Oxide Interface Nanostructures Generated ... - ACS Publications

Letter pubs.acs.org/NanoLett

Metal/Oxide Interface Nanostructures Generated by Surface Segregation for Electrocatalysis Zhe Weng,†,‡ Wen Liu,†,‡ Li-Chang Yin,∥ Ruopian Fang,∥ Min Li,§ Eric I. Altman,§ Qi Fan,†,‡ Feng Li,∥ Hui-Ming Cheng,∥ and Hailiang Wang*,†,‡ †

Energy Sciences Institute, ‡Department of Chemistry, and §Department of Chemical and Environmental Engineering, Yale University, New Haven, Connecticut 06520, United States ∥ Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China S Supporting Information *

ABSTRACT: Strong metal/oxide interactions have been acknowledged to play prominent roles in chemical catalysis in the gas phase, but remain as an unexplored area in electrocatalysis in the liquid phase. Utilization of metal/oxide interface structures could generate high performance electrocatalysts for clean energy storage and conversion. However, building highly dispersed nanoscale metal/oxide interfaces on conductive scaffolds remains a significant challenge. Here, we report a novel strategy to create metal/oxide interface nanostructures by growing mixed metal oxide nanoparticles on carbon nanotubes (CNTs) and then selectively promoting migration of one of the metal ions to the surface of the oxide nanoparticles and simultaneous reduction to metal. Employing this strategy, we have synthesized Ni/CeO2 nanointerfaces coupled with CNTs. The Ni/CeO2 interface promotes hydrogen evolution catalysis by facilitating water dissociation and modifying the hydrogen binding energy. The Ni/CeO2−CNT hybrid material exhibits superior activity for hydrogen evolution as a result of synergistic effects including strong metal/oxide interactions, inorganic/carbon coupling, and particle size control. KEYWORDS: metal/oxide interface, electrocatalysis, surface segregation, hydrogen evolution

N

Ni/NiO nanoparticles attached on carbon nanotubes (CNTs).24 In another study, Co/Fe3O4 interface has been reported to facilitate four-electron transfer and increase catalytic stability for oxygen reduction reaction in alkaline media.25 To explore the potential of utilizing metal/oxide interfaces to improve electrocatalysis for challenging energy reactions, we believe it is of paramount importance to rationally design and controllably synthesize a variety of metal/oxide interface nanostructures on conductive scaffolds and probe the roles played by metal/oxide interactions in electrocatalysis. Here we report a novel general strategy to create highperformance electrocatalytic metal/oxide interfaces by growing mixed metal oxide nanoparticles on CNTs and then selectively promoting migration of one of the metals to the surface of the oxide nanoparticles. We have taken a Ni/CeO2−CNT material as a model system to demonstrate our approach (Figure 1). In the first stage of material synthesis, solution-phase reactions are used to directly and selectively grow Ni doped CeO2 (NixCeO2+x) nanoparticles on CNTs. The nanotubes not only provide anchoring sites for size control and uniform

anostructuring and metal/oxide interface engineering are emerging as two effective routes to enhance performances of heterogeneous catalyst materials.1−5 On the one hand, reducing the size of catalyst particles to the nanoscale drastically increases the percentage of surface atoms compared to the bulk counterparts.6−8 On the other hand, oxide supports play key roles in catalytic processes by providing dual active sites at metal/oxide interfaces and/or strongly influencing the physical and chemical properties of metal nanoparticles, leading to better catalytic activity, selectivity, and durability.9,10 For example, various metals such as Pt, Pd, Au, Cu, and Ni supported on CeO2 exhibit improved catalytic properties for a number of important reactions including CO oxidation, CO2 hydrogenation, and water−gas shift reactions, as a result of the strong metal/oxide interactions.11−21 While reinforced catalytic effects from metal/oxide interfaces have widely been recognized and applied in chemical catalysis in the gas phase, only a handful of such reports exist for electrocatalysis in the liquid phase.2,22−25 For instance, combining Pt with Ni(OH)2 enhances hydrogen evolution reaction (HER) in alkaline solution, as the Ni(OH)2 sites may promote water dissociation by interacting with the hydroxyl groups, and the hydrogen atoms adsorbed on the nearby Pt sites can subsequently combine to give molecular hydrogen.22,23,26 A similar enhancing effect is observed for core/shell © 2015 American Chemical Society

