Fine-tuning the activity of oxygen evolution catalysts: The effect of oxidation pre-treatment on size-selected Ru nanoparticles

Fine-tuning the activity of oxygen evolution catalysts: The effect of oxidation pre-treatment on size-selected Ru nanoparticles

Catalysis Today 262 (2016) 57–64 Contents lists available at ScienceDirect Catalysis Today journal homepage: www.elsevier.com/locate/cattod Fine-tu...

2MB Sizes 0 Downloads 22 Views

Catalysis Today 262 (2016) 57–64

Contents lists available at ScienceDirect

Catalysis Today journal homepage: www.elsevier.com/locate/cattod

Fine-tuning the activity of oxygen evolution catalysts: The effect of oxidation pre-treatment on size-selected Ru nanoparticles Elisa Antares Paoli a , Federico Masini a , Rasmus Frydendal a , Davide Deiana b , Paolo Malacrida a , Thomas W. Hansen b , Ib Chorkendorff a,∗ , Ifan E.L. Stephens a,∗ a b

Center for Individual Nanoparticle Functionality, Department of Physics, Building 312, Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark Center for Electron Nanoscopy, DTU Danchip, Building 307, Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark

a r t i c l e

i n f o

Article history: Received 30 July 2015 Received in revised form 2 October 2015 Accepted 4 October 2015 Available online 31 October 2015 Keywords: Oxygen evolution Electrocatalysis Nanoparticle Electrolysis Electrochemistry Corrosion

a b s t r a c t Water splitting is hindered by the sluggish kinetics of the oxygen evolution reaction (OER). The choice of materials for this reaction in acid is limited to the platinum group metals; high loading required of these scarce and expensive elements severely limit the scalability of such technology. Ruthenium oxide is among the best catalysts for OER, however the reported activity and stability can vary tremendously depending on the preparation conditions and pre-treatment. Herein, we investigate the effect of oxidation treatment on mass-selected Ru nanoparticles in the size range between 2 and 10 nm. The effect of two distinct oxidation pre-treatments on the activity and stability have been investigated: (1) thermal oxidation; and (2) oxidation with an oxygen plasma under vacuum. We report that activity and stability can be tuned by using different oxidation pre-treatments. Thermally oxidized particles exhibited the lowest activity, although over an order of magnitude higher than the state of the art, and the highest stability. Plasma-treated particles showed intermediate performance between as-deposited and thermally oxidized NPs. © 2015 Elsevier B.V. All rights reserved.

1. Introduction The exponential growth of population and catch up of living standards in developing countries has increased the global demand for power significantly. In 2014 it was assessed to be around 17 terawatt and it is predicted to reach 30 TW by 2050 [1]. This means that any technology, in order to have a significant impact on the global energy scenario, will need to be scaled to the TW level [2,3]. This puts stringent requirements on the materials used as components for such technologies. At the moment, we completely rely on fossil fuels, which are a limited resource and introduce carbon dioxide in the atmosphere when burned. Renewable energy sources, such as solar and wind power, can play a key role in substituting coal and gas. Furthermore, as highlighted in a recent economic perspective, the costs of wind and photovoltaics sources are decreasing significantly, approaching the price of petroleum, making them more economically competitive [4].

∗ Corresponding authors. Tel.: +45 22979924. E-mail addresses: [email protected] (I. Chorkendorff), [email protected] (I.E.L. Stephens). http://dx.doi.org/10.1016/j.cattod.2015.10.005 0920-5861/© 2015 Elsevier B.V. All rights reserved.

On the other hand, the major drawback of renewable energy sources is their variability over time. This leads to the need for efficient ways of energy storage. Electrolysis of water allows hydrogen to be produced from excess renewable energy when wind or solar plants exceed consumption [5–9]. The hydrogen can then be converted back into water in fuel cells, producing electricity, whenever it is required. Among the commercially available electrolyzers, Polymer Electrolyte Membrane (PEM) electrolyzers are highly suitable towards applications as small-scale delocalized storage units, e.g. hydrogen refuelling stations. Their advantages over traditional alkaline electrolyzers include; (a) the absence of liquid water, as they use a humidified proton-conducting polymer (b) they can operate at much higher current densities with better efficiency (1–3 A/cm2 in comparison to the 0.2 A/cm2 typical of traditional alkaline electrolyzers) (c) they can manage fluctuating power input and (d) have faster start up times [8]. On the other hand, PEM electrolyzers still suffer from durability issues and high costs, mainly related to the membrane and the electrode materials. In fact, the extremely acidic conditions limit the choice of the electrode materials to precious metals (Pt and Ir). Significant advances are required in order to make this technology cost competitive. In particular, minimising the energy losses at the anode side, where oxygen is evolved,

