Cobalt Boron Imidazolate Framework Derived Cobalt

FULL PAPER Water Splitting

www.afm-journal.de

Cobalt Boron Imidazolate Framework Derived Cobalt Nanoparticles Encapsulated in B/N Codoped Nanocarbon as Efficient Bifunctional Electrocatalysts for Overall Water Splitting Mei-Rong Liu, Qin-Long Hong, Qiao-Hong Li, Yonghua Du, Hai-Xia Zhang, Shumei Chen, Tianhua Zhou,* and Jian Zhang*

The development of efficient and low-cost bifunctional electrocatalysts for oxygen evolution reaction (OER) and hydrogen evolution reaction (HER) is highly desirable for electrochemical energy conversion. Herein, this study puts forward a new Co decorated N,B-codoped interconnected graphitic carbon and carbon nanotube materials (Co/NBC) synthesized by direct carbonization of a cobalt-based boron imidazolate framework. It is demonstrated that the carbonization temperature can tune the surface structure and component of the resultant materials and optimize the electrochemically active surface area to expose more accessible active sites, effectively boosting the electrocatalytic activity. As a result, the optimized Co/NBC shows superior bifunctional catalytic activity and stability toward OER and HER in 1.0 m KOH solution. Furthermore, the catalyst can serve as both the anode and cathode for water splitting to achieve a current density of 10 mA cm−2 at a cell voltage of 1.68 V. Experimental results and theoretical calculations indicate that the excellent activity of Co/NBC catalyst benefits from the synergistic effect of partial oxidation of metallic cobalt, conductive N,B-codoped graphitic carbon and carbon nanotube, and the coupled interactions among these components. This work opens a promising avenue toward the exploration of boron imidazolate frameworks as efficient heteroatom-doped catalysts for electrocatalysis.

M.-R. Liu, Q.-L. Hong, Prof. Q.-H. Li, Prof. H.-X. Zhang, Prof. T. Zhou, Prof. J. Zhang State Key Laboratory of Structural Chemistry Fujian Institute of Research on the Structure of Matter Chinese Academy of Sciences Fuzhou, Fujian 350002, P. R. China E-mail: [email protected]; [email protected] M.-R. Liu, Prof. S. Chen College of Chemistry Fuzhou University Fuzhou, Fujian 350108, P. R. China Dr. Y. Du Institute of Chemical and Engineering Sciences A*STAR Singapore 1 Pesek Road, Jurong Island 627833, Singapore

DOI: 10.1002/adfm.201801136

Adv. Funct. Mater. 2018, 1801136

1. Introduction

Electrochemical splitting of water to produce hydrogen and oxygen is one promising way for providing sustainable clean energy.[1] The key to successfully realizing this target is the discovery of highly efficient electrocatalysts for driving oxygen evolution reaction (OER) and hydrogen evolution reaction (HER).[2] Currently, noble-metal-based electrocatalysts (Ir or Ru-based for OER and Pt-based for HER) are the state-of-the-art electrocatalysts to catalyze water-splitting, but their high price and scarcity limit their large-scale use in electrolyzes.[2,3] Consequently, some earth abundant 3d transition metalbased electrocatalysts are intensively investigated as potential substitutes for precious metal catalysts for water splitting application.[4] Nevertheless, most of materials only catalyze half-reaction of water splitting (OER or HER). A few of them can serve as efficient bifunctional electrocatalysts for overall water splitting.[5] It still remains a challenge to develop an active and low-cost bifunctional catalyst for both OER and HER. Among well-developed bifunctional catalysts, 3d transitionmetal nonoxides have