RuO 2 Nanosheet Modiﬁed Pt 3 Co/C Cathode: Mitigating Activity Loss at High Temperature and High Potential Conditions

Pt 3 Co/C is known as an active catalyst for the oxygen reduction reaction (ORR). The stability of this catalyst under high temperature and potential conditions is a concern for practical application in fuel cell systems. In this study, we have pursued the possibility of improving the performance of Pt 3 Co/C under such severe conditions by the addition of ruthenium oxide nanosheets. The initial mass activity of RuO 2 nanosheets modiﬁed Pt 3 Co/C was higher compared to that of pristine Pt 3 Co/C. The improved mass activity is attributed to the increase in accessible active sites for ORR for the composite catalyst. The mass activity after accelerated durability test by potential step cycling between 0.6 and 1.2 V vs. RHE at 60 ◦ C for RuO 2 ns-Pt 3 Co/C was 243 A (g-Pt) − 1 , which is 1.9 times higher than non-modiﬁed Pt 3 Co/C (127 A (g-Pt) − 1 ). Transmission electron microscopy analysis and energy dispersive X-ray spectroscopy of the catalyst after durability test showed that the particle growth and dissolution of cobalt was inhibited by the addition of RuO 2 nanosheets. Platinum-cobalt alloy nanoparticles supported on carbon (PtCo/C) reaction. ◦ C to yield the RuO 2 ns-Pt 3 Co/C catalyst. The catalyst morphology char- acterized by electron operated at an accelerating voltage of 200 kV. Particle size distri- bution were determined from at least 10 randomly chosen typical TEM images (total of 200 to 300 particles). The atomic ratio (Pt/Co) in the catalysts was estimated from energy dispersive X-ray spectroscopy (EDX, Horiba EX-200) operated at an accelerating voltage of 20 kV.Electrochemical measurements were conducted with a rotating disk electrode (Nikko Keisoku). Pt 3 Co/C and RuO 2 ns-Pt 3 Co/C cata- lyst suspensions were prepared by dispersing 18.5 mg of catalyst in 25 mL of 2-propanol/water solution (75/25 volume ratio). A 5 wt% Naﬁonsolution(100 μ L)wasaddedtothecatalystsuspensionasapro-ton conducting binder to ensure adhesion. The working electrode was prepared by depositing 17.3 μ g-Pt cm − 2 on a mirror-polished glassy carbon rod (6 mm diameter) and vacuum dried at 60 ◦ C for 30 min. A carbon ﬁber (Toho Tenax Co., HTA40 E13 3K 200tex) was used as a counter electrode, and a reversible hydrogen electrode (RHE) was used as a reference electrode. Initial break-in cycles were conducted in de-aerated 0.5 M H 2 SO 4 (25 ◦ C) at 50 mV s − 1 for 70 cycles between 0.05and1.2Vvs.RHEtoelectrochemicallystabilizetheactivePtskinsurfacelayer.Theelectrochemicallyactivesurfacearea(ECSA)wasestimatedfromthehydrogenadsorptioncharge,excludingthehydro-genevolutionarea,andtheCoulombicchargenecessaryforoxidationonpolycrystallinePtas210 μ C cm − 2 . Linear sweep voltammetry was measured in O 2 -saturated 0.5 M H 2 SO 4 at rotation rates of ω = 2500, 2200, 1600, 1200, 800, and 400 rpm. Background subtraction was conducted by subtracting the linear sweep voltammograms collected under de-aerated conditions (N 2 gas) at the respective rotation rates from the voltammograms taken in O 2 -saturated 0.5 M H 2 SO 4 . The potential was swept at 10 mV s − 1 from 0.05 to 1.2 V vs. RHE. Oxygen reduction reaction (ORR) activity was obtained by extrapolation of Koutecky-Levich plots at 0.85 V vs. RHE to ω = ∞ , which gives the kinetically controlled current. Mass activity will be reported as the kinetically controlled current at 0.85 V vs. RHE divided by the mass of Pt. Three different Accelerated durability test (ADT) conditions were applied to elucidate the impact of stepping to high potentials. After ac-cumulation of initial ORR activity, the potential was stepped between 0.6 and 1.0 V vs. RHE (3 s hold each) at 25 or 60 ◦ C for 2000 cycles ( ω = 0). For the second protocol, 2000 cycles were applied between 0.6 and 1.1 V vs. RHE. The upper potential limit was increased to 1.2 V vs. RHE for the third protocol. The three durability test protocols will be abbreviated as and The values are 3 individual

Platinum-cobalt alloy nanoparticles supported on carbon (PtCo/C) shows high oxygen reduction reaction (ORR) activity, and is used as a practical catalyst in polymer electrolyte fuel cells for both residential and vehicle applications. [1][2][3][4] Surface cobalt atoms of the PtCo alloy easily dissolve in hot acid solution, forming a Pt skeleton layer enriched with Pt, which is then either thermally 5 or electrochemically treated 6 to stabilize an active Pt skin surface layer. Such catalysts have higher ORR activity than Pt alone, 5,[7][8][9] owing in part to the modification of the electronic state of Pt by the alloying with cobalt. The ORR activity and durability of PtCo alloy are affected by various parameters including the surface structure, 5,7-14 the atomic ratio of platinum and cobalt, 6,[15][16][17][18][19] as well as the nanoparticle size, 20,21 which have been finely optimized to give maximum performance.
