), correction-of-edge energy (ΔE0) (ΔE0 (Pt−Pt) = ΔE0 (Pt−O)), and Debye−Waller factors (σ2) for Pt−Pt and Pt−O. The Debye−Waller factors (σ2) for Pt−Pt and Pt−O determined by the curve-fitting processes for the series of QEXAFS data were averaged and fixed to be 0.007 and 0.004 Å2, respectively, to obtain convincing data in the series of the AGEX experiments. The phase shifts and amplitude functions for Pt−Pt and Pt− O were calculated from the FEFF 8.4 code using structural parameters obtained from the crystal structures of Pt and PtO.58 The amplitude reduction factor (S02) for Pt−Pt bonds in this study was estimated to be 0.836 by analyzing Pt foil. Error ranges of the curve-fitting analysis of QEXAFS Fourier transforms were based on the definition of the Larch code.56 The quality of the observed 100 ms time-resolved QEXAFS data was good enough for achieving the curve-fitting analysis similar to the previous one.52 2.4. Rate Constant Decision. The white line peak intensity (area) of the Pt LIII-edge QXANES and the coordination numbers (CN) of Pt−Pt and Pt−O determined by QEXAFS analysis were plotted against time (t) in the transient-response processes under the potential jumps of 0.4 → 1.4 VRHE and 1.4 → 0.4 VRHE. The rate constants determined from the time change in the above parameters were estimated by data fitting using an exponential function for the QXAFS analysis data or a linear combination of two exponential functions, taking into account the error weighting given by the inverse of the error.46,52 The data fitting with one or two exponential functions for the change of each parameter against

(RHE). The electrochemical conditioning and measurements were similar to the previous report as follows.19 The gas flows of H 2 (99.99999%; 150 sccm) for the anode and N 2 (99.99995%; 300 sccm) or air (atmospheric composition; 900 sccm) for the cathode were regulated by mass-flow controllers and were bubbled through humidifiers at 351 K. The high pressure gases (purity grade 1) were purchased from Taiyo Nippon Sanso Corp. The humidified gases were supplied to the in situ XAFS cell heated at 353 K, resulting in ∼93% relative humidity (RH). The MEA in the XAFS cell was conditioned by 150 conditioning (aging) cycles with a sequence of stepwise galvanostatic current steps every 6 s from the open circuit voltage (OCV) to a potential near 0.3 VRHE in H2 (anode) and air (cathode) operating atmospheres. The cathode degradation was carried out by repeated AGEX cycles (Figure 1).19 Electrochemical experiments in the AGEX treatments were controlled and monitored using a combination of AUTOLAB302N PGSTAT and BOOSTER20A with NOVA software (Metrohm Autolab B.V.). The passage of a gas-front between either hydrogen and air or air and hydrogen in the AGEX causes temporarily a steep increase in the difference between the cathode potential and electrolyte potential as shown in the potential vs time plots of Figure 1. Starting from the PEFC operating conditions (H2−air) and when the anode H2 gas was switched to air, which corresponds to the shut-down conditions (air−air), cathode potential spikes to about 1.3−1.5 VRHE were observed (RCD effect).19 After switching back to the operating conditions, the anode potential immediately returned to the standard value, and the cathode potential showed a negative potential spike down to about 0.7 VRHE before recovering. Both the cathode and cell potentials initially overshot and decreased to the standard OCV potentials relatively slowly. It should be noted that in our experimental setup, Ru/C was used as the anode catalyst material to avoid interference with the Pt signals in in situ QXAFS measurements as mentioned above. Since the overpotential for Ru−oxide formation is smaller compared to Pt−oxide formation, the Nafion−electrolyte potential ϕ will be less effected than in the case of a Pt/C anode. Therefore, the measured potential difference Ecathode − ϕ is smaller compared to MEAs in which both cathode and anode consist of Pt/C.19 Cyclic voltammograms (CVs) in MEA after the aging treatment and after every 100 AGEX cycles were conducted between 0.05 and 0.9 VRHE at 20 mV s−1 in H2 (anode) and N2 (cathode) operating atmospheres. The hydrogen adsorption charge determined from the CV was used to calculate the electrochemical surface area (ECSA) after correction for the double-layer charge in the potential region between 0.05−∼0.4 V, assuming 210 μC cm−2 as hydrogen adsorption charge for polycrystalline Pt.19,36 Following the CVs, MEA performances were examined under the galvanostatic conditions by applying stepwise increasing constant currents for 1 min followed by a frequency response analysis (FRA) at 10 mA AC amplitude and 1000 Hz frequency to measure the real part of the fuel-cell impedance to obtain the iR-free cell potentials. Performance tests were carried out using air as cathode feed gas as described previously.19 2.3. In Situ Time-Resolved QXAFS Measurements under Transient Response Conditions and QXAFS Data Analysis. The series of in situ (operando) time-resolved QXAFS spectra at the Pt LIII-edge for Pt/C in PEFC MEA under transient potential operations (anode: H2; cathode: N2; 353 K; ∼93% relative humidity) were measured in a C


Article


2.8 ± 1.0, and 3.0 ± 1.1 nm, respectively. In the bottom window area they were 2.8 ± 0.9, 2.8 ± 0.8, and 2.8 ± 0.8, respectively. These mean sizes are similar to 2.8 nm for the averaged Pt nanoparticle size in Pt/C after aging (no AGEX). The Pt nanoparticles with the 2.8 nm mean size are regarded to constitute approximately six Pt layers (the sixth, fifth, fourth, third, second, and first layers as well as a central atom), assuming approximate sphere structures with the fcc arrangement (d111 (interplanar spacing) ∼0.226 nm).37,56 These results demonstrate that there is no significant particle growth in the whole Pt/C cathode layer up to 600 AGEX cycles under the present AGEX conditions, though a little particle growth may occur at the minor part of the cathode layer during the repeated AGEX cycles.19 Thus, the analysis of the rate constants for each elementary step (Pt valence charging/discharging, Pt−O bond formation/dissociation, and Pt−Pt bond dissociation/reformation) on Pt cathode nanoparticles in the transient-response processes by the in situ time-resolved QXAFS analysis can directly document the transformations of those elementary reaction steps without considering the particle growth contribution to the QXAFS analysis. 3.2. In Situ Time-Resolved Pt LIII-Edge QXAFS Analysis under Transient Potential Operations. Figure 3 shows the typical k2-weighted Pt LIII-edge QEXAFS oscillations and their associated Fourier transforms (100 ms acquisition time) for the MEA Pt/C cathode catalysts after aging (no AGEX) at the top and bottom positions under the H2 (anode)−N2 (cathode) operating conditions at 0.4 and 1.4 VRHE. The structural parameters of the local coordination of the Pt/C in the MEA were determined by the curve fitting analysis of the Pt LIII-edge QEXAFS Fourier transforms (Figure 3). We followed our previous approach to the estimation of Pt nanoparticle structures at 0.4 and 1.4 VRHE8,9,35−37,52 The resultant estimated structures at 0.4 and 1.4 VRHE are also shown in Figure 3 as described in detail in the previous reports.8,9,35−37,52 The Pt nanoparticles in the Pt/C at 0.4 VRHE are essentially metallic, while the Pt nanoparticles at 1.4 VRHE constitute a surface PtO layer and a Pt metallic core, irrespective of the top, middle, and bottom positions (Figure 3). In situ time-resolved QXAFS allows the monitoring of the change of Pt catalysts at different cathode areas of an MEA in operando conditions. The temporal variations in the electronic and structural parameters of the Pt/C cathode catalyst in MEA samples after aging (no AGEX), 100 AGEX cycles, 200 AGEX cycles, and 300 AGEX cycles during the transient voltage cycling processes (0.4 → 1.4 → 0.4 VRHE) under H2 (anode)− N2 (cathode) were investigated by the in situ time-resolved QXAFS at the Pt LIII-edge. The temporal variations of the white line intensity (peak area) vs the reaction time at the top (a), middle (b), and bottom (c) window positions of the MEA in the transient-response process under the voltage operation 0.4 → 1.4 VRHE are shown in Figure 4 (left) and also are shown in Figure S4 (each curve is separately shown typically at the top position). In the voltage change from 0.4 to 1.4 VRHE, the white line peak intensity (Pt valence) in the Pt/C catalyst increased with reaction time, which indicates that Pt atoms were positively charged. The rate constants (k) of the Pt valence change (Pt charging) processes for the Pt/C catalysts after the aging and 100−300 AGEX cycles were determined by the fitting analysis using an exponential function (eq 1).

time (t) was performed for the period of 0−15 s after the voltage was jumped (0.4 → 1.4 VRHE and 1.4 → 0.4 VRHE). 2.5. Transmission Electron Microscope. Field-emission transmission electron microscope (FE-TEM, JEM-2100F, JEOL) at 200 kV equipped with an energy dispersive spectrometer (EDS) was used for estimation of the catalyst nanoparticle sizes. The average particle sizes in the top, middle, and bottom positions of MEA were estimated from 260−400 particles.

