Exclusive Ni–N4 Sites Realize Near-Unity CO Selectivity for

Communication pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. 2017, 139, 14889-14892

Exclusive Ni−N4 Sites Realize Near-Unity CO Selectivity for Electrochemical CO2 Reduction Xiaogang Li,†,∥ Wentuan Bi,†,∥ Minglong Chen,‡,∥ Yuexiang Sun,† Huanxin Ju,§ Wensheng Yan,§ Junfa Zhu,§ Xiaojun Wu,‡ Wangsheng Chu,§ Changzheng Wu,*,† and Yi Xie† †

Hefei National Laboratory for Physical Sciences at the Microscale, iChEM (Collaborative Innovation Center of Chemistry for Energy Materials), and CAS Key Laboratory of Mechanical Behavior and Design of Materials, ‡CAS Key Lab of Materials for Energy Conversion, Department of Materials Science and Engineering, and §National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei 230026, P.R. China S Supporting Information *

metal coordinated with nitrogen (M−Nx) in carbon-based materials has attracted much attention for its superiority in electrocatalysis.13,14 The unique structure and coordination environment endow M−Nx with favored kinetics and thus bring about excellent activity.15,16 However, the M−N x structure has seldom been explored for CO2 reduction.17 This provides a potential approach for designing high-efficiency catalysts for electrochemical CO2 reduction. Nevertheless, M− Nx species will undergo uncontrolled agglomeration during the high-temperature synthetic process, leading to loss of active M−Nx sites, and eventually degrading electrochemical performance.18,19 Thus, keeping the exclusive structure of M−Nx sites is challenging yet significant for high-activity electrocatalytic CO2 systems. Herein, we demonstrated exclusive Ni−N4 sites through a topo-chemical transformation strategy, bringing unprecedented high activity and selectivity for CO2 reduction. Topo-chemical transformation by carbon layer coating successfully ensures the preservation of the Ni−N4 structure to a maximized extent and avoids agglomeration of Ni atoms to particles, providing abundant active sites for catalytic reaction. The exclusive Ni− N4 structure as reaction site boosts the catalytic reaction, achieving a maximum conversion efficiency of 99% for CO with high current density. We anticipate this work will pave a new avenue for design of high-efficiency electrocatalytic systems and benefit further development for CO2 conversion. Electrochemical reduction of CO2 on as-prepared catalyst was measured by linear sweep voltammetry (LSV). As seen in Figure 1a, in CO2-saturated KHCO3 solution, Ni−N4−C (1.41 wt % Ni loading) exhibits remarkably excellent activity, generating a current density of 36.2 mA cm−2 at −0.91 V, much higher than that of the sample without Ni decoration (denoted as N−C). While in Ar-saturated KHCO3 solution Ni−N4−C shows a much smaller current density (Figure S1) than that in CO2-saturated KHCO3 solution, indicating the excellent activity of Ni−N4−C indeed derives from CO2 reduction. In addition, SCN−, a commonly adopted ion to poison metal sites,20 was used as an indicator for active sites. As seen in Figure S2, the remarkable depression of catalytic activity for Ni−N4−C is observed which could be attributed to blocking of Ni atoms by SCN−, thus confirming the active site

ABSTRACT: Electrochemical reduction of carbon dioxide (CO2) to value-added carbon products is a promising approach to reduce CO2 levels and mitigate the energy crisis. However, poor product selectivity is still a major obstacle to the development of CO2 reduction. Here we demonstrate exclusive Ni−N4 sites through a topochemical transformation strategy, bringing unprecedentedly high activity and selectivity for CO2 reduction. Topochemical transformation by carbon layer coating successfully ensures preservation of the Ni−N4 structure to a maximum extent and avoids the agglomeration of Ni atoms to particles, providing abundant active sites for the catalytic reaction. The Ni−N4 structure exhibits excellent activity for electrochemical reduction of CO2 with particularly high selectivity, achieving high faradaic efficiency over 90% for CO in the potential range from −0.5 to −0.9 V and gives a maximum faradaic efficiency of 99% at −0.81 V with a current density of 28.6 mA cm−2. We anticipate exclusive catalytic sites will shed new light on the design of high-efficiency electrocatalysts for CO2 reduction.

