Nanoscale Engineering of Efficient Oxygen Reduction Electrocatalysts

Engineering of Efficient Oxygen Reduction Electrocatalysts by Tailoring the Local Chemical Environment of Pt Surface Sites. Tim Van Cleve† , Sam...
0 downloads 0 Views 1MB Size
Subscriber access provided by University of Otago Library

Article

Nanoscale Engineering of Efficient Oxygen Reduction ElectroCatalysts by Tailoring Local Chemical Environment of Pt Surface Sites Tim Van Cleve, Saman Moniri, Gabrielle Belok, Karren L More, and Suljo Linic ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b01565 • Publication Date (Web): 16 Nov 2016 Downloaded from http://pubs.acs.org on November 17, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Nanoscale Engineering of Efficient Oxygen Reduction ElectroCatalysts by Tailoring Local Chemical Environment of Pt Surface Sites Tim Van Cleve1, Saman Moniri1, Gabrielle Belok1, Karren L. More2, Suljo Linic1* 1 2

Department of Chemical Engineering, University of Michigan, Ann Arbor, Michigan 48109, USA Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA

ABSTRACT: Oxygen reduction reaction is the limiting half-reaction in hydrogen fuel cells. While Pt is the most active single component electro-catalyst for the reaction, it is hampered by high cost and low reaction rates. Most research to overcome these limitations has focused on Pt/3d alloys, which offer higher rates and lower cost. Herein, we synthesized, characterized and tested alloy materials belonging to a multilayer family of electro-catalysts. The multilayer alloy materials contain AuCu alloy core of precise composition, surrounded by Au layers and covered by a catalytically active Pt surface layer. Their performance relative to that of the commercial Pt standards reaches up to four times improved area-specific activity. Characterization studies support the hypothesis that the activity improvement originates from a combination of Au-Pt ligand effect and local strain effect manipulated through the AuCu alloy core. The presented approach to control the strain and ligand effects in the synthesis of Pt-based alloys for ORR is very general and it could lead to promising alloy materials.

KEYWORDS: oxygen reduction, Platinum alloy catalysts, core-shell nanoparticles, fuel cells, electro-catalysis

INTRODUCTION Low temperature proton exchange membrane fuel cells are promising devices for portable power generation and transportation applications.1 A critical obstacle to the commercialization of this technology is the significant voltage loss associated with the oxygen reduction reaction (ORR), O2 + 4(H+ + e -) → 2 H2O, even when costly Pt nanoparticle electro-catalysts are used.2,3 In practical terms this means that significant amounts of Pt metal are required to achieve high current and power densities, making the overall cost prohibitively high.2 Techno-economic analysis suggests that the technology would be commercially viable if the rates of ORR on Pt-based nanoparticle electro-catalysts were higher by ~ 2-10 times on Pt mass basis compared to best performing pure Pt electrocatalysts which contain Pt nanoparticles with 3 – 5 nm diamter.2,4–6 One way to accomplish this objective is by designing Pt alloy electro-catalysts that are more active that pure Pt. Alternatively, significant efforts have also been made to develop Pt-free ORR electro-catalysts including metal oxides, nitrides, oxynitrides, carbonitrides, chalcogenides, as well as solid carbon based materials.7–9 While these materials offer some advantages, in general their activities are not comparable to the Pt-based electrocatalysts and they exhibit limited stability. Studies have shown that a key descriptor of the ORR activity of Pt-based electro-catalysts is the relative binding energy of the OH adsorbate.10–15 It has been demonstrated that the optimal Pt-based catalysts (meeting the previously mentioned techno-economic target) should have a high concentration of catalytic surface sites that bind OH approximately 0.1 eV more weakly compared to the Pt(111) surface.3,11,13,14,15,16 It has been shown that one approach to accomplish this optimal OH binding is to synthesize Pt/3d metal alloys.17–21 In particular, Pt/Co and Pt/Ni alloys have received considerable atten-

tion.17,19,21–23 The models used to describe the active Pt/3d surface sites include a monolayer of pure Pt on a Pt/3d alloy as shown in Figure 1.18 Various nano-architectures of these alloys, including nanoparticles, de-alloyed mesoporous networks, skeletal nano-structures and continuous metal films have been tested showing improved performances.24,25 Due to a different interplay of the kinetic ORR rates, the O2 transport to the electro-catalyst surface, the diffusion of O2 to the active centers through the electro-catalysts’ pores and potentially even different geometries of active surface sites (2-D vs. 3-D) for the different Pt/3d nano-architectures, it is difficult to conclusively compare their inherent (kinetic) performance with the performance of pure Pt commercial standards (on per active site basis). On the other hand, it has been demonstrated that when nanoparticles of Pt/3d materials are compared to the nanoparticles of pure Pt of similar size, the Pt/3d alloys show 2 – 3 times enhancements in the ORR rates on a per active surface site basis at potentials of interest (between 0.7 and 0.9 V vs. the reversible hydrogen electrode, RHE).24–30 It has also been discussed that a possibly significant obstacle to the deployment of Pt/3d metals is their potential lack of stability in highly acidic, high potential conditions of ORR.31 The extensive focus on the studies of the Pt/3d metal alloys has been stimulated by the lack of other Pt alloy compositions that can meet the required criterion of the elevated ORR rates compared to pure Pt. It would be beneficial to expand the family of the Pt alloy materials that can meet or exceed the abovementioned techno-economic targets beyond the Pt/3d alloys. It is well established that the activity of the surface sites of metal nanoparticles (e.g. Pt sites) can be modified by a manipulation of the local ligand and strain effects, which can be accomplished by changing the local chemical environment of the surface site through the formation of specific metal alloy nanostructure. Herein, we apply this concept to design alloy