3. RESULTS AND DISCUSSION 3.1. Pt Nanoparticle Sizes at the MEA Pt/C Cathode during the AGEX Treatments. Due to the spatiotemporal properties of the RCD effect, the degree of the MEA deterioration should vary depending on the locations in the MEA with respect to the anode gas inlet and outlet. In situ time-resolved QXAFS allows the monitoring of changes of the Pt/C catalysts themselves in operando conditions while choosing each particular MEA area of interest. Since in the QXAFS analysis for the Pt/C catalysts, the Pt valence and the coordination number of Pt−Pt and Pt−O bonds at increasing and decreasing operations of the potentials depend on the Pt nanoparticle size, at first we examined the change in the Pt nanoparticle sizes during the AGEX treatments. The Pt nanoparticle sizes in the top, middle, and bottom positions of the MEA Pt/C cathode layer (Figure S1) were estimated by FE-TEM, discriminating the following three areas of the cathode layer: close to the GDL (gas diffusion layer; SI, page S1), around the middle, and close to the electrolyte. The results for the MEA Pt/C cathode layer after 600 AGEX cycles are shown in the histograms of Figure 2. Typical TEM images for

Figure 2. Histograms of Pt nanoparticle sizes in the top, middle, and bottom positions of the Pt/C cathode layer in the MEA after 600 AGEX cycles. The particle size distributions were estimated by 260− 400 nanoparticles measured by FE-TEM (JEM-2100F, JEOL).

the MEA Pt/C cathode after aging and after 600 AGEX cycles are shown in Figure S3. In the top window area, the mean sizes of Pt nanoparticles close to the GDL, in the middle of Pt/C cathode layer, and close to the electrolyte were estimated to be 3.0 ± 0.9, 2.6 ± 0.7, and 2.6 ± 0.7 nm, respectively. In the middle window area, the estimated mean sizes were 3.0 ± 1.0, D


Article


Figure 3. Typical Pt LIII-edge EXAFS oscillations and associated Fourier transforms for the MEA Pt/C cathode catalyst after aging (no AGEX) at different positions and potentials. The EXAFS data for Pt foil are also shown as reference. Phase shift: uncorrected. Potential-dependent surface structures of Pt nanoparticles in the MEA Pt/C at 0.4 VRHE (a) and 1.4 VRHE (b) are also illustrated, similar to ref 37 and ref 52

By using the k1 and k2 thus determined, we fitted the transient-response curves for the 100−300 AGEX cycle samples in Figure 4 (left) by eq 2 again to determine the fractions of A1‑x00 and A2‑x00 more adequately. The rate constants k1 and k2 for the Pt/C cathode catalysts after the aging and 100−300 AGEX cycles are listed in Table 1 (shown later). The k1 processes are completed within a few seconds (relaxation time = 0.45−0.52 s), and the k2 processes are completed within ten seconds (relaxation time = 2.44−3.57 s). The sum of the k1 and k2 processes should reach the similar white line peak areas (Pt valences) because the amount of Pt nanoparticles remains unchanged during the AGEX cycles as the average particle sizes do not change, which means A1−0 = A1−100 + A2−100 = A1−200 + A2−200 = A1−300 + A2−300. However, the experimental results on the amplitudes (fractions) of A1 and A2 in eq 2 under the voltage operation 0.4 → 1.4 VRHE in Figure S6 were not the case. The fraction A1 decreased from A1−0 (fraction = 1) at AGEX = 0 to the fractions A1−100 at AGEX = 100, A1−200 at AGEX = 200, and A1−300 at AGEX = 300, while the fraction A2 increased from A2−0 (fraction = 0) at AGEX = 0 to the fractions A2−100 at AGEX = 100, A2−200 at AGEX = 200, and A2−300 at AGEX = 300, and the total fractions of A1 and A2 definitely decreased with increasing AGEX cycles. The results in Figure S6 cannot be explained by the two kinds of active (k1) and less active (k2) Pt species because the total amount of Pt species in the Pt/C cathode should be constant. The results demonstrate the formation of inactive Pt species as the third kind of Pt species, which has no rate constant for the ORR event, by the repeated AGEX treatments. As the result, the final levels of the white line peak areas deceased with the increase of the AGEX cycles as shown in Figure 4 (left). The fractions (A3) of the inactive Pt species were calculated as A3 = 1−(A1/A1−0 + A2/ A1−0), which are also shown in Figure S6. It was found for the first time that we can quantitatively discriminate the presence and behavior of three different Pt species in the MEA Pt/C cathode catalyst under the AGEX conditions by the in situ

⎧ y t < 0⎫ ⎪ 0 ⎪ ⎬ f (x) = ⎨ ⎪ ⎪ ⎩ y1 + A1 − 0exp(− k1t ) t ≥ 0 ⎭ A1 − 0 = y1 − y0

(1)

The first-order rate constant (k1) determined by the exponential fitting decreased with increasing AGEX cycles as shown in Figure S5. However, this is strange because the firstorder rate constant (k1) should be independent of the initial amount (A1−0) of active Pt species in MEAs even if the initial amount of the active Pt nanoparticles and the initial Pt utilization fraction due to the carbon corrosion and porosity decrease by the AGEX treatment59 reduced by the transformation to inactive Pt species. In other words, the first-order rate constants for the transient-response process should remain unchanged with the AGEX cycles unlike Figure S5. Thus, the reduction of the rate constants in Figure S5 indicates the formation of the less active Pt nanoparticles with the smaller rate constant (k2) than the rate constant (k1) for the aging (activated) Pt/C catalyst. Hence, the transient-response curves in Figure 4 (left) were analyzed by two exponential functions (eq 2). In the fitting procedure using eq 2 for degraded MEA samples after 100, 200, and 300 AGEX cycles, we fixed the rate constant (k1) at the value for the aging sample and determined A1−x00, the rate constant (k2), and A2−x00 where (x00 = 100, 200, and 300). ⎧ y t < 0⎫ ⎪ 0 ⎪ ⎬ f (x) = ⎨ ⎪ ⎪ exp( ) exp( ) 0 y A k t A k t t + − + − ≥ 1 − x 00 1 2 − x 00 2 ⎩ 1 ⎭ A1 − x 00 + A 2 − x 00 = y1 − y0

x 00 = 100, 200, 300 (2)

The k2 values determined for each of the 100−300 AGEX samples were averaged to estimate a convincing k2, minimizing the error range for k2 in the fitting. E


Article


Figure 4. Transient response curves of the white line peak area (Pt valence) for the MEA Pt/C cathode catalysts at the top (a,A), middle (b,B), and bottom (c,C) positions under the voltage operation 0.4 → 1.4 VRHE (left) and under the voltage operation 1.4 → 0.4 VRHE (right). (1: blue) after aging (no AGEX); (2: red) 100 AGEX cycles; (3: green) 200 AGEX cycles; (4: purple) 300 AGEX cycles. Under H2 (anode)−N2 (cathode); cell temp.: 353 K, relative humidity: ∼93%. Data acquisition: every 100 ms.

different levels of carbon deterioration at the top, middle, and bottom positions. Local differences in the carbon corrosion after application of repeated start-up/shut-down cycles have been observed by ex situ techniques like TEM and SEM.23,60 The Pt/C cathode carbon layer around the top position among the MEA areas may be most damaged by the RCD mechanism.21 Particularly, the inactive Pt species was produced more at the top position than at the middle and bottom positions. In the latter MEA areas, the less active Pt species was formed more than the inactive Pt species. The temporal variations of the white line intensity (peak area) vs the reaction time at the top (a), middle (b), and bottom (c) window positions of the MEA in the transientresponse process under the voltage operation 1.4 → 0.4 VRHE

time-resolved QXAFS transient-response technique. The three different Pt species were also discriminated at the middle and bottom positions of the MEA (Figure S6). The rate constants (k1) for the Pt valence change at the top, middle, and bottom positions were determined to be 1.98 (± 0.19), 2.22 (± 0.34), and 1.93 (± 0.29) s−1, respectively, as listed in Tables 1(a), 2(a), and 3(a), which are nearly independent of the positions within the error range. The rate constants (k2) in these tables were also similar to each other; 0.28 (± 0.14), 0.41 (± 0.12), and 0.34 (± 0.17) s−1. In contrast to the rate constants, the reduction of the active Pt fraction (A1) and the increase of the inactive Pt fraction (A3) at the top position were more remarkable compared to those at the middle and bottom positions (Figure S6) probably due to F


