In Situ Raman Spectroscopy of Copper and Copper

0 downloads 0 Views 4MB Size Report
Feb 24, 2016 - during Electrochemical Oxygen Evolution Reaction: Identification of CuIII Oxides as .... plating solution at +5 mA/cm2 onto a mechanically polished copper disk for 200 s. ..... ization data for NaCuO2 (PDF). □ AUTHOR .... (62) Pourbaix, M. In Atlas of Electrochemical Equilibria in Aqueous. Solutions; 1st ...
Research Article pubs.acs.org/acscatalysis

In Situ Raman Spectroscopy of Copper and Copper Oxide Surfaces during Electrochemical Oxygen Evolution Reaction: Identification of CuIII Oxides as Catalytically Active Species Yilin Deng,† Albertus D. Handoko,† Yonghua Du,‡ Shibo Xi,‡ and Boon Siang Yeo*,†,§ †

Department of Chemistry, Faculty of Science, National University of Singapore, 3 Science Drive 3, Singapore 117543 Institute of Chemical and Engineering Sciences, A*STAR, 1 Pesek Road, Jurong Island, Singapore 627833 § Solar Energy Research Institute of Singapore, National University of Singapore, 7 Engineering Drive 1, Singapore 117574 ‡

S Supporting Information *

ABSTRACT: Scanning electron microscopy, X-ray diffraction, cyclic voltammetry, chronoamperometry, in situ Raman spectroscopy, and X-ray absorption near-edge structure spectroscopy (XANES) were used to investigate the electrochemical oxygen evolution reaction (OER) on Cu, Cu2O, Cu(OH)2, and CuO catalysts. Aqueous 0.1 M KOH was used as the electrolyte. All four catalysts were oxidized or converted to CuO and Cu(OH)2 during a slow anodic sweep of cyclic voltammetry and exhibited similar activities for the OER. A Raman peak at 603 cm−1 appeared for all the four samples at OER-relevant potentials, ≥1.62 V vs RHE. This peak was identified as the Cu−O stretching vibration band of a CuIII oxide, a metastable species whose existence is dependent on the applied potential. Since this frequency matches well with that from a NaCuIIIO2 standard, we suggest that the chemical composition of the CuIII oxide is CuO2−-like. The four catalysts, in stark contrast, did not oxidize the same way during direct chronoamperometry measurements at 1.7 V vs RHE. CuIII oxide was observed only on the CuO and Cu(OH)2 electrodes. Interestingly, these two electrodes catalyzed the OER ∼10 times more efficiently than the Cu and Cu2O catalysts. By correlating the intensity of the Raman band of CuIII oxide and the extent of the OER activity, we propose that CuIII species provides catalytically active sites for the electrochemical water oxidation. The formation of CuIII oxides on CuO films during OER was also corroborated by in situ XANES measurements of the Cu K-edge. The catalytic role of CuIII oxide in the O2 evolution reaction is proposed and discussed. KEYWORDS: oxygen evolution reaction, electrocatalyst, copper oxide, water, Raman spectroscopy, X-ray absorption near edge structure spectroscopy, linear sweep voltammetry, chronoamperometry

1. INTRODUCTION The electrochemical reduction of carbon dioxide to alcohols and hydrocarbons, as well as water splitting to hydrogen gas, using solar-generated electricity is a sustainable and green way of producing these important transportation fuels.1 A limiting factor constraining the wide scale application of these processes is the energetic inefficiency associated with the anodic oxygen evolution reaction (OER) (4OH− → O2 + 2H2O + 4e−). RuO2 and IrO2 are currently the benchmarking catalysts for this reaction.2 However, these metal oxides typically require overpotentials of at least several hundreds of millivolts in order to produce O2 currents of ∼10 mA/cm2.3 Furthermore, they corrode and dissolve during the OER process.4,5 The low natural abundance of ruthenium and iridium also makes these metals expensive and renders their large-scale applications in the industry impractical. Hence, there has been an urgent search for alternate OER catalysts that are efficient and economical. © XXXX American Chemical Society

During the past few years, the use of monometallic copper materials as catalysts for the OER has gained attention. Crystalline copper nanoparticles,6,7 Cu2O nanoparticles,8 Cu/ (Cu(OH)2−CuO) core/shell nanorods arrays,9 nanostructured copper oxide films,10 and Cu nanocomposites11 have been found active and stable for the OER. Among these, the Cu/(Cu(OH)2−CuO) core/shell nanorods arrays required a relatively low overpotential of 417 mV to produce an O2 current of 10 mA/cm2. Furthermore, they are catalytically stable for at least 22 h in 0.1 M KOH.9 Giving even more impressive figuresof-merit are Cu atoms attached to a graphene/phenanthroline scaffold.11 In KOH electrolyte, these nanocomposites were reported to catalyze OER starting from 1.34 V vs RHE (RHE = reversible hydrogen electrode; all potentials cited hereafter in Received: January 21, 2016 Revised: February 24, 2016