Despite these tremendous efforts, binary catalysts still suffer from durability issues. Although cobalt dissolution is protected by the Ptskin layer, [22][23][24][25][26][27] studies have underlined that dissolution of cobalt from PtCo core is accelerated at high temperature leading to severe loss in ORR performance. 16,28 The stability of electrocatalyst under high load or high potential conditions is also problematical as a result of dissolution of both Pt and Co. [29][30][31][32] In the case of PtCo alloy catalysts, once the protective Pt skin layer is lost, severe de-alloying of cobalt eventually leads to a thick Pt layer that has activity comparable to pure platinum. 16,[28][29][30][31][32] Based on these studies, electrocatalyst degradation under high temperature and severe potential cycling conditions may be one of the few problems associated with using PtCo alloys as an ORR catalyst in practical systems.
Many approaches have been proposed to mitigate electrocatalyst degradation, with one of the most well-studied methods being the addition of metal oxide to Pt/C. Oxide additives such as TiO 2 , 33,34 SiO 2 , 35,36 CeO 2 , 37 and WO 3 , 34 have been shown to inhibit particle growth and/or dissolution of the catalyst. For example, TiO 2 -Pt/C exhibited higher durability compared to Pt/C, which was attributed to the suppression of the aggregation of Pt nanoparticles. 33,34 Similarly, particle growth and sintering of Pt nanoparticles was suppressed for Pt-CeO x /C composite catalyst. 37 Although these are effective cocatalysts, one must take into consideration that the co-catalysts are poor electronic conductors and care must be taken so that the oxide additive does not cover or block the metal surface, which would result in loss of accessible surface area.
We have reported that the catalytic performance of Pt/C could be improved by the addition of RuO 2 nanosheets. [38][39][40] RuO 2 nanosheets are 1 nm thick nanosheets derived by exfoliation of layered H 0.2 RuO 2.1 · nH 2 O, and is high electronically conductive and stable. 41,42 Owing to the water-swelling property of RuO 2 nanosheets, access of electrolyte to the Pt surface should not be disturbed in the presence of water. An increase in initial ORR activity and enhancement in long term performance was observed by the addition of RuO 2 nanosheets to Pt/C. [38][39][40] Here, we extend our studies on the use of RuO 2 nanosheets as a co-catalyst to the Pt 3 Co/C alloy system. We anticipate that this will allow us to take advantage of both the high ORR activity of Pt 3 Co alloy, and the durability/activity enhancing effect of RuO 2 nanosheets to overcome the durability issue of Pt 3 Co/C under high temperature and high potential conditions. It should be noted that although much effort has been placed on increasing the performance of Pt/C with oxide additives, studies on oxide-PtCo composites are scarce. To the best the authors knowledge, only one study has been reported, namely the CeO x -PtCo/C system, which exhibited enhanced activity. 43 The durability of this catalyst was unfortunately not described. In this study, emphasis is placed on increasing the durability of Pt 3 Co/C at high temperature and high potential conditions. The ORR activity of RuO 2 nanosheet modified Pt 3 Co/C (RuO 2 ns-Pt 3 Co/C) after accelerated durability tests at three different potential windows (0.6-1.0, 0.6-1.1, 0.6-1.2 V vs. RHE) was investigated at 60 • C. We will show that the composite catalyst exhibited improved performance in terms of both initial ORR activity and stability especially at high potential regions.