Article


Table 1. Rate Constants (k1 and k2) and Amplitudes (A1 and A2) for the Active and Less Active Pt/C Catalysts as Well as A3Valuesa (a) 0.4 → 1.4 VRHE 0.4 → 1.4 V Pt valence

CN (Pt−Pt)

CN (Pt−O)

k A ratio k A ratio k A ratio

k1 or A1

k2 or A2

A3

k1 or A1

1.98 2.15 1.00 2.24 4.04 1.00 2.69 1.27 1.00

0.00 0.00 0.00

0.00 0.00 0.00

1.98 1.14 0.53 2.24 2.46 0.61 2.69 0.63 0.50

1.4 → 0.4 V

CN (Pt−Pt)

CN (Pt−O)

k′ A′ ratio k′ A′ ratio k′ A′ ratio

k2 or A2

after AGEX #200 A3

0.28 0.37 0.64 0.17 0.30 0.32 0.75 0.83 0.19 0.21 0.34 0.40 0.24 0.31 0.19 (b) 1.4 → 0.4 VRHE

after AGEX #300

k1 or A1

k2 or A2

A3

k1 or A1

k2 or A2

A3

1.98 0.77 0.36 2.24 1.33 0.33 2.69 0.48 0.38

0.28 0.49 0.23 0.32 1.52 0.38 0.34 0.36 0.28

0.89 0.41 1.19 0.29 0.43 0.34

1.98 0.49 0.23 2.24 1.32 0.33 2.69 0.32 0.25

0.28 0.55 0.26 0.32 1.12 0.28 0.34 0.39 0.31

1.11 0.52 1.60 0.40 0.56 0.44

after AGEX #100

after aging

top window Pt valence

after AGEX #100

after aging

top window

after AGEX #200

after AGEX #300

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

2.80 2.34 1.00 2.64 3.89 1.00 2.40 1.32 1.00

0.00 0.00 0.00

0.00 0.00 0.00

2.80 1.31 0.56 2.64 2.37 0.61 2.40 0.60 0.45

1.50 0.32 0.14 1.47 0.86 0.22 0.78 0.37 0.28

0.71 0.30 0.66 0.17 0.35 0.27

2.80 0.98 0.42 2.64 1.64 0.42 2.40 0.52 0.39

1.50 0.36 0.15 1.47 1.27 0.33 0.78 0.56 0.42

1.00 0.43 0.98 0.25 0.24 0.18

2.80 0.48 0.21 2.64 1.17 0.30 2.40 0.33 0.25

1.50 0.54 0.23 1.47 1.00 0.26 0.78 0.40 0.30

1.32 0.56 1.72 0.44 0.59 0.45

A3 = 1 − (A1/A1−0 + A2/A1−0) at the top position of the MEA sample under the voltage operations 0.4 → 1.4 VRHE (a) and 1.4 → 0.4 VRHE (b) for Pt valence, CN (Pt−Pt), and CN (Pt−O). H2 (anode)−N2 (cathode); cell temp.: 353 K, relative humidity: ∼93%. Data acquisition: every 100 ms.

a

that the Pt discharging process is relatively insensitive to the Pt/C deterioration in MEA compared to the Pt charging process. The temporal variations of the coordination number (CN) of Pt−Pt bonds (CN (Pt−Pt)) for the samples after aging (no AGEX) (1), 100 AGEX (2), 200 AGEX (3), and 300 AGEX (4) at the top (a), middle (b), and bottom (c) window positions of the MEA in the transient-response process under the voltage operation 0.4 → 1.4 VRHE are shown in Figure 5. The plots of the CN (Pt−Pt) determined by the in situ timeresolved (100 ms) QEXAFS analysis are scattered compared to the plots of the Pt white line intensity (Pt valence) in Figure 4. Each transient-response curve was separately shown in Figure 5, where the transient-response curves (a)-(1)−(a)-(4) are each transient-response curve at the top position (a), and the transient-response curves (c)-(1)−(c)-(4) are each transientresponse curve at the bottom position (c). These in situ timeresolved (100 ms) QEXAFS analysis data reveal that the temporal QEXAFS analysis quality is good enough to determine the rate constants (k1 and k2) under the voltage operation; for example, the rate constants (k1) at the top, middle, and bottom positions are 2.24 (± 0.66) and 0.32 (± 0.08) s−1, 2.18 (± 0.89) and 0.37 (± 0.01) s−1, and 2.64 (± 0.89) and 0.44 (± 0.11) s−1, respectively. The determined rate constants (k1 and k2) for the transient-response processes are listed in Table 1(a), Table 2(a), and Table 3(a). The fractions A1, A2, and A3 of the active, less active, and inactive Pt nanoparticles, respectively, are also shown in these tables. The temporal variations of CN (Pt−O) vs the reaction time at the top (A), middle (B), and bottom (C) window positions of the MEA in the transient-response process under the voltage operation 0.4 → 1.4 VRHE are shown in Figure 6. The plots of

are shown in Figures 4 (right) and S7 (each curve for the aging (no AGEX), 100 AGEX, 200 AGEX, and 300 AGEX is separately shown typically at the top position). In the voltage change from 1.4 to 0.4 VRHE, the white line peak intensity (Pt valence) in the Pt/C catalyst decreased with reaction time to the initial metallic intensity level at 0.4 VRHE in Figure 4 (right) (a−c), respectively, that is, positively charged Pt nanoparticles were discharged back to metallic Pt nanoparticles again. The rate constants (k′1 and k′2) for the Pt valence change (Pt discharging) processes after the aging and after 100−300 AGEX cycles were determined by the fitting analysis using eq 2, similar to the case of the voltage operation 0.4 → 1.4 VRHE. The rate constants (k′1 and k′2) for the Pt valence change at the top, middle, and bottom positions were determined to be 2.80 (± 0.30) and 1.50 (± 0.31) s−1, 2.49 (± 0.33) and 1.49 (± 0.42) s−1, and 2.99 (± 0.53) and 1.03 (± 0.46) s−1, respectively, as listed in Tables 1(b), 2(b), and 3(b), which are also nearly independent of the positions within the error range. The rate constants (k′1 and k′2) and fractions (A′1 and A′2) for the active and less active Pt species and the fraction A′3 of the inactive Pt species, estimated from the temporal profiles of Pt valence, CN (Pt−Pt), and CN (Pt−O), are also listed in Tables 1(b), 2(b), and 3(b) for the top, middle, and bottom positions, respectively. The rate constants (k′) in the transient process 1.4 → 0.4 VRHE were larger than the rate constants (k) in the transient process 0.4 → 1.4 VRHE. Particularly, the rate constants (k′2) for the less active Pt species (1.50, 1.49, and 1.03 s−1 for the top, middle, and bottom positions, respectively) were remarkably larger than the rate constants (k2) (0.28, 0.41, and 0.34 s−1 for the top, middle, and bottom positions, respectively), and as the result, the difference between k′1 and k′2 is smaller than that between k1 and k2 The results indicate G


Article


Figure 5. Transient response curves of the coordination number (CN) of Pt−Pt bonds for the MEA Pt/C cathode catalysts at the top (a), middle (b), and bottom (c) positions under the voltage operation 0.4 → 1.4 VRHE. (1: blue) after aging (no AGEX); (2: red) 100 AGEX cycles; (3: green) 200 AGEX cycles; (4: purple) 300 AGEX cycles. Under H2 (anode)−N2 (cathode); cell temp.: 353 K, relative humidity: ∼93%. Data acquisition: every 100 ms. (a)-(1)−(a)-(4) are each transient-response curve in (a). (c)-(1)−(c)-(4) are each transient-response curve in (c).

the CN (Pt−O) determined by the in situ time-resolved QEXAFS analysis are also scattered compared to the plots of the Pt white line intensity (Pt valence), but the transientresponse curves (A)-(1)−(A)-(4) at the top position and (C)(1)−(C)-(4) at the bottom position as typical examples show that the temporal QEXAFS analysis quality for the CN (Pt−O) is also good enough to determine the rate constants (k1 and k2) and the fractions (A1, A2, and A3), which are also listed in Table 1(a), Table 2(a), and Table 3(a). Typically, the rate constants (k2) at the top, middle, and bottom positions are 0.34 (± 0.04), 0.31 (± 0.06), and 0.44 (± 0.10) s−1, respectively. Here, the effect of Pt bands produced near the boundary in the AGEX degradation process on the white line peak intensity (Pt valence), CN (Pt−Pt), and CN (Pt−O) should be addressed because the existence of metallic Pt band decreases the Pt valence, increases the CN (Pt−Pt), and decreases the CN(Pt−O) compared to the real values for the Pt/C cathode. However, the Pt quantity of Pt bands relative to the whole Pt quantity was only 1.8% after 300 AGEX cycles. Thus, we can assume a negligible contribution of the Pt bands in the polymer electrolyte near the boundary of the cathode to the present estimation by in situ time-resolved QXAFS analysis.