Received: September 14, 2015 Revised: October 23, 2015 Published: October 28, 2015 7704

DOI: 10.1021/acs.nanolett.5b03709 Nano Lett. 2015, 15, 7704−7710

Letter

Nano Letters

nanoparticles diffuse to the surface and are reduced to metallic Ni to form the desired Ni/CeO2 interfaces. The segregation process leads to fine Ni nanoparticles or even possibly dispersed Ni atoms or clusters growing out of the CeO2 nanoparticles, generating abundant and intimate metal/oxide interfaces. Owing to the combined effects of strong metal/oxide interactions, inorganic/carbon coupling, and particle size control, our Ni/CeO2−CNT material exhibits superior catalytic activity and stability for HER in alkaline solution, close to that seen for Pt precious metal catalysts. The Ni/CeO2 interface not only promotes the dissociation of water molecules to form hydrogen adatoms but also lowers the hydrogen binding energy (HBE) to accelerate the desorption of molecular hydrogen, as revealed by density functional theory (DFT) calculations.

Figure 1. Schematic synthesis strategy for creating nanoscale Ni/CeO2 interfaces on CNTs.

dispersion of the oxide nanoparticles but also can serve as electron conduction pathways during electrocatalytic processes. In the second stage, a reductive gas atmosphere is applied for surface segregation to take place. The Ni atoms in NixCeO2+x

Figure 2. (a) XRD pattern, (b) XPS spectrum, (c) low-magnification TEM image, (d) high-resolution TEM image, and (e) STEM image of NixCeO2+x−CNT. (f−j) EDS maps of Ce, Ni, C, and O elements in NixCeO2+x−CNT and their overlay with the STEM image. 7705


Letter

Nano Letters

Figure 3. (a) XRD pattern, (b) XPS spectrum, (c) low-magnification TEM image, (d) high-resolution TEM image, and (e) STEM image of Ni/ CeO2−CNT. (f−j) EDS maps of Ce, Ni, C, and O elements in Ni/CeO2−CNT and their overlay with the STEM image.

The Ni/CeO2−CNT material was synthesized in two steps (Figure 1). In the first step, NixCeO2+x nanoparticles were directly grown on mildly oxidized multiwall CNTs by solutionphase synthesis. Before solvothermal treatment at 180 °C, Ni(OAc)2 and Ce(NO3)3 (Ni/Ce = 1/1 in mole) were hydrolyzed in aqueous N,N-dimethylformamide (DMF) solution in the presence of CNTs and hexamethylenetetramine (HMTA) at 90 °C. The mixed solvent and relatively low reaction temperature allowed sufficient interactions between the metal ions and the oxygen-containing functional groups on CNTs so that selective nucleation and uniform growth of nanoparticles on CNTs could be achieved.27,28 The resulting nanoparticles, having an average diameter of ∼4 nm, were found to be selectively anchored and uniformly distributed on the surface of CNTs (Figure 2c,d). The X-ray diffraction (XRD) pattern of the product was found to be almost identical to that of pure CeO2 (Figure 2a), which is consistent with

literature results that Ni doping does not significantly change the lattice parameters of CeO2.29 High-resolution transmission electron microscopy (TEM) imaging revealed the lattice fringes of the NixCeO2+x nanoparticles (Figure 2d), which agrees well with the XRD result. X-ray photoelectron spectroscopy (XPS) confirmed that Ce and Ni coexist in the material, and they are in the 4+ and 2+ oxidation states, respectively (Figure 2b).30,31 The distribution of the two elements in the material was analyzed by energy dispersive spectroscopy (EDS) coupled with scanning transmission electron microscopy (STEM). The EDS maps of Ce, Ni, and O almost overlap, clearly indicating the nanoparticles are mixed metal oxides (Figure 2e−j). It is also evident from the EDS mapping results that the mixed oxide nanoparticles are selectively attached to the CNTs. A bulk Ni/Ce molar ratio close to 1:1 can be derived from the EDS spectrum, which is almost the same as that revealed by 7706


Letter

Nano Letters

Figure 4. (a) Polarization curves of Ni/CeO2−CNT, NixCeO2+x−CNT, Ni−CNT, and commercial Pt/C in 1 M KOH solution. Catalyst mass loading is 0.14 and 0.1 mg cm−2 based on the mass of Ni and Pt metal elements, respectively. (b) Amperometric I−t curve of the Ni/CeO2−CNT at a constant overpotential of 0.153 V.