Experimental
Pt/C (TEC10E50E, 47 mass% Pt) and Pt 3 Co/C (TEC36E52, 52 mass% Pt 3 Co) were purchased from Tanaka Kikinzoku Kogyo K.K.. Pt 3 Co/C was used after acid-treatment with 0.1 M HClO 4 (60 • C) for 12 hours to remove excess cobalt. The composite catalyst was synthesized following our previous recipe for RuO 2 nanosheet modified Pt/C. [38][39][40] RuO 2 nanosheets were derived via exfoliation of layered H 0.2 RuO 2.1 ·nH 2 O as reported previously. 41,42 Briefly, 0.1 g of layered H 0.2 RuO 2.1 ·nH 2 O was added to an aqueous solution of 10% tetrabutylamonium hydroxide (TBA + OH − ), adjusted to a solid-liquid ratio of 4 g L −1 (TBA + /H + = 1.5). The mixed solution was shaken for 10 days and then centrifuged to remove any non-exfoliated material at 2000 rpm. RuO 2 nanosheet colloid was then mixed with a suspension of commercial Pt 3 Co/C catalyst to prepare the composite catalyst. In a typical synthesis, the RuO 2 nanosheet colloid, diluted to 1 g-RuO 2 L −1 , was added to an aqueous Pt 3 Co/C suspension with a molar ratio of RuO 2 /Pt/Co = 0.3/1/0.3 (9.0 mass% RuO 2 , 42.8 mass% Pt, 4.5 mass% Co). The dispersion was magnetically stirred and ultrasonificated to ensure homogenous reaction. After sedimentation, the composite was washed with water and dried at 120 • C to yield the RuO 2 ns-Pt 3 Co/C catalyst. The catalyst morphology was characterized by transmission electron microscopy (TEM, JEOL 2010) operated at an accelerating voltage of 200 kV. Particle size distribution were determined from at least 10 randomly chosen typical TEM images (total of 200 to 300 particles). The atomic ratio (Pt/Co) in the catalysts was estimated from energy dispersive X-ray spectroscopy (EDX, Horiba EX-200) operated at an accelerating voltage of 20 kV.
Electrochemical measurements were conducted with a rotating disk electrode (Nikko Keisoku). Pt 3 Co/C and RuO 2 ns-Pt 3 Co/C catalyst suspensions were prepared by dispersing 18.5 mg of catalyst in 25 mL of 2-propanol/water solution (75/25 volume ratio). A 5 wt% Nafion solution (100 μL) was added to the catalyst suspension as a proton conducting binder to ensure adhesion. The working electrode was prepared by depositing 17.3 μg-Pt cm −2 on a mirror-polished glassy carbon rod (6 mm diameter) and vacuum dried at 60 • C for 30 min. A carbon fiber (Toho Tenax Co., HTA40 E13 3K 200tex) was used as a counter electrode, and a reversible hydrogen electrode (RHE) was used as a reference electrode. Initial break-in cycles were conducted in de-aerated 0.5 M H 2 SO 4 (25 • C) at 50 mV s −1 for 70 cycles between 0.05 and 1.2 V vs. RHE to electrochemically stabilize the active Pt skin surface layer. The electrochemically active surface area (ECSA) was estimated from the hydrogen adsorption charge, excluding the hydrogen evolution area, and the Coulombic charge necessary for oxidation on polycrystalline Pt as 210 μC cm −2 . Linear sweep voltammetry was measured in O 2 -saturated 0.5 M H 2 SO 4 at rotation rates of ω = 2500, 2200, 1600, 1200, 800, and 400 rpm. Background subtraction was conducted by subtracting the linear sweep voltammograms collected under de-aerated conditions (N 2 gas) at the respective rotation rates from the voltammograms taken in O 2 -saturated 0.5 M H 2 SO 4 . The potential was swept at 10 mV s −1 from 0.05 to 1.2 V vs. RHE. Oxygen reduction reaction (ORR) activity was obtained by extrapolation of Koutecky-Levich plots at 0.85 V vs. RHE to ω = ∞, which gives the kinetically controlled current. Mass activity will be reported as the kinetically controlled current at 0.85 V vs. RHE divided by the mass of Pt. Three different Accelerated durability test (ADT) conditions were applied to elucidate the impact of stepping to high potentials. After accumulation of initial ORR activity, the potential was stepped between 0.6 and 1.0 V vs. RHE (3 s hold each) at 25 or 60 • C for 2000 cycles (ω = 0). For the second protocol, 2000 cycles were applied between 0.6 and 1.1 V vs. RHE. The upper potential limit was increased to 1.2 V vs. RHE for the third protocol. The three durability test protocols will be abbreviated as ADT(0.6-1.0), ADT(0.6-1.1) and ADT(0.6-1.2). The reported initial ORR activity values are averaged from 3 individual runs. The ORR activity after ADT(0.6-1.0), ADT(0.6-1.1) and ADT(0.6-1.2) are averaged from two measurements for each ADT condition. All water used was ultrapure grade (Milli-Q, > 18 M cm).