58

E.A. Paoli et al. / Catalysis Today 262 (2016) 57–64

should be addressed. The oxygen evolution reaction (OER) exhibits sluggish kinetics, which limits the efficiency of an electrolyzer [6]. Furthermore, the amount of precious metal required is high. At the moment the lowest metal loading reported is based on platinum iridium nanowhiskers on a non-conducting support, which required 0.30 mgPt group /cm2 [10]. Conversely, the half reaction at the cathode side, the hydrogen evolution reaction (HER), has been greatly investigated in the past and advances in the electrode material have been achieved [11–16]. Specifically, Gasteiger and co-workers have shown that HER can be carried out at overpotentials lower than 2 mV employing low amounts of platinum (0.05 mg/cm2 ) in a membrane electrolyte assembly (MEA) [11]. Moreover, a number of non-precious metal compounds such as sulfides and phosphides now have HER activity approaching that of Pt, although the stability needs to be improved [13,14,17–20]. Switching to hydroxide conducting membranes may eventually allow a much wider range of active and stable compounds to be used at the electrodes. However, at present such membranes constitute an immature technology: they are less stable, less conducting, and typically impose a higher overpotential at the hydrogen electrode [21–24]. In a previous publication we estimated the amount of precious metals required per terawatt of hydrogen storage capacity, should Pt and PtIr be employed in a real electrolyzer as cathode and anode, respectively [10]. We calculated that half a year of platinum and 10 years of iridium annual production would be needed per 1 TW of hydrogen storage capacity. Hence, for widespread use of electrolyzers, it is critical to decrease the amount of precious metal at the anode. Strikingly, a lot of efforts have been dedicated to finding new catalysts for the HER that are stable and active in acid. Unfortunately if no improvements are made to the oxygen electrode, these efforts would be in vain. A first approach to minimize the amount of precious metal is to prepare catalysts in nanoparticulate form, where the surfaceto-bulk ratio is maximized. Specifically, it would be of great technological relevance to investigate nanoparticles in the size range below 10 nm, where the effect of particle size is enhanced [25–27]. Alternatively, the precious metals could be combined with a non-noble metal [28–35], analogous to the Pt-alloys used for Oxygen Reduction Reaction (ORR) [36], or replaced altogether. At present, non-noble metal oxides with reasonable activity and stability in acid have not been found. However, if they were to be discovered it would significantly improve the scalability of PEM electrolyzers. The current state of the art for the oxygen evolution reaction in acidic media, expressed as mass activity, is shown in Fig. 1. Ruthenium dioxide is known to be the most active catalyst for OER and has been extensively investigated in the past [40–42]. However, the reported activity varies drastically, depending on the preparation method and experimental conditions, as evidenced by the plot [39–41,43–48]. The oxidation pre-treatment plays a particularly strong role in controlling activity and stability under OER conditions [37,41,44,48–50]. The surface will always form an oxide during oxygen evolution. Even so, the oxidation pre-treatment can have dramatic effects on the activity and stability of a material. Electrochemically grown oxides tend to be significantly more active, but at the same time highly unstable, in comparison to thermally grown oxides [48,49,51]. The reason behind this behaviour is not well understood. Recently, Danilovic et al. have proposed an explanation based on differences in surface density of defects, suggesting a functional link between activity and stability [49]. However, the correlation between activity and stability is not universal, and seems to be somewhat dependent on the measurement conditions; Rh, Ir and Au seem to be particular outliers [52,53].

Fig. 1. Overview of the state of the art for the oxygen evolution reaction in acid, expressed in mass activity. Data taken from this work for 5 nm (P)-RuO2 , from [37] for 3 nm RuO2 NP, from [33] for IrNi3.3 NP, from [35] for Ir nanodentrites (ND) on Antimonium Tin Oxide (ATO), from [38] for Ru0.5 Ir0.5 Thin Film (TF), from [10] for PtIr NanoStructured Thin Film (NSTF), from [39] for rutile (r)-RuO2 NP and (r)-IrO2 NP.

The establishment of common procedures to evaluate the catalytic performance has become more impelling, allowing for straightforward comparison among catalysts and among different experimental groups. There are several recent examples which suggest that there is a general trend towards standardization of the oxygen evolution reaction, both in terms of activity and stability [5,37,39,43,54–64]. Although numerous efforts have been dedicated to estimate the true active area, reports of intrinsic catalytic activities are largely missing. Hence, there is a general call for studies on welldefined surfaces such as single crystals [56,60,65,66]. However, under oxidative conditions most single crystals undergo drastic changes due to surface reconstruction, making their investigation very challenging [50]. Alternatively, other model systems such as size selected and well-defined nanoparticles can offer interesting insights into the reaction and allow accurate evaluation of the mass activity [37]. By adapting a methodology previously used in our laboratory to investigate Pt and Pt alloys for ORR [25,26,67,68], we recently investigated the activity and stability of mass selected ruthenium nanoparticles prepared by magnetron sputtering in the range size between 2 and 9 nm. The particles exhibited over an order of magnitude improvement in mass activity compared to the state of the art. We also investigated the effect of the oxidation treatment, which showed, in accordance with literature, that as-prepared nanoparticles are more active than thermally oxidized particles. Herein, we further examine how various oxidation methods can affect the performance of ruthenium nanoparticles. The activity and stability of as-prepared, thermally oxidized and oxygen plasma treated nanoparticles are compared. The mass selected nanoparticles were prepared by a physical method and then mass-selected using a Time of Flight (TOF) filter. This technique allows us to precisely control several parameters, such as particle size and coverage, which in turn allows us to estimate accurately the mass activity. Furthermore, differently from chemical synthesis, it is possible to avoid inherent artefacts from precursors and surfactants [45]. 2. Material and methods 2.1. As-deposited Nanoparticles preparation Ruthenium nanoparticles (NPs) were prepared using an Ultra High Vacuum (UHV) compatible mass aggregation source (Birmingham Instruments Inc.) from a Ru target (Kurt J. Lesker) and

E.A. Paoli et al. / Catalysis Today 262 (2016) 57–64

mass- selected using a Time of Flight (TOF) filter. The substrate was mounted in a multichamber UHV system (Omicron, Multiscan) with a base pressure of 10−10 mbar, where Ru atoms are sputtered away using Ar+ ions and condense into nanoparticles of different sizes and masses by collision with cooled Ar and He gas. After separation, based on their mass to charge ratio, the particles were deposited onto either a glassy carbon disk (6 mm dia, 4 mm thick, HTV Hochtemperatur Werkstoffe GmbH) for the electrochemical measurements or onto a Si3 N4 grid for Transmission Electron Microscopy (TEM) characterisation. Prior to deposition, the glassy carbon disks were heated up to 550 K for 8 h, in order to eliminate any residue of water. The mass loading was determined directly from the deposition current and the particle coverage was set to 15%. 2.2. Thermal oxidation The as-deposited nanoparticles were then transferred to a furnace, where they were thermally oxidized in a flow of oxygen at 400 ◦ C for one minute, as described in our previous study on massselected Ru nanoparticles. Throughout this study, we will refer to these particles as (T)-RuO2 NPs. 2.3. Oxygen plasma treatment Alternatively, in order to investigate the effect of the oxidation treatment on the OER activity, Ru NPs were oxidized using an oxygen plasma. The as-deposited particles were moved, through exposure to air, to a sputter deposition chamber (AJA international, Inc.), where it was possible to form an oxygen plasma by RF bias on the substrate holder. The following conditions were carried out for 2 min: RF power of 30 W, 20 mTorr as base pressure and 20 sccm of O2 flow. These catalysts will be identified from now on as (P)RuO2 , in order to differentiate them from the as-deposited Ru and thermally oxidized (T)-RuO2 NPs. 2.4. Characterization In order to compare the different oxidation treatments, the nanoparticles were characterised with X-Ray Photoelectron Spectroscopy (XPS) and High Resolution–Transmission Electron Microscopy (HR-TEM). XPS was performed using a Theta Probe instrument (Thermo Scientific) with a base pressure of 5 × 10−10 mbar. The X-Ray source was a monochromatised Al K˛ (1486.7 eV). The fitting of the data was made using a GaussianLorentzian functions mixed with exponential tail. The carbon signal was fitted using three different features, which we attribute to carbon and to surface oxides. Ruthenium was fitted using two peaks for the metallic part, two peaks for ruthenium dioxide (low binding energy B.E) and two peaks for the satellite features of ruthenium dioxide (high binding energy B.E.) [69–72]. HR-TEM images were acquired using a FEI Titan Analytical 80–300 electron microscope, equipped with a CEOS Cescor probe spherical aberration correction. 2.5. Electrochemical measurements All the electrochemical tests were performed with a Rotating Ring Disk Electrode (RRDE) assembly (Pine Instruments Corporation), using a Bio-Logic Instruments VMP2 multichannel potentiostat/galvanostat. The data was acquired using the Bio-Logic EC-Lab software. The electrochemical cell was a standard three electrode glass cell, equipped with a Luggin capillary. The counter electrode was a carbon rod, while the reference electrode was a Hg/HgSO4 electrode. The reference electrode was calibrated against

