Ternary FexCo1–xP Nanowire Array as a Robust Hydrogen Evolution

Sep 27, 2016 - Density functional theory calculations further reveal that Fe substitution of Co in CoP leads to more optimal free energy of hydrogen a...
1 downloads 13 Views 2MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Ternary FexCo1−xP Nanowire Array as a Robust Hydrogen Evolution Reaction Electrocatalyst with Pt-like Activity: Experimental and Theoretical Insight Chun Tang,† Linfeng Gan,‡ Rong Zhang,† Wenbo Lu,† Xiue Jiang,‡ Abdullah M. Asiri,§ Xuping Sun,*,† Jin Wang,*,‡,∥ and Liang Chen*,⊥ †

College of Chemistry, Sichuan University, Chengdu 610064, Sichuan China State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, Jilin, China § Chemistry Department, King Abdulaziz University, Jeddah 21589, Saudi Arabia ∥ Department of Chemistry and Physics, State University of New York at Stony Brook, New York, New York 11794-3400, United States ⊥ Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo 315201, Zhejiang, China ‡

S Supporting Information *

ABSTRACT: Replacement of precious Pt with earth-abundant electrocatalysts for the hydrogen evolution reaction (HER) holds great promise for clean energy devices, but the development of low-cost and durable HER catalysts with Ptlike activity is still a huge challenge. In this communication, we report on the development of self-standing ternary FexCo1−xP nanowire array on carbon cloth (FexCo1−xP/CC) as a Pt-free HER catalyst with activities being strongly related to Fe substitution ratio. Electrochemical tests show that Fe0.5Co0.5P/ CC not only possesses Pt-like activity with the need of overpotential of only 37 mV to drive 10 mA cm−2, outperforming all nonnoble-metal HER catalysts reported to date, but demonstrates superior long-term durability in 0.5 M H2SO4. Density functional theory calculations further reveal that Fe substitution of Co in CoP leads to more optimal free energy of hydrogen adsorption to the catalyst surface. This study offers us a promising flexible monolithic catalyst for practical applications. KEYWORDS: Ternary, CoP, hydrogen evolution reaction, electrocatalyst, density functional theory

W

reaction.7 Because both HDS and HER work in the same way8 and the good electrical conductivity of TMPs offers a great benefit to electrochemical performances of catalysts, TMPs are expected to be active for the HER. Indeed, the past three years have witnessed the rapid development of TMPs for efficient HER catalysis in acids.9,10 Among such catalysts, Co phosphides have received considerable attention due to the superior catalytic activity.5,11−19 Although with such big success, it still remains a huge challenge to design and develop TMPs catalysts with Pt-like HER activity and superior long-term electrochemical durability, which is critical for practical applications. Here, we report our recent effort toward this direction in developing ternary FexCo1−xP nanowire array on conductive carbon cloth (FexCo1−xP/CC) as a monolithic HER catalyst in 0.5 M H2SO4. Fe substitution greatly enhances the HER activity which is strongly related to its substitution ratio.

ith the increasing global energy demands and concerns over climate change caused by greenhouse gases, great effort has been put to exploit clean renewable energy sources and carriers. Hydrogen is an ideal alternative to fossil fuels.1 Steam methane reforming as a primary technique for current hydrogen production suffers from high-energy heat input and CO2 emission. Photoelectrochemical water splitting or electrolysis coupled to renewable energy sources enables large-scale production of CO2-free hydrogen, promising its application for sustainable energy conversion.2 In acidic electrolytes, hydrogen evolution reaction (HER) entails electrochemical reduction of protons to form molecular hydrogen: 2H+ + 2e−H2. Active HER catalysts with low overpotential (η) is needed for successful implementation of water splitting.3,4 Precious Pt is the most active HER catalyst,5 but the high cost and scarcity severely hinder its widespread uses. It is thus highly desirable to develop cost-effective, active, and durable HER catalysts made of earth-abundant elements. Transition-metal phosphides (TMPs) are an important class of compounds with metalloid characteristics6 and have been extensively utilized to catalyze the hydrodesulfurization (HDS) © XXXX American Chemical Society

Received: August 8, 2016 Revised: September 13, 2016

A

DOI: 10.1021/acs.nanolett.6b03332 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

