Durability of Unsupported Pt-Ni Aerogels in PEFC Cathodes

Thecommercialsuccessofpolymerelectrolytefuelcells(PEFCs)dependsonthedevelopmentofPt-basedoxygenreductionreaction(ORR)catalystswithgreateractivityandstabilitytoreducetheamountofexpensivenoblemetalperdevice.Toadvancetowardthisgoal,wehavetestedanovelclassofunsupportedbimetallicalloycatalysts(aerogels)asthecathodematerialinPEFCsundertwoacceleratedstresstestconditionsandcomparedittoastate-of-the-artcarbon-supportedbenchmark(Pt/C).TheinvestigatedPt 3 Ni aerogel shows little degradation under high potential conditions ( > 1.0 V) which can occur during fuel starvation and start-up/shut- down of the cell. If tested under the same conditions, the Pt/C benchmark displays signiﬁcant losses of electrochemical surface area and ORR activity due to carbon support corrosion as observed in cross section and transmission electron microscopy analysis. When testing the durability upon extended load cycling (0.6–1.0 V), Pt 3 Ni aerogel demonstrates less stability than Pt/C which is related to the severe Ni leaching from the alloy under such conditions. These ﬁndings highlight the advantages of using unsupported ORR catalysts in PEFCs and point to the reduction

Polymer electrolyte fuel cells (PEFCs) currently rely on large amounts of carbon-supported platinum (Pt/C) catalysts (≈0.4 mg Pt /cm 2 electrode ) to reduce the voltage losses due to the sluggish kinetics of the cathodic oxygen reduction reaction (ORR). 1 Recent advancements in minimizing the Pt loading and associated costs were achieved by alloying platinum with other metals like Ni, Cu and Co which increases the Pt mass-specific ORR activity. 2 On the other hand, these catalysts suffer from significant corrosion of the carbon support and Pt nanoparticles during PEFC operation which compromises their long-term efficiency and reliability. 3 To specifically mitigate the issue of support stability, researchers have developed alternative corrosionresistant supports (e.g. conductive metal oxides 4-7 ), extended metal surfaces (e.g. 3 M nanostructured thin film catalysts 8 ) and unsupported materials (e.g. Pt-coated Ni, Co or Cu nanowires [9][10][11]. Pursuing this last strategy, unsupported bimetallic Pt-Ni electrocatalysts with high specific surface area (≈30 m 2 /g Pt ) and nanochain network structure, referred to as aerogels, were synthesized in a previous work. 12 These materials reach the U.S. Department of Energy target (DOE; i.e. 440 A/g Pt at 0.9 V vs. the reversible hydrogen electrode (V RHE )) for automotive PEFC application when tested as thin films by the rotating disk electrode (RDE) technique, 13 which is the standard tool employed by the majority of researchers in this field for initial assessment of catalyst activities. Considering that performance figures derived from such RDE experiments, often do not translate fully to the technical system, it is fundamental to also assess activity and durability in PE-FCs to evaluate the real application potential of new catalysts. 14,15 At present, few groups have reported promising durability in PE-FCs for spray-coated membrane electrode assemblies (MEAs) prepared from unsupported catalysts. Studies by Tamaki et al. 16 (on hollow Pt-Fe nanocapsules) and Lee et al. 17 (on FePt nanotubes) showed outstanding retention of electrochemical surface area (ECSA) and performance in H 2 /O 2 polarization (I/E) curves for 10000 potential cycles between 1.0 and 1.5 V RHE , and upon constant exposure to a high potential of 1.4 V RHE for three hours, respectively. However, none of those studies focused on the PEFC-performance of these materials under application-relevant conditions, i.e., using H 2 and air at the anode and cathode inlet feeds, respectively. With this motivation, in a recent report we have demonstrated how the increase of porosity in the catalyst layer (CL) of Pt 3 Ni aerogel MEAs caused by the addition of a removable filler material greatly improves the performance in H 2 /air I/E curves. 18 Following this recent study, in this work the durability of Pt 3 Ni aerogels in the PEFC is investigated for two different accelerated stress tests (ASTs) proposed by the DOE and compared to a commercial Pt/C benchmark. The first AST exposes the catalyst to the potential regime of 1.0 to 1.5 V RHE , which can occur in an operating PEFC during fuel starvation and start-up/shut-down of the cell and triggers C-support corrosion, 5,[19][20][21][22] and that will be referred to as 'start-stop degradation' in the following. 23 The second AST which has not been investigated in the studies cited above simulates the variation in power output present during automotive application that results in potential fluctuations between ≈0.6 and ≈1.0 V RHE causing Pt dissolution (and re-deposition); 21,24-26 it will be denoted 'load-cycle degradation'. As we will demonstrate in this article, Pt 3 Ni electrodes show superior durability for start-stop degradation compared to Pt/C due to the absence of a corrodible carbon-support. Additionally, severe Ni-leaching from the Pt 3 Ni aerogel in load-cycle degradation experiments is identified as the major cause for their activity loss and inferior stability when compared to Pt/C under these conditions.