59

the reversible hydrogen electrode (RHE) using the Pt ring as working electrode. All the potentials were then converted to the RHE scale. The electrochemical measurements were carried out in a N2 saturated 0.05 M H2 SO4 solution at room temperature and 1600 rpm. The acid solution was prepared by dilution with Millipore water (18 M cm) from 98% Merck Suprapur sulphuric acid. The catalysts were all tested with the same protocol. At first, cyclic voltammetry in the capacitive region (between 1.00 and 1.23 V vs. RHE) were recorded at 20 mV/s, while the Pt ring was cleaned by cycling the potential between 0.05 and 1.35 V vs. RHE at 100 mV/s. Once the platinum ring was cleaned, the Ohmic losses at the disk were evaluated using electrochemical impedance spectroscopy (EIS) over the range 1–200,000 Hz at a DC potential of 10 mV at open circuit potential [73]. Next, the OER activity at the disk was measured by cycling the potential between 1.0 V and 1.5 V vs. RHE at 20 mV/s, while the ring was held at 1.1 V in order to follow the RuO4 species formation [74]. The CVs were performed using the bipotentiostat mode of the EC-Lab program. Due to deposition of the dissolved Ru on the ring, only the first CV was used to estimate the Faradaic efficiency towards ruthenium corrosion. The cyclic voltammetry was first corrected for the capacitance and then for Ohmic losses [39]. During all tests the Ohmic drop was compensated at 85% using the EC-Lab program MIR. The collection efficiency was calculated using a ferricyanide redox couple and was found to be 0.2 ± 0.01, based on three independent measurements. The corrosion efficiency was estimated by assuming a one electron process for the ruthenium oxide dissolution. 3. Results and discussion The aim of this work is to investigate the effect of a different oxidation pre-treatment on the activity and stability of ruthenium oxide nanoparticles, using an oxygen plasma created in an ultra-high vacuum system. By investigating mass-selected mono-dispersed nanoparticles, it is possible to provide a model investigation on the oxygen evolution reaction and precisely define the intrinsic catalytic activity. In turn, this allows us to objectively compare particles that undergo to different oxidation treatments. The nanoparticles were prepared in the range between 0.035 and 4 × 106 unit (u), which corresponds to a range in size between 2 and 10 nm, assuming the particles to be spherical. From previous studies in our laboratory on as-deposited NPs, it was observed that smaller particles (2–6 nm) possess a more regular shape, whereas larger particles are more corrugated and irregular [75]. The equivalent particle diameter will be used throughout the paper instead of the single particle mass for simplicity. 3.1. Nanoparticles characterisation The surface oxidation state of the NPs was investigated by XPS. In a previous report, we compared the XPS spectra of the Ru 3d core level of as deposited and thermally oxidized 9 nm Ru NPs [37]. We observed that the metallic component dominates the spectrum of as deposited Ru NPs, suggesting only a surface oxidation and a metallic core; whereas thermally oxidized NPs are fully converted to RuO2 . The bulk oxidation was confirmed by Grazing Angle–Xray Diffraction (GA-XRD). Fig. 2 shows the XPS spectra of 4 nm and 7 nm plasma treated NPs. The metallic component is present in both spectra. While the smaller particles are almost completely oxidized, at least at the surface (Fig. 2a), the larger particles exhibit a bigger contribution from the metallic component. This may be simply related to difference in particle size. Fig. 3 shows HR-TEM micrographs of 7 nm (T)-RuO2 and 5 nm (P)-RuO2 NPs, together with Fast Fourier Transform (FFT). In the

60

E.A. Paoli et al. / Catalysis Today 262 (2016) 57–64

Fig. 2. XPS spectra of Ru 3d core level of oxygen plasma treated nanoparticles. (a) 4 nm (P)-RuO2 NPs; and (b) 7 nm (P)-RuO2 NPs.

Fig. 3. (a) HR-TEM micrograph of 7 nm (T)-RuO2 NPsR-TEM highlighting an atomic resolution region within the red circle. (b) FFT of the area within the red circle in Fig. 3(a). (c) HR-TEM micrograph of 5 nm (P)-RuO2 NPs HR-TEM highlighting an atomic resolution region within the red circle. (d) FFT of the area within the red circle in Fig. 3(c).

micrograph of (T)-RuO2 (Fig. 3a) a high atomic resolution area is highlighted in the red circle. Of this region, a Fast Fourier Transform is represented in Fig. 3b, where the values uniquely match with ruthenium dioxide, confirming again the full oxidation of the thermally annealed nanoparticles, in accordance with previous XPS and GA-XRD. Fig. 3c,d represent a micrograph of an atomically resolved (P)-RuO2 nanoparticle and its Fast Fourier Transform, respectively. The FFT of the plasma treated nanoparticles could match only with metallic Ru. From this result and from XPS, it seems that there is a metallic core and an oxidized surface.