The temporal variations of CN (Pt−Pt) and CN (Pt−O) under the voltage operation 1.4 → 0.4 VRHE for the MEA Pt/C samples after the aging (no AGEX) and 100−300 AGEX cycles are shown in Figure S9 at the top, middle, and bottom positions. Each transient curve for the aging, 100 AGEX, 200 AGEX, and 300 AGEX at the top position is also shown in Figures S10 and S11 for the CN (Pt−Pt) and CN(Pt−O), respectively. The CN(Pt−Pt) and the CN(Pt−O) increased and decreased, respectively, with the transient-response time due to the Pt−O bond dissociation and Pt−Pt bond reformation by the reduction of the PtO surface layer to the metallic Pt state reversely to the voltage operation 0.4 → 1.4 VRHE.37,52 As the temporal variations under the transient voltage operation 0.4 → 1.4 VRHE were well characterized by the double exponential functions (eq 2) to reveal the Pt species with different rate constants (k1 and k2) and hence different activities as discussed above, we also analyzed the temporal variations under the transient voltage operation 1.4 → 0.4 VRHE by eq 2. The determined rate constants (k′1 and k′2) and fractions (A′1, A′2, and A′3) are shown in Table 1(b), Table 2(b), and Table 3(b). The similarity and difference among the rate constants for the temporal profiles of the Pt valence, CN H


Article


Table 2. Rate Constants (k1 and k2) and Amplitudes (A1 and A2) for the Active and Less Active Pt/C Catalysts as Well as A3 Valuesa (a) 0.4 → 1.4 VRHE 0.4 → 1.4 V Pt valence

CN (Pt−Pt)

CN (Pt−O)


k1 or A1

k2 or A2

A3

k1 or A1

2.22 2.33 1.00 2.18 3.99 1.00 2.84 1.35 1.00

0.00 0.00 0.00

0.00 0.00 0.00

2.22 1.07 0.46 2.18 2.18 0.55 2.84 0.88 0.65

1.4 → 0.4 V

CN (Pt−Pt)

CN (Pt−O)


k2 or A2

after AGEX #200 A3

0.41 0.97 0.29 0.42 0.12 0.37 1.59 0.22 0.40 0.06 0.31 0.45 0.02 0.33 0.01 (b) 1.4 → 0.4 VRHE

after AGEX #300

k1 or A1

k2 or A2

A3

k1 or A1

k2 or A2

A3

2.22 0.83 0.36 2.18 2.04 0.51 2.84 0.78 0.58

0.41 1.09 0.47 0.37 1.82 0.46 0.31 0.52 0.39

0.41 0.18 0.13 0.03 0.05 0.04

2.22 0.60 0.26 2.18 1.65 0.41 2.84 0.46 0.34

0.41 1.13 0.48 0.37 1.68 0.42 0.31 0.75 0.56

0.60 0.26 0.66 0.17 0.14 0.10

after AGEX #100

after aging

middle window Pt valence

after AGEX #100

after aging

middle window

after AGEX #200

after AGEX #300

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

2.49 2.45 1.00 2.15 3.75 1.00 2.01 1.40 1.00

0.00 0.00 0.00

0.00 0.00 0.00

2.49 1.34 0.55 2.15 2.87 0.77 2.01 0.74 0.53

1.49 0.61 0.25 1.33 0.80 0.21 0.82 0.46 0.33

0.50 0.20 0.08 0.02 0.20 0.14

2.49 0.78 0.32 2.15 1.79 0.48 2.01 0.69 0.49

1.49 1.27 0.52 1.33 1.61 0.43 0.82 0.46 0.33

0.40 0.16 0.35 0.09 0.25 0.18

2.49 0.67 0.27 2.15 1.81 0.48 2.01 0.23 0.16

1.49 1.26 0.51 1.33 1.84 0.49 0.82 1.01 0.72

0.52 0.21 0.10 0.03 0.16 0.11

a A3 = 1 − (A1/A1−0 + A2/A1−0) at the middle position of the MEA sample under the voltage operations 0.4 → 1.4 VRHE (a) and 1.4 → 0.4 VRHE (b) for Pt valence, CN (Pt−Pt), and CN (Pt−O). H2 (anode)−N2 (cathode); cell temp.: 353 K, relative humidity: ∼93%. Data acquisition: every 100 ms.

Table 3. Rate Constants (k1 and k2) and Amplitudes (A1 and A2) for the Active and Less Active Pt/C Catalysts as Well as A3 Valuesa (a) 0.4 → 1.4 VRHE 0.4 → 1.4 V bottom window Pt valence

CN (Pt−Pt)

CN (Pt−O)


k1 or A1

k2 or A2

A3

k1 or A1

1.93 2.16 1.00 2.64 4.06 1.00 2.17 1.65 1.00

0.00 0.00 0.00

0.00 0.00 0.00

1.93 1.68 0.78 2.64 2.54 0.63 2.17 1.20 0.73

1.4 → 0.4 V

CN (Pt−Pt)

CN (Pt−O)


k2 or A2

after AGEX #200 A3

0.34 0.43 0.05 0.20 0.02 0.44 1.38 0.14 0.34 0.03 0.44 0.23 0.22 0.14 0.13 (b) 1.4 → 0.4 VRHE

after AGEX #300

k1 or A1

k2 or A2

A3

k1 or A1

k1 or A2

A3

1.93 0.74 0.34 2.64 1.72 0.42 2.17 0.98 0.59

0.34 0.97 0.45 0.44 1.59 0.39 0.44 0.42 0.25

0.45 0.21 0.75 0.18 0.25 0.15

1.93 0.43 0.20 2.64 1.14 0.28 2.17 0.67 0.41

0.34 1.12 0.52 0.44 1.82 0.45 0.44 0.54 0.33

0.61 0.28 1.10 0.27 0.44 0.27

after AGEX #100

after aging

bottom window Pt valence

after AGEX #100

after aging

after AGEX #200

after AGEX #300

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

k′1 or A′1

k′2 or A′2

A′3

2.99 2.46 1.00 2.38 4.46 1.00 2.25 1.79 1.00

0.00 0.00 0.00

0.00 0.00 0.00

2.99 1.68 0.68 2.38 3.31 0.74 2.25 0.85 0.47

1.03 0.46 0.19 0.83 0.79 0.18 0.77 0.83 0.46

0.32 0.13 0.36 0.08 0.11 0.06

2.99 0.84 0.34 2.38 1.73 0.39 2.25 0.80 0.45

1.03 1.10 0.45 0.83 1.68 0.38 0.77 0.38 0.21

0.52 0.21 1.05 0.24 0.61 0.34

2.99 0.53 0.22 2.38 1.05 0.24 2.25 0.29 0.16

1.03 1.09 0.44 0.83 1.99 0.45 0.77 0.87 0.49

0.84 0.34 1.42 0.32 0.63 0.35

A3 = 1 − (A1/A1−0 + A2/A1−0) at the bottom position of the MEA sample under the voltage operations 0.4 → 1.4 VRHE (a) and 1.4 → 0.4 VRHE (b) for Pt valence, CN (Pt−Pt), and CN (Pt−O). H2 (anode)−N2 (cathode); cell temp.: 353 K, relative humidity: ∼93%. Data acquisition: every 100 ms. a