Experimental
Pt 3 Ni aerogel was synthesized according to the procedure described in Reference 12. In brief, 0.585 ml of a 0.205 M H 2 PtCl 6 solution (8 wt% in H 2 O, Sigma Aldrich) and 4 ml of a freshly prepared 10 mM NiCl 2 solution (NiCl 2 * 6H 2 O, 99%, Sigma Aldrich) were dissolved in 790 ml of ultrapure water (18.2 M cm, Millipore) and stirred until the mixing was complete. Subsequently, 7.0 ml of freshly prepared 0.1 M NaBH 4 solution (granular, 99.99%, Sigma Aldrich) were added while stirring vigorously. A brown solution was obtained that was kept stirring for another 30 min. Afterwards, the reaction solution was distributed among several 100 ml vials. After about four days, black Pt 3 Ni hydrogel was formed at the bottom of the containers. The hydrogel was washed with water and the solvent was exchanged with acetone afterwards. The resulting anhydrous gels were subjected to critical point drying in CO 2 (Critical Point Dryer 13200J-AB, SPI Supplies).
Catalyst inks for Pt 3 Ni electrodes were prepared as described in Reference 18 by mixing 5 mg of catalyst, 0.7 mg of K 2 CO 3 (99.995% trace metals basis, Sigma Aldrich), 18 mg of Na + -exchanged Nafion solution (prepared from a 1:2 mixture of 0.1 M NaOH and Nafion solution, 27 and equal to a Nafion-to-catalyst-ratio of 0.12) and 1.0 ml of an 8 wt% aqueous isopropanol solution (ultrapure water, 18.2 M cm, Elga Purelab Ultra and isopropanol, 99.9%, Chromasolv Plus for HPLC, Sigma Aldrich). After ultrasonication (USC100T, 45 kHz, VWR) for 30 minutes and spray coating (using a frame to confine the coating to the active area of 1 cm 2 ), the resulting catalyst coated membranes (CCMs) were immersed into 1 M H 2 SO 4 solution (96%, Suprapur, Merck) overnight (≈16 hours), followed by rinsing with ultrapure water and drying under ambient conditions. The acid washing step was introduced to remove the filler material K 2 CO 3 and thus to create a CL with increased porosity (for details see Reference 18). For Pt/C (Pt/C graphitized ) cathodes, 50 mg of catalyst was mixed with 500 (650) mg of 5 wt% Nafion solution (equal to a Nafion-to-carbon-ratio of 1.0) and 4.5 ml of a 20 wt% aqueous isopropanol solution, followed by the steps described above without the acid washing. In a last step, the CCMs were hotpressed at 120 • C and 1 bar/cm 2 geom for 5 minutes to a gas diffusion layer (GDL 25 BC, Sigracet) and a commercial gas diffusion electrode (see above) on the cathode and anode side, respectively.