T

he excessive emission of CO2 has triggered serious climatic problems and environmental issues.1,2 Thus, efficient conversion of CO2 is critically needed to mitigate CO2 levels and close the anthropogenic carbon cycle. Electrochemical CO2 reduction reaction with water as reaction medium has been shown to be a promising approach not only for decreasing CO2 accumulation but also its conversion to value-added carbon products.3 Despite many efforts devoted to electrochemical conversion of CO2, low efficiency limited by competitive reduction of water itself to hydrogen is still a major hurdle for further development of CO2 reduction reactions.4,5 Hence, robust catalysts with high activity and particularly high selectivity for electrochemical CO2 reduction are much desired. Recently, design strategies for high-performance catalysts have been boosted, including surface modification, construction of heterostructure, atomic-ordering transformations, and so on.6−9 For economic consideration, development of non-noble metal based catalysts, especially with maximum atomic efficiency, has become urgent for high-efficiency catalytic systems.10−12 In particular, the unit of single-atom transition © 2017 American Chemical Society

Received: August 25, 2017 Published: October 9, 2017 14889

DOI: 10.1021/jacs.7b09074 J. Am. Chem. Soc. 2017, 139, 14889−14892

Communication

Journal of the American Chemical Society

Figure 1. (a) Linear sweep voltammetric curves. (b) Faradaic efficiencies for CO. (c) Tafel plots for producing CO. (d) Nyquist plots of the samples.

role of the Ni−N4 structure. In contrast, in spite of the existence of Ni, the sample prepared by directly pyrolyzing Nidoped g-C3N4 (denoted as Ni@N−C) and the sample made by pyrolyzing homogeneous mixture of glucose and precursor for Ni-doped g-C3N4 (denoted as Ni@N−C−Glu) show poor activity similar to that of N−C, as seen in Figure 1a, demonstrating the superiority of Ni−N4−C. The measurement of faradaic efficiency (FE) was further carried out to evaluate the selectivity of Ni−N4−C for CO2 reduction. The results show that H2 and CO are the reduction product, and there is no liquid product detected by 1H nuclear magnetic resonance spectroscopy (Figure S3). Figure 2b and Figure S4 present the FE for CO and H2, respectively. Ni−N4− C shows significantly suppressed H2 evolution, achieving high FE over 90% for CO and giving a maximum FE of 99% at −0.81 V with a current density of 28.6 mA cm−2, while N−C, Ni@N−C, and Ni@N−C-Glu only show maximum FE of 64%, 65%, and 63%, respectively, demonstrating the excellent activity and selectivity of Ni−N4−C. To uncover this high performance of Ni−N4−C, the electrochemical active surface area (ECSA) was first measured. The measured double-layer capacitance is presented in Figure S5, in which the slope could be a reference of ECSA.21 We could see that Ni−N4−C and N−C share a similar ECSA, while Ni@N−C and Ni@N−C−Glu give an extremely small ECSA, indicative of a limited area to support active sites, which may be one of the reasons Ni@N−C and Ni@N−C-Glu show poor catalytic performance. Considering similar ECSA but much different activity of Ni−N4−C and N−C, the kinetics may be the key role for final performance. Tafel analysis is presented in Figure 1c. The Tafel slope of Ni−N4−C shows 103 mV/ decade, much smaller than that of N−C, Ni@N−C, and Ni@ N−C-Glu with a Tafel slope of 444, 169, and 124 mV/decade, respectively, indicating favorable kinetics for the formation of CO.22 To gain further insight into CO2 reduction reaction

Figure 2. (a) Schematic illustration of the topo-chemical transformation strategy (Ni atoms, green; N atoms, blue; C atoms, gray; O atoms, red). (b) FT of the Ni K-edge EXAFS oscillations of Ni-doped g-C3N4. (c) FT of the Ni K-edge EXAFS oscillations of Ni−N4−C. (d) TEM image of Ni−N4−C. Scale bar is 500 nm. (e) HAADF-STEM image of Ni−N4−C. (e) Element mapping images of Ni−N4−C.