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(a)

(c)

(e)

(b)

(d)

(f)

Page 2 of 9

Figure 1. Representations of Pt3Ni and AuCu@Au2ML@Pt electrocatalysts as nanoparticles (a,c) and (111) surface slab (b, d). Gray: Pt. Green: Ni. Brown: Cu. Yellow: Au. e) DFT-calculated OH binding energies as a function of the x-y lattice constant on various slab terminated by the (111) surface of Pt. The AuxCuy@Au2ML@Pt structure corresponds to materials with AuxCuy alloy core with two atomic layers of Au adsorbed directly on top of alloy layers and with a Pt monolayer adsorbed on the surface (directly above subsurface Au layers). f) Relationship between molar composition of Au in AuCu alloys and the experimental32 and DFT-calculated lattice parameters. In (e) and (f), the shaded regions highlight alloy surfaces that bind OH by ~0.1 - 0.2 eV weaker than Pt. These are the primary materials of interest.

materials belonging to a new family of alloy electro-catalysts for ORR. These alloys contained an AuCu core of controlled composition surrounded by two Au layers covered by a Pt surface layer. We rigorously tested the ORR polarization behavior of these alloy nanoparticle electro-catalysts using rotating disk electrode (RDE) voltammetry, showing that these materials exhibit up to 4 times higher ORR rates on per surface area basis compared to the commercial Pt nanoparticles (~5 nm diameter). Characterization studies showed that the activity improvement originated from a combination of the local Au-Pt ligand effect and the strain effect manipulated through the AuCu alloy core. The presented approach to control the strain and ligand effects in the synthesis of Pt-based alloys for ORR is very general and it could lead to novel alloy materials. In addition to high activity, another potentially appealing feature of these materials is that the Au interlayers between the Pt surface layer and the core could act as a chemically stable buffer that improves the material stability by protecting the elements in the internal core structure from leaching.

RESULTS AND DISCUSSION DFT Calculations of OH Adsorption Energies on Model and Alloy Systems. The electrochemical reduction of O2 involves four electron/proton transfer steps to produce water. It has been demonstrated that in this process O2 undergoes the

initial electron/proton transfer to form an adsorbed OOH intermediate.33 This intermediate dissociates to generate O and OH intermediates, which are further reduced on the electrocatalyst surface by the subsequent proton/electron transfer steps and removal from the surface as H2O. On Pt electrodes at high operating ORR potentials, OH is the most abundant surface intermediate, effectively blocking catalytic surface sites and preventing O2 activation, thereby lowering the overall rate of ORR. For this reason, a more active metal catalyst would bind OH less strongly relative to Pt without making the other steps kinetically slower. As described above, detailed kinetic analysis has shown that this objective can be accomplished if the Pt surface is changed in a way that it binds OH by approximately 0.1 – 0.2 eV less strongly than Pt(111).11–13,15 We used Density Functional Theory (DFT) calculations to compute the binding energy of OH on various Pt alloy surface sites relative to the Pt(111) surface. The Pt(111) surface was modeled as five layers of Pt with the equilibrium Pt-Pt lattice spacing. The alloys were modeled as the (111) surface terminated slabs containing a surface monolayer of Pt, covering two Au layers on top of two additional Pt layers (Pt2ML@Au2ML@Pt). The top three layers were fully relaxed in the z-direction in all systems (alloys and pure Pt). Data in Figure 1e show the differences in the OH adsorption energy on the Pt2ML@Au2ML@Pt model systems relative to Pt (111) (∆∆EOH=Ealloy+OH-Ealloy-(EPt+OH-EPt)) as a function of the xy