2473

DOI: 10.1021/acscatal.6b00205 ACS Catal. 2016, 6, 2473−2481

Research Article

ACS Catalysis

Cu(OH)2 electrodes during CV and chronoamperometry in aqueous 0.1 M KOH electrolyte. Potentials relevant to the OER were applied. Measurements made using scanning electron microscopy (SEM), X-ray diffraction (XRD), and in situ X-ray absorption near-edge structure spectroscopy (XANES) were also used to corroborate our findings.

this work are with respect to the RHE), which represents an overpotential of 0.11 V from the thermodynamic potential of 1.23 V. The high efficiencies of these copper-based catalysts have been largely attributed to their high electrochemical surface areas.6−9 More recently, we discovered that H2O2-treated CuO nanostructures could achieve turnover frequencies for O2 evolution which are comparable or better than many other manganese- and cobalt-based OER catalysts.12 In situ Raman spectroscopy of the CuO nanostructures at OER-relevant potentials recorded evidence of CuIII oxides. This feature was not found on electrodeposited copper metal, which exhibited a comparatively weaker OER activity. We thus postulated that the enhanced catalysis of O2 evolution by the CuO nanostructures can be attributed not only to its larger surface area, but also to the higher population of CuIII active sites present on its surface. Based on the above discussion, we envisage that the judicious selection of starting Cu materials and careful design of its morphology (for example, maximizing the surface density of active sites) could lead to the engineering of a new class of efficient, Earth-abundant, and economical water oxidation catalysts. To do so, it is essential to understand the changes in composition of the copper anodes as a function of the applied potentials and the influence of these changes on the rate of OER. Identification of the catalytically active species is also necessary. The electrochemical behavior of copper metal during cyclic voltammetry (CV) or linear sweep voltammetry (LSV) in alkaline electrolytes has been investigated and can be summarized as follows (see Figure 1).13−21 A Cu2O layer is first

2. EXPERIMENTAL SECTION 2.1. Preparation of Catalysts. For the XRD, SEM, and Raman spectroscopy experiments, the substrates of the working electrodes were prepared from copper disks (10 mm diameter, 99.99%, Goodfellow, Inc.). Four types of working electrodes were prepared. 2.1.1. Metallic Copper (Cu). The copper disks were first polished using diamond slurries (9 and 3 μm, Struers) and then rinsed with ultrapure water (18.2 MΩ cm, Barnstead, ThermoFisher Scientific). 2.1.2. Cu2O Films (Cu2O/Cu). A two-electrode setup, which comprised of a mechanically polished copper disk as the working electrode and a platinum wire (99.99%, Sigma− Aldrich) as the counter electrode, was used. The Cu2O film was galvanostatically deposited onto the copper substrate from a plating solution at −1.4 mA/cm2 for 150 s. The plating solution consisted of aqueous 0.3 M CuSO4·5H2O (98.5−101%, GCE), 3.2 M NaOH (99%−100%, Chemicon) and 2.3 M lactic acid (85%, Sigma−Aldrich). A water bath was used to maintain the temperature of the electrochemical cell at 60 °C.26 2.1.3. Cu(OH)2 Films (Cu(OH)2/Cu). The Cu(OH)2 film was prepared by anodizing a mechanically polished copper disk at a constant potential of 0.81 V for 15 min in 1 M KOH (99.99%, Sigma−Aldrich) electrolyte using a three-electrode setup.27 A platinum wire and a Hg/HgO electrode (1 M KOH, CHI 152, CH Instruments) were used as counter and reference electrodes, respectively. 2.1.4. CuO Films (CuO/Cu and CuO/Carbon Cloth). A twoelectrode setup containing a platinum wire counter electrode was used. The CuO film was galvanostatically deposited from a plating solution at +5 mA/cm2 onto a mechanically polished copper disk for 200 s. The plating solution contained aqueous 0.2 M CuSO4·5H2O and 0.2 M tartaric acid (99%, Alfa Aesar), which was adjusted to pH 13 using 4 M NaOH solution.28 The bath temperature was 30 °C. For the XANES experiments, a strip of carbon cloth (W0S1002, CoTech) was used as substrate. The CuO film was deposited on it at a current density of +10 mA/cm2 for 75 s. A 1.5 cm × 1 cm strip was exposed to the deposition solution. NaCuO2 powder was synthesized by heating a mixture of Na2O2 (93+%, Sigma−Aldrich) and CuO (45 °C) in the same highly concentrated alkaline electrolyte. The currents of these four samples poised at 1.7 V have a different trend from those obtained earlier from cyclic voltammetry. CuO/Cu and Cu(OH)2/Cu samples exhibited currents at 1.7 mA/cm2 and 0.8 mA/cm2, respectively (Figure 4a). In contrast, the Cu2O/Cu and Cu samples showed ∼1 order of magnitude lower currents at ∼0.1−0.2 mA/cm2. The smaller current was also reflected by our observation of a smaller amount of oxygen bubbles on the Cu and Cu2O/Cu, compared to the Cu(OH)2/Cu and CuO/Cu electrodes. Surface oxidation of the Cu and Cu2O/Cu samples contributed ≤1%−2% to the observed currents in the chronoamperograms, especially at the initial stage of the polarization (see section S6 in the Supporting Information).33 The Tafel slopes of Cu, Cu2O/Cu, CuO/Cu, and Cu(OH)2/Cu, each measured from a single linear sweep voltammogram of a post-chronoamperometry sample, were found to be 100, 102, 60, and 69 mV/dec, respectively (see section S5.2 in the Supporting Information). The lower Tafel slope values of CuO/Cu and Cu(OH)2/Cu catalysts indicate a kinetically more favorable charge transfer process on these two samples, which is consistent with their superior OER activity.