CNT material was examined by TEM. It was observed that the Ni/CeO2 nanoparticles were still selectively attached to and well dispersed on the CNTs (Figure 3c). High-resolution TEM imaging further confirmed the presence of metallic Ni nanoparticles in direct contact with CeO2 nanoparticles (Figure 3d). The average size of the CeO2 nanoparticles (∼7 nm) is larger than that of the original NixCeO2+x nanoparticles (∼4 nm). The Ni nanoparticles are in the size range of 2−5 nm. EDS was employed to reveal the distribution of elements in the material structure. The Ce map overlaps well with the O map, consistent with the presence of a CeO2 phase (Figure 3f,i). The Ni map does not overlap with the Ce or O map, indicating the presence of a separate metallic Ni phase (Figure 3g). The EDS mapping clearly illustrated that the Ni had indeed migrated to the CeO2 surface and the newly formed Ni/CeO2 heterostructured nanoparticles were still firmly anchored on the CNT surface (Figure 3e−j). The Ni/CeO2−CNT material showed superior catalytic activity for HER in 1 M aqueous KOH solution. Indeed, the HER onset potential was similar to that of a commercial Pt/C catalyst (Figure 4a). The Ni/CeO2 was identified as the active phase for HER because the NixCeO2+x−CNT material was almost inactive at any overpotential lower than 450 mV (Figure 4a, black trace). At a Ni mass loading of 0.14 mg/cm2, the Ni/ CeO2−CNT catalyst was able to deliver a current density of 10 mA/cm2 at an overpotential less than 100 mV (Figure 4a, red trace), representing one of the most active HER catalyst materials yet seen in alkaline solutions (Table S1).22,24,32−36 The Ni/CeO2−CNT catalyst also showed good durability. The catalytic current density was maintained at around 30 mA cm−2 during 10 h of continuous hydrogen evolution operation at an overpotential of ∼150 mV (Figure 4b). As a control, the Ni−CNT material showed an HER onset potential that was >100 mV more negative than the Ni/CeO2− CNT. At an overpotential of 150 mV, the catalytic current density of the Ni/CeO2−CNT (∼32 mA cm−2) was much higher than that of the Ni−CNT (∼4 mA cm−2). The superior HER catalytic activity of Ni/CeO2−CNT arises from two aspects. The Ni nanoparticles of the Ni/CeO2−CNT are considerably smaller in size than those of the Ni−CNT material (Figures 2d and S3c), which could provide more surface sites for the same mass of Ni. Importantly, the Ni/CeO2 interface plays a key role in enhancing the catalytic performance. In a separate control experiment, we deposited CeO2 onto cleaned