ACS Paragon Plus Environment

Page 3 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

lattice spacing in the alloy (pure Pt slab was at the equilibrium Pt lattice constant). The DFT calculations showed that the OH adsorption energy on these alloys at the equilibrium lattice of Pt is ~ 0.3 eV weaker (less exothermic) than on pure Pt. The change in the OH binding energy in these systems relative to pure Pt is due to the Au-Pt ligand effect. Here electronic interactions between the Pt surface atoms and Au atoms under the Pt surface layer impact the electronic structure of Pt and therefore the OH adsorption energy.12,13,15,34,35,36 We have discussed this phenomenon in our previous studies showing that charge transfer between the surface Pt atoms and subsurface Au atoms induces additional electron-electron repulsion between the OH adsorbate and the Pt surface atoms.37 The problem is that the destabilization of OH in these systems, due to the AuPt ligand effect is greater than desired (> 0.2 eV). This can be modulated by slightly expanding the lattice of the alloy compared to the pure Pt lattice. The data in Figure 1e show that the lattice should be expanded by approximately 2 % relative to pure Pt to reach the optimal OH binding environment. We find that the number of subsurface Au layers used in our model systems and the exact composition of the bottom two layers (pure Pt or a metal alloy) do not affect our conclusions (Figure 1e). The main reason for this is that the sub-surface Au layer effective shields the direct electronic communication between the surface Pt sites and the metal atoms under the Au layer. The bottom two layers control the OH binding energy through the strain effect (i.e., through the control of the Pt-Pt spacing in the Pt surface layer), and their direct effect on the electronic structure of the top Pt layers is shielded by the two Au layers. The question is how to design a practical electro-catalyst that is inspired by the model alloys used in the DFT calculations, i.e., that contains the desired Pt surface sites. To accomplish this, we focused on alloy nanoparticles, where the core of the nanoparticle is designed to control the strain effect (i.e., the lattice spacing) and where Au and Pt are subsequently deposited on this core. We note that the lowest thermodynamic state of spherical metal nanoparticles is characterized by a high fraction of the (111) surface facet.38 To control the lattice spacing we used alloys of highly miscible metals, which can span a large range of lattice spacing, as the core of nanoparticles. A promising alloy to accomplish this objective is AuCu. These two metals form an almost ideal alloy mixture where the lattice spacing scales linearly with the relative Au and Cu concentrations in the alloy. The data in Figure 1f show the relative lattice expansion for AuCu alloys compared to pure Pt (a)

for the experimentally-measured and DFT-calculated lattice parameters as a function of Au mole fraction in the AuCu alloy.32 The data suggest that ~ 2 % expansion of the lattice constant compared to pure Pt is accomplished for AuCu alloys with the Au content between 70 and 90 %. Catalyst Preparation and Characterization. We used a thermal reduction approach to synthesize AuCu alloy core nanoparticles, containing 75 and 85% molar fraction of Au with the balance Cu, supported on carbon (AuCu/C).39 The complete synthesis procedure is described in the Supplementary Information. Prior to the deposition of Au and Pt overlayers, the AuCu core nanoparticles were characterized. The UV-Vis extinction spectra for various compositions of AuCu alloy nanoparticles are shown in Figure 2a. The high wavelength extinction feature at ~510 nm is due to the excitation of localized surface plasmon resonance (LSPR) which is characteristic of Au nanoparticles of ~ 10 nm diameter.40,41 As the content of Cu in the nanoparticles increases, the LSPR peak is red-shifted to higher wavelengths. This LSPR red-shift is an unambiguous characteristic of the formation of AuCu alloy nanoparticles.42 The red shift is accompanied by the increase in the low wavelength extinction due to inter-band (d → sp) electronic transitions in Cu atoms which are of lower energy (higher wavelength) than the corresponding transitions in Au.42 AuCu nanoparticles supported on carbon (AuCu/C) were also characterized using X-ray diffraction (XRD). The diffraction patterns for Au75Cu25, as well as pure Au and Pt nanoparticles, are shown in Figure 2b. The XRD data show no evidence of distinct Au or Cu phases. The diffraction peaks of Au75Cu25 are located between equivalent peaks of Au and Pt samples, suggesting that the alloy has an intermediate lattice parameter. The analysis of (111), (200), and (220) peak locations using Bragg’s law indicated that the lattice constants of Au75Cu25 and Pt samples were ~ 4.03 Å and 3.91 Å, respectively, making the Au75Cu25 lattice about 3 % larger than Pt. Finally, transmission electron microscopy was used to show that alloy samples such as AuCu/C contained relatively monodisperse spherical nanoparticles with an average diameter of ~ 11 nm (Figure S1). The spherical shape of alloy nanoparticles was maintained following the electrochemical deposition of the Au layers and Pt surface layers. The molar composition of the AuCu electro-catalysts cores was measured using ion coupled plasma optical emission spectroscopy, and found to match the nominal metal loadings. (b)

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 9

Figure 2. a) Normalized UV-Vis extinction spectra for various AuCu nanoparticle suspensions in hexane. b) X-ray diffraction patterns for Au, Au75Cu25, and Pt nanoparticles supported on Vulcan XC72R carbon support. (a)

(c)

(b)

Figure 3. a) BF-STEM micrograph depicting a collection of Au75Cu25@Au2ML@Pt nanoparticles on carbon support. b,c) EDS elemental maps show the distribution of Pt, Au, and Cu throughout sample.