(020) planes of Fe0.5Co0.5P phase. Figure 1k shows the scanning TEM (SETM) image and EDX elemental mapping images of Fe, Co, and P for Fe0.5Co0.5P nanowire, confirming the uniform distribution of all elements in the whole nanowire. Figure S3a−d shows the X-ray photoelectron spectroscopy (XPS) spectra for Fe0.5Co0.5P. The survey spectrum confirms the presence of Fe, Co, and P elements (Figure S3a) with an elemental composition of Fe:Co:P close to 1:1:2. The spectrum in the Fe 2p region exhibits two spin−orbit doublets. The first doublet at 711.5 and 723.6 eV and the second doublet at 715.6 and 727.3 eV are assigned to Fe2+ and Fe3+ (Figure S3b),20 respectively. Note that the Fe 2p region shows no peaks characteristic of FeP,21 suggesting the formation of a ternary Fe0.5Co0.5P compound rather than a mixture of two solid phases. The Co 2p spectrum (Figure S3c) is fitted with two spin−orbit doublets at 778.6 and 793.5 eV assigned to the binding energies (BEs) for Co in Fe0.5Co0.5P, and the second at 781.9 and 798.4 eV are assigned to oxidized Co resulting from superficial oxidation of Fe0.5Co0.5P,5 with two shakeup satellites (identified as “Sat.”). In Figure S 3d, the high-resolution P 2p region shows two peaks at 130.0 and 129.4 eV reflecting the binding energies (BEs) of P 2p1/2 and P 2p3/2, respectively, which can be assigned to the phosphide.5 A broad feature at approximately 134.1 eV is assigned to oxidized P species arising from superficial oxidation of Fe0.5Co0.5P exposed to air.21 From the fitted XPS of Fe0.5Co0.5P, the BEs for Co 2p are higher than that of corresponding metallic Co (778.4 eV) and the BEs for P 2p are lower than that of elemental P 10 (130.2 eV). It is note that the BE (778.6 eV) for Co in Fe0.5Co0.5P is negatively shifted compared with the vale (779.1 eV) for Co in CoP,5 suggesting strong electron interactions between Fe and Co. This may have important implications in promoting the HER catalysis.20 We have simulated the XPS spectra of CoP and Fe0.5Co0.5P by using first-principles DFT calculations based on Koopmans’ theory, in which the ionization energy (IE) of one electron is approximately equal to the negative of the eigenvalue of its orbital in closed-shell Hartee-Fock theory. It is known that the BE is relevant to IE of the core electrons. Note that the energy shift rather than the absolute value is meaningful in our calculations, since the calculated IE is generally smaller than the experimentally measured XPS value. In our calculations, we found that the Co 2p and P 2p peaks in Fe0.5Co0.5P are negatively shifted by 0.41 and 0.35 eV compared to CoP, respectively. This is consistent with the XPS experimental observation in the present study. The difference can be attributed to the slightly lower electron-negativity of Fe than Co (1.83 vs 1.88). Indeed, on the basis of Bader charge analysis, each Co atom in Fe0.5Co0.5P donates 0.2 less electrons than in CoP. We examined the electrocatalytic HER activity of FexCo1−xP/CC (x = 0, 0.25, 0.33, and 0.50) with catalyst loading of 2.2 mg cm−2 in 0.5 M H2SO4 in a typical threeelectrode setup with a scan rate of 2 mV s−1. Commercial Pt/C deposited on CC with Pt loading of 2.2 mg cm−2 and bare CC were also examined for comparison. Figure 2a and b shows the linear sweep voltammetry (LSV) curves on the reversible hydrogen electrode (RHE). Pt/C on CC shows excellent activity, while bare CC has no HER activity within the examined potential window. FexCo1−xP/CC are highly efficient toward the HER. Remarkably, Fe0.5Co0.5P/CC electrode exhibits Pt-like HER activity with the lowest onset potential close to that of commercial Pt/C and only needs overpotential

Notably, Fe0.5Co0.5P/CC electrode shows outstanding Pt-like HER activity needing overpotential of only 37 mV to drive 10 mA cm−2, outperforming all non-noble-metal HER catalysts reported to date, and density function theory (DFT) calculations reveal its optimal free energy of hydrogen adsorption. It also shows remarkably superior long-term durability. A series of FexCo1−xP nanowire arrays were grown on CC via conversion reaction from their corresponding Fex Co 1−x precursors (see SI for preparation details). We investigated the morphology, structure, and composition of the catalysts (Figure 1, S1 and S2). Figure 1a shows a low-magnification