I


Article


Figure 6. Transient response curves of the coordination number (CN) of Pt−O bonds for the MEA Pt/C cathode catalysts at the top (A), middle (B), and bottom (C) positions under the voltage operation 0.4 → 1.4 VRHE. (1: blue) after aging (no AGEX); (2: red) 100 AGEX cycles; (3: green) 200 AGEX cycles; (4: purple) 300 AGEX cycles. Under H2 (anode)−N2 (cathode); cell temp.: 353 K, relative humidity: ∼93%. Data acquisition: every 100 ms. (A)-(1)−(A)-(4) are each transient-response curve in (A). (C)-(1)−(C)-(4) are each transient-response curve in (C).

to estimate the ECSAs and MAs, respectively, while measuring the in situ time-resolved QXAFS spectra at the top, middle, and bottom positions (Figures S12−S15). At the top position of the MEA Pt/C catalyst layer, a half of the inactive Pt nanoparticles seem to contribute to the ECSA (hydrogen adsorption/ desorption) event though they are not active for the ORR catalysis (Figure S12). On the other hand, the change of the ECSA by the AGEX cycles coincided with the change in the sum of the fractions A1 and A2 or A′1 and A′2 at the middle position and particularly at the bottom position as shown in Figure S14. The electrochemical data are values in the average over the whole MEA cathode layer. So, the fractions A1, A2, and A3 or A′1, A′2, and A′3 were averaged at the top, middle, and bottom positions in the MEA Pt/C cathode layer, and those positionaveraged fractions were plotted against the AGEX cycles in Figure S16. The ECSAs were also averaged with those in Figure S12−14, and the ECSAaveraged values were also plotted in Figure S16. Figure S16 shows the relationship between the ECSAaveraged and the position-averaged fractions under the transient voltage changes 0.4 → 1.4 VRHE and 1.4 → 0.4 VRHE, respectively. It was found that the ECSAaveraged correlates with

(Pt−Pt), and CN(Pt−O) have been demonstrated in relation to the MEA performance and durability,45,52 and the issues are not discussed in this study. 3.3. Relationships of the QXAFS Analysis Data with the Electrochemical Surface Areas and Mass Activities in the AGEX Degradation Process. The in situ timeresolved QXAFS kinetics identified the three kinds of Pt nanoparticles with different activities during the AGEX treatments. The active Pt nanoparticles in the aging (no AGEX) Pt/C were transformed to the less active and inactive Pt nanoparticles by the AGEX cycles. To find the relationship between the fractions of the three different Pt nanoparticles in the MEA Pt/C and the electrochemical data, the changes in the electrochemical surface area (ECSA) were plotted against the changes in the fractions of the three kinds of Pt nanoparticles (A1−A3 and A′1−A′3 under the transient voltage changes 0.4 → 1.4 VRHE and 1.4 → 0.4 VRHE, respectively) by the AGEX cycles in Figures S12−S14 for the top, middle, and bottom positions of the MEA. Previously, we reported the change in the ECSA by the AGEX cycles,19 and here we show the typical change in the ORR activity (MA (mass activity): A mgPt−1) by the AGEX cycles in Figure S15. The CVs and Tafel plots were measured J


Article


Figure 7. Relationship between the ECSAaveraged (blue outlined circles) or the MAaveraged (red outlined circles) and the averaged fractions of the active (A1: green), less active (A2: orange), and inactive (A3: white) Pt nanoparticles. The averaged fractions denote the averaged values of the three position-averaged fractions determined by the WL intensity, CN (Pt−Pt), and CN (Pt−O) for the MEA Pt/C after the aging and 100−300 AGEX cycles under the transient voltage changes (1) 0.4 → 1.4 VRHE and (2) 1.4 → 0.4 VRHE. A1, A2, and A3 were determined by the in situ time-resolved QXAFS. H2 (anode)−N2 (cathode); cell temp.: 353 K, relative humidity: ∼93%.

the sum of the fractions of the active (A1, A′1) and less active (A2, A′2) Pt nanoparticles. Figure S17 shows the relationship between the MAaveraged determined electrochemically and the position-averaged fractions A1, A2, and A3 or A′1, A′2, and A′3 determined by the in situ time-resolved QXAFS. It was also found that the MAaveraged correlates with the sum of the fractions of the active (A1, A′1) and less active (A2, A′2) Pt nanoparticles. Finally, we compared the electrochemical data (ECSAaveraged and MAaveraged) for the MEA Pt/C cathode catalysts after the aging (no AGEX), 100 AGEX, 200 AGEX, and 300 AGEX with the averaged fractions of active, less active, and inactive Pt nanoparticles determined by the in situ time-resolved QXAFS. Figure 7 shows the relationship between the ECSAaveraged (blue outlined circle) or the MAaveraged (red outlined circle) and the averaged fractions of the active (A1, A′1), less active (A2, A′2), and inactive (A3, A′3) Pt nanoparticles. The averaged fractions denote the averaged values of the three position-averaged fractions determined by the white line peak intensity (Pt valence), CN (Pt−Pt), and CN (Pt−O) (Figures S16 and S17). The fractions of the three kinds of Pt nanoparticles determined from the three XAFS parameters should be a similar value, whereas the rate constants k1 (k1(Pt valence), k1(CN(Pt−Pt)), and k1(CN(Pt−O)) for the three elementary steps determined from the temporal variations of Pt valence, CN (Pt−Pt), and CN (Pt− O), respectively, under the transient voltage changes 0.4 → 1.4 VRHE may not naturally be similar to each other, and the rate constants k′1 (k′1(Pt valence), k′1(CN(Pt−Pt)), and k′1(CN(Pt−O))) for the three elementary steps determined from the temporal variations of Pt valence, CN (Pt−Pt), and CN (Pt−O), respectively, under the transient voltage change 1.4 → 0.4 VRHE may not naturally be similar to each other, and so on (Tables 1−3).37,52 It is to be noted that the decreases in the ECSA and MA did not trace the decrease in the active Pt nanoparticles but coincided with the decrease in the sum of the two fractions of the active and less active Pt nanoparticles (Figure 7). The less active Pt particles contribute to the overall electrochemical activity (cell performance), but the electrochemical parameter can only be measured for the whole assemble of Pt particles, where the maximum electrochemical performance of the MEA reflecting the high reactivity of the active Pt particles may not be achieved. The difference of the reactivity between the active and less active Pt particles might be insensitive at the electrochemical MEA measurements (Tafel plots for MA in

Figure S15). In this case, the MEA performance would be possible to maximize to the higher level of the MEA Pt/C performance than the present performance level due to the high reactivity of the active Pt particles by a device of the way to fabricate MEA Pt/C cathodes. In Figure 7, there is a little inconsistency between the MA and the sum of the two fractions, but the difference may be due to the error ranges of the MA and fractions. This sort of detailed discussion would not be valid for the present analysis. The difference between the electrochemical maximum performance and the kinetically highest reactivity of the cathode catalyst in MEA is still an important issue to be tackled for development of nextgeneration PEFCs. Anyhow, the transient-response kinetic analysis for the Pt nanoparticles in the MEA Pt/C revealed the existence and fractions of three different kinds of Pt particles during the AGEX cycles, determining the rate constants and fractions for six elementary steps for the cathode catalysis. The active Pt nanoparticles in the MEA Pt/C cathode were more remarkably transformed to the less active and particularly inactive Pt nanoparticles at the top position than at the middle and bottom positions as shown in Figures S12−S14. It suggests that the formation of the inactive Pt nanoparticles is related to the carbon corrosion more in the top area derived from the high cathode voltages by the RCD mechanism for the AGEX. Figure S18 shows a STEM image and a Pt EDS map at the top position in an MEA Pt/C cathode after the aging treatment (no AGEX), where there are few holes, and the Pt amount distribution is not biased so much in the whole region. After 300 AGEX cycles, considerable carbon corrosion with deteriorated places and holes is observed to be accompanied by detached Pt nanoparticles from the carbon support, and the Pt amount distributions are remarkably biased as shown in the STEM image−Pt EDS map of Figure S19. The inactive Pt nanoparticles may be assigned to Pt nanoparticles detached from the carbon support due to carbon vanishing. Recently, we identified the degraded places by the same-view nano-XAFS/ STEM-EDS measurement technique. Pt nanoparticles were detached from carbon and also oxidized to Pt2+ ions with tetrahedral coordination.53−55 If the Pt2+ species is major in the degraded Pt species, the Pt valence observed at 0.4 VRHE should increase with increasing AGEX cycles, but it remained almost unchanged as shown in Figures 4 (left) and S4. Thus, the majority of the inactive Pt nanoparticles is suggested to be metallic detached Pt nanoparticles under the present AGEX K