kinetics, electrochemical impedance spectroscopy (EIS) was carried out. The Nyquist plots in Figure 1d also demonstrate that Ni−N4−C owns the much faster charge-transfer process during the CO2 reduction process, eventually leading to remarkably enhanced activity and selectivity for CO2 reduction. The long-time CO2 reduction test in Figure S6 demonstrates the relative stability of Ni−N4−C. 14890

DOI: 10.1021/jacs.7b09074 J. Am. Chem. Soc. 2017, 139, 14889−14892

Communication

Journal of the American Chemical Society

to the higher energy side, in accordance with N K-edge XAS spectra, demonstrating again Ni bonding with pyridinic-N. The C K-edge XAS spectra in Figure S11 show no difference between Ni−N4−C and N−C, excluding formation of the Ni− C bond. Thus, we could conclude single-atom Ni sites only bond to pyridinic-N in Ni−N4−C. Benefiting from topo-chemical transformation strategy, Ni− N4−C morphology was reminiscent of Ni-doped g-C3N4 precursor. The transmission electron microscope (TEM) image in Figure 2d shows the morphology of Ni−N4−C, similar to that of N−C and Ni-doped g-C3N4 in Figure S12a,b. The high-angle annular dark field scanning transmission electron microscope (HAADF-STEM) image for Ni−N4−C in Figure 2e shows isolated bright spots corresponding to Ni atoms, demonstrating the single-atom form of Ni in Ni−N4−C, agreeing well with the Ni−N4 structure. Element mapping images of Ni−N4−C in Figure 2f indicates that Ni and N distribute homogeneously in carbon sheets. Interestingly, without the outer carbon layer, Ni particles warped by carbon nanotubes were observed for both Ni@N−C and Ni@N−C− Glu, as shown by TEM images and element mapping images in Figure S13. For Ni−N4−C and N−C, Raman spectra in Figure S14a both exhibit peaks at 1359 and 1595 cm−1, assigned to disordered sp3 carbon (D band) and graphitic sp2 carbon (G band), respectively, demonstrating formation of graphene.21 Meanwhile, X-ray diffraction (XRD) patterns of Ni−N4−C and N−C in Figure S14b exhibit only two characteristic peaks at approximately 24° and 44°, corresponding to (002) and (100) planes of graphite. While the XRD patterns of Ni@N−C and Ni@N−C−Glu in Figure S15 show (111) and (200) planes of cubic nickel (JCPDS Card No. 65-2865), demonstrating formation of Ni particles in Ni@N−C and Ni@N−C−Glu. Notably, more loading of Ni for Ni−N4−C also leads to formation of Ni particles as seen in the TEM image and HAADF-STEM image in Figure S16, in which the distinct lattice fringe of 0.175 and 0.204 nm agree well with (200) and (111) lattice planes of Ni particle. LSV, EIS, and FE results for Ni−N4−C in Figures S17−S19 indicate the more loading of Ni the better catalytic performance of CO2 reduction until the formation of Ni particles. Thus, we could see the detrimental effect of Ni particles and the superiority of the Ni−N4 structure. To understand the high selectivity of Ni−N4−C, density functional theory (DFT) calculations were performed. The free energy diagrams of CO2 reduction for Ni−N4−C and N−C are present in Figure 3a. The formation of adsorbed intermediate COOH* was found as the potential limiting step for both Ni− N4−C and N−C. From a thermodynamic point of view, reaction free energy can be linked to reaction energy barrier, so the trend of free energy can be associated with activity of CO2