3.2. Oxygen evolution activity The activity was evaluated from the first cyclic voltammogram, corrected for capacitance and Ohmic losses. The corrected cyclic voltammetry of the thermally oxidized and plasma treated NPs are shown in Fig. 4a. In agreement with literature, the pre-treated catalysts exhibit lower activity in comparison to the freshly prepared catalysts [48]. The OER mass activity extracted from the CV at 0.25 V overpotential (1.48 V vs. RHE) is displayed as a function of the equivalent diameter and the single particle mass in Fig. 4b. The error bars are based on three independent measurements, whereas the other values are based on single points.

E.A. Paoli et al. / Catalysis Today 262 (2016) 57–64

61

Fig. 4. (a) Capacitance and Ohmic corrected cyclic voltammogram, first cycle, recorded at 20 mV/s between 1.0 and 1.5 V vs. RHE in N2 saturated 0.05 M H2 SO4 solution, for 5 nm oxygen plasma treated RuO2 NPs and 5 nm thermally oxidized RuO2 NPs. (b) Mass Activity of as-deposited, plasma treated (P) and thermal treated (T) nanoparticles as a function of nanoparticle equivalent diameter. The mass activity was evaluated at 1.48 V vs. RHE (0.25 overpotential) from the first capacitance corrected cyclic voltammogram.

Fig. 5. (a) OER specific activity for thermally oxidised RuO2 NPs vs. the particle equivalent diameter. The activity was evaluated from the first capacitance and Ohmic corrected cyclic voltammogram at 1.48 V (vs. RHE). The CVs were recorded at 20 mV/s in N2 saturated 0.05 M H2 SO4 . The surface area was estimated assuming the particles to be spherical. b) Specific activity comparison for RuO2 NPs from this work and the current state of the art for well-defined RuO2 surfaces. Data adapted from [39] for RuO2 NPs, from [56] for epitaxially grown RuO2 (1 0 0) in alkaline solution.

As deposited Ru NPs are almost an order of magnitude more active than both types of pre-treated nanoparticles, (T)-RuO2 and (P)-RuO2 . However, due to their more irregular shape, it is harder to precisely evaluate the intrinsic catalytic activity of as deposited nanoparticles. The plasma treated NPs exhibit activity closer to the thermally oxidized NPs, in particular for 3 and 9 nm the activities are similar. Plasma treated particles between 4 and 8 nm instead show better performance than the thermally treated particles of same size, although still below the activity of the as-deposited NPs. Specifically, 5 nm seems to be the most active. The great advantage of investigating mono-dispersed mass–selected nanoparticles is the possibility of precisely controlling the particle coverage and accurately estimating the mass loading. Previous studies on similar Ru particles in our laboratory showed that smaller particles, below 5–7 nm [37,75] tend to be spherical; this allows us to estimate the surface area of the surface area. Fig. 5a shows the surface area normalized current at an overpotential of 0.25 V for thermally oxidized nanoparticles as a function of equivalent diameter (9 nm NPs are also included for comparison). The activity increases as the particle size increases. An analogous trend, together with the trend of the mass activity,

Fig. 6. Turnover frequency comparison of 3 nm thermally oxidized RuO2 NPs and 5 nm plasma treated RuO2 NPs from this work with the current state of the art for well-defined surfaces. Data adapted from [33] for Ir NPs and IrNi3.3 NPs; from [39] for RuO2 NPs and IrO2 NPs; and from [56] for epitaxially grown RuO2 (1 0 0) in alkaline.

62

E.A. Paoli et al. / Catalysis Today 262 (2016) 57–64

where we tentatively identify a maximum at around 3–5 nm, was observed and predicted on Pt nanoparticles catalysing ORR [25,76,77]. This resemblance suggests that, similarly to ORR, for the oxygen evolution reaction the terrace sites are the most active. The number of terrace sites would follow the same trend and exhibit a maximum for a specific particle size. In Fig. 5b the specific activity of thermally oxidized and plasma treated particles are plotted together with the current state of the art for well-defined RuO2 [39,56]. The specific activity of the pretreated nanoparticles from this work exhibit an order of magnitude improvement in comparison to epitaxially grown RuO2 (1 1 0) in alkaline media and rutile RuO2 nanoparticles prepared by chemical synthesis [39,56]. Among the different pre-treated NPs, the plasma treated NPs have a slightly higher specific activity and lower Tafel slope, although the activity for the 3 nm are almost overlapping. Moreover, the surface area and mass loading can be used to estimate a range for the turnover frequency (TOF). It is challenging to precisely determine the number of active sites; however a lower and upper limit can be identified. TOFmin is calculated assuming all the Ru atoms are involved in the reaction; TOFmax is obtained assuming only the surface is active. This is a cautious estimation since all the atoms at the surface are considered active, although the measured current may be completely dominated by the most active sites. Fig. 6 compares the TOF of the most active pre-treated NPs from this work with the current state of the art for precious metal catalysts [33,39,56]. Again, the particles from this work exhibit the highest performance, with the 5 nm plasma treated NPs being the most active. Recently, Stoerzinger et al. reported an order of magnitude activity improvement for the more open (1 0 0) structure compared to the (1 1 0) facet for rutile RuO2 and IrO2 films, and proposed that it was due to a higher number of active sites.[56] The improved turnover frequency of the nanoparticles suggests that they possess a site that is more active than the oriented thin films; whether or not this is due to pH effects (the films were tested in alkaline electrolytes) or the inherent catalytic properties of the different surfaces is an open question. 3.3. Stability measurements While measuring the activity on the disk electrodes, the corrosion of ruthenium dioxide to RuO4 species was followed on a platinum ring [74]. By holding the potential at 1.1 V (vs. RHE) the corrosion product can be collected and reduced at the ring. The Rotating Ring Disk Electrode measurements for the as-deposited, thermally oxidized and plasma treated nanoparticles are shown in Fig. 7. As expected, the stability of the as-deposited nanoparticles, is extremely poor, and the RuO4 species formation accounts for 15% of the total disk current (Fig. 7a). This is in agreement with recent reports by Reier et al., who observed that the durability of nanoparticles in the range between 4 and 6 nm is much lower than extended surfaces [57]. The plasma-treated nanoparticles exhibit intermediate stability, as well as activity (Fig. 7b). Specifically, the current due to corrosion is around 8%, which is almost half the value of the as-deposited catalysts. Finally, the thermally oxidized NPs display the lowest activity among the three, but at the same time they are the most stable ones (Fig. 7c). The ring current is low, however a corrosion efficiency of 1% ca can be estimated. This stability performance are not good enough for real applications in PEM electrolyzers and further improvements will need to be addressed. Nevertheless, these results suggest that a proper oxidation pre-treatment can have drastic effect on both stability and activity and a trade-off between the two parameters could be found.