The MEAs were placed in a differential fuel cell that allows studying the MEA under homogeneous, well-defined conditions in the absence of along-the-channel effects such as changing temperature, relative humidity (RH) and gas concentration. 28 The fuel cell used for this study was developed inhouse, featuring 5 parallel channels of 1 mm width over an active area of 1 cm 2 . 29 Using steel spacers with defined thickness, cell compression was set such that ≈25% compression of the gas diffusion media was obtained. 29,30 The MEA break-in started by drawing the maximum current that would yield cell potentials > 0.6 V in H 2 /O 2 at 1.5 bar abs and a relative humidity (RH) of 100% between 25 and 80 • C for 2 hours (flow rates anode/cathode: 300/750 ml/min 31 , stoichiometries ≥ 30/≥ 30), followed by cooling down of the cell, activating potential cycles and a repetition of the first step (at 80 • C). It must be noted here that even as the applied stoichiometry ratios are significantly higher than for technical cells, the gas flow velocities remain in the same order of magnitude due to the reduced size of the device. 28 Cyclic voltammograms (CVs) were measured after break-in at 25 • C and 100% RH, scanning the potential between 0.075 and 1.0 V RHE at 50 mVs −1 with a H 2 anode flow rate of 50 ml/min and the N 2 cathode flow halted just prior to the measurement. The corresponding ECSA value was averaged from the H-adsorption and H-desorption charges between 0.09 and 0.4 V RHE after double-layer correction, assuming a conversion factor of 210 μC/cm 2 Pt . 32 H 2 -crossover tests were conducted by a linear potential sweep from 0.6 to 0.1 V RHE with a scan rate of 1 mVs −1 at 80 • C, 100% RH, 1.5 bar abs , an anode H 2 flow rate of 300 ml/min and a cathode N 2 flow rate of 750 ml/min, respectively; 33 the H 2 -crossover current densities typically amounted to ≈2 mA/cm 2 MEA . All I/E curves were recorded at 80 • C and 100% RH with anode/cathode flow rates of 300/750 ml/min (stoichiometries ≥ 30/≥ 30) at 1.5 bar abs for either H 2 /O 2 or H 2 /air, using a Biologic VSP-300 potentiostat with a 10 A/5 V current booster. The measurement was done galvanostatically, whereby the cell current was stabilized for 3 minutes at each data point and the data was averaged from the last 2 minutes. Concomitantly, the cell resistance (R ) was determined for each data point by galvanostatic electrochemical impedance spectroscopy (1 MHz to 1 Hz). Mass-and surface-specific activites for H 2 /O 2 operation were extracted at 0.9 V RHE after correcting potential and current for cell resistance and H 2 -crossover, respectively.
The accelerated stress tests were performed at 80 • C, 100% RH, ambient pressure, anode H 2 flow of 100 ml/min and cathode N 2 flow of 100 ml/min following shortened AST protocols established by the DOE for automotive PEFC application. 23 For start-stop and load- cycle degradation, the potential was cycled 10000 times between 1.0 and 1.5 V RHE at 500 mVs −1 and 0.6 and 1.0 V RHE at 50 mVs −1 , respectively. 5,16,23 At designated times (1000, 5000 cycles), the ASTs were interrupted to record I/E curves and CVs (cf. above) to determine the ECSA.
Transmission electron microscopy (TEM) images and elemental composition of the catalysts were obtained on a TECNAI F30 operated at 300 kV and equipped with an energy dispersive X-ray spectroscopy (EDX) detector. For tomography and cross section preparation with a focused ion beam-scanning electron microscope (FIB-SEM), a Zeiss NVision 40 microscope with a Ga + beam source and an EDX detector was employed.

Results and Discussion
Figures 1A, 1C and 1E (adapted from Reference 18) show the evolution of mass-specific activities (MAs), surface-specific activities (SAs) and ECSAs for the start-stop degradation test of 10000 potential cycles between 1.0 and 1.5 V RHE at 500 mVs −1 , normalized to the respective beginning-of-life (BOL) values. The latter are summarized in Table I, whereby the experimental data for the benchmark catalyst is in good agreement with literature reports for the same material. 32 After the start-stop AST, Pt 3 Ni aerogel cathodes show only a minor decrease of the MA, SA and ECSA. The aerogel thus behaves similarly to the unsupported hollow Pt-Fe nanocapsule catalyst that was recently investigated by Tamaki et al. following the same protocol. 