The topo-chemical transformation strategy provides a feasible way to maintain active sites during pyrolysis as depicted in Figure 2a, thus endowing exclusive Ni−N4 sites with high catalytic performance. Synchrotron-based extended X-ray absorption fine structure (EXAFS) was further used to determine the chemical configurations around Ni sites. The local chemical configuration of the precursor, Ni-doped g-C3N4, has been determined by Fourier transform (FT) of Ni K-edge EXAFS oscillation as shown in Figure 2b. Its FT curve only presents a doublet feature in the distance shorter than 2.0 Å, which is generally ascribed to interaction between Ni and lowatomic number (low-z) atoms (C, N, and O). Beyond this distance, no obvious FT peaks present for the precursor, especially at a distance of Ni−Ni interaction, identified by that of nickel foil as reference. One dominant feature of the FT curve indicates Ni has been atomically dispersed into g-C3N4 framework. The doublet feature at the lower distance demonstrates local chemical environment around Ni in Nidoped g-C3N4 is a bit complicated, involving two kinds of chemical bonds. One bond length is about 1.95 Å and the other 2.10 Å. According to the reported works, we know that the metal center in the N-doped carbon material easily bonds oxygen and is controlled by temperature.23 The quantitative simulation for the EXAFS in Figure 2b reveals the Ni in the Nidoped g-C3N4 has been coordinated by two O atoms at 1.95 Å and four N atoms at 2.09 Å. The fit parameters are given in Table S1. Figure 2c shows FT of Ni K-edge EXAFS oscillation of Ni− N4−C. The FT curve of Ni−N4−C only presents a dominant peak at about 1.3 Å (no phase correction), ascribed to the bond between Ni and the low-z atom. In the distance of Ni−Ni interaction beyond 2.0 Å, no obvious FT peaks are present for Ni−N4−C, suggesting atomic dispersed Ni species were confined in the N−C framework. Moreover, the single FT peak corresponding to the interaction between Ni and the lowz element exhibits a reduced intensity compared with the doublet of the Ni-doped g-C3N4. This suggests that the topochemical transformation process preserved nitrogen chemical bond and removed the oxygen bond. The FT-EXAFS curve of the Ni−N4−C sample has been fitted in r space to extract quantitatively atomic structure around Ni. The fit result shows Ni has been coordinated by four N atoms at a distance of 1.86 Å. The quality of the best fit is shown in Figure S7, and the fit parameters are given in Table S1. In addition, the Ni L-edge soft X-ray absorption spectroscopy (XAS) in Figure S8 gives a single peak at the L3-edge, indicative of delocalized 3d-electrons shared by N-coordinated environments.24 Thus, we see that the topo-chemical transformation strategy guarantees the exclusive Ni−N4 structure in Ni−N4−C. To further confirm the type of the chemical bond related to Ni sites, N K-edge and C K-edge soft XAS for Ni−N4−C and N−C were performed. The N K-edge presents three obvious resonances (Figure S9) attributed to the π*-transition in the aromatic C−N−C portion of the pyridinic site (peak a), N-3C bridging of the graphitic site (peak b), and the σ*-transition of C−N bond (peak c).25 Obviously, peak a of Ni−N4−C shifts to higher energy compared with that of N−C while other peaks keep the same, indicating pyridinic-N formed a chemical bond with Ni sites. This is also confirmed by X-ray photoelectron spectroscopy (XPS) results in Figure S10. N 1s spectra could be deconvolved into three peaks, corresponding to pyridinic-N, graphitic-N, and quaternary N+−O−, respectively.26 Compared with N−C, the peak assigned to pyridinic-N in Ni−N4−C shifts

Figure 3. (a) Calculated free energy diagram. (b) Difference in limiting potentials for CO2 reduction and H2 evolution. 14891