Fig. 7. Rotating Ring Disk Electrode (RRDE) measurement of (a) as-deposited 9 nm Ru NPs; (b) plasma treated 8 nm RuO2 nanoparticles; (c) thermally oxidised 9 nm RuO2 NPs. All measurements were carried out at 20 mV/s in N2 saturated 0.05 M H2 SO4 , and based on the first cycle. The potential at the disk was scanned between 1.0 V and 1.5 V (vs. RHE), while the ring was held at 1.1 V (vs. RHE). The voltammogram was then normalised for the capacitance and Ohmic losses.

4. Conclusion In this paper we have focused on the effect of the oxidation treatment on the activity and stability of mass-selected nanoparticles. Mono-disperse mass-selected nanoparticles provide a means by which we can measure the intrinsic catalytic activity of OER catalysts. We have compared the activity and stability of as-deposited, thermally oxidized and oxygen plasma treated ruthenium nanoparticles, which have been prepared by a physical method. We observed that as deposited NPs are the most active but at the same time the most unstable particles, in accordance with literature [43,48,51]. On the other hand, the thermally oxidized particles are the most stable but least active one; even so, they exhibit over an order of magnitude improvement over the current state of the art. Finally, the plasma treated nanoparticles display intermediate performance, with an activity slightly higher than the thermally oxidized nanoparticles and an improved stability compared to the freshly prepared NPs. These findings suggest that depending on the oxidation method and condition, the activity and stability of Ru can be tuned. This system shows that there is further scope to improve the activity of a well-known and widely investigated catalyst, such as ruthenium. In particular, the future challenges will be related to

E.A. Paoli et al. / Catalysis Today 262 (2016) 57–64

replicate such high activity as the one reported here at greater overpotentials, relevant for applications in real electrolysers. To do so, the stability of ruthenium dioxide needs to be improved, e.g. by addition of a more stable metal such as Ir or Ti [78–82]. A different strategy would be to discover compounds based on nonnoble metals, which would be stable in acidic media and exhibit reasonable OER activity [83]. This would ultimately improve the widespread market of PEM electrolyzers, and any other type of device that is hindered by this reaction. Acknowledgments The authors gratefully acknowledge financial support from the Danish Ministry of Science’s UNIK initiative, Catalysis for Sustainable Energy. The Center for Individual Nanoparticle Functionality is supported by the Danish National Research Foundation (DNRF54). References [1] BP, BP Statistical Review of World Energy, 2014. [2] R.E. Smalley, Future Global Energy Prosperity: The Terawatt Challenge, MRS Bull. 30 (2004) 412–417. [3] P.C.K. Vesborg, T.F. Jaramillo, Addressing the terawatt challenge: scalability in the supply of chemical elements for renewable energy, RSC Adv. 2 (2012) 7933–7947. [4] J. Newman, P.G. Hoertz, C.a. Bonino, J.a. Trainham, Review: an economic perspective on liquid solar fuels, J. Electrochem. Soc. 159 (2012) A1722–A1729. [5] E. Fabbri, A. Habereder, K. Waltar, R. Kotz, T. Schmidt, Developments and perspectives of oxide-based catalysts for the oxygen evolution reaction, Catal. Sci. Technol. 4 (2014) 38000. [6] H. Dau, C. Limberg, T. Reier, M. Risch, S. Roggan, P. Strasser, The mechanism of water oxidation: from electrolysis via homogeneous to biological catalysis, ChemCatChem 2 (2010) 724–761. [7] M.G. Walter, E.L. Warren, J.R. McKone, S.W. Boettcher, Q. Mi, E. a Santori, N.S. Lewis, Solar water splitting cells, Chem. Rev. 110 (2010) 6446–6473. [8] S. Kotrel, S. Bräuninger, Industrial electrocatalysis, in: G. Ertl, H. Knôzinger, F. Schuth, J. Weitkamp (Eds.), Handbook of Heterogenous Catalysis, 2nd ed., Wiley-CPH, 2008, pp. 1936–1958. [9] K. Maeda, K. Domen, Photocatalytic water splitting: recent progress and future challenges, J. Phys. Chem. Lett. 1 (2010) 2655–2661. [10] M.K. Debe, S.M. Hendricks, G.D. Vernstrom, M. Meyers, M. Brostrom, M. Stephens, Q. Chan, J. Willey, M. Hamden, C.K. Mittelsteadt, C.B. Capuano, K.E. Ayers, E.B. Anderson, Initial Performance, Durability of ultra-low loaded NSTF electrodes for PEM electrolyzers, J. Electrochem. Soc. 159 (2012) K165–K176. [11] K.C. Neyerlin, W. Gu, J. Jorne, H.A. Gasteiger, Study of the exchange current density for the hydrogen oxidation and evolution reactions, J. Electrochem. Soc. 154 (2007) B631. [12] J. Kibsgaard, T.F. Jaramillo, F. Besenbacher, Building an appropriate active-site motif into a hydrogen-evolution catalyst with thiomolybdate [Mo3S13]2clusters, Nat. Chem. 6 (2014) 248–253. [13] P.C.K. Vesborg, B. Seger, I. Chorkendorff, Recent development in hydrogen evolution reaction catalysts and their practical implementation, J. Phys. Chem. Lett. (2015) 951–957. [14] E.J. Popczun, C.G. Read, C.W. Roske, N.S. Lewis, R.E. Schaak, Highly active electrocatalysis of the hydrogen evolution reaction by cobalt phosphide nanoparticles, Angew. Chem. Int. Ed. 53 (2014) 5427–5430. [15] E.J. Popczun, J.R. Mckone, C.G. Read, A.J. Biacchi, A.M. Wiltrout, N.S. Lewis, R.E. Schaak, Nanostructured nickel phosphide as an electrocatalyst for the hydrogen evolution reaction, J. Am. Soc. 135 (2013) 9267–9270. [16] A.B. Laursen, K.R. Patraju, M.J. Whitaker, M. Retuerto, T. Sarkar, N. Yao, K.V. Ramanujachary, M. Greenblatt, G.C. Dismukes, Nanocrystalline Ni 5 P 4: a hydrogen evolution electrocatalyst of exceptional efficiency in both alkaline and acidic media, Energy Environ. Sci. 8 (2015) 1027–1034. [17] B. Hinnemann, P.G. Moses, J. Bonde, K.P. Jørgensen, J.H. Nielsen, S. Horch, I. Chorkendorff, J.K. Nørskov, Biomimetic hydrogen evolution: MoS2 nanoparticles as catalyst for hydrogen evolution, J. Am. Chem. Soc. 127 (2005) 5308–5309. [18] W. Sheng, A.P. Bivens, M. Myint, Z. Zhuang, R.V. Forest, Q. Fang, J.G. Chen, Y. Yan, Non-precious metal electrocatalysts with high activity for hydrogen oxidation reaction in alkaline electrolytes, Energy Environ. Sci. 7 (2014) 1719. [19] D. Merki, H. Vrubel, L. Rovelli, S. Fierro, X. Hu, Fe, Co, and Ni ions promote the catalytic activity of amorphous molybdenum sulfide films for hydrogen evolution, Chem. Sci. 3 (2012) 2515. [20] J.D. Benck, T.R. Hellstern, J. Kibsgaard, P. Chakthranont, T.F. Jaramillo, Catalyzing the Hydrogen Evolution Reaction (HER) with Molybdenum Sul fi de Nanomaterials, ACS Catal. (2014). [21] K.E. Ayers, E.B. Anderson, C. Capuano, B. Carter, L. Dalton, G. Hanlon, J. Manco, M. Niedzwiecki, Research advances towards low cost, high efficiency PEM electrolysis, ECS Trans. 33 (2010) 3–15.