After synthesizing the AuCu nanoparticles on carbon, two layers of Au and subsequently one layer of Pt were deposited electrochemically on these nanoparticles by the galvanic replacement of under-potentially deposited Cu layers (Cu UPD).43–45 The polarization behavior of these Pt alloys was tested under ORR conditions as described below. The bright field (BF) scanning transmission electron microscopy (STEM) micrograph in Figure 3a shows a collection of used (Au75Cu25@Au2ML@Pt)/C nanoparticles (i.e., after the performance and surface area measurements). The best estimate of electrode age of the “used” samples is ~100 cycles between 0.05 VRHE and 1 VRHE limits, with approximately equal number of cycles under O2 and Ar (see Supporting Information). Images in Figures 3b and 3c display elemental maps of Au, Pt, and Cu in the used nanoparticle electro-catalysts acquired using energy-dispersive X-ray spectroscopy (EDS). The elemental maps show the nanoparticles maintain their structural integrity after electrochemical testing; Pt atoms are detected at the surface of the nanoparticles with Au and Cu occupying the core of the nanoparticles. The data also shows that Pt atoms are coated to the surface of the nanoparticles with high degree of uniformity, with only those parts of the nanoparticles that are in the direct contact with carbon and not accessible to the Pt electro-deposition not coated. Individual Au, Cu, and Pt maps are also shown in Figure S3. We note that based on the 2D integral projection images in Figure 3, it was impossible to distinguish the two layers of Au atoms between the Pt surface atoms and the alloy core. However, our synthesis approach, which included multiple cycles of the Cu galvanic replacement as well as electrochemical characterization studies (which showed no events consistent with Cu oxidation) suggest that Cu atoms are absent from surface and subsurface layers. We also note there was no significant structure changes observed for these materials during the activity and surface area measurements. Electrochemical Performance. Cyclic voltammograms (CV) of Au75Cu25@Au2ML@Pt/C and Au85Cu15@Au2ML@Pt/C, along with pure Pt/C for comparison, are shown in Figure 4a. All samples exhibit the main features of the Pt surface: H adsorption/desorption at low potentials (0.0 – 0.4 VRHE), increasing OH coverage around 0.7 VRHE, and further surface oxidation at higher potentials. Data in Figure 4b show the electro-

chemical ORR polarization behavior of nanoparticle electrocatalysts without Pt and Au layers (only the AuCu/C), with the Au layers but without Pt (AuCu@Au2ML/C), and with both Pt and Au layers deposited onto the AuCu cores (AuCu@Au2ML@Pt/C) normalized by electrode geometric surface area. The data show that both bare AuCu and Aucoated AuCu nanoparticles exhibited poor ORR activity and selectivity, manifested in high overpotential losses and low limiting current densities. The data also show that upon the deposition of a Pt layer, the ORR rate (the current) is dramatically increased. Analysis of limiting current densities as a function of RDE rotation rates using Levich approach showed that the Pt alloys exhibited 4-electron pathway with almost exclusive H2O selectivity. Kinetic current densities were calculated using the Koutecky-Levich equation. First, kinetic current was computed using |ik(V)| = (1/|i(V)| -1/|iL|)-1, where i is the measured current, ik is the kinetic current, and iL is the limiting current measured between 0.3 and 0.6 VRHE. To quantitatively assess the kinetic activities of the alloys and compare these to the systems containing commercial Pt electro-catalysts, it is critical to rigorously measure the surface area of the nanoparticle electro-catalysts. The electrochemical surface area (ECSA) of the electro-catalysts was determined using two approaches, CO stripping and hydrogen under potential deposition (HUPD) voltammetry. Data in Figure 4c show the kinetic current density (kinetic current normalized by the ECSA) as the function of potential for pure Pt and the alloy materials. The data show that across the entire potential range of interest the alloys outperform the commercial Pt standard. The data also show that pure Pt and the Pt alloy electrocatalysts exhibit similar Tafel behavior, suggesting that ORR proceeds through a similar mechanism on these electro-catalysts.33 Data in Figure 4d show the kinetic current densities of Au75Cu25@Au2ML@Pt/C, Au85Cu15@Au2ML@Pt/C, and Pt/C measured at 0.9 VRHE. The data show that the current for the commercial Pt/C (5 nm) is roughly 1.0 mA/cm2, which is consistent with other reports in literature showing that if rigorously measured, the 5 nm Pt nanoparticle electro-catalysts exhibit ORR activities between 0.8-1.2 mA/cm2 at these conditions.43 The data also show that both alloy samples exhibit superior ORR performance compared to the commercial Pt standard.

ACS Paragon Plus Environment

Page 5 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

For example, the ORR rates on Au85Cu15@Au2ML@Pt/C nanoparticle electro-catalysts are ~ 2.5 and 3.7 times greater than on pure Pt based on the CO-stripping and HUPD ECSAs, respectively. Figure 4d also shows the measured kinetic current densities for pure Pt nanoparticle electro-catalysts prepared in-

house using identical precursors and carbon support used in the synthesis of alloy samples. These catalysts exhibit ORR rates similar to commercial 3 nm Pt standards, which are consistent with their size (see Figure S2).2,46,47

(a)

(b)

(c)

(d)

Figure 4. a) Cyclic voltammograms of Au75Cu25@Au2ML@Pt/C, Au85Cu15@Au2ML@Pt/C, and pure Pt/C recorded in Ar-purged 0.1 M HClO4 at 50 mV s-1 (no electrode rotation) between 0.05-1.0 VRHE. The voltammograms have been corrected for both uncompensated resistance and capacitive current contributions. b) Positive-going RDE polarization curves of Au85Cu15/C, Au85Cu15@Au2ML/C, and Au85Cu15@Au2ML@Pt/C as a function of potential (V vs. RHE) taken in O2-saturated 0.1 M HClO4 at rotation rates of 400, 900, 1600, and 2500 rpm at a scan rate of 50 mV s-1 between 0.05-1.0 VRHE. In (a) and (b), the current has been normalized by the geometric surface area of the electrode (~0.196 cm2). c) Tafel plots of specific kinetic current densities of the alloys and Pt electrocatalysts. d) Kinetic current density of Pt and Pt alloy electrocatalysts at 0.9 VRHE normalized by HUPD and CO surface areas. Error bars indicate standard deviation of activity calculated from three independent experiments. The 5 and 3 nm Pt/C samples were commercially available and were thermally treated identically to all the alloys prior to electrochemical tests (see Supplementary Information); the sample ‘PtHouse/C’ was synthesized using identical precursors and carbon support used in the synthesis of all alloys (full description provided in Supplementary Information).