Figure 1. SEM images for (a, b) Fe0.5Co0.5-precursor on CC and (c, d) Fe0.5Co0.5P/CC. TEM images of one single (e) Fe0.5Co0.5-precursor and (f) Fe0.5Co0.5P nanowire. (g) EDX spectrum of Fe0.5Co0.5P/CC. (h) XRD pattern for Fe0.5Co0.5P. (i) HRTEM image and (j) SAED pattern taken from Fe0.5Co0.5P nanowire. (k) STEM image and EDX elemental mapping of Fe, Co, and P for Fe0.5Co0.5P nanowires.

scanning electron microscopy (SEM) image of Fe0.5Co0.5precursor, indicating uniform coverage of the entire CC with nanowire array. The high-magnification SEM image (Figure 1b) further reveals that such nanowires have diameters in the range of 100−200 nm. The transmission electron microscopy (TEM) image shows Fe0.5Co0.5-precursor nanowire has a smooth surface (Figure 1e). After phosphidation, the array remains intact (Figure 1c and d) with the preservation of 1D morphology and integration feature but with a rough surface (Figure 1f). The energy-dispersive X-ray (EDX) spectrum reveals the existence of Fe, Co, and P (Figure 1g) and inductively coupled plasma mass spectrometry (ICP-MS) analysis suggests nearly a 1:1:2 atomic ratio for Fe:Co:P (Table S1). The X-ray diffraction (XRD) pattern for Fe0.5Co0.5P is consistent with a CoP standard pattern (JCPDS No. 29-0497), as shown in Figure 1h. The high-resolution TEM (HRTEM) image taken from the nanowire (Figure 1i) reveals well-resolved lattice fringes with interplanar distance of 2.85 Å, slightly larger than CoP (2.83 Å) due to the substitution of Fe for Co, which is indexed to the (011) plane of Fe0.5Co0.5P. The selected area electron diffraction (SAED) pattern (Figure 1j) exhibits discrete spots indexed to the (011), (111), (211), and B

DOI: 10.1021/acs.nanolett.6b03332 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

electrochemical surface areas, by a simple cyclic voltammetry method25 (Figure S4). From Figure 2e, it is clear that the Cdl of Fe0.5Co0.5P/CC (32 mF cm−2) is higher than that of CoP/CC (18 mF cm−2), Fe0.25Co0.75P/CC (29 mF cm−2), and Fe0.33Co0.66P/CC (30 mF cm−2). It suggests that increasing Fe content from 0 to 0.5 leads to an increase in surface roughness, which is beneficial to enhance the HER activity.25 Figure 2f shows the current density at overpotential of 100 mV as a function of electrochemical capacitance for all four electrodes. Although there is almost a linear correlation between the activity and double layer capacitance for the Fe− Co−P catalysts, the activity of CoP/CC is significantly lower than predicted by this correlation. Also note that, as protons were adsorbed and desorbed during the CV cycling, there might be pseudocapacitance contribution on the top of double layer capacitance.26 Such pseudocapacitance could potentially lead to an overestimation of electrochemical surface areas. DFT calculations reveal that the strong electron interactions involving the ions and synergetic effects of Fe, Co, and P are also responsible for the superior HER activity for Fe−Co−P catalysts. Stability is also a key concern for all catalysts. For this purpose, we tested the stability of Fe0.5Co0.5P/CC electrode by cyclic voltammetry scan between 0.20 and −0.20 V vs RHE at an accelerated scan rate of 100 mV s−1. After 3000 cycles in 0.5 M H2SO4, the polarization curve overlays almost exactly with the initial one (Figure 3a). Besides, SEM images (Figure S5)

Figure 2. (a, b) LSV curves for Pt/C on CC, bare CC, and FexCo1−xP/CC with a scan rate of 2 mV s−1 for HER. (c) Tafel plots for Pt/C on CC and FexCo1−xP/CC. (d) Overpotentials at 10 mA cm−2 vs RHE (left) and Tafel slopes (right). (e) Capacitive currents as a function of scan rate for FexCo1−xP/CC. (f) Current density at overpotential of 100 mV as a function of electrochemical capacitance of FexCo1−xP/CC.