63

[22] K.E. Ayers, E.B. Anderson, M. Hickner, C.-Y. Wang, Characterization of anion exchange membrane technology for low cost electrolysis by K.E Ayers, ECS Trans. 45 (2013) 121–130. [23] J.R. Varcoe, P. Atanassov, D.R. Dekel, A.M. Herring, M.A. Hickner, P.A. Kohl, R. Kucernak, W.E. Mustain, K. Nijmeijer, K. Scott, T. Xu, L. Zhuang, Anion-exchange membranes in electrochemical energy systems, Energy Environ. Sci. 7 (2014) 3135–3191. [24] J. Durst, A. Siebel, C. Simon, F. Hasché, J. Herranz, H.a. Gasteiger, New insights into the electrochemical hydrogen oxidation and evolution reaction mechanism, Energy Environ. Sci. 7 (2014) 2255–2260. [25] F.J. Perez-Alonso, D.N. McCarthy, A. Nierhoff, P. Hernandez-Fernandez, C. Strebel, I.E.L. Stephens, J.H. Nielsen, I. Chorkendorff, The effect of size on the oxygen electroreduction activity of mass-selected platinum nanoparticles, Angew. Chem. Int. Ed. Engl. 51 (2012) 4641–4643. [26] P. Hernandez-Fernandez, F. Masini, D.N. McCarthy, C.E. Strebel, D. Friebel, D. Deiana, P. Malacrida, A. Nierhoff, A. Bodin, A.M. Wise, J.H. Nielsen, T.W. Hansen, A. Nilsson, I.E.L. Stephens, I. Chorkendorff, Mass-selected nanoparticles of PtxY as model catalysts for oxygen electroreduction, Nat. Chem. 6 (2014) 732–738. [27] B.R. Cuenya, S.-H. Baeck, T.F. Jaramillo, E.W. McFarland, Size- and support-dependent electronic and catalytic properties of Au0/Au3+ nanoparticles synthesized from block copolymer micelles, J. Am. Chem. Soc. 125 (2003) 12928–12934. [28] R. Forgie, G. Bugosh, K.C. Neyerlin, Z. Liu, P. Strasser, Bimetallic Ru electrocatalysts for the OER and electrolytic water splitting in acidic media, Electrochem. Solid-State Lett. 13 (2010) B36–B39. ´ M. Makarova, P. Krtil, Particle size dependence of oxygen [29] J. Jirkovsky, evolution reaction on nanocrystalline RuO2 and Ru0.8 Co0.2 O2−x , Electrochem. Commun. 8 (2006) 1417–1422. ´ M.V. Makarova, J. Franc, P. Krtil, Oxygen evolution [30] K. Macounová, J. Jirkovsky, on Ru1−x Nix O2−y nanocrystalline electrodes, J. Solid State Electrochem. 13 (2009) 959–965. [31] V. Petrykin, K. Macounova, O.a. Shlyakhtin, P. Krtil, Tailoring the selectivity for electrocatalytic oxygen evolution on ruthenium oxides by zinc substitution, Angew. Chem. Int. Ed. Engl. 49 (2010) 4813–4815. [32] V. Petrykin, K. Macounová, M. Okube, S. Mukerjee, P. Krtil, Local structure of Co doped RuO2 nanocrystalline electrocatalytic materials for chlorine and oxygen evolution, Catal. Today. 202 (2013) 63–69. [33] H.N. Nong, L. Gan, E. Willinger, D. Teschner, P. Strasser, IrOx core-shell nanocatalysts for cost- and energy-efficient electrochemical water splitting, Chem. Sci. 5 (2014) 2955–2963. [34] H.N. Nong, H.-S. Oh, T. Reier, E. Willinger, M.-G. Willinger, V. Petkov, D. Teschner, P. Strasser, Oxide-supported IrNiOx core-shell particles as efficient, cost-effective, and stable catalysts for electrochemical water splitting, Angew. Chem. 127 (2015) 3018–3022. [35] H. Oh, H.N. Nong, T. Reier, M. Gliech, P. Strasser, Oxide-supported Ir nanodendrites with high activity and durability for the oxygen evolution reaction in acid PEM water electrolyzers, Chem. Sci. (2015), in press. [36] J. Greeley, I.E.L. Stephens, a.S. Bondarenko, T.P. Johansson, H. a Hansen, T.F. Jaramillo, J. Rossmeisl, I. Chorkendorff, J.K. Nørskov, Alloys of platinum and early transition metals as oxygen reduction electrocatalysts, Nat. Chem. 1 (2009) 552–556. [37] E.A. Paoli, F. Masini, R. Frydendal, D. Deiana, C. Schlaup, M. Malizia, T.W. Hansen, S. Horch, I.E.L. Stephens, I. Chorkendorff, Oxygen evolution on well-characterized mass-selected Ru and RuO 2 nanoparticles, Chem. Sci. 6 (2015) 190–196. [38] N. Danilovic, R. Subbaraman, K.C. Chang, S.H. Chang, Y. Kang, J. Snyder, A.P. Paulikas, D. Strmcnik, Y.T. Kim, D. Myers, V.R. Stamenkovic, N.M. Markovic, Using surface segregation to design stable Ru-Ir oxides for the oxygen evolution reaction in acidic environments, Angew. Chem. (2014). [39] Y. Lee, J. Suntivich, K.J. May, E.E. Perry, Y. Shao-horn, Synthesis and activities of rutile IrO2 and RuO2 nanoparticles for oxygen evolution in acid and alkaline solutions, J. Phys. Chem. Lett. 3 (2012) 399–404. [40] S. Trasatti, Electrocatalysis in the anodic evolution of oxygen and chlorine, Electrochim. Acta 29 (1984) 1503–1512. [41] D. Galizzioli, F. Tantardini, S. Trasatti, Ruthenium dioxide: a new electrode material. I. Behaviour in acid solutions of inert electrolytes, J. Appl. Electrochem. 4 (1974) 57–67. [42] L.D. Burke, O.J. Murphy, J.F. O´ıNeill, S. Venkatesan, The Oxygen Electrode. Part 8- Oxygen evolution at ruthenium dioxide anodes, J. Chem. Soc. 73 (1977) 1659–1671. [43] N. Danilovic, R. Subbaraman, K.-C. Chang, S.H. Chang, Y. Kang, J.D. Snyder, A.P. Paulikas, D. Strmcnik, Y.-T. Kim, D.J. Myers, V.R. Stamenkovic, N.M. Markovic, Activity-stability trends for the oxygen evolution reaction on monometallic oxides in acidic environments, J. Phys. Chem. Lett. 5 (2014) 2474–2478. [44] C. Iwakura, K. Hirao, H. Tamura, Preparation of ruthenium dioxide electrodes and their anodic polarization characteristics in acidic solutions, Electrochim. Acta 22 (1977) 335–340. [45] S. Ardizzone, M. Falciola, S. Trasatti, Effect of the nature of the precursor on the electrocatalytic properties of thermally prepared ruthenium oxide, J. Electrochem. Soc. 136 (1989) 3–8. [46] E. Tsuji, A. Imanishi, K. Fukui, Y. Nakato, Electrocatalytic activity of amorphous RuO2 electrode for oxygen evolution in an aqueous solution, Electrochim. Acta 56 (2011) 2009–2016.