16 The Pt/C benchmark, however, suffers from ≈50% loss of MA and ECSA at the end-of-life (EOL), in agreement with observations in Reference 5 using the same protocol. Additionally, the Pt/C benchmark displays drastic performance reduction in H 2 /air I/E curves, whereas Pt 3 Ni again maintains the BOL performance as discernable from Figure 2A. Moreover, we conducted the same start-stop AST on a commercial Pt-catalyst implementing a graphitized carbon support (Pt/C graphitized , see Experimental section for details). As showcased in Figure SI-1 of the Supplementary Information, in the early stages of the AST, this Pt/C graphitized underwent a less severe decrease of MA and ECSA when compared to Pt/C, but by the end of the 10000 potential cycles both variables reached ≈50% of their BOL-values (as compared to ≈40% for Pt/C, cf. Figures 1A and 1E). Therefore, this result further highlights the enhanced stability of the Pt 3 Ni-aerogel, even when compared to a commercial catalyst implementing a graphitized carbon support with a short-term durability better than that of Pt/C. For the rest of this study, though, we limit the comparison among degradation behaviors to the Pt 3 Ni aerogel and the standard (i.e., with a nongraphitized support) Pt/C benchmark, which has received much more attention in the existing literature. Upon load-cycle degradation tests of 10000 potential cycles between 0.6 and 1.0 V RHE at 50 mVs −1 (see Figures 1B, 1D and 1F) on the other hand, Pt 3 Ni MEAs display a severe MA reduction of ≈60%, as opposed to moderate losses of ≈30% for the Pt/C benchmark. Concomitantly, the SA of Pt 3 Ni aerogel decreases by ≈40% whereas for Pt/C electrodes the EOL value exceeds the BOL one by 50%. Moreover, the ECSAs for both catalysts evolve similarly (≈50% reduction, in agreement with literature reports on Pt/C 25 ) and the potential in H 2 /air I/E curves at high current densities decrease moderately (≈50 mV at 1.5 A/cm 2 , cf. Figure 2B). To explain the observed degradation of Pt 3 Ni aerogel vs. Pt/C MEAs during the ASTs, beginning-and end-of-life samples were analyzed at the micro-and nanoscale to complement the electrochemical characterization in Figures 1 and 2. First, cross sections of the catalyst layers (CLs) were cut with a focused-ion beam (FIB) and captured by scanning electron microscopy (SEM). Qualitative comparison of the cross section images for the Pt 3 Ni CL at BOL ( Figure 3A) with the ones at EOL after start-stop ( Figure 3B) and after load-cycle degradation ( Figure 3C) does not reveal any significant morphology or porosity changes. To support this observation by a quantitative parameter, the thickness of the CL was measured in at least three different locations and normalized to the Pt loading to allow for a comparison  between the different electrodes (BOL, EOL start-stop, EOL loadcycle). As discernable from Figure 4, the EOL CL thicknesses for Pt 3 Ni aerogel are comparable to the one at beginning-of-life, indicating negligible variations in porosity throughout the ASTs. Changes of the latter porosity can affect the mass transport of reactants by molecular and Knudsen diffusion within the catalyst layer and alter the gas diffusion overpotential caused by O 2 concentration gradients. 34,35 Variations of this overpotential can e.g. be deduced from the overall cell potential at high current densities (HCDs) in H 2 /air I/E curves, i.e. in conditions of strong concentration gradients. Thus, the negligible changes in HCD performance throughout the ASTs of Pt 3 Ni aerogel CLs in Figure 2 are consistent with a steady gas diffusion overpotential, and support the hypothesis that the porosity regime crucial to reactant and product mass transport remains unmodified.
The corresponding SEM cross section images for Pt/C are displayed in Figures 3D, 3E and 3F for BOL and following the startstop and load cycle ASTs, respectively. At the end of the start-stop degradation test, one can identify domains in the CL (highlighted by red circles in Figure 3E) that appear far less porous than at the BOL. This observation is concomitant with a ≈40% decrease in thickness (cf. Figure 4D) from a BOL value of ≈38 μm/(mg Pt cm −2 ) (or ≈33 μm/(mg C cm −2 ), which agrees with the 28 ± 2 μm/(mg C cm −2 ) reported in literature) 36 to ≈23 μm/(mg Pt cm −2 ) and points to a collapse of the CL structure which can inhibit effective mass transport of reactants and that is supported by the drastic performance decline in the H 2 /air I/E curves in Figure 2A. The CL collapse can be attributed to carbon support corrosion in the applied potential regimes [20][21][22]37,38 and the resulting CL porosity decrease has been demonstrated by Schulenburg et al. 39 in a FIB-SEM tomography study. On the other hand, representative cross section images before/after the load-cycle AST (Figures 3D/3F) do not indicate significant morphology/porosity changes concomitant with the negligible decrease in thickness to ≈34 μm/(mg Pt cm −2 ) discernable from Figure 4B. Again, this agrees well with the electrochemical results at the cell level, where only a minor alteration of the HCD performance in H 2 /air curves is observed (cf. Figure 2B).