Article


due to carbon corrosion/degradation derived from the high cathode voltages by the RCD mechanism for the AGEX. The degradation probability and the consecutive with parallel reaction mechanism were determined by the transient-response kinetics. The electrochemical surface area (ECSA) and mass activity (MA) of the Pt/C cathode catalyst also decreased with increasing AGEX cycles. It was found that the change in the sum of the fractions of the active and less active Pt nanoparticles correlates with the change in the ECSA and MA during the AGEX treatments. The in situ time-resolved QXAFS study provides a new insight into understanding of the degradation of MEA Pt/C cathode catalysts.

conditions. There are also lots of Pt nanoparticles on thinner amorphous carbon layers with worse electronic Pt−C interaction (Figure S19). These Pt nanoparticles are regarded as the less active sites. It has been demonstrated that the ORR performance of the MEA Pt/C cathode catalyst linearly correlates with the rate constants k′(CN(Pt−O)) and k′(Pt valence). The averaged k′1(CN(Pt−O)) and k′1(Pt valence) for the active Pt nanoparticles are 2.22 and 2.76 s−1, respectively, while the averaged k′2(CN(Pt−O)) and k′2(Pt valence) for the less active Pt nanoparticles are 0.79 and 1.34 s−1, respectively. Thus, the ORR performance of the less active Pt nanoparticles is estimated to be about 41% of the active Pt nanoparticles. In consequence, the active Pt nanoparticles on carbon in the MEA Pt/C (denoted as Ptact) degrade to the less active Pt nanoparticles (Ptlac) and the inactive Pt nanoparticles (Ptina) in a parallel reaction mechanism,

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.7b07264. MEA Pt/C catalyst, QXAFS measurement setup, TEM images, electrochemical data, transient-response profiles of the white line intensity, CN(Pt−Pt) and CN(Pt−O) under cyclic voltage operations, variation of the fractions of active, less active, and inactive Pt nanoparticles during AGEX degradation (PDF)

a consecutive reaction mechanism, or their mixture mechanism (consecutive with parallel reaction mechanism).

■

The inactive Pt particles cannot be directly characterized by the time-resolved XAFS technique, which prevents the determination of the transformation pathways. To determine the transformation pathways we used the change in the fractions (amounts) of the active, less active, and inactive Pt particles by the AGEX cycles, determined by using eq 2. From the change in the fractions (A1, A2, and A3) of Ptact, Ptlac, and Ptina by the AGEX cycles in Figure 7, we can estimate the ratio of the degradation probability (pdg) and the degradation mechanism. It was concluded that the degradation proceeds by eq 3, and the ratio of the degradation probabilities pdg1, pdg2, and pdg3 for each step was estimated to be approximately 1.0:0.91:0.39, and hence, the transformation of Ptact to Ptina via Ptlac is the major degradation pathway, but the direct degradation of Ptact to Ptina also proceeds under the AGEX (start-up/shut-down) conditions.

AUTHOR INFORMATION

Corresponding Author

*Tel: +81-42-443-5921; E-mail: [email protected] (Y.I.) ORCID

Yasuhiro Iwasawa: 0000-0002-5222-5418 Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the New Energy and Industrial Technology Development Organization (NEDO) of the Ministry of Economy, Trade, and Industry (METI), Japan. XAFS measurements were conducted at BL36XU station in SPring-8 (No. 2013A7802, 2013B7801, 2013B7802, 2013B7803, 2014A7802, 2014A7805, 2014B7801, 2014B7802, 2015A7802, 2015B7801, 2015B7802, 2016A7801, 2016A7802, 2016B7801, 2016B7802, 2017A7801, 2017A7802, 2017A7808).

4. CONCLUSIONS We have investigated the transient-response kinetics of Pt nanoparticles in a Pt/C cathode catalyst of PEFC MEA under cyclic voltage operations (0.4 → 1.4 → 0.4 VRHE) by in situ time-resolved QXAFS (100 ms/spectrum) for thorough understanding of the Pt/C degradation process by AGEX (start-up/shut-down) toward PEFC improvement. The transient-response analysis of the kinetics of the transformations (six elementary steps) of the white line peak intensity (Pt valence), CN (Pt−Pt), and CN (Pt−O) in the AGEX degradation process, while measuring the electrochemical data such as ECSA and MA, identified the three kinds of Pt nanoparticles with different activities during the AGEX treatments. The active Pt nanoparticles degraded to less active and inactive Pt nanoparticles by the AGEX cycles. The fractions of the less active and inactive Pt nanoparticles were more at the top position of MEA than at the middle and bottom positions

■

REFERENCES

(1) Mistry, H.; Varela, A. S.; Kühl, S.; Strasser, P.; Cuenya, B. R. Nanostructured Electrocatalysts with Tunable Activity and Selectivity. Nature Rev. Materials [Online] 2016, 110.1038/natrevmats.2016.9. (2) Stamenkovic, V. R.; Markovic, N. M. Nanosegregated Cathode Catalysts with Ultra-Low Platinum Loading. Presented at 2015 Annual Merit Review and Peer Evaluation Meeting, Arlington, VA, 2015. (3) Zhao, X.; Takao, S.; Higashi, K.; Kaneko, T.; Samjeskè, G.; Sekizawa, O.; Sakata, T.; Yoshida, Y.; Uruga, T.; Iwasawa, Y. Simultaneous Improvements of Performance and Durability of An Octahedral PtNix/C Electrocatalyst for Next−Generation Fuel Cells by Continuous, Compressive and Concave Pt Skin Layers. ACS Catal. 2017, 7, 4642−4654. (4) Debe, K. Electrocatalyst Approaches and Challenges for Automotive Fuel Cells. Nature 2012, 486, 43−51. L