Results and Discussion
ORR activity of fresh Pt 3 Co/C and RuO 2 ns-Pt 3 Co/C catalyst.- Figure 1 shows typical TEM images of Pt 3 Co/C and RuO 2 ns-Pt 3 Co/C. The RuO 2 nanosheets have lateral dimension of a few hundred nanometers in size with ∼1.0 nm thickness. The corresponding particle size distribution are also shown. The average diameter for Pt 3 Co/C and RuO 2 ns-Pt 3 Co/C were 4.2±0.7 nm and 4.2±0.8 nm, respectively. The addition of RuO 2 nanosheets has no negative impact on the Pt 3 Co particle size distribution, in other words, aggregation of Pt 3 Co nanoparticles is negligible.
The ECSA of fresh RuO 2 ns-Pt 3 Co/C was 47 m 2 (g-Pt) −1 , which is comparable or slightly higher than pristine Pt 3 Co/C (43 m 2 (g-Pt) −1 ) ( Table I). The mass activity of the catalysts increased with increasing temperature, as shown in Figure 2. This is attributed to an increase in ORR kinetics with increasing temperature, agreeing well with previous reports. 12,16,28 The mass activity of RuO 2 ns-Pt 3 Co/C was higher than Pt 3 Co/C for all temperatures studied (Table II). The enhancement effect was more pronounced at higher temperatures; the   mass activity at 60 • C for RuO 2 ns-Pt 3 Co/C was 467A (g-Pt) −1 , 1.2 times higher than that of Pt 3 Co/C (385 A (g-Pt) −1 ) and 1.6 times higher than that of Pt/C (285 A (g-Pt) −1 ). The higher mass activity of RuO 2 ns-Pt 3 Co/C may be interpreted as an increase in accessible sites for ORR compared to pristine Pt 3 Co/C. This is also supported by the fact that the limiting current for ORR in the linear sweep voltammograms was higher for RuO 2 ns-Pt 3 Co/C compared to Pt 3 Co/C at 60 • C (Fig. 3). The reason for the increase in ORR accessible sites is not clear at the moment, but the addition of conducting RuO 2 nanosheets may have a positive impact on the overall conductivity of the catalyst.

ORR activity of RuO 2 ns-Pt 3 Co/C after accelerated durability tests under different potential windows.-
The durability of the catalysts was evaluated by potential step cycling tests with different potential  windows (0.6-1.0, 0.6-1.1, and 0.6-1.2 V vs RHE) at 60 • C. The retention in ECSA and mass activity of Pt 3 Co/C after the three different ADT protocols decrease with the increase in the upper potential limit ( Table I). The durability of Pt 3 Co/C is higher than Pt/C which is evident from the slower loss of ECSA for Pt 3 Co/C compared to Pt/C (Fig. 4a). The decrease in ECSA was further retarded for RuO 2 ns-Pt 3 Co/C (Fig. 4b). The ECSA after ADT(0.6-1.0) and ADT(0.6-1.2) for Pt 3 Co/C was 36 and 20 m 2 (g-Pt) −1 , respectively. The ECSA after ADT(0.6-1.0) and ADT(0.6-1.2) for RuO 2 ns-Pt 3 Co/C was 44 and 33 m 2 (g-Pt) −1 , respectively. The ECSA of RuO 2 ns-Pt 3 Co/C after both ADT protocols was higher than that of Pt 3 Co/C, suggesting that the addition of RuO 2 nanosheets contributes to the enhancement in catalyst durability. The mass activity of RuO 2 ns-Pt 3 Co/C after all ADT protocols was higher than that of Pt/C and Pt 3 Co/C, agreeing with the trend in ECSA retention. The degradation of the catalyst is accelerated by the high potential durability conditions; the mass activity after ADT(0.6-1.2) for all catalysts were lower than after ADT(0.6-1.0) (Fig. 5). Although, the mass activity of Pt 3 Co/C after ADT(0.6-1.0) and ADT(0.6-1.1) was higher than that of Pt/C, the superiority becomes less evident for Pt 3 Co/C at a high potential durability condition (ADT(0.6-1.2)). As shown in Fig. 5 and Table I, this drawback is mitigated by the addition of RuO 2 nanosheets. The mass activity after ADT(0.6-1.2) for RuO 2 ns-Pt 3 Co/C was 243 A (g-Pt) −1 , which is 1.9 times higher than that of Pt 3 Co/C. As shown in Figure 6, the addition of RuO 2 nanosheets is particularly effective for enhancing the performance at high temperature. The results show that the addition of RuO 2 nanosheets is especially effective to improve the catalyst durability under high potential conditions and high temperatures.  Temperature / °C RuO 2 ns-Pt 3 Co/C Pt 3 Co/C Pt/C Figure 6. The ORR activity of (circles) Pt/C, (squares) Pt 3 Co/C, and (triangles) RuO 2 ns-Pt 3 Co/C after ADT(0.6-1.2) (refer to text for detailed measurement conditions).
The average particle size after ADT(0.6-1.2) at 60 • C for Pt 3 Co/C was 5.9±1.4 nm, whereas it was 4.6±1.3 nm for RuO 2 ns-Pt 3 Co/C (Fig. 7). After ADT (0.6-1.2), the metal ratio for Pt 3 Co/C was Pt/Co = 1/0.14, while for RuO 2 ns-Pt 3 Co/C this was Pt/Co = 1/0.19. By the ADT(0.6-1.2) treatment, 34% of cobalt was leached out from Pt 3 Co/C (Pt/Co = 1/0.23 after acid treatment). This was suppressed to 25% for RuO 2 ns-Pt 3 Co/C. The results indicate that the addition of RuO 2 nanosheets is effective to inhibit cobalt dissolution from catalyst. From TEM and EDX analysis, the enhancement in durability of RuO 2 ns-Pt 3 Co/C can be explained by the inhibition of particle size growth and Co dissolution. The better durability of the RuO 2 ns-Pt 3 Co/C composite catalyst is strongly correlated to the retention in the particle size and Pt/Co composition. RuO 2 nanosheet model electrode studies have shown that there are chemical interactions with the RuO 2 nanosheets and Pt n+ ions as well as metallic Pt. 44,45 Assuming a similar mechanism for the RuO 2 ns-Pt 3 Co/C system, the enhancement in durability for Pt 3 Co/C can be attributed to the suppression of the diffusion of dissolved ions from the catalyst and a strong metal support interaction. Finally, it should be noted that the initial activity of RuO 2 ns-Pt 3 Co/C is 1.6 times higher than a typi-cal Pt/C catalyst. Moreover, the higher retention of mass activity is obtained for all ADT conditions, which leads to a 2 times higher ORR activity of RuO 2 ns-Pt 3 Co/ compared to Pt/C after different ADTs.

Conclusions
The durability of Pt 3 Co/C at high temperature and high potential condition was successfully improved by the addition of RuO 2 nanosheets. The initial mass activity of Pt 3 Co/C at various temperatures between 25 and 60 • C were improved by the addition of RuO 2 nanosheets. The enhancement effect was more pronounced at higher temperatures. The catalyst durability was evaluated by potential cycling at 60 • C with three different potential windows 0.6-1.0, 0.6-1.1, and 0.6-1.2 V vs. RHE. The mass activity of RuO 2 ns-Pt 3 Co/C was 1.9 times higher than Pt 3 Co/C after the potential cycling tests. Based on the TEM analysis, the increase in particle size was suppressed by the addition of RuO 2 nanosheets. Additionally, the modification with RuO 2 nanosheets also inhibited cobalt dissolution, which should also contribute to the enhanced durability. Based on our results on the impact of the lateral size of nanosheets to the ORR activity 38 and influence of the nanosheet content, 40 further enhancement in catalyst performance may be anticipated by optimization of catalyst structure and composition.