The in situ Raman spectra acquired in tandem with the CV measurements enables us to understand the chemical compositional changes of the four catalysts during anodic oxidation (Figure 3b). The Raman spectra of metallic Cu during CV are shown in Figure 3b(i). At 0.42 V, the spectrum was featureless, which indicates that the Cu electrode remained metallic. As the potential increased anodically to 0.72 V (past the first anodic peak at 0.55 V), two Raman peaks at 523 and 623 cm−1 appeared, indicating the formation of Cu2O (see Table 1).21 Above 1.32 V, CuO and Cu(OH)2 were formed as evidenced by their characteristic signals at 298/347 and 490 cm−1 respectively. Above 1.62 V, where OER commences, a peak was observed at 603 cm−1. This peak disappeared just outside the OER range at 1.55 V during the reverse cathodic sweep, after which only CuO and Cu(OH)2 Raman bands were recorded. This is an interesting phenomenon, because it suggests the existence of a transient, Cu-related species that exists only at OER-relevant potentials. The Cu2O/Cu, Cu(OH)2/Cu, and CuO/Cu samples oxidize similarly to the Cu sample during CV (Figures 3b(ii−iv)). For these samples, the 603 cm−1 peak also evolved at ≥1.62 V and disappeared as the applied potential decreased below 1.55 V. After the CV and chronoamperometry experiments, all four samples were characterized by SEM (Figure 3d). The morphology of the CuO/Cu electrode was similar to its pre-OER morphology, while the appearance of the Cu(OH)2/Cu electrode changed partly from needlelike to more irregularly shaped particles. The morphologies of the Cu and Cu2O/Cu surfaces underwent the most dramatic changes, and now consist of flakelike particles. These significant changes corroborate their extensive chemical transformation during the CV experiment. The present results indicate that all four Cu samples converted to CuO and Cu(OH)2 during the anodic CV sweep, and catalyzed O2 evolution with similar efficiencies. A Raman peak at 603 cm−1 also appeared in the spectra of all the samples at O2-evolving potentials. This means that it is not possible to 2477