surface-sensitive XPS, further supporting our conclusion that Ni is uniformly distributed in the CeO2 lattice. The Ni/Ce molar ratio was varied in control experiments. In the case where no Ni precursor was added in the synthesis, ∼10 nm sized CeO2 nanocubes formed on CNTs (Figure S1d). When the Ni/Ce ratio increased to 1/3, smaller (∼5 nm) nanoparticles of irregular polyhedral shapes were produced on CNTs (Figure S1c). The obvious differences in both the size and shape of the NixCeO2+x nanoparticles compared to the pure CeO2 case are strong indications of Ni doping. The single NixCeO2+x phase continued to be observed as the Ni/Ce ratio was further increased to 1/1 (Figures S2 and 2a). This is different from previous reports that no more than 20% of Ni can be incorporated into the CeO2 structure.20,29 At a Ni/Ce ratio of 3/1, a second phase of α-Ni(OH)2 appeared, having a nanoplate morphology, in addition to the NixCeO2+x phase (Figures S1b and S2). In the case of 100% Ni, only nanoplates of α-Ni(OH)2 were formed on CNTs (Figures S1a and S2). The Ni/CeO2 interface was created in the second step during which the products from the first step were treated in H2/Ar atmosphere at 500 °C. During this treatment metallic Ni nanoparticles grew out of the original NixCeO2+x nanoparticles as the Ni ions migrated to the surface and were reduced by H2. Such a migration and reduction process led to the formation of Ni/CeO2 heterostructured nanoparticles on CNTs, a finding that was verified by multiple characterization techniques. XRD was used to provide phase information for the final material. Diffraction peaks corresponding to Ni metal were clearly observed in addition to the CeO2 peaks (Figure 3a). XPS measurements were then performed to characterize surface composition and oxidation states. The Ni 2p3/2 peak at the binding energy of 852.7 eV is attributed to Ni(0) (Figure 3b), supporting the existence of metallic Ni on the surface of the material.21,31 Ni 2p3/2 and 2p1/2 peaks corresponding to Ni(II) were also observed at binding energies of 855.7, 861.3, and 873.5 eV (Figure 3b), which is probably due to surface oxidation of Ni under ambient conditions. The same phenomenon also occurred in a Ni−CNT control sample (Figure S3b); this was derived from the Ni(OH)2−CNT material by reduction under the same conditions as in the synthesis of the Ni/CeO2−CNT. No significant change in the oxidation state of the Ce takes place during the surface segregation, with Ce(IV) being the dominant valence state throughout (Figure 3b). The microstructure of the Ni/CeO2− 7707


Letter

Nano Letters

Figure 5. (a) DFT calculated reaction energy diagram of water dissociation for Ni1/CeO2(111) and Ni(111). (b) DFT calculated HBE for Ni1/ CeO2(111), Ni(111), and CeO2(111) systems. (c) The volcano plot of the experimentally measured exchange current density versus the DFT calculated HBE for Ni/CeO2−CNT (red star), Ni−CNT (blue star), commercial Pt/C (green star), and common metal catalysts (black squares; data collected from ref 41).

CNTs. We found that the segregation process could even happen at a temperature as low as 300 °C, which was corroborated by the HER activity of the resulted material (Figure S6). EDS mapping suggested that the Ni had very similar distribution as the Ce in the material (Figure S7). It is thus likely that the segregated Ni exists in the form of dispersed atoms or nanoclusters on the mixed oxide surface without forming detectable Ni nanoparticles. Such a synthetic methodology could potentially be utilized to generate similar structures with atomically distributed metals having potentially useful catalytic performance.41−43 On increasing the temperature to 400−500 °C, the resulting materials exhibited incrementally higher catalytic activity for HER (Figure S6), likely due to an incrementally larger amount of Ni segregated to the surface increasing the number of active sites. Nevertheless, a further increase of the temperature to 600 °C caused significant sintering of the nanoparticles (Figure S8), and the catalytic activity of the material decreased drastically (Figure S6). The presence of CNTs is important: free NixCeO2+x nanoparticles without CNTs severely agglomerated even at 500 °C (Figure S9). This shows the vital importance of the CNTs in providing anchoring sites for the nanoparticles and keeping them small and dispersed during the segregation process. In addition, the CNTs are highways for electron transport, and their onedimensional structures can create abundant macropores for improved mass transport kinetics, both of which may facilitate electrocatalysis.44,45 In summary, we have synthesized size-controlled Ni/CeO2 metal/oxide interface nanostructures on highly conductive CNTs by a novel solvation−segregation method. The Ni/ CeO2−CNT hybrid material exhibits superior HER catalytic activity attributed to the nanoscale Ni/CeO2 interface that promotes water dissociation and renders optimum hydrogen adsorption energy. Our strategy can be deployed to synthesize a wide series of metal/oxide interface nanostructures for exploring metal/oxide interactions in liquid-phase electrocatalysis as well as affording high-performance electrocatalyst materials.