of 37 mV to drive 10 mA cm−2. This overpotential is only 7 mV more than that needed by Pt catalyst and compares favorably to the behaviors of Co phosphides catalysts, including CoP/CC (η10mAcm−2 = 67 mV),5 hollow CoP nanoparticles (η10mAcm−2 = 75 mV),11 CoP/CNT (η10mAcm−2 = 122 mV),12 CoP/Ti (η10mAcm−2 = 90 mV),13 Co2P/Ti (η10mAcm−2 = 95 mV),14 urchin-like CoP nanocrystals (η10mAcm−2 = ∼ 95 mV),15 Co2P branched nanostructures (η10mAcm−2 = 120 mV),16 CoP/rGO-400 (η10mAcm−2 = 105 mV),17 CoP hollow polyhedron (η10mAcm−2 = 159 mV),18 and CoP2/RGO (η10mAcm−2 = 70 mV).19 This catalyst rivals the performance of Co-based ternary TMPs including Fe0.5Co0.5P (η10mAcm−2 ≈ 130 mV)22 and Co0.59Fe0.41P (η10mAcm−2 ≈ 72 mV)23 as well as most active non-noble-metal CoPS (η10mAcm−2 = 48 mV).24 It also compares favorably to the behaviors of other non-noble-metal HER catalysts in acids (Table S1). Figure 2c shows the Tafel plots. Pt/C shows a Tafel slope of 28 mV dec−1. CoP/CC, Fe0.25Co0.75P/CC, Fe0.33Co0.66P/CC, and Fe0.5Co0.5P/CC exhibit Tafel slopes of 55, 38, 37, and 30 mV dec−1, respectively. Such Tafel slopes suggest that the HER occurs on FexCo1−xP/CC via a Volmer−Heyrovsky mechanism. The HER performances were compared quantitatively using the required overpotentials for 10 mA cm−2 (η10mAcm−2, left of Figure 2d). The overpotential for Fe0.5Co0.5P/CC is much smaller than those for CoP/CC (103 mV), Fe0.25Co0.75P/ CC (64 mV), and Fe0.33Co0.66P/CC (52 mV). The lowest overpotential and smallest Tafel slope (right of Figure 2d) for Fe0.5Co0.5P/CC imply its superior HER activity. To make a better comparison of the activity of FexCo1−xP/ CC described here, we tested the electrochemical double layer capacitances (Cdl) of the catalysts, proportional to their

Figure 3. (a) LSV curves for Fe0.5Co0.5P/CC before and after 3000 CV cycles. (b) Time-dependent current density curves of Fe0.5Co0.5P/ CC at fixed overpotentials of 37, 72, and 98 mV to drive 10, 50, and 100 mA cm−2, respectively.

indicate that this catalyst electrode still maintains its morphology after 3000 cycles. ICP-MS analysis of this catalyst electrode further suggests nearly a 1:1:2 atomic ratio for Fe:Co:P. This confirms stable performance among non-noble metal HER electrocatalysts. The long-term electrochemical stability of this electrode was also examined in bulk electrolysis of water. The overpotentials at fixed of 37, 72, and 98 mV to drive 10, 50, and 100 mA cm−2, respectively, are negligibly declined over 100 h, as shown in Figure 3b. However, a further increase of Fe content leads to decreased activity and stability (Figure S6). Hydrogen evolution activity is strongly correlated with the free energy of hydrogen adsorption to the catalyst surface, ΔGH.27 ΔGH is a natural parameter used to quantify the hydrogen−metal bond strength, and a ΔGH value of 0.0 eV leads to optimal HER activity with intermediate binding energies due to the balance between the rate of proton reduction and the ease of removal of adsorbed hydrogen from the surface.28 ΔGH value determined by DFT calculations is a reasonable descriptor of hydrogen evolution activity for HER catalysts. We thus used DFT to calculate the ΔGH on C

DOI: 10.1021/acs.nanolett.6b03332 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

hydrogen fuels but would open an exciting new direction to the rational design and scalable fabrication of ternary TMPs nanoarrays as a cost-effective, high-active, and robust 3D electrode for energy and sensing applications.