DOI: 10.1021/acscatal.6b00205 ACS Catal. 2016, 6, 2473−2481

Research Article

ACS Catalysis The difference in OER activities of the electrodes measured using CV and chronoamperometry can be understood by comparing their in situ Raman spectra acquired during these experiments (Figures 3b and 4b). During chronoamperometry at 1.7 V, both Cu and Cu2O/Cu electrodes did not oxidize to CuO and Cu(OH)2, as demonstrated by the absence of their Raman peaks at 298/347 and 490 cm−1, respectively (see Figures 4b(i and ii)). A broad band at 594 cm−1 appeared instead. This band is different from the 603 cm−1 peak, because of its different wavenumber and peak shape. The broadness of the 594 cm−1 peak implies that it originates from amorphous Cu oxide structures. This amorphous structure is not OER active, as indicated by the low currents exhibited by these samples. In contrast, the Raman spectra acquired for CuO/Cu and Cu(OH)2/Cu were similar to those collected earlier during CV (Figures 4b(iii and iv)). A sharp singlet peak at 603 cm−1 was clearly detected on both samples, along with Raman peaks at 298, 347, and 490 cm−1, which belong to CuO and Cu(OH)2. The direct correlation between the superior OER activities of the Cu(OH)2/Cu and CuO/Cu and the presence of the 603 cm−1 peak in these samples strongly suggest that the 603 cm−1 peak belongs to a catalytically active Cu species for water oxidation. Raman spectra of the samples were acquired after the chronoamperometric measurements (Figure 4b). On the Cu(OH)2/Cu and CuO/Cu surfaces, the 603 cm−1 peak had disappeared, which is consistent with the Raman spectroscopy data presented in section 3.2 (see Figures 3b(iii and iv)). A mixture of CuO and Cu(OH)2 still remained, as identified by their Raman peaks at 298, 347, and 490 cm−1. On the Cu and Cu2O/Cu surfaces, peaks belonging to Cu2O at 523 and 623 cm−1 were observed (Figures 4b(i and ii)). We propose that the amorphous Cu oxide formed during chronoamperometry at 1.7 V could have (partially) transformed to Cu2O. Neither CuO nor Cu(OH)2 was detected. SEM images of the catalysts taken after chronoamperometry showed that the Cu and Cu2O/Cu samples largely retained their original surface morphologies (Figure 4c). These results are expected since Cu and Cu2O/Cu samples could not be deeply oxidized, as the bulk of the Cu oxidation currents at the Cu active regions have been avoided by the direct chronoamperometry measurement at 1.7 V. CuO/Cu also retained its morphology. The Cu(OH)2/Cu surface changed partly from needlelike particles to irregular-shaped particles. Because of the low concentration of alkaline electrolyte used (0.1 M KOH) and the avoidance of excessively high overpotentials, dissolution of the copper electrodes during the electrochemical measurements is expected to be minimal.17 This is consistent with the generally stable currents measured in the chronoamperograms presented in Figures 3c and 4a. 3.4. Identification of CuIII Oxides as Catalytically Active Species for the Oxygen Evolution Reaction. The formation of CuIII oxides at the onset of the OER has been postulated.13−15,17 To ascertain its presence, we used NaCuO2 as the standard CuIII oxide (the existence of Cu2O3 is unclear).25 The successful preparation of NaCuO2, which is a bluish-black colored powder, was confirmed by XRD and Raman spectroscopy (see Figures 5a and 5b).29,45,46 The Raman spectrum of NaCuO2 displayed a strong, sharp peak at 603 cm−1 with two other considerably weaker peaks at 500 and 1145 cm−1. This 603 cm−1 peak could thus be considered as a characteristic marker of NaCuO2. Significantly, its frequency matches well with the 603 cm−1 peak observed on CuO/Cu and Cu(OH)2/Cu

Figure 5. (a) X-ray diffractogram and (b) Raman spectrum of NaCuO2. The expected XRD pattern of NaCuO2 (vertical lines) is included for comparison. A small quantity of CuO was detected by XRD.

electrodes during OER (see Figures 3 and 4, as well as section S7 in the Supporting Information). This suggests that CuIIIO2−-like species formed on the CuO and Cu(OH)2 surfaces at O2-evolving potentials. We further note that the 603 cm−1 peak did not exhibit frequency shifts upon substitution of 1 H in the electrochemical system with deuterium atoms (see section S8 in the Supporting Information). This demonstrates that the vibration does not involve an −OH (−OD) bond,21 and the peak can be confidently ascribed to CuIII−O stretching vibration. This assignment is also consistent with the CuIII−O vibration peak of bis-(μ-oxo)-CuIII2 at ∼608 cm−1.47 Further evidence for the formation of CuIII oxides during O2 evolution was obtained from the XANES spectroscopy of its K-edge. XANES is a powerful analytical technique that provides element-specific information about the valence structure of a catalyst. The absorption edge of Cu generally blue shifts as the metal center becomes more oxidized, because of its higher effective nuclear charge, which increases the binding energy of the 1s electron.48 NaCuIIIO2 and CuIIO standards were first measured (see section S9.1 in the Supporting Information). As expected, the NaCuO2 K-edge (1s−4pπ transition) blue shifts by ∼2.2 eV, relative to that of the CuO standard.49−52 The in situ XANES spectrum of a CuO/carbon cloth poised at 2.0 V in 0.1 M KOH electrolyte was then recorded. It is notable that the in situ CuO/carbon cloth K-edge shows a ∼0.3 eV blue shift toward that of the NaCuO2 standard (second derivative spectrum). The shift in edge energy indicates an increase in the average valence of copper in the CuO sample, which we attribute to the formation of CuIII oxides. The extent of the blue shift is expected based on the estimated ∼7% CuIII content in the film (see section S9.2 in the Supporting Information). 3.5. Role of CuIII Oxide in the Oxygen Evolution Reaction. To understand the catalytic role of CuIII oxide on water oxidation, the cyclic voltammograms of the Cu, Cu2O, CuO, and Cu(OH)2 electrodes post-chronoamperometry at 1.7 V were acquired (cathodic sweep first, Figure 6a). Cathodic peaks at ∼1.55 V were observed on the samples. This peak, 2478