It is also important to comment on the impact of the size of nanoparticles on ORR activity. As discussed above, the Ptalloy nanoparticles used in these studies are larger (~ 11 nm diameter for Au/Cu cores and ~ 14 nm for Au85Cu15@Au2ML@Pt) than the commercial Pt standards (~ 5 nm). To demonstrate that the particle size is not critically influencing our findings, we measured the ORR rates on polycrystalline Pt electrodes under identical condition to be below 1.0 A/cm2 based on HUPD and CO surface area, which is ~ 3.5 times lower that the rates on our alloy nanoparticles. We note that the reported ORR rates on polycrystalline Pt are between 0.7 and 2.0 mA/cm2) at 0.9 VRHE based on the HUPD surface areas.48,49 Even this high limit of the Pt performance is well

below the activities reported on our alloy nanoparticle electrocatalysts, which illustrates that the particle size is not the driving factor behind the observed enhanced ORR rates on the alloys. We hypothesized that the main reason for the improved performance of the alloy electro-catalysts compared to pure Pt was weaker OH binding on alloys. The data in Figure 5a shows the HUPD and OH oxidation regions (normalized by HUPD ECSA) for pure Pt and the Pt alloy electro-catalysts measured in Ar-purged 0.1 M HClO4. Compared to pure Pt, both alloys exhibit weaker surface OH features and the onset of surface oxidation at higher potential, which is consistent with a weaker binding of oxygenated surface intermediates.

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The data in Figure 5b shows the CO stripping curves for the electro-catalyst measured in 0.1 M HClO4. On both Pt alloys, the CO oxidation takes place at slightly lower potentials relative to pure Pt, which is also consistent with weaker binding of CO and other intermediates (potentially including OH) involved in the CO stripping process on the Pt surface.50,51 The durability of Pt alloy and commercial Pt catalysts were assessed using a DOE protocol for accelerated durability test (ADT) as described in the supplemental section; we note that voltage cycling was done in O2-saturated electrolyte.52 We found that kinetic current densities of Pt/C and Au85Cu15@Au2ML@Pt/C at 0.9 VRHE were relatively stable; however, the loss of ECSA was higher in alloys compared to pure Pt (see Figures S6 and S7). Several factors may contribute to observed losses in ECSA. To test whether impurity surface adsorbates -- potentially from the electrolyte, reference (a)

Page 6 of 9

electrode, or catalyst film itself -- poisoned the active sites during the stability testing, we paused the potential cycling to wash and refill the electrochemical cell with fresh electrolyte. This process was done after 5,000, 10,000, 20,000, and 30,000 potential cycles and immediately before ORR activity and ECSA measurements. The Pt-monolayer electro-catalyst reasonably maintained its limiting current (Figure S7b) up to 10,000 cycles, but lower limiting currents were observed upon additional cycling, suggesting delamination of either the catalyst film from the glassy carbon electrode substrate or of Pt surface sites from the catalyst. We note that Nafion (a conductive binder) was employed to secure the catalyst to the glassy carbon substrate. Subsequent tests suggest the initial loss in nominal ORR activity results from the dissolution or restructuring of Pt surface atoms, however a detailed investigation is outside the scope the scope of this work. (b)

Figure 5. a) Current due to HUPD and OH oxidation on pure Pt and the alloy electro-catalysts measured in Ar-purged 0.1 M HClO4 electrolyte normalized on an equal HUPD area basis. Scan rate: 50 mV s-1. b) CO stripping voltammograms for pure Pt and Pt alloy electrocatalysts measured in Ar-purged 0.1 M HClO4 electrolyte normalized on an equal CO stripping area basis. Only the positive-going sweeps are shown for clarity. The CO adlayer was adsorbed by bubbling CO gas into the electrolyte solution at 30 sccm while fixing the potential of the working electrode at 0.05 VRHE. Scan rate: 10 mV s-1.

CONCLUSIONS Herein, we demonstrate that the activity of Pt alloy electrocatalysts in ORR can be improved by systematically tuning the ligand and strain effect to produce Pt surface sites that optimize the strength of interaction between the reacting adsorbates and functioning electro-catalyst. In this particular case, we designed optimal Pt surface sites by combining the Au-Pt ligand effect with the strain effect that was manipulated by engineering of the core of nanoparticles with atomically precise composition and lattice constant. This approach should be very general and could apply even for materials that use significantly less costly metals than Au/Cu to construct the core of the nanoparticles. We believe that this family of materials great promise in the design of the next generation of ORR electro-catalysts that could represent an alternative to wellestablished Pt/3d alloy electro-catalysts.