64

E.A. Paoli et al. / Catalysis Today 262 (2016) 57–64

[47] G. Lodi, E. Sivieri, A. De Battisti, S. Trasatti, Ruthenium dioxide-based film electrodes. III. Effect of chemical composition and surface morphology on oxygen evolution in acid solutions, J. Appl. Electrochem. 8 (1978) 135–143. [48] M. Vukovic, Oxygen evolution on an electrodeposited ruthenium electrode in acid solution - the effect of thermal treatment, Electrochim. Acta 34 (1989) 287–291. [49] N. Danilovic, R. Subbaraman, K.-C. Chang, S.H. Chang, Y.J. Kang, J.D. Snyder, A.P. Paulikas, D. Strmcnik, Y.-T. Kim, D.J. Myers, V.R. Stamenkovic, N.M. Markovic, Activity–stability trends for the oxygen evolution reaction on monometallic oxides in acidic environments, J. Phys. Chem. Lett. 5 (2014) 2474–2478. [50] M.B. Vukmirovic, R.L. Sabatini, R.R. Adzic, Growth of RuO2 by electrochemical and gas-phase oxidation of an Ru(0001) surface, Surf. Sci. 572 (2004) 269–276. [51] R. Kötz, H.J. Lewerenz, S. Stucki, XPS studies of oxygen evolution on Ru and Ru02 anodes, J. Electrochem. Soc. 130 (1983) 825–828. [52] S. Cherevko, A.R. Zeradjanin, A.A. Topalov, N. Kulyk, J.J. Mayrhofer, Dissolution of noble metals during oxygen evolution in acidic media, ChemCatChem Commun. 6 (2014) 2219–2223. [53] S. Cherevko, T. Reier, A.R. Zeradjanin, Z. Pawolek, P. Strasser, K.J.J. Mayrhofer, Stability of nanostructured iridium oxide electrocatalysts during oxygen evolution reaction in acidic environment, Electrochem. Commun. 48 (2014) 81–85. [54] C.C.L. McCrory, S. Jung, J.C. Peters, T.F. Jaramillo, Benchmarking heterogeneous electrocatalysts for the oxygen evolution reaction, J. Am. Chem. Soc. 135 (2013) 16977–16987. [55] C.C.L. McCrory, S. Jung, I.M. Ferrer, S. Chatman, J.C. Peters, T.F. Jaramillo, H.E.R. Benchmarking, OER Electrocatalysts for Solar Water Splitting Devices, J. Am. Chem. Soc. (2015). [56] K.A. Stoerzinger, L. Qiao, M.D. Biegalski, Y. Shao-horn, Orientation-dependent oxygen evolution activities of rutile IrO2 , J. Phys. Chem. Lett. 5 (2014) 1636–1641. [57] T. Reier, M. Oezaslan, P. Strasser, Electrocatalytic oxygen evolution reaction (OER) on Ru, Ir, and Pt catalysts: a comparative study of nanoparticles and bulk materials, ACS Catal. 2 (2012) 1765–1772. [58] S. Cherevko, A.R. Zeradjanin, A.A. Topalov, N. Kulyk, I. Katsounaros, K.J.J. Mayrhofer, Dissolution of noble metals during oxygen evolution in acidic media, ChemCatChem. 6 (2014) 2219–2223. [59] W. Hong, M. Risch, K. a Stoerzinger, A.J.L. Grimaud, J. Suntivich, Y. Shao-Horn, Toward the rational design of non-precious transition metal oxides for oxygen electrocatalysis, Energy Environ. Sci. 8 (2015) 1404–1427. [60] I. Katsounaros, S. Cherevko, A.R. Zeradjanin, K.J.J. Mayrhofer, Oxygen electrochemistry as a cornerstone for sustainable energy conversion, Angew. Chem. Int. Ed. 52 (2013) 2–22. [61] S. Cherevko, A.A. Topalov, A.R. Zeradjanin, I. Katsounaros, K.J.J. Mayrhofer, Gold dissolution: towards understanding of noble metal corrosion, RSC Adv. 3 (2013) 16516. [62] R. Frydendal, E.A. Paoli, B.P. Knudsen, B. Wickman, P. Malacrida, I.E.L. Stephens, I. Chorkendorff, Benchmarking stability for oxygen evolution reaction catalysts: the importance of monitoring mass losses, ChemElectroChem. 1 (2014) 2075–2080. [63] O. Diaz-Morales, F. Calle-Vallejo, C. de Munck, M.T.M. Koper, Electrochemical water splitting by gold: evidence for an oxide decomposition mechanism, Chem. Sci. 4 (2013) 2334–2343. [64] K.J. May, C.E. Carlton, K.a. Stoerzinger, M. Risch, J. Suntivich, Y.-L. Lee, A. Grimaud, Y. Shao-Horn, Influence of oxygen evolution during water oxidation on the surface of perovskite oxide catalysts, J. Phys. Chem. Lett. 3 (2012) 3264–3270.