DOI: 10.1021/acscatal.6b00205 ACS Catal. 2016, 6, 2473−2481

Research Article

ACS Catalysis

CuIII (in LaCuIIIO3) was found to lie further away from the peak of the volcano, i.e., it was predicted to have a lower activity toward OER, compared to Cu and CuO (due to the weaker adsorption of O(ad), OH(ad), and OOH(ad) on the CuIII sites). This computation is based on the surface CuIII atoms being surrounded by a surface adsorbate and five O ligands (four in the X−Y plane and one in the subsurface layer).56 However, from our Raman spectroscopy results, CuIIO was oxidized to a CuIII species that has a structure similar to NaCuIIIO2. The CuIII centers therein are coordinated in a square planar geometry, which has a different orbital splitting.57 More importantly, a decrease in coordination number of the transition metal atom will increase its possibility of creating stronger bonds.56 We thus propose that this difference in coordination may have allowed the CuIII centers in this work to form stronger bonds with the OH(ad), O(ad), and OOH(ad) intermediates (the scaling relations) than that predicted from theory. This would effectively move it closer toward the peak of the volcano, and lower its overpotential for evolving O2. Defects on the catalyst surface could also enhance the adsorption of these intermediates. DFT calculations on a CuO2−-like structure would be helpful in testing this hypothesis. Another OER mechanism involves O2 gas evolution from the decomposition of the higher valence oxides. Here, the CuO or Cu(OH)2 films could be oxidized to CuIII oxides at sufficiently anodic potentials. Being unstable, these are then decomposed to CuIIO and O2 gas (see section S10 in the Supporting Information).24 Therefore, repeated CuII oxide−CuIII oxide cycling forms the basis of this OER mechanism. This process was shown for O2 evolution occurring on RuOx catalysts and exhibited small Tafel slopes of 32−44 mV/dec.4,58 Rapid dissolution of the Cu oxides and deactivation of the catalysts (as had been observed for the Ru oxides) are expected with this mechanism.59 However, the currents exhibited by the Cu catalysts in our chronoamperograms are relatively stable for at least 30 min (see Figures 3c and 4a). Their Tafel slopes are also higher at 60−66 mV/dec (section S5 in the Supporting Information). Thus, we believe that this mechanism is unlikely to occur on the Cu catalysts. Metal cations in high oxidation states such as MnIV and CoIV have been previously proposed as active centers for OER electrocatalysts.60,61 This is usually based on observations of these oxidized species when O2 evolution is occurring for a particular metal oxide. However, their presence is not unexpected, since these are usually the thermodynamically stable species at OER potentials (predicted using the Pourbaix diagrams).62 Therefore, it is difficult to ascertain if these oxidized metal cations are OER-active catalytic species or spectators. More specifically, previous electrochemical investigations on Cu substrates suggest the presence of CuIII species during O2 evolution. However, its identity and role has never been clearly identified. Here, through in situ and ex situ Raman spectroscopy, XANES, and judicious choice of electrochemical measurements, we provided extensive spectroscopic evidence for the formation of CuIII oxide. By comparing its Raman signature with that of the NaCuO2 standard, we suggest that its chemical composition is CuO2−-like. We have furthermore correlated its presence with the O2 current, and could thus propose it as a catalytically active species/site for the electrochemical water oxidation. CuO or Cu(OH)2 surfaces was found to be a crucial prerequisite for the formation of CuIII oxides. Consequently, Cu-based catalysts containing CuII or CuIII ions would be interesting materials to investigate in the quest to develop a new generation of Earth-abundant and efficacious catalysts for the electrochemical oxidation of water.

Figure 6. (a) Cyclic voltammograms (cathodic sweep first) of the Cu, Cu2O/Cu, Cu(OH)2/Cu, and CuO/Cu samples (after chronoamperometry at 1.7 V) in 0.1 M KOH electrolyte. Scan rate: 10 mV/s. The starting potential of all the CV measurements was at 1.7 V. (b) The OER current densities of the four samples at 1.7 V (from Figure 4a) vs the amounts of charges beneath the reduction peaks at ∼1.55 V. The inset shows an enlarged view of the Cu2O/Cu and Cu data points.

which can be attributed to the reduction of CuIII to CuII, has been previously observed under similar experimental conditions (for example, on Cu electrodes at a 25 mV/s scan rate in 0.1 M NaOH electrolyte).15,17 The population of CuIII species present in the samples can thus be estimated from the charges belonging to these cathodic peaks. We are aware that it is unlikely for CV to fully detect all the CuIII species present. However, a consistent use of this method will ensure a selfconsistent estimation of its population on all electrodes. Interestingly, the measured charges correlate linearly to the OER current densities exhibited by the Cu electrodes (Figure 6b). This indicates that the CuIII species strongly influences the OER activity of the Cu electrocatalyst. Considering this positive correlation and the appearance of the 603 cm−1 Raman peak belonging to CuO2−-like oxides during OER (Figures 3b and 4b), we propose that CuIII oxides are catalytically active for evolving O2. However, we note that the participation of CuII (present together with CuIII) in the catalysis of the OER cannot be definitively ruled out. An OER mechanism involving three surface intermediates, OH(ad), O(ad), and OOH(ad), has been proposed as follows:53 M + OH− → M−OH + e−