EXPRIMENTAL METHODS Theoretical Methods. All DFT calculations were carried out with the GPAW plane wave code (https://wiki.fysik.dtu.dk/gpaw/) using the generalized gradient approximation and revised Perdew, Burke and Ernzerhof exchange correlation functional. All surface calculations used

a five-layer 2 × 2 fcc [111] periodic unit cell separated by 15 Å of vacuum space in the [111] direction and a dipole layer to decouple the slabs electrostatically. The bottom two layers were fixed and the top three layers and all adsorbates were relaxed until the sum of forces was below 0.05 eV Å−1. Ultrasoft pseudopotentials were used to represent the ionic cores, with the valence electron density determined through iterative diagonalization of the Kohn–Sham Hamiltonian using Pulay mixing. Unit cells were sampled with a 6 × 6 × 1 Monkhorst– Pack k-point grid, and the plane-wave basis-set energy cutoff was 340 eV. An electronic temperature of 0.1 kBT was used, with final energies extrapolated to 0 K. Lattice constants for Au, Cu, Pt, and AuCu alloys were calculated using 4 atom unit cell with 8 x 8 x 8 k-point grid. Characterization. UV-visible extinction spectroscopy (UV-vis) experiments were performed using a Thermo Scientific Evolution 300 UV-Vis spectrophotometer with a Xenon lamp source to measure the extinction spectra of dilute AuCu nanoparticle suspensions in hexane. UV-vis spectra were collected between 300 and 1000 nm at a rate of 240 nm min-1 and were normalized by the extinction peak signal intensity to facilitate comparison of samples with different concentrations. XRD measurements were conducted in a Rigaku rotating anode diffractometer with a monochromated Cu Kα X-ray source

ACS Paragon Plus Environment

Page 7 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

at a scan rate of 5° min−1. Ion coupled plasma optical emission spectroscopy was performed on samples dissolved in aqua regia (3:1 mixture of concentrated HCl and HNO3) using a Perkin-Elmer Optima 2000DV with Winlab software to determine the molar composition of bulk AuCu nanoparticles. Au and Cu signals were normalized by 2 ppm Yttrium internal standard and concentrations were measured three times using calibrated standards of 0, 1, 2, 5, 10, 20, and 40 ppm Au and Cu solutions. Aberration-corrected STEM imaging and sub-nm resolution STEM-EDS were performed at the Center for Nanophase Materials Sciences at Oak Ridge National Laboratory with a JEOL 2200FS TEM/STEM equipped with a CEOS aberration (probe) corrector and operated at a 200 kV accelerating voltage. The microscope was operated in high angle annular dark field (HAADF)-STEM imaging mode and wasequipped with a Bruker AXS X-Flash 5030 silicon drift detector. The probe size was ~0.7 Å and the probe current was ~30 pA during HAADF-STEM imaging. When collecting the EDS spectrum image data, the probe current was increased to ~280 pA and the probe size was ~2 Å. Electrochemical Testing. Electrochemical measurements were performed at room temperature in a custom-made, allTelfon three-electrode cell with a Gamry Instruments Reference 3000 potentiostat/galvanostat/frequency response analyzer. The working electrodes were prepared by sonicating the catalysts powders in absolute ethanol (Fisher) at 0.75 mg mL-1 for > 1 hr. A uniform film of catalysts was achieved by depositing four 10 µL droplets onto a 5 mm glassy carbon electrode insert (Pine Instruments) at a rotation around 600 rpm using the inverted RDE. All electrochemical measurements were performed in 0.1 M HClO4. The reference electrode was Ag/AgCl in 3 M KCl with saturated AgCl (Radiometer Analytical) and counter electrodes (Pt wire, Alfa Aesar) were both in isolated compartments connected by capillaries to the working electrode chamber. During the long-term stability tests, H2 gas was bubbled at the counter electrode to prevent Pt dissolution. Electrolyte solutions of 0.1 M HClO4 were prepared from ultrapure water and 70% perchloric acid (Merck Suprapur). A consistent uncompensated resistance of ~25 Ω was measured with high-frequency impedance, and was corrected for in the polarization curves. All potentials are reported relative to the reversible hydrogen electrode which is calibrated against the H2 oxidation equilibrium at the pH of the solution. Reported currents have been corrected for capacitance and uncompensated solution resistance. Sample capacitance was measured by comparing the differences in the limiting current (around 0.3-0.5 VRHE) between forward and reverse scans during the electrode conditioning scans in O2-saturated electrolyte at 100 mV s-1 between 0.05 and 1.0 VRHE. Cyclic voltammetry was performed in Ar-purged electrolyte at 50 mV s-1. ORR polarization curves were measured in O2saturated electrolyte at rotation rates of 400, 900, 1600, and 2500 rpm. CO-stripping voltammetry was performed immediately following post-polarization CV scans by holding the working electrode at 0.05 VRHE as CO was bubbled (5 minutes) followed by an Ar gas purge (20 minutes) to remove excess CO from solution. The electrode potential was then scanned to 1.0 VRHE at 10 mV s-1. The electrochemical methods and determination of ECSA are discussed in greater detail in the Supplementary Information.

ASSOCIATED CONTENT Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/ Description of catalyst synthesis, detailed characterization and experimental methods, and additional activity and stability results.

AUTHOR INFORMATION Corresponding Author *[email protected]

Author Contributions TVC and SL devised and developed the project. TVC, SM, and GB carried out experimental work and data analysis. TVC performed theoretical calculations. KLM performed STEM and EDS imaging at ORNL. All the authors wrote the manuscript.