Article

The Journal of Physical Chemistry C (5) Zhang, L.; Roling, L. T.; Wang, X.; Vara, M. V.; Chi, M.; Liu, J.; Choi, S.; Park, J.; Herron, J. A.; Xie, Z.; et al. Platinum-Based Nanocages with Subnanometer-Thick Walls and Well-Defined, Controllable Facets. Science 2015, 349, 412−416. (6) Escudero-Escribano, M.; Malacrida, P.; Hansen, M. H.; VejHansen, U. G.; Velázquez-Palenzuela, A.; Tripkovic, V.; Schiøtz, J.; Rossmeisl, J.; Stephens, I. E. L.; Chorkendorff, I. Tuning the Activity of Pt Alloy Electrocatalysts by Means of the Lanthanide Contraction. Science 2016, 352, 73−76. (7) Oezaslan, M.; Hasché, F.; Strasser, P. Pt-Based Core−Shell Catalyst Architectures for Oxygen Fuel Cell Electrodes. J. Phys. Chem. Lett. 2013, 4, 3273−3291. (8) Nagasawa, K.; Takao, S.; Nagamatsu, S.; Samjeské, G.; Sekizawa, O.; Kaneko, T.; Higashi, K.; Yamamoto, T.; Uruga, T.; Iwasawa, Y. Surface-Regulated Nano-SnO2/Pt3Co/C Cathode Catalysts for Polymer Electrolyte Fuel Cells Fabricated by a Selective Electrochemical Sn Deposition Method. J. Am. Chem. Soc. 2015, 137, 12856− 12864. (9) Tada, M.; Uruga, T.; Iwasawa, Y. Key Factors Affecting the Performance and Durability of Cathode Electrocatalysts in Polymer Electrolyte Fuel Cells Characterized by In Situ Real Time and Spatially Resolved XAFS Techniques. Catal. Lett. 2015, 145, 58−70. (10) Chen, C.; Kang, Y.; Huo, Z.; Zhu, Z.; Huang, W.; Huolin, W.; Xin, L.; Snyder, J. D.; Li, D.; Herron, J. A.; et al. Highly Crystalline Multimetallic Nanoframes with Three-Dimensional Electrocatalytic Surfaces. Science 2014, 343, 1339−1343. (11) Rabis, A.; Rodriguez, P.; Schmidt, T. J. Electrocatalysis for Polymer Electrolyte Fuel Cells: Recent Achievements and Future Challenges. ACS Catal. 2012, 2, 864−890. (12) Wu, G.; More, K. L.; Johnston, C. M.; Zelenay, P. HighPerformance Electrocatalysts for Oxygen Reduction Derived from Polyaniline, Iron, and Cobalt. Science 2011, 332, 443−447. (13) Borup, R.; Meyers, J.; Pivovar, B.; Kim, Y. S.; Mukundan, R.; Garland, N.; Myers, D.; Wilson, M.; Garzon, F.; Wood, D.; et al. Scientific Aspects of Polymer Electrolyte Fuel Cell Durability and Degradation. Chem. Rev. 2007, 107, 3904−3951. (14) Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G.; Ross, P. N.; Lucas, C. A.; Markovic, N. M. Improved Oxygen Reduction Activity on Pt3Ni(111) via Increased Surface Site Availability. Science 2007, 315, 493−497. (15) Strasser, P.; Koh, S.; Anniyev, T.; Greeley, J.; More, K.; Yu, C.; Liu, Z.; Kaya, S.; Nordlund, D.; Ogasawara, H.; et al. Lattice-Strain Control of the Activity in Dealloyed Core−Shell Fuel Cell Catalysts. Nat. Chem. 2010, 2, 454−460. (16) Wu, J.; Yang, H. Platinum-Based Oxygen Reduction Electrocatalysts. Acc. Chem. Res. 2013, 46, 1848−1857. (17) Weber, A. Z.; Newman, J. Modeling Transport in PolymerElectrolyte Fuel Cells. Chem. Rev. 2004, 104, 4679−4726. (18) Stamenkovic, V.; Mun, B. S.; Mayrhofer, K. J. J.; Ross, P. N.; Markovic, N. M.; Rossmeisl, J.; Greeley, J.; Norskov, J. K. Changing the Activity of Electrocatalysts for Oxygen Reduction by Tuning the Surface Electronic Structure. Angew. Chem., Int. Ed. 2006, 45, 2897− 2901. (19) Samjeské, G.; Higashi, K.; Takao, S.; Nagamatsu, S.; Nagasawa, K.; Sekizawa, O.; Kaneko, T.; Uruga, T.; Iwasawa, Y. Effects of Anode or Cathode Gas Exchange Cycles on Deterioration of Pt/C Cathode Catalysts in PEFCs Characterized by In-Situ Time-resolved XAFS, TEM, and Electrochemical Techniques. ChemElectroChem 2015, 2, 1595−1606. (20) Brightman, E.; Hinds, G. J. Power Sources 2014, 267, 160−170. (21) Reiser, C. A.; Bregoli, L.; Patterson, T. W.; Yi, J. S.; Yang, J. D. L.; Perry, M. L.; Jarvi, T. D. A Reverse-Current Decay Mechanism for Fuel Cells. Electrochem. Solid-State Lett. 2005, 8, A273−A276. (22) Yu, Y.; Li, H.; Wang, H.; Yuan, X.-Z.; Wang, G.; Pan, M. A Review on Performance Degradation of Proton Exchange Membrane Fuel Cells During Startup and Shutdown Processes: Causes, Consequences, and Mitigation Strategies. J. Power Sources 2012, 205, 10−23.

(23) Ishigami, Y.; Takada, K.; Yano, H.; Inukai, J.; Uchida, M.; Nagumo, Y.; Hyakutake, T.; Nishide, H.; Watanabe, M. Corrosion of Carbon Supports at cathode during hydrogen/air replacement at anode studied by visualization of oxygen partial pressures in a PEFC Start-up/shut-down simulation. J. Power Sources 2011, 196, 3003− 3008. (24) Iwasawa, Y.; Asakura, K.; Tada, M., Eds. XAFS Techniques for Catalysts, Nanomaterials, and Surfaces; Springer: New York, 2016. (25) Iwasawa, Y., Ed. X-ray Absorption Fine Structure for Catalysts and Surfaces; World Scientific Publishing: Singapore, 1996. (26) Tada, M.; Murata, S.; Asakoka, T.; Hiroshima, K.; Okumura, K.; Tanida, H.; Uruga, T.; Nakanishi, H.; Matsumoto, S.; Inada, Y.; et al. In Situ Time-Resolved Dynamic Surface Events on the Pt/C Cathode in a Fuel Cell under Operando Conditions. Angew. Chem., Int. Ed. 2007, 46, 4310−4315. (27) Sekizawa, O.; Uruga, T.; Higashi, K.; Kaneko, T.; Yoshida, Y.; Sakata, T.; Iwasawa, Y. Simultaneous Operando Time-Resolved XAFSXRD Measurements of a Pt/C Cathode Catalyst in Polymer Electrolyte Fuel Cell under Transient Potential Operations. ACS Sustainable Chem. Eng. 2017, 5, 3631−3636. (28) Iwasawa, Y.; Asakura, K.; Ishii, H.; Kuroda, H. Dynamic Behaviour of Active Sites of a SiO2-Attached Mo(VI)-Dimer Catalyst during Ethanol Oxidation Observed by Means of EXAFS. Z. Phys. Chem. 1985, 144, 105−115. (29) Iwasawa, Y. Chemical Design Surfaces for Active Solid Catalysts. Adv. Catal. 1987, 35, 187−264. (30) Mukerjee, S.; Srinivasan, S.; Soriaga, M. P.; McBreen, J. Role of Structural and Electronic Properties of Pt and Pt Alloys on Electrocatalysis of Oxygen Reduction: An in Situ XANES and EXAFS Investigation. J. Electrochem. Soc. 1995, 142, 1409−1422. (31) Weckhuysen, B. M. Determining the Active Site in a Catalytic Process: Operando Spectroscopy is More than a Buzzword. Phys. Chem. Chem. Phys. 2003, 5, 4351−4360. (32) Singh, J.; Lamberti, C.; van Bokhoven, J. A. Advanced X-ray Absorption and Emission Spectroscopy: In Situ Catalytic Studies. Chem. Soc. Rev. 2010, 39, 4754−4766. (33) Friebel, D.; Viswanathan, V.; Miller, D. J.; Anniyev, T.; Ogasawara, H.; Larsen, A. H.; O’Grady, C. P.; Nørskov, J. K.; Nilsson, A. Balance of Nanostructure and Bimetallic Interactions in Pt Model Fuel Cell Catalysts: In Situ XAS and DFT Study. J. Am. Chem. Soc. 2012, 134, 9664−9671. (34) Adzic, R. R.; Wang, J. X.; Ocko, B. M.; McBreen, J. EXAFS, XANES, SXS. Handbook of Fuel Cells [Online]; John Wiley & Sons, 2010. (35) Nagamatsu, S.; Arai, T.; Yamamoto, M.; Ohkura, T.; Oyanagi, H.; Ishizaka, T.; Kawanami, H.; Uruga, T.; Tada, M.; Iwasawa, Y. Potential-Dependent Restructuring and Hysteresis in the Structural and Electronic Transformations of Pt/C, Au(Core)-Pt(Shell)/C, and Pd(Core)-Pt(Shell)/C Cathode Catalysts in Polymer Electrolyte Fuel Cells Characterized by in Situ X-ray Absorption Fine Structure. J. Phys. Chem. C 2013, 117, 13094−13107. (36) Nagasawa, K.; Takao, S.; Higashi, K.; Nagamatsu, S.; Samjeské, G.; Imaizumi, Y.; Sekizawa, O.; Yamamoto, T.; Uruga, T.; Iwasawa, Y. Performance and Durability of Pt/C Cathode Catalysts with Different Kinds of Carbons for Polymer Electrolyte Fuel Cells Characterized by Electrochemical and in Situ XAFS Techniques. Phys. Chem. Chem. Phys. 2014, 16, 10075−10087. (37) Nagamatsu, S.; Takao, S.; Samjeské, G.; Nagasawa, K.; Sekizawa, O.; Kaneko, T.; Higashi, K.; Uruga, T.; Gayen, S.; Velaga, S.; et al. Structural and Electronic Transformations of Pt/C, Pd@Pt(1 ML)/C and Pd@Pt(2 ML)/C Cathode Catalysts in Polymer Electrolyte Fuel Cells during Potential-Step Operating Processes Characterized by inSitu Time-resolved XAFS. Surf. Sci. 2016, 648, 100−113. (38) Iwasawa, Y. In Situ Characterization of Supported Metal Catalysts and Model Surfaces by Time-Resolved and Three-Dimensional XAFS Techniques. J. Catal. 2003, 216, 165−177. (39) Suzuki, A.; Inada, Y.; Yamaguchi, A.; Chihara, T.; Yuasa, M.; Nomura, M.; Iwasawa, Y. Time Scale and Elementary Steps of COM