commercial Ni foam with the result that improved catalytic activity for HER was also observed (Figure S4), further confirming the promotion effect of the Ni/CeO2 interface. The high electrocatalytic performance of the Ni/CeO2−CNT material is a direct consequence of our designed synthesis, which enables the production of abundant and intimate metal/ oxide interfaces while maintaining the small size of the nanoparticles and their attachment to highly conductive CNTs. To further understand the mechanistic roles played by the Ni/CeO2 interface during electrocatalysis of HER, we performed a series of DFT calculations to simulate the energetics of the key reaction steps on a model Ni1/ CeO2(111) catalyst. As reported in previous work,23,26,37,38 catalytic activity of HER in alkaline solution is controlled both by the activation energy barrier for water dissociation as well as the adsorption and combination rate of reactive hydrogen intermediates (H*). Since water is the hydrogen source, its dissociation is considered as the first rate-determining step.23 As shown in Figure 5a, Ni/CeO2 interactions lower the energy barrier of OH−H bond cleavage to 0.21 eV, which is only a quarter of that on the Ni(111) surface, consistent with data reported earlier.39 The Ni/CeO2 interface promotes water dissociation and thus increases the rate of H* formation. In the subsequent step that converts H* to H2, there exists an optimal HBE that balances the energetics of H* adsorption and H2 desorption.40 Our DFT results show that the HBE (ΔEH) on either CeO2(111) or Ni(111) surface is stronger than optimum (Figure 5b), which would suppress H* desorption and thus hinder H2 production. Interestingly, combining Ni with CeO2 can tune the HBE to a more moderate value (−0.33 eV), which is almost equal to that on a Pt catalyst.40 For comparison, we incorporated the calculated HBE values along with the measured HER activities (represented by exchange current densities derived from Figure S5) for the catalyst materials we tested into the volcano plot established by Nørskov et al. (Figure 5c).40 It is clear that the Ni/CeO2−CNT displays nearly optimum HBE and excellent HER activity just like Pt. In contrast, the Ni−CNT shows significantly stronger HBE and lower activity. Hence, the Ni/CeO2 interface is able to facilitate the HER pathway in the Ni/CeO2−CNT catalyst by lowering the energy barrier for water dissociation and tuning the HBE near to the optimum value that balances hydrogen adsorption and desorption. As it is the critical step that permits formation of the Ni/ CeO2 interface, we investigated the surface segregation of NixCeO2+x nanoparticles at various temperatures anchored on