Notes The authors declare no competing financial interests.

ACKNOWLEDGMENT We acknowledge support from the US DOE Office of Basic Energy Sciences, Division of Chemical Sciences (FG-0205ER15686). We also acknowledge the University of Michigan X-ray Micro-Analysis Laboratory and Chemistry Technical Services for use of characterization and analytic facilities. Research also supported as part of a user project by Oak Ridge National Laboratory (ORNL)'s Center for Nanophase Materials Sciences, which is an Office of Science User Facility (KLM). Finally, we acknowledge H. Xin and A. Holewinski for helpful discussions and experimental assistance.

REFERENCES (1) Debe, M. K. Nature 2012, 486, 43–51. (2) Gasteiger, H. A.; Kocha, S. S.; Sompalli, B.; Wagner, F. T. Appl. Catal. B Environ. 2005, 56, 9–35. (3) Stephens, I. E. L.; Bondarenko, A. S.; Grønbjerg, U.; Rossmeisl, J.; Chorkendorff, I. Energy Environ. Sci. 2012, 5, 6744 – 6762. (4) Wagner, F. T.; Lakshmanan, B.; Mathias, M. F. J. Phys. Chem. Lett. 2010, 1, 2204–2219. (5) Gasteiger, H. A.; Marković, N. M. Science 2009, 324, 48–49. (6) Wiley: Hydrogen and Fuel Cells - Detlef Stolten http://www.wiley.com/WileyCDA/WileyTitle/productCd3527327118.html (accessed Mar 23, 2016). (7) Shao, M.; Chang, Q.; Dodelet, J.-P.; Chenitz, R. Chem. Rev. 2016, 116, 3594–3657. (8) Ishihara, A.; Ohgi, Y.; Matsuzawa, K.; Mitsushima, S.; Ota, K. Electrochimica Acta 2010, 55 (27), 8005–8012. (9) Qu, L.; Liu, Y.; Baek, J.-B.; Dai, L. ACS Nano 2010, 4, 1321 – 1326. (10) Anderson, A. B. Electrochimica Acta 2002, 47, 3759 –3763. (11) Nørskov, J. K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J. R.; Bligaard, T.; Jónsson, H. J. Phys. Chem. B 2004, 108, 17886–17892. (12) Xin, H.; Holewinski, A.; Schweitzer, N.; Nikolla, E.; Linic, S. Top. Catal. 2012, 55, 376–390. (13) Xin, H.; Holewinski, A.; Linic, S. ACS Catal. 2012, 2, 12 –16. (14) Viswanathan, V.; Hansen, H. A.; Rossmeisl, J.; Nørskov, J. K. ACS Catal. 2012, 2 (8), 1654–1660. (15) Holewinski, A.; Xin, H.; Nikolla, E.; Linic, S. Curr. Opin. Chem. Eng. 2013, 2, 312–319. (16) Markovic, N. M.; Gasteiger, H. A.; Ross, P. N. J. Phys. Chem. 1995, 99, 3411–3415. (17) Paulus, U. A.; Wokaun, A.; Scherer, G. G.; Schmidt, T. J.; Stamenkovic, V.; Radmilovic, V.; Markovic, N. M.; Ross, P. N. J. Phys. Chem. B 2002, 106, 4181–4191.