DOI: 10.1021/jacs.7b09074 J. Am. Chem. Soc. 2017, 139, 14889−14892

Communication

Journal of the American Chemical Society reduction.27 Obviously, introduction of Ni−N4 sites lowers the formation energy of COOH* compared with that for N−C, facilitating followed formation of CO and thus showing higher activity. Considering that H2 evolution is the major competitive reaction and previous works have suggested the difference between thermodynamic limiting potentials for CO2 reduction and H2 evolution (denoted as UL(CO2) − UL(H2), in which UL = −ΔG0/e) can reflect the selectivity in CO2 reduction reactions,28,29 the value of UL(CO2) − UL(H2) was calculated and a more positive value means a higher selectivity. As seen in Figure 3b and Figure S20, Ni−N4−C shows a significantly more positive value for UL(CO2) − UL(H2) than that of N−C, Ni@N−C, and Ni@N−C-Glu, corresponding to higher selectivity for reduction of CO2 to CO. This also coincides with experimental results. From thermochemical DFT and experimentally analysis, the Ni−N4 structure shows a lowered energy barrier and accelerated charge transfer for the CO2 reduction, thus eventually boosting activity and selectivity for conversion of CO2 to CO. In summary, we have successfully constructed exclusive Ni− N4 sites for unprecedented high activity and selectivity for electrocatalytic CO2 reduction by topo-chemical transformation strategy. The Ni−N4 sites endow the catalyst with high faradaic efficiency over 90% for CO and give a maximum faradaic efficiency of 99% with high current density. Topo-chemical transformation strategy provides a feasible way to construct abundant and exclusive active sites, avoiding aggregation of metal atoms and the resulting loss of active sites. We anticipate this work will lead to promising path for design of highefficiency catalysts and should inject new vitality not only to CO2 reduction research but also to the extensive electrocatalytic field.

■

Development Foundation of Hefei Center for Physical Science and Technology and USTC Center for Micro and Nanoscale Research and Fabrication.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b09074. Experimental procedures, electrochemical test, and other additional figures (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Junfa Zhu: 0000-0003-0888-4261 Xiaojun Wu: 0000-0003-3606-1211 Changzheng Wu: 0000-0002-4416-6358 Yi Xie: 0000-0002-1416-5557 Author Contributions ∥