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.5b03709. 7708


Letter

Nano Letters

■

(16) Cargnello, M.; Jaen, J. J. D.; Garrido, J. C. H.; Bakhmutsky, K.; Montini, T.; Gamez, J. J. C.; Gorte, R. J.; Fornasiero, P. Science 2012, 337, 713−717. (17) Fu, Q.; Saltsburg, H.; Flytzani-Stephanopoulos, M. Science 2003, 301, 935−938. (18) Rodriguez, J. A.; Ma, S.; Liu, P.; Hrbek, J.; Evans, J.; Perez, M. Science 2007, 318, 1757−1760. (19) Graciani, J.; Mudiyanselage, K.; Xu, F.; Baber, A. E.; Evans, J.; Senanayake, S. D.; Stacchiola, D. J.; Liu, P.; Hrbek, J.; Sanz, J. F.; Rodriguez, J. A. Science 2014, 345, 546−550. (20) Zhou, G.; Barrio, L.; Agnoli, S.; Senanayake, S. D.; Evans, J.; Kubacka, A.; Estrella, M.; Hanson, J. C.; Martinez-Arias, A.; Fernandez-Garcia, M.; Rodriguez, J. A. Angew. Chem., Int. Ed. 2010, 49, 9680−9684. (21) Zhou, Y. H.; Zhou, J. J. Phys. Chem. C 2012, 116, 9544−9549. (22) Yin, H.; Zhao, S.; Zhao, K.; Muqsit, A.; Tang, H.; Chang, L.; Zhao, H.; Gao, Y.; Tang, Z. Nat. Commun. 2015, 6, 6430. (23) Subbaraman, R.; Tripkovic, D.; Strmcnik, D.; Chang, K. C.; Uchimura, M.; Paulikas, A. P.; Stamenkovic, V.; Markovic, N. M. Science 2011, 334, 1256−1260. (24) Gong, M.; Zhou, W.; Tsai, M. C.; Zhou, J. G.; Guan, M. Y.; Lin, M. C.; Zhang, B.; Hu, Y. F.; Wang, D. Y.; Yang, J.; Pennycook, S. J.; Hwang, B. J.; Dai, H. J. Nat. Commun. 2014, 5, 4695. (25) Wang, C. H.; Yang, C. W.; Lin, Y. C.; Chang, S. T.; Chang, S. L. Y. J. Power Sources 2015, 277, 147−154. (26) Subbaraman, R.; Tripkovic, D.; Chang, K. C.; Strmcnik, D.; Paulikas, A. P.; Hirunsit, P.; Chan, M.; Greeley, J.; Stamenkovic, V.; Markovic, N. M. Nat. Mater. 2012, 11, 550−557. (27) Wang, H. L.; Dai, H. J. Chem. Soc. Rev. 2013, 42, 3088−3113. (28) Wang, H. L.; Liang, Y. Y.; Gong, M.; Li, Y. G.; Chang, W.; Mefford, T.; Zhou, J. G.; Wang, J.; Regier, T.; Wei, F.; Dai, H. J. Nat. Commun. 2012, 3, 917. (29) Qi, W.; Xie, K.; Liu, M.; Wu, G.; Wang, Y.; Zhang, Y.; Wu, Y. RSC Adv. 2014, 4, 40494−40504. (30) Mullins, D. R. Surf. Sci. Rep. 2015, 70, 42−85. (31) Senanayake, S. D.; Evans, J.; Agnoli, S.; Barrio, L.; Chen, T. L.; Hrbek, J.; Rodriguez, J. A. Top. Catal. 2011, 54, 34−41. (32) McKone, J. R.; Sadtler, B. F.; Werlang, C. A.; Lewis, N. S.; Gray, H. B. ACS Catal. 2013, 3, 166−169. (33) Jiang, N.; You, B.; Sheng, M. L.; Sun, Y. J. Angew. Chem., Int. Ed. 2015, 54, 6251−6254. (34) Hao, J.; Yang, W.; Zhang, Z.; Tang, J. Nanoscale 2015, 7, 11055−11062. (35) Xiao, P.; Sk, M. A.; Thia, L.; Ge, X. M.; Lim, R. J.; Wang, J. Y.; Lim, K. H.; Wang, X. Energy Environ. Sci. 2014, 7, 2624−2629. (36) Yu, X. Y.; Yu, L.; Wu, B. H.; Lou, X. W. Angew. Chem., Int. Ed. 2015, 54, 5331−5335. (37) Danilovic, N.; Subbaraman, R.; Strmcnik, D.; Chang, K. C.; Paulikas, A. P.; Stamenkovic, V. R.; Markovic, N. M. Angew. Chem., Int. Ed. 2012, 51, 12495−12498. (38) Sheng, W. C.; Myint, M.; Chen, J. G. G.; Yan, Y. S. Energy Environ. Sci. 2013, 6, 1509−1512. (39) Carrasco, J.; Lopez-Duran, D.; Liu, Z. Y.; Duchon, T.; Evans, J.; Senanayake, S. D.; Crumlin, E. J.; Matolin, V.; Rodriguez, J. A.; Ganduglia-Pirovano, M. V. Angew. Chem., Int. Ed. 2015, 54, 3917− 3921. (40) Norskov, J. K.; Bligaard, T.; Logadottir, A.; Kitchin, J. R.; Chen, J. G.; Pandelov, S.; Norskov, J. K. J. Electrochem. Soc. 2005, 152, J23− J26. (41) Bruix, A.; Lykhach, Y.; Matolinova, I.; Neitzel, A.; Skala, T.; Tsud, N.; Vorokhta, M.; Stetsovych, V.; Sevcikova, K.; Myslivecek, J.; Fiala, R.; Vaclavu, M.; Prince, K. C.; Bruyere, S.; Potin, V.; Illas, F.; Matolin, V.; Libuda, J.; Neyman, K. M. Angew. Chem., Int. Ed. 2014, 53, 10525−10530. (42) Flytzani-Stephanopoulos, M.; Gates, B. C. Annu. Rev. Chem. Biomol. Eng. 2012, 3, 545−574. (43) Kwak, J. H.; Kovarik, L.; Szanyi, J. ACS Catal. 2013, 3, 2094− 2100.