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(18) Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G.; Ross, P. N.; Lucas, C. A.; Marković, N. M. Science 2007, 315, 493–497. (19) Ball, S. C.; Hudson, S. L.; Theobald, B. R.; Thompsett, D. ECS Trans. 2007, 11, 1267–1278. (20) Neyerlin, K. C.; Srivastava, R.; Yu, C.; Strasser, P. J. Power Sources 2009, 186, 261–267. (21) Stamenković, V.; Schmidt, T. J.; Ross, P. N.; Marković, N. M. J. Phys. Chem. B 2002, 106, 11970–11979. (22) Koh, S.; Leisch, J.; Toney, M. F.; Strasser, P. J. Phys. Chem. C 2007, 111, 3744–3752. (23) Cui, C.; Gan, L.; Heggen, M.; Rudi, S.; Strasser, P. Nat. Mater. 2013, 12, 765–771. (24) Wang, D.; Xin, H. L.; Hovden, R.; Wang, H.; Yu, Y.; Muller, D. A.; DiSalvo, F. J.; Abruña, H. D. Nat. Mater. 2013, 12, 81–87. (25) van der Vliet, D. F.; Wang, C.; Tripkovic, D.; Strmcnik, D.; Zhang, X. F.; Debe, M. K.; Atanasoski, R. T.; Markovic, N. M.; Stamenkovic, V. R. Nat. Mater. 2012, 11, 1051–1058. (26) Stamenkovic, V. R.; Mun, B. S.; Arenz, M.; Mayrhofer, K. J. J.; Lucas, C. A.; Wang, G.; Ross, P. N.; Markovic, N. M. Nat. Mater. 2007, 6, 241–247. (27) Choi, S.-I.; Xie, S.; Shao, M.; Odell, J. H.; Lu, N.; Peng, H.C.; Protsailo, L.; Guerrero, S.; Park, J.; Xia, X.; Wang, J.; Kim, M. J.; Xia, Y. Nano Lett. 2013, 13, 3420–3425. (28) Choi, S.-I.; Xie, S.; Shao, M.; Lu, N.; Guerrero, S.; Odell, J. H.; Park, J.; Wang, J.; Kim, M. J.; Xia, Y. ChemSusChem 2014, 7, 1476–1483. (29) Chen, C.; Kang, Y.; Huo, Z.; Zhu, Z.; Huang, W.; Xin, H. L.; Snyder, J. D.; Li, D.; Herron, J. A.; Mavrikakis, M.; Chi, M.; More, K. L.; Li, Y.; Markovic, N. M.; Somorjai, G. A.; Yang, P.; Stamenkovic, V. R. Science 2014, 343, 1339–1343. (30) Stamenkovic, V.; Mun, B. S.; Mayrhofer, K. J. J.; Ross, P. N.; Markovic, N. M.; Rossmeisl, J.; Greeley, J.; Nørskov, J. K. Angew. Chem. Int. Ed. 2006, 45, 2897–2967. (31) Menning, C. A.; Hwu, H. H.; Chen, J. G. J. Phys. Chem. B 2006, 110, 15471–15477. (32) Okamoto, H.; Chakrabarti, D. J.; Laughlin, D. E.; Massalski, T. B. J. Phase Equilibria 1987, 8, 454–474. (33) Holewinski, A.; Linic, S. J. Electrochem. Soc. 2012, 159, H864–H870. (34) Xin, H.; Schweitzer, N.; Nikolla, E.; Linic, S. J. Chem. Phys. 2010, 132, 111101. (35) Schweitzer, N.; Xin, H.; Nikolla, E.; Miller, J. T.; Linic, S. Top. Catal. 2010, 53, 348–356. (36) Wang, G; Huang, B.; Xiao, L.; Ren, Z.; Chen, H.; Wang, D.; Abruña, H. D. Zhuang, L. JACS, 2014, 136, 9643–9649. (37) Xin, H.; Linic, S. J. Chem. Phys., 2010, 132, 221101. (38) Van Cleve, T; Gibara, E; Linic, S. ChemCatChem, 2016, 8 (1), 256–261. (39) Yang, J.; Chen, X.; Yang, X.; Ying, J. Y. Energy Environ. Sci. 2012, 5, 8976–8981.

Page 8 of 9

(40) Linic, S.; Aslam, U.; Boerigter, C.; Morabito, M. Nat. Mater. 2015, 14, 567–576. (41) Huang, X.; El-Sayed, M. A. J. Adv. Res. 2010, 1, 13–28. (42) Motl, N. E.; Ewusi-Annan, E.; Sines, I. T.; Jensen, L.; Schaak, R. E. J. Phys. Chem. C 2010, 114, 19263–19269. (43) Sasaki, K.; Wang, J. X.; Naohara, H.; Marinkovic, N.; More, K.; Inada, H.; Adzic, R. R. Electrochimica Acta 2010, 55, 2645 – 2652. (44) Zhang, J.; Sasaki, K.; Sutter, E.; Adzic, R. R. Science 2007, 315, 220–222. (45) Wang, J. X.; Inada, H.; Wu, L.; Zhu, Y.; Choi, Y.; Liu, P.; Zhou, W.-P.; Adzic, R. R. J. Am. Chem. Soc. 2009, 131, 17298– 17302. (46) Pedersen, C. M.; Escudero-Escribano, M.; VelázquezPalenzuela, A.; Christensen, L. H.; Chorkendorff, I.; Stephens, I. E. L. Electrochimica Acta 2015, 179, 647–657. (47) Perez-Alonso, F. J.; McCarthy, D. N.; Nierhoff, A.; Hernandez-Fernandez, P.; Strebel, C.; Stephens, I. E. L.; Nielsen, J. H.; Chorkendorff, I. Angew. Chem. Int. Ed. 2012, 51, 4641–4643. (48) Garsany, Y.; Singer, I. L.; Swider-Lyons, K. E. J. Electroanal. Chem. 2011, 662, 396–406. (49) Garsany, Y.; Baturina, O. A.; Swider-Lyons, K. E.; Kocha, S. S. Anal. Chem. 2010, 82, 6321–6328. (50) van der Vliet, D. F.; Wang, C.; Li, D.; Paulikas, A. P.; Greeley, J.; Rankin, R. B.; Strmcnik, D.; Tripkovic, D.; Markovic, N. M.; Stamenkovic, V. R. Angew. Chem. Int. Ed Engl. 2012, 51, 3139– 3196. (51) Bandarenka, A. S.; Varela, A. S.; Karamad, M.; Calle-Vallejo, F.; Bech, L.; Perez-Alonso, F. J.; Rossmeisl, J.; Stephens, I. E. L.; Chorkendorff, I. Angew. Chem. Int. Ed Engl. 2012, 51, 11845–11848. (52) Koenigsmann, C.; Santulli, A. C.; Gong, K.; Vukmirovic, M. B.; Zhou, W.; Sutter, E.; Wong, S. S.; Adzic, R. R. J. Am. Chem. Soc. 2011, 133, 9783 –9795.

ACS Paragon Plus Environment

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis O2

H2O

ORR Rate at 0.9 VRHE

Page 9 of 9

AuCux@Au@Pt alloy nanostructure leads to improved catalytic activity

ACS Paragon Plus Environment

9