(M represents a Cu III active site) (1)

M−OH + OH− → M−O + e− + H 2O −

M−O + OH → M−OOH + e

(2)





(3) −

M−OOH + OH → M−OO + e + H 2O

M−OO → M + O2

(4) (5)

II

III

III

Using this mechanism, Cu, Cu O, Cu (in LaCu O3 perovskite), and CuIV (in SrCuIVO3) have been simulated using density functional theory (DFT) to lie on the weak binding side of the Sabatier-type volcano plot.54−56 The more-oxidized 2479

DOI: 10.1021/acscatal.6b00205 ACS Catal. 2016, 6, 2473−2481

Research Article

ACS Catalysis

(5) McCrory, C. C. L.; Jung, S. H.; Peters, J. C.; Jaramillo, T. F. J. Am. Chem. Soc. 2013, 135, 16977−16987. (6) Ahmed, J.; Trinh, P.; Mugweru, A. M.; Ganguli, A. K. Solid State Sci. 2011, 13, 855−861. (7) Kumar, B.; Saha, S.; Basu, M.; Ganguli, A. K. J. Mater. Chem. A 2013, 1, 4728−4735. (8) Kumar, B.; Saha, S.; Ganguly, A.; Ganguli, A. K. RSC Adv. 2014, 4, 12043−12049. (9) Cheng, N.; Xue, Y.; Liu, Q.; Tian, J.; Zhang, L.; Asiri, A. M.; Sun, X. Electrochim. Acta 2015, 163, 102−106. (10) Liu, X.; Jia, H.; Sun, Z.; Chen, H.; Xu, P.; Du, P. Electrochem. Commun. 2014, 46, 1−4. (11) Wang, J.; Wang, K.; Wang, F.-B.; Xia, X.-H. Nat. Commun. 2014, 5, 5285. (12) Handoko, A. D.; Deng, S.; Deng, Y.; Cheng, A. W. F.; Chan, K. W.; Tan, H. R.; Pan, Y.; Tok, E. S.; Sow, C. H.; Yeo, B. S. Catal. Sci. Technol. 2016, 6, 269−274. (13) Muller, E. Z. Elektrochem. Angew. Phys. Chem. 1907, 13, 133− 145. (14) El Din, A. M. S.; El Wahab, F. M. A. Electrochim. Acta 1964, 9, 113−121. (15) Miller, B. J. Electrochem. Soc. 1969, 116, 1675−1680. (16) Ambrose, J.; Barradas, R. G.; Shoesmith, D. W. J. Electroanal. Chem. Interfacial Electrochem. 1973, 47, 47−64. (17) Abd el Haleem, S. M.; Ateya, B. G. J. Electroanal. Chem. Interfacial Electrochem. 1981, 117, 309−319. (18) Hamilton, J. C.; Farmer, J. C.; Anderson, R. J. J. Electrochem. Soc. 1986, 133, 739−745. (19) Schwartz, D. T.; Muller, R. H. Surf. Sci. 1991, 248, 349−358. (20) Mayer, S. T.; Muller, R. H. J. Electrochem. Soc. 1992, 139, 426− 434. (21) Chan, H. Y. H.; Takoudis, C. G.; Weaver, M. J. J. Phys. Chem. B 1999, 103, 357−365. (22) Landolt, D. In Corrosion and Surface Chemistry of Metals; 1st Edition; EFPL Press: Lausanne, Switzerland, 2007; pp 227−274. (23) Fredj, N.; Burleigh, T. D. J. Electrochem. Soc. 2011, 158, C104− C110. (24) Popova, T. V.; Aksenova, N. V. Russ. J. Coord. Chem. 2003, 29, 743−765. (25) Levason, W.; Spicer, M. D. Coord. Chem. Rev. 1987, 76, 45−120. (26) Mukhopadhyay, A. K.; Chakraborty, A. K.; Chatterjee, A. P.; Lahiri, S. K. Thin Solid Films 1992, 209, 92−96. (27) Reyter, D.; Odziemkowski, M.; Bélanger, D.; Roué, L. J. Electrochem. Soc. 2007, 154, K36−K44. (28) Wang, L.; Han, K.; Tao, M. J. Electrochem. Soc. 2007, 154, D91− D94. (29) Ono, Y.; Yui, Y.; Asakura, K.; Nakamura, J.; Hayashi, M.; Takahashi, K. I. Am. J. Phys. Chem. 2014, 3, 61−66. (30) Yeo, B. S.; Klaus, S. L.; Ross, P. N.; Mathies, R. A.; Bell, A. T. ChemPhysChem 2010, 11, 1854−1857. (31) Du, Y.; Zhu, Y.; Xi, S.; Yang, P.; Moser, H. O.; Breese, M.; Borgna, A. J. Synchrotron Radiat. 2015, 22, 839−843. (32) Ravel, á.; Newville, M. J. Synchrotron Radiat. 2005, 12, 537−541. (33) Shoesmith, D. W.; Rummery, T. E.; Owen, D.; Lee, W. J. Electrochem. Soc. 1976, 123, 790−799. (34) Paracchino, A.; Brauer, J. C.; Moser, J.-E.; Thimsen, E.; Graetzel, M. J. Phys. Chem. C 2012, 116, 7341−7350. (35) Niaura, G. Electrochim. Acta 2000, 45, 3507−3519. (36) Singhal, A.; Pai, M. R.; Rao, R.; Pillai, K. T.; Lieberwirth, I.; Tyagi, A. K. Eur. J. Inorg. Chem. 2013, 2013, 2640−2651. (37) Debbichi, L.; Marco de Lucas, M. C.; Pierson, J. F.; Krüger, P. J. Phys. Chem. C 2012, 116, 10232−10237. (38) Wang, W.; Liu, Z.; Liu, Y.; Xu, C.; Zheng, C.; Wang, G. Appl. Phys. A: Mater. Sci. Process. 2003, 76, 417−420. (39) Xu, J. F.; Ji, W.; Shen, Z. X.; Li, W. S.; Tang, S. H.; Ye, X. R.; Jia, D. Z.; Xin, X. Q. J. Raman Spectrosc. 1999, 30, 413−415. (40) Chen, X. K.; Irwin, J. C.; Franck, J. P. Phys. Rev. B: Condens. Matter Mater. Phys. 1995, 52, R13130−R13133.