To analyze the materials' change on the nanoscale, TEM images of the catalysts at BOL and after ASTs were taken and are summarized in Figure 5 (for additional images at other magnifications see Figures SI-2 and SI-3 in the Supplementary Information). At beginning-oflife, Pt 3 Ni consists of a well-defined 3D network of interconnected nanoparticles with an average nanochain diameter d nanochain of 5-6 nm (cf. Figure 5A). 12 After the start-stop test (Figure 5B), the nanochain diameter seems to be of similar size, yet the nanochains appear somewhat "smoothed", i.e. individual nanoparticle segments are less distinct. Unfortunately, it was not feasible to determine an accurate average nanochain diameter by measuring the size at multiple locations along the nanochain as in the case of the BOL sample since the catalyst layer was difficult to disperse in the process of TEM grid preparation (cf. 3D nanochain 'stacks' in low magnification images of Figure  SI-2). Regardless, the finding of comparable nanochain diameters agrees well with the electrochemical results in Figure 1E which displays the almost constant ECSA of the Pt 3 Ni aerogel after 10000 degradation cycles. On the other hand, after load-cycle degradation, Pt 3 Ni aerogel displays a significant increase in the nanochain diameter to a value ≥ 10 nm (see Figure 5C), albeit this finding cannot be quantified either due to the issue of poor CL dispersibility discussed above. Assuming d nanochain ∝ 1/ECSA, 32 the increase in nanochain size by about a factor of two is in good agreement with the halving of the ECSA values in Figure 1F at EOL.
As for the Pt/C benchmark, Figure 5D vs. Figure 5E illustrate the carbon support corrosion during start-stop degradation that was introduced above to explain the porosity decrease of the CL (Figures 3E  and 4B) and the decline of high current density performance (cf. Figure 2A). This corrosion leads to domains which are completely void of Pt nanoparticles (especially at the carbon edges) and to changes in the morphology of the carbon support that, as reported in previous studies, 40,41 becomes preponderantly amorphous during these startstop stress tests (see Figure SI-3). The dramatic ECSA reduction for Pt/C in Figure 1E (estimated on the basis of the initial mass of Pt in the catalyst layer) is thus related to particle detachment due to carbon corrosion, whereby smaller nanoparticles exhibit an increased probability to detach, 21 and not to other ECSA decreasing mechanisms leading to Pt nanoparticle growth (see below). Specifically, the negligible impact of these nanoparticle-aggregation routes is likely related to the potential regime applied in the start-stop AST (i.e., 1.0-1.5 V RHE ), which does not include sufficiently low potentials that would lead to the reduction (and dissolution/redeposition) 42-44 of the Pt-(hydr)oxides passivating the nanoparticles' surface.
Complementarily, the loss of ECSA during the load-cycle degradation tests (cf. Figure 1F) is most likely caused by Ostwald ripening and particle migration/coalescence, 21,25,26 which trigger the formation of larger Pt nanoparticles as discernable by comparison of Figures  5D and 5F. It has been observed that such an increase in particle size increases the surface-specific ORR activity of Pt-based catalysts, in terms caused by a positive shift of the adsorption energy of ORR impeding, oxygen-containing species (OH ads ) and the subsequent presence of more sites available for the catalysis of the reaction on the nanoparticles' surface. 3,32,45 Considering the universal ORR activity -ECSA relationships introduced in References 3,32,46 and 47 to express this particle size effect, the ECSA reduction in the load-cycle AST from ≈70 m 2 /g Pt to ≈30 m 2 /g Pt should result in a SA increase to ≈140% and a MA reduction to ≈60% of the respective BOL values. These estimates are very close to the actual EOL SA and MA values of ≈155% and ≈65% (vs. BOL), respectively. The remaining difference could arise from the fact that the ORR activity -ECSA relations in the above References 3,32,46,47 are based on RDE experiments in 0.1 M HClO 4 or H 2 SO 4 between 20 and 60 • C, whereas the activity values reported in this work were obtained in a PEFC at 80 • C. While the particle size effect can account for the load-cycle degradation results of the Pt/C benchmark, it falls short to explain the behavior of the Pt 3 Ni aerogel. To be precise, the ECSA reduction from ≈30 m 2 /g Pt to ≈15 m 2 /g Pt observed in this AST is expected to increase the SA to ≈135% and reduce the MA to ≈55% of the initial values, yet the final SA and MA amount to a mere ≈65% and ≈30% of their BOL counterparts, respectively. Consequently, other factors that can contribute to the evolution of the ORR activity need to be considered. In this respect, the leaching of the non-noble metal from Pt alloy ORR catalysts during ASTs has been reported as a cause for the concomitant activity decrease; 32,48 as an example, the resembling, ≈70% SA loss reported for a PtNi 3 /C catalyst cycled between 0.4 and 1.4 V RHE in Ref. 49 was partially related to the dissolution of ≈60% of its initial Ni-content. Moreover, previous works on Pt-Cu aerogels 50 and carbon-supported Pt-Cu and Pt-Ni electrocatalysts 49,51,52 have unveiled a correlation between the bulk phase non-noble metal content and the ORR activity that could explain this decrease of catalytic activity upon extended potential cycling.