Article

The Journal of Physical Chemistry C Induced Disintegration of Surface Rhodium Clusters. Angew. Chem., Int. Ed. 2003, 42, 4795−4799. (40) Yamamoto, T.; Suzuki, A.; Nagai, Y.; Tanabe, T.; Dong, F.; Inada, Y.; Nomura, M.; Tada, M.; Iwasawa, Y. Origin and Dynamics of Oxygen Storage/Release in a Pt/Ordered CeO2−ZrO2 Catalyst Studied by Time-Resolved XAFS Analysis. Angew. Chem., Int. Ed. 2007, 46, 9253−9256. (41) Tada, M.; Uemura, Y.; Bal, R.; Inada, Y.; Nomura, M.; Iwasawa, Y. In Situ Time-Resolved DXAFS for the Determination of Kinetics of Structural Changes of H-ZSM-5-Supported Active Re-Cluster Catalyst in the Direct Phenol Synthesis from Benzene and O2. Phys. Chem. Chem. Phys. 2010, 12, 5701−5706. (42) Uemura, Y.; Inada, Y.; Bando, K. K.; Sasaki, T.; Kamiuchi, N.; Eguchi, K.; Yagishita, A.; Nomura, M.; Tada, M.; Iwasawa, Y. Core− Shell Phase Separation and Structural Transformation of Pt3Sn Alloy Nanoparticles Supported on γ-Al2O3 in the Reduction and Oxidation Processes Characterized by In Situ Time-Resolved XAFS. J. Phys. Chem. C 2011, 115, 5823−5833. (43) Uemura, Y.; Inada, Y.; Bando, K. K.; Sasaki, T.; Kamiuchi, N.; Eguchi, K.; Yagishita, A.; Nomura, M.; Tada, M.; Iwasawa, Y. In Situ Time-Resolved XAFS Study on the Structural Transformation and Phase Separation of Pt3Sn and PtSn Alloy Nanoparticles on Carbon in the Oxidation Process. Phys. Chem. Chem. Phys. 2011, 13, 15833− 15844. (44) Ishiguro, N.; Saida, T.; Uruga, T.; Nagamatsu, S.; Sekizawa, O.; Nitta, K.; Yamamoto, T.; Ohkoshi, S.; Yokoyama, T.; Tada, M.; et al. Operando Time-Resolved X-ray Absorption Fine Structure Study for Surface Events on a Pt3Co/C Cathode Catalyst in a Polymer Electrolyte Fuel Cell during Voltage-Operating Processes. ACS Catal. 2012, 2, 1319−1330. (45) Ishiguro, N.; Kityakarn, S.; Sekizawa, O.; Uruga, T.; Sasabe, T.; Nagasawa, K.; Yokoyama, T.; Tada, M. Rate Enhancements in Structural Transformations of Pt−Co and Pt−Ni Bimetallic Cathode Catalysts in Polymer Electrolyte Fuel Cells Studied by in Situ TimeResolved X-ray Absorption Fine Structure. J. Phys. Chem. C 2014, 118, 15874−15883. (46) Ishiguro, N.; Saida, T.; Uruga, T.; Sekizawa, O.; Nagasawa, K.; Nitta, K.; Yamamoto, T.; Ohkoshi, S.; Yokoyama, T.; Tada, M. Structural Kinetics of a Pt/C Cathode Catalyst with Practical Catalyst Loading in an MEA for PEFC Operating Conditions Studied by in Situ Time-Resolved XAFS. Phys. Chem. Chem. Phys. 2013, 15, 18827− 18834. (47) Li, Y.; Frenkel, A. I. Metal Nanocatalysts. In XAFS Techniques for Catalysts, Nanomaterial, and Surfaces; Iwasawa, Y.; Asakura, K.; Tada, M., Eds.; Springer: New York, 2016. (48) Jentys, A. Estimation of Mean Size and Shape of Small Metal Particles by EXAFS. Phys. Chem. Chem. Phys. 1999, 1, 4059−4063. (49) Friebel, D.; Miller, D. J.; O’Grady, C. P.; Anniyev, T.; Bargar, J.; Bergmann, U.; Ogasawara, H.; Wikfeldt, K. T.; Pettersson, L. G. M.; Nilsson, A. In Situ X-ray Probing Reveals Fingerprints of Surface Platinum Oxide. Phys. Chem. Chem. Phys. 2011, 13, 262−266. (50) Frahm, R. New method for time dependent x-ray absorption studies. Rev. Sci. Instrum. 1989, 60, 2515−2518. (51) Ganzler, A. M.; Casapu, M.; Boubnov, A.; Muller, O.; Conrad, S.; Lichtenberg, H.; Frahm, R.; Grunwaldt, J.-D. Operando spatially and time-resolved X-ray absorption spectroscopy and infrared thermography during oscillatory CO oxidation. J. Catal. 2015, 328, 216−224. (52) Kaneko, T.; Samjeské, G.; Nagamatsu, S.; Higashi, K.; Sekizawa, O.; Takao, S.; Yamamoto, T.; Zhao, X.; Sakata, T.; Uruga, T.; Iwasawa, Y. Key Structural Kinetics for Carbon Effects on the Performance and Durability of Pt/Carbon Cathode Catalysts in Polymer Electrolyte Fuel Cells Characterized by In Situ Time-Resolved XAFS. J. Phys. Chem. C 2016, 120, 24250−24264. (53) Takao, S.; Sekizawa, O.; Nagamatsu, S.; Kaneko, T.; Yamamoto, T.; Samjeské, G.; Higashi, K.; Nagasawa, K.; Tsuji, T.; Suzuki, M.; Kawamura, N.; Mizumaki, M.; Uruga, T.; Iwasawa, Y. Mapping Platinum Species in Polymer Electrolyte Fuel Cells by Spatially

Resolved XAFS Techniques. Angew. Chem., Int. Ed. 2014, 53, 14110− 14114. (54) Takao, S.; Sekizawa, O.; Samjeské, G.; Nagamatsu, S.; Kaneko, T.; Yamamoto, T.; Higashi, K.; Nagasawa, K.; Uruga, T.; Iwasawa, Y. Same-View Nano-XAFS/STEM-EDS Imagings of Pt Chemical Species in Pt/C Cathode Catalyst Layers of a Polymer Electrolyte Fuel Cell. J. Phys. Chem. Lett. 2015, 6, 2121−2126. (55) Takao, S.; Sekizawa, O.; Samjeské, G.; Nagamatsu, S.; Kaneko, T.; Higashi, K.; Yamamoto, T.; Nagasawa, K.; Zhao, X.; Uruga, T.; Iwasawa, Y. Spatially Non-uniform Degradation of Pt/C Cathode Catalysts in Polymer Electrolyte Fuel Cells Imaged by Combination of Nano XAFS and STEM-EDS Techniques. Top. Catal. 2016, 59, 1722− 1731. (56) Newville, M.; Ravel, B.; Haskel, D.; Rehr, J. J.; Stern, E. A.; Yacoby, Y. Analysis of Multiple-Scattering XAFS Data Using Theoretical Standards. Phys. B 1995, 208−209, 154−156. (57) Ravel, B.; Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: Data Analysis for X-ray Absorption Spectroscopy Using IFEFFIT. J. Synchrotron Radiat. 2005, 12, 537−541. (58) Rehr, J. J.; Albers, R. C. Theoretical Approaches to X-ray Absorption Fine Structure. Rev. Mod. Phys. 2000, 72, 621−654. (59) Schulenburg, H.; Schwanitz, B.; Linse, N.; Scherer, G. G.; Wokaun, A.; Krbanjevic, J.; Grothausmann, R.; Manke, I. 3D Imaging of Catalyst Support Corrosion in Polymer Electrolyte Fuel Cells. J. Phys. Chem. C 2011, 115, 14236−14243. (60) Yu, Y.; Tu, Z.; Zhang, H.; Zhan, Z.; Pan, M. Comparison of Degradation Behaviors for Open-Ended and Closed Proton Exchange Membrane Fuel Cells During Startup and Shutdown Cycles. J. Power Sources 2011, 196, 5077−5083.

N


The Relationship between the Active Pt Fraction in ... - ACS Publications

Recommend Documents