REFERENCES

(1) Li, C. W.; Ciston, J.; Kanan, M. W. Nature 2014, 508, 504. (2) Qiao, J.; Liu, Y.; Hong, F.; Zhang. Chem. Soc. Rev. 2014, 43, 631. (3) Li, Q.; Fu, J.; Zhu, W.; Chen, Z.; Shen, B.; Wu, L.; Xi, Z.; Wang, T.; Lu, G.; Zhu, J.-j.; Sun, S. J. Am. Chem. Soc. 2017, 139, 4290. (4) Lin, S.; Diercks, C. S.; Zhang, Y.-B.; Kornienko, N.; Nichols, E. M.; Zhao, Y.; Paris, A. R.; Kim, D.; Yang, P.; Yaghi, O. M. Science 2015, 349, 1208. (5) Varela, A. S.; Ranjbar Sahraie, N.; Steinberg, J.; Ju, W.; Oh, H. S.; Strasser, P. Angew. Chem. Int. Ed. 2015, 54, 10758. (6) Bi, W.; Li, X.; Zhang, L.; Jin, T.; Zhang, L.; Zhang, Q.; Luo, Y.; Wu, C.; Xie, Y. Nat. Commun. 2015, 6, 8647. (7) Chen, G.; Xu, C.; Huang, X.; Ye, J.; Gu, L.; Li, G.; Tang, Z.; Wu, B.; Yang, H.; Zhao, Z.; Zhou, Z.; Fu, G.; Zheng, N. Nat. Mater. 2016, 15, 564. (8) Fan, Z.; Huang, X.; Han, Y.; Bosman, M.; Wang, Q.; Zhu, Y.; Liu, Q.; Li, B.; Zeng, Z.; Wu, J. Nat. Commun. 2015, 6, 6571. (9) Kuld, S.; Thorhauge, M.; Falsig, H.; Elkjær, C. F.; Helveg, S.; Chorkendorff, I.; Sehested, J. Science 2016, 352, 969. (10) Qiao, B.; Wang, A.; Yang, X.; Allard, L. F.; Jiang, Z.; Cui, Y.; Liu, J.; Li, J.; Zhang, T. Nat. Chem. 2011, 3, 634. (11) Li, X.; Bi, W.; Zhang, L.; Tao, S.; Chu, W.; Zhang, Q.; Luo, Y.; Wu, C.; Xie, Y. Adv. Mater. 2016, 28, 2427. (12) Ge, J.; He, D.; Chen, W.; Ju, H.; Zhang, H.; Chao, T.; Wang, X.; You, R.; Lin, Y.; Wang, Y. J. Am. Chem. Soc. 2016, 138, 13850. (13) Lefèvre, M.; Proietti, E.; Jaouen, F.; Dodelet, J.-P. Science 2009, 324, 71. (14) Malko, D.; Kucernak, A.; Lopes, T. Nat. Commun. 2016, 7, 13285. (15) Liang, J.; Zhou, R. F.; Chen, X. M.; Tang, Y. H.; Qiao, S. Z. Adv. Mater. 2014, 26, 6074. (16) Wang, Q.; Zhou, Z.-Y.; Lai, Y.-J.; You, Y.; Liu, J.-G.; Wu, X.-L.; Terefe, E.; Chen, C.; Song, L.; Rauf, M. J. Am. Chem. Soc. 2014, 136, 10882. (17) Zhao, C.; Dai, X.; Yao, T.; Chen, W.; Wang, X.; Wang, J.; Yang, J.; Wei, S.; Wu, Y.; Li, Y. J. Am. Chem. Soc. 2017, 139, 8078. (18) Wu, Z. S.; Chen, L.; Liu, J.; Parvez, K.; Liang, H.; Shu, J.; Sachdev, H.; Graf, R.; Feng, X.; Müllen, K. Adv. Mater. 2014, 26, 1450. (19) Kramm, U. I.; Herrmann-Geppert, I.; Behrends, J.; Lips, K.; Fiechter, S.; Bogdanoff, P. J. Am. Chem. Soc. 2016, 138, 635. (20) Thorum, M. S.; Hankett, J. M.; Gewirth, A. A. J. Phys. Chem. Lett. 2011, 2, 295. (21) Lei, F.; Liu, W.; Sun, Y.; Xu, J.; Liu, K.; Liang, L.; Yao, T.; Pan, B.; Wei, S.; Xie, Y. Nat. Commun. 2016, 7, 12697. (22) Li, C. W.; Kanan, M. W. J. Am. Chem. Soc. 2012, 134, 7231. (23) Niu, W.; Li, L.; Liu, X.; Wang, N.; Liu, J.; Zhou, W.; Tang, Z.; Chen, S. J. Am. Chem. Soc. 2015, 137, 5555. (24) Chen, P.; Zhou, T.; Xing, L.; Xu, K.; Tong, Y.; Xie, H.; Zhang, L.; Yan, W.; Chu, W.; Wu, C. Angew. Chem. Int. Ed. 2017, 56, 610. (25) Zheng, Y.; Jiao, Y.; Zhu, Y.; Li, L. H.; Han, Y.; Chen, Y.; Du, A.; Jaroniec, M.; Qiao, S. Z. Nat. Commun. 2014, 5, 3783. (26) Lin, L.; Zhu, Q.; Xu, A.-W. J. Am. Chem. Soc. 2014, 136, 11027. (27) Wang, S.; Petzold, V.; Tripkovic, V.; Kleis, J.; Howalt, J. G.; Skulason, E.; Fernandez, E.; Hvolbæk, B.; Jones, G.; Toftelund, A. Phys. Chem. Chem. Phys. 2011, 13, 20760. (28) Kim, D.; Xie, C.; Becknell, N.; Yu, Y.; Karamad, M.; Chan, K.; Crumlin, E. J.; Norskov, J. K.; Yang, P. J. Am. Chem. Soc. 2017, 139, 8329. (29) Shi, C.; Hansen, H. A.; Lausche, A. C.; Nørskov, J. K. Phys. Chem. Chem. Phys. 2014, 16, 4720.

X.L., W.B., and M.C. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the National Basic Research Program of China (2015CB932302), National Natural Science Foundation of China (U1432133, 11621063, 21701164), National Program for Support of Top-notch Young , and the Fundamental Research Funds for the Central Universities (WK2060190084, WK2060190058). We also appreciate support from the Major/Innovative Program of 14892

DOI: 10.1021/jacs.7b09074 J. Am. Chem. Soc. 2017, 139, 14889−14892