Complete experimental details, DFT calculation details, additional characterizations, and HER data (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The work is partially supported by Yale University and the Global Innovation Initiative from Institute of International Education. L.C.Y., F.L., and H.M.C. thank the National Natural Science Foundation of China (Nos. 51202255, 51472249, 51221264, 51172239, 51272051, and U1401243) and the Ministry of Science and Technology of China (Nos. 2011CB932604, 2014CB932402) for financial support. M.L. and E.I.A. acknowledge the support of the U.S. Department of Energy through Basic Energy Sciences grant DE-FG0298ER14882 and the use of facilities supported by the National Science Foundation through the Yale Materials Research Science and Engineering Center (Grant No. MRSEC DMR1119826). The theoretical calculations were performed on TianHe-1 (A) at National Supercomputer Center in Tianjin, China. The authors thank Prof. Fei Wei (Tsinghua University) for providing the CNTs and Prof. Robert H. Crabtree (Yale University) for proofreading the manuscript.

■

REFERENCES

(1) Yu, T. Y.; Zeng, J.; Lim, B.; Xia, Y. N. Adv. Mater. 2010, 22, 5188−5192. (2) Lee, Y. M.; Garcia, M. A.; Huls, N. A. F.; Sun, S. H. Angew. Chem., Int. Ed. 2010, 49, 1271−1274. (3) Cuenya, B. R. Thin Solid Films 2010, 518, 3127−3150. (4) Chen, G. X.; Zhao, Y.; Fu, G.; Duchesne, P. N.; Gu, L.; Zheng, Y. P.; Weng, X. F.; Chen, M. S.; Zhang, P.; Pao, C. W.; Lee, J. F.; Zheng, N. F. Science 2014, 344, 495−499. (5) Tao, F.; Grass, M. E.; Zhang, Y. W.; Butcher, D. R.; Renzas, J. R.; Liu, Z.; Chung, J. Y.; Mun, B. S.; Salmeron, M.; Somorjai, G. A. Science 2008, 322, 932−934. (6) Li, Y. M.; Somorjai, G. A. Nano Lett. 2010, 10, 2289−2295. (7) Zhou, Z. Y.; Tian, N.; Li, J. T.; Broadwell, I.; Sun, S. G. Chem. Soc. Rev. 2011, 40, 4167−4185. (8) Roldan Cuenya, B.; Behafarid, F. Surf. Sci. Rep. 2015, 70, 135− 187. (9) Fu, Q.; Yang, F.; Bao, X. H. Acc. Chem. Res. 2013, 46, 1692− 1701. (10) Wu, B. H.; Zheng, N. F. Nano Today 2013, 8, 168−197. (11) Yamada, Y.; Tsung, C. K.; Huang, W.; Huo, Z. Y.; Habas, S. E.; Soejima, T.; Aliaga, C. E.; Somorjai, G. A.; Yang, P. D. Nat. Chem. 2011, 3, 372−376. (12) Vayssilov, G. N.; Lykhach, Y.; Migani, A.; Staudt, T.; Petrova, G. P.; Tsud, N.; Skala, T.; Bruix, A.; Illas, F.; Prince, K. C.; Matolin, V.; Neyman, K. M.; Libuda, J. Nat. Mater. 2011, 10, 310−315. (13) Alayoglu, S.; An, K. J.; Melaet, G.; Chen, S. Y.; Bernardi, F.; Wang, L. W.; Lindeman, A. E.; Musselwhite, N.; Guo, J. H.; Liu, Z.; Marcus, M. A.; Somorjai, G. A. J. Phys. Chem. C 2013, 117, 26608− 26616. (14) Divins, N. J.; Angurell, I.; Escudero, C.; Perez-Dieste, V.; Llorca, J. Science 2014, 346, 620−623. (15) Cargnello, M.; Doan-Nguyen, V. V. T.; Gordon, T. R.; Diaz, R. E.; Stach, E. A.; Gorte, R. J.; Fornasiero, P.; Murray, C. B. Science 2013, 341, 771−773. 7709


Letter

Nano Letters (44) Liu, C.; Li, F.; Ma, L. P.; Cheng, H. M. Adv. Mater. 2010, 22, E28−E62. (45) Weng, Z.; Su, Y.; Wang, D. W.; Li, F.; Du, J. H.; Cheng, H. M. Adv. Energy Mater. 2011, 1, 917−922.

7710