4. CONCLUSION This work provides the first systematic study of the intriguing oxygen evolution reaction (OER) on copper surfaces with different oxidation states. Extensive materials and electrochemical characterizations were performed using X-ray diffraction, scanning electron microscopy, cyclic voltammetry, chronoamperometry, in situ X-ray absorption near-edge structure spectroscopy (XANES), and Raman spectroscopy. The electrochemical behaviors of Cu, Cu2O/Cu, Cu(OH)2/Cu, and CuO/ Cu samples were different during CV and direct chronoamperometry measurements. During CV, all four samples were either oxidized to or remained as CuO and Cu(OH)2, and expectedly showed similar OER performances. A Raman peak at 603 cm−1 appeared at OER-relevant potentials of ≥1.62 V on all four samples. This peak was identified as the marker band of a CuIII oxide (proposed to be CuO2−-like), which is a metastable species. In contrast, during chronoamperometry at 1.7 V, Cu or Cu2O/Cu were not oxidized to CuO or Cu(OH)2, and their OER currents were 1 order of magnitude lower than that of the CuO/Cu and Cu(OH)2/Cu samples. Interestingly, the CuIII Raman band only appeared on the CuO/Cu or Cu(OH)2/Cu catalysts. These observations allow us to propose CuIII oxides as catalytically active species for the electrochemical oxidation of water. The formation of CuIII oxides on CuO film during O2 evolution was further demonstrated by the in situ XANES of the Cu K-edge.



ASSOCIATED CONTENT

S Supporting Information *

This Supporting Information is available free of charge on the ACS Publication Web site at The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscatal.6b00205. Schematic diagrams of Raman spectroscopy and XAS setups; additional electrochemical, Raman, AFM and XANES data; isotopic labeling experiments; characterization data for NaCuO2 (PDF)



AUTHOR INFORMATION

Corresponding Author

*Tel.: +65 6516 2836. Fax: +65 6779 1691. E-mail: chmyeos@ nus.edu.sg. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by an academic research fund (No. R-143-000-587-112) from the National University of Singapore. We thank Laura-Lynn Liew (Institute of Materials Research and Engineering, Singapore) for performing the atomic force microscopy measurements.