FexCo1−xP (101) surface. Note that Pt as the most active HER catalyst has a ΔGH value of nearly −0.09 eV. We investigated all possible active sites of FexCo1−xP, including different Co, Fe, and P sites (Figure 4a−d). For CoP, hydrogen atom adsorbed

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b03332. Experimental details, SEM images, EDX and XPS spectra, Tabled S1 and S2, cyclic voltammograms, LSV and time-dependent current density curves (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (X.S.). *E-mail: [email protected] (J.W.). *E-mail: [email protected] (L.C.). Figure 4. Schematic structural representations for hydrogen adsorption at (a) CoP, (b) Fe0.25Co0.75P, (c) Fe0.33Co0.66P, and (d) Fe0.5Co0.5P. (e) Free energy diagram of HER under favorable Co site of hydrogen coverage on surface of FexCo1−xP.

Author Contributions

at P site is unstable, but adsorbed at Co site is stable, with ΔGH value of −0.14 eV (Figure 4e). On the Fe0.25Co0.75P (101) surface, the hydrogen adsorption strength is slightly weakened in the presence of Fe, which has a lower electron negativity than Co. The resulting ΔGH value is calculated to be −0.132 eV (Figure 4e). On the Fe0.33Co0.66P (101) surface, the hydrogen atom can be adsorbed at Co, Fe, and P atom sites, as well as the Co−Fe bridge site. The ΔGH value for the favorable Co site is determined to be −0.106 eV (Figure 4e). On the Fe0.5Co0.5P (101) surface, hydrogen shows a similar adsorption behavior with that of Fe0.33Co0.66P surface, but with a ΔGH value of −0.088 eV for the favorable Co site (Figure 4e). Above computing results suggest that the Fe substitute in the catalyst leads to more optimal ΔGH. CoP has the lowest ΔGH value relatively far from the ideal value zero. Upon the doping of Fe, which has slightly lower electron-negativity than Co, the binding strength of hydrogen on Co sites would be weakened. On the (101) facet, H prefers the Co site so that the negative shift of BE for Co would weaken the H−Co binding strength and accordingly increase the ΔGH toward zero. Besides, only Co sites absorb hydrogen atom, making hydrogen coverage less than that on FexCo1−xP. These two reasons explain the inferior activity of CoP among all four catalysts. Although Fe0.5Co0.5P has the same hydrogen coverage as CoP, its ΔGH value is closest to zero. Thus, Fe0.5Co0.5P has the highest HER activity. Fe0.25Co0.75P and Fe0.33Co0.66P have ΔGH values between CoP and Fe0.5Co0.5P, implying their moderate HER activity. In summary, self-standing ternary FexCo1−xP nanowire array has been demonstrated as a superior catalyst for electrochemical hydrogen evolution in acids. Fe substitution of Co in CoP has a heavy influence on the HER performances, and Fe0.5Co0.5P/CC electrode shows Pt-like activity and only needs overpotential of 37 mV to drive 10 mA cm−2 with excellent long-term durability. DFT calculations suggest that Fe substitution leads to more optimal free energy of hydrogen adsorption. This study not only offers us a flexible monolithic catalyst in technological devices for large-scale production of

ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (No. 21575137) and Science and Technology Innovation Foundation of Jilin Province for Talents Cultivation (Grants 20150519014JH). The Network Center of Changchun Normal University is also appreciated for the provision of computing resource.

C. Tang and L. Gan contributed equally. Notes

The authors declare no competing financial interest.