[65] S.H. Chang, N. Danilovic, K.-C. Chang, R. Subbaraman, A.P. Paulikas, D.D. Fong, M.J. Highland, P.M. Baldo, V.R. Stamenkovic, J.W. Freeland, J.a. Eastman, N.M. Markovic, Functional links between stability and reactivity of strontium ruthenate single crystals during oxygen evolution, Nat. Commun. 5 (2014) 4191. [66] A.S. Bandarenka, E. Ventosa, A. Maljusch, J. Masa, W. Schuhmann, Techniques and methodologies in modern electrocatalysis: evaluation of activity, selectivity and stability of catalytic materials, Analyst 139 (2014) 1274–1291. [67] F. Masini, P. Hernández-Fernández, D. Deiana, C.E. Strebel, D.N. McCarthy, A. Bodin, P. Malacrida, I. Stephens, I. Chorkendorff, Exploring the phase space of time of flight mass selected Pt x Y nanoparticles, Phys. Chem. Chem. Phys. (2014). [68] A. Velázquez-palenzuela, F. Masini, A.F. Pedersen, M. Escudero-escribano, D. Deiana, P. Malacrida, T.W. Hansen, D. Friebel, A. Nilsson, I.E.L. Stephens, I. Chorkendorff, The enhanced activity of mass-selected Pt x Gd nanoparticles for oxygen electroreduction 328 (2015) 297–307. [69] H. Over, A.P. Seitsonen, E. Lundgren, M. Smedh, J.N. Andersen, On the origin of the Ru-3d5/2 satellite feature from RuO2 (110), Surf. Sci. 504 (2002) L196–L200. [70] W. Chen, M. Zhang, D.W. Zhang, S.-J. Ding, J.-J. Tan, M. Xu, X.-P. Qu, L.-K. Wang, Growth of high-density Ru- and RuO2 -composite nanodots on atomic-layer-deposited Al2 O3 film, Appl. Surf. Sci. 253 (2007) 4045–4050. [71] S. Bhaskar, P.S. Dobal, S.B. Majumder, R.S. Katiyar, X-Ray photoelectron spectroscopy and micro-Raman analysis of conductive RuO2 thin films, J. Appl. Phys. 89 (2001) 2987–2992. [72] D. Rochefort, P. Dabo, D. Guay, P.M.A. Sherwood, XPS investigations of thermally prepared RuO2 electrodes in reductive conditions, Electrochim. Acta 48 (2003) 4245–4252. ˇ c, ´ J. Tymoczko, A. Maljusch, A. Ganassin, W. Schuhmann, A.S. [73] V. Coli Bandarenka, Experimental aspects in benchmarking of the electrocatalytic activity, ChemElectroChem 2 (2015) 143–149. [74] M. Vukovic, Rotating ring-disc electrode study of the enhanced oxygen evolution on an activated ruthenium electrode, J. Chem. Soc. Faraday Trans. 86 (1990) 3743–3746. [75] F. Masini, C.E. Strebel, D.N. McCarthy, A.U.F. Nierhoff, J. Kehres, E.M. Fiordaliso, J.H. Nielsen, I. Chorkendorff, Methanation on mass-selected Ru nanoparticles on a planar SiO2 model support: The importance of under-coordinated sites, J. Catal. 308 (2013) 282–290. [76] G.A. Tritsaris, J. Greeley, J. Rossmeisl, J.K. Nørskov, Atomic-scale modeling of particle size effects for the oxygen reduction reaction on Pt, Catal. Lett. 141 (2011) 909–913. [77] J. Greeley, R.J., A. Hellman, J.K. Nørskov, Theoretical trends in particle size effects for teh oxygen reduction reaction, Zeitschrift Fur Phys. Chemie-International J. Res. Phys. Chem. Chem. Phys. 221 (2007) 1209–1220. [78] R. Kötz, S. Stucki, Stabilization of RuO2 by IrO2 for anodic oxygen evolution in acid media, Electrochim. Acta 31 (1986) 1311–1316. [79] H.B. Beer, The invention and industrial development of metal anodes, J. Electrochem. Soc. 127 (1980) 303C–307C. [80] S. Trasatti, Electrocatalysis: understanding the success of DSA® , Electrochim. Acta 45 (2000) 2377–2385. [81] C. Comninellis, P.G. Vercesi, Problems in DSA | coating deposition by thermal decomposition, J. Appl. Electrochem. 21 (1991) 136–142. [82] C. Comninellis, G.P. Vercesi, Characterization of DSA | electrodes: choice of a coating, J. Appl. Electrochem. 21 (1991) 335–345. [83] R. Frydendal, E.A. Paoli, I. Chorkendorff, J. Rossmeisl, I.E.L. Stephens, Towards an active and stable catalyst for oxygen evolution in acidic media: Ti-stabilized MnO2 , Adv. Energy Mater. (2015), http://dx.doi.org/10.1002/ aenm.201500991, accepted.