To evaluate the possibility that the loss of Ni during the load-cycle AST may lead to the aerogel's ORR activity decrease, the catalyst layers and individual Pt 3 Ni nanochains were analyzed by SEM-EDX and TEM-EDX (three different locations each) to determine their Ni content (see Figure 6). The accuracy of the analysis is supported by the agreement of SEM and TEM data and the good match of the beginning-of-life Ni content values with the one derived from inductively coupled plasma-optical emission spectroscopy (ICP-OES) analysis in Reference 12. As discernable from Figure 6, Pt 3 Ni aerogel suffers from the loss of ≈35% of its Ni content during start-stop AST vs. ≈60% in the load-cycle degradation test. The lower Ni loss in the start-stop AST is related to the passivation of Pt by the formation of a PtO surface layer that mitigates the dissolution of surface Pt and subsurface Ni when cycling the potential above ≈1.0 V RHE . 24,53 The severe Ni leaching in the load-cycle test is a consequence of the observed Ostwald ripening phenomenon which constantly exposes Ni atoms from the nanochain interior to the corrosive PEFC environment. This effect can explain the mismatch between the ORR activity evolution of the Pt 3 Ni aerogel (cf. Figures 1D/1F) and the predictions from the particle size effect introduced above. 25,26 More precisely, Ni dissolution can lead to a positive shift of the catalyst's d-band center, thereby increasing the binding energy of oxygenated species and decreasing the ORR activity. 54,55 To study other potential degradation effects associated with Ni leaching, such as ionomer poisoning in the membrane and catalyst layer, 32 we performed BOL/EOL impedance measurements in H 2 /N 2 56,57 which showed a constant high frequency resistance with non-measurable proton transfer resistance along the catalyst layer (not shown here), 18,58 thus preventing the quantification of possible changes in this last variable. Alternatively, Ni-loss may also lead to an increase in O 2 -transport resistance 59 tentatively derivable from limiting current measurements 60 -a possibility that will be within the focus of our forthcoming works.

Conclusions
In summary, we have investigated the behavior of unsupported Pt 3 Ni aerogel cathodes under start-stop and load-cycle ASTs in the PEFC. Under start-stop degradation conditions, Pt 3 Ni MEAs retain their BOL mass-specific activity, ECSA and high current density performance in H 2 /air I/E curves after 10000 cycles, whereas Pt/C shows a ≈50% MA and ECSA loss concomitant with a drastic performance reduction in H 2 /air I/E curves. This superior durability of Pt 3 Ni aerogel vs. Pt/C is related to the severe corrosion of the carbon support in the latter catalyst that leads to Pt nanoparticle detachment and collapse of the catalyst layer structure. For the load-cycle AST, both Pt 3 Ni and Pt/C display a ≈50% reduction of the ECSA which is related to nanochain/nanoparticle size growth due to Ostwald ripening and particle migration/coalescence. The simultaneous decrease of MA and increase of SA for the Pt/C benchmark was found to be in excellent agreement with the well-known particle size effect on the ORR activity of Pt nanoparticles. For the Pt 3 Ni aerogel, though, the activity loss does not follow the same trend since, in addition to the increase in particle sizes, the catalyst also undergoes significant Ni dissolution (≈60%) from the alloy phase which further decreases both MA and SA. Nevertheless, based on our results and the overall fuel cell power output in the course of both ASTs, Pt 3 Ni aerogel is a promising PEFC catalyst for automotive application that warrants high durability in case of fuel starvation and cell start-up/shut-down events. To further improve its stability under normal PEFC operation, future work will focus on tuning the aerogel nanochain structure and surface vs. bulk composition, as to minimize Ni dissolution from the alloy catalyst. 61