■ ■

REFERENCES

(1) Hoffert, M. I.; Caldeira, L.; Benford, G.; Criswell, D. R.; Green, C.; Herzog, H.; Jain, A. K.; Kheshgi, H. S.; Lackner, K. S.; Lewis, J. S.; Lightfoot, H. D.; Manheimer, W.; Mankins, J. C.; Mauel, M. E.; Perkins, L. J.; Schlesinger, M. E.; Volk, T.; Wigley, T. M. L. Science 2002, 298, 981−987. (2) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15729−15735. (3) Walter, M. G.; Warren, E. L.; Mckone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis, N. S. Chem. Rev. 2010, 110, 6446−6473. (4) Wang, J.; Cui, W.; Liu, Q.; Xing, Z.; Asiri, A. M.; Sun, X. Adv. Mater. 2016, 28, 215−230. (5) Tian, J.; Liu, Q.; Asiri, a. M.; Sun, X. J. Am. Chem. Soc. 2014, 136, 7587−7590. (6) Oyama, S. T.; Gott, T.; Zhao, H.; Lee, Y. K. Catal. Today 2009, 143, 94−107. (7) Carenco, S.; Portehault, D.; Boissière, C.; Mèzailles, N.; Sanchez, C. Chem. Rev. 2013, 113, 7981−8065. (8) Liu, O.; Rodriguez, J. A. J. Am. Chem. Soc. 2005, 127, 14871− 14878. (9) Xiao, P.; Chen, W.; Wang, X. Adv. Energy Mater. 2015, 5, 1500985. (10) Shi, Y.; Zhang, B. Chem. Soc. Rev. 2016, 45, 1529−1541. (11) Popczun, E. J.; Read, C. G.; Roske, C. W.; Lewis, N. S.; Schaak, R. E. Angew. Chem., Int. Ed. 2014, 53, 5427−5430. (12) Liu, Q.; Tian, J.; Asiri, A. M.; Sun, X. Angew. Chem., Int. Ed. 2014, 53, 6710−6714. (13) Pu, Z.; Liu, Q.; Jiang, P.; Asiri, A. M.; Obaid, A. Y.; Sun, X. Chem. Mater. 2014, 26, 4326−4329. (14) Callejas, J. F.; Read, C. G.; Popczun, E. J.; McEnaney, J. M.; Schaak, R. E. Chem. Mater. 2015, 27, 3769−3774. D

DOI: 10.1021/acs.nanolett.6b03332 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters (15) Yang, H.; Zhang, Y.; Hu, F.; Wang, Q. Nano Lett. 2015, 15, 7616−7620. (16) Popczun, E. J.; Roske, C. W.; Read, C. G.; Crompton, J. C.; McEnaney, J. M.; Callejas, J. F.; Lewis, N. S.; Schaak, R. E. J. Mater. Chem. A 2015, 3, 5420−5425. (17) Jiao, L.; Zhou, Y.; Jiang, H. Chem. Sci. 2016, 7, 1690−1695. (18) Liu, M.; Li, J. ACS Appl. Mater. Interfaces 2016, 8, 2158−2165. (19) Wang, J.; Yang, W.; Liu, J. J. Mater. Chem. A 2016, 4, 4686− 4690. (20) Long, X.; Li, J.; Wang, Z.; Zhu, H.; Zhang, T.; Xiao, S.; Guo, W.; Yang, S. J. Am. Chem. Soc. 2015, 137, 11900−11903. (21) Jiang, P.; Liu, Q.; Liang, Y.; Tian, J.; Asiri, A. M.; Sun, X. Angew. Chem., Int. Ed. 2014, 53, 12855−12859. (22) Kibsgaard, J.; Tsai, C.; Chan, K.; Benck, J. D.; Norskov, J. K.; Abild-Pedersen, F.; Jaramillo, T. F. Energy Environ. Sci. 2015, 8, 3022− 3029. (23) Hao, J.; Yang, W.; Zhang, Z.; Tang, J. Nanoscale 2015, 7, 11055−11062. (24) Cabán-Acevedo, M.; Stone, M. L.; Schmidt, J. R.; Thomas, J. R.; Ding, Q.; Chang, H.-C.; Tsai, M.-L.; He, J.-H.; Jin, S. Nat. Mater. 2015, 14, 1245−1251. (25) Tian, J.; Liu, Q.; Cheng, N.; Asiri, A. M.; Sun, X. Angew. Chem., Int. Ed. 2014, 53, 9577−9581. (26) Wang, G.; Zhang, L.; Zhang, J. Chem. Soc. Rev. 2012, 41, 797− 828. (27) Norskov, J. K.; Bligaard, T.; Logadottir, A.; Kitchin, J. R.; Chen, J. G.; Pandelov, S.; Stimming, U. J. Electrochem. Soc. 2005, 152, J23− J26. (28) Greeley, J.; Norskov, J. K. Surf. Sci. 2007, 601, 1590−1598.

E

DOI: 10.1021/acs.nanolett.6b03332 Nano Lett. XXXX, XXX, XXX−XXX