REFERENCES

(1) Hori, Y. In Modern Aspects of Electrochemistry, Vol. 42; Vayenas, C., White, R., Gamboa-Aldeco, M., Eds.; Springer: New York, 2008; pp 89−189. (2) Trasatti, S. J. Electroanal. Chem. Interfacial Electrochem. 1980, 111, 125−131. (3) Matsumoto, Y.; Sato, E. Mater. Chem. Phys. 1986, 14, 397−426. (4) Reier, T.; Oezaslan, M.; Strasser, P. ACS Catal. 2012, 2, 1765− 1772. 2480

DOI: 10.1021/acscatal.6b00205 ACS Catal. 2016, 6, 2473−2481

Research Article

ACS Catalysis (41) Kliche, G.; Popovic, Z. V. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 42, 10060−10066. (42) Fazle Kibria, A. K. M.; Tarafdar, S. A. Int. J. Hydrogen Energy 2002, 27, 879−884. (43) Brossard, L.; Marquis, B. Int. J. Hydrogen Energy 1994, 19, 231− 237. (44) Liu, Y. C.; Koza, J. A.; Switzer, J. A. Electrochim. Acta 2014, 140, 359−365. (45) Steiner, P.; Kinsinger, V.; Sander, I.; Siegwart, B.; Hüfner, S.; Politis, C.; Hoppe, R.; Müller, H. P. Z. Phys. B: Condens. Matter 1987, 67, 497−502. (46) Ruther, R. E.; Zhou, H.; Dhital, C.; Saravanan, K.; Kercher, A. K.; Chen, G.; Huq, A.; Delnick, F. M.; Nanda, J. Chem. Mater. 2015, 27, 6746−6754. (47) Henson, M. J.; Mukherjee, P.; Root, D. E.; Stack, T. D. P.; Solomon, E. I. J. Am. Chem. Soc. 1999, 121, 10332−10345. (48) DuBois, J. L.; Mukherjee, P.; Collier, A. M.; Mayer, J. M.; Solomon, E. I.; Hedman, B.; Stack, T.; Hodgson, K. O. J. Am. Chem. Soc. 1997, 119, 8578−8579. (49) Alp, E. E.; Shenoy, G. K.; Hinks, D. G.; Ii, D. W. C.; Soderholm, L.; Schuttler, H. B.; Guo, J.; Ellis, D. E.; Montano, P. A.; Ramanathan, M. Phys. Rev. B: Condens. Matter Mater. Phys. 1987, 35, 7199−7202. (50) Pratesi, A.; Giuli, G.; Cicconi, M. R.; Della Longa, S.; Weng, T.C.; Ginanneschi, M. Inorg. Chem. 2012, 51, 7969−7976. (51) Akeyama, K.; Kuroda, H.; Kosugi, N. Jpn. J. Appl. Phys. 1993, 32, 98−100. (52) Guo, J.; Ellis, D. E.; Goodman, G. L.; Alp, E. E.; Soderholm, L.; Shenoy, G. K. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 41, 82− 95. (53) Man, I. C.; Su, H.-Y.; Calle-Vallejo, F.; Hansen, H. A.; Martínez, J. I.; Inoglu, N. G.; Kitchin, J.; Jaramillo, T. F.; Nørskov, J. K.; Rossmeisl, J. ChemCatChem 2011, 3, 1159−1165. (54) Calle-Vallejo, F.; Díaz-Morales, O. A.; Kolb, M. J.; Koper, M. T. M. ACS Catal. 2015, 5, 869−873. (55) Diaz-Morales, O.; Ledezma-Yanez, I.; Koper, M. T. M.; CalleVallejo, F. ACS Catal. 2015, 5, 5380−5387. (56) Calle-Vallejo, F.; Inoglu, N. G.; Su, H.-Y.; Martinez, J. I.; Man, I. C.; Koper, M. T. M.; Kitchin, J. R.; Rossmeisl, J. Chem. Sci. 2013, 4, 1245−1249. (57) Pickardt, J.; Paulus, W.; Schmalz, M.; Schoellhorn, R. J. Solid State Chem. 1990, 89, 308−314. (58) Wohlfahrt-Mehrens, M.; Heitbaum, J. J. Electroanal. Chem. Interfacial Electrochem. 1987, 237, 251−260. (59) Cherevko, S.; Geiger, S.; Kasian, O.; Kulyk, N.; Grote, J.-P.; Savan, A.; Shrestha, B. R.; Merzlikin, S.; Breitbach, B.; Ludwig, A.; Mayrhofer, K. J. J. Catal. Today 2016, 262, 170−180. (60) Yeo, B. S.; Bell, A. T. J. Am. Chem. Soc. 2011, 133, 5587−5593. (61) Seitz, L. C.; Hersbach, T. J. P.; Nordlund, D.; Jaramillo, T. F. J. Phys. Chem. Lett. 2015, 6, 4178−4183. (62) Pourbaix, M. In Atlas of Electrochemical Equilibria in Aqueous Solutions; 1st Edition (Engl.); Pergamon Press: Oxford, New York, 1966; pp 286−392.

2481

DOI: 10.1021/acscatal.6b00205 ACS Catal. 2016, 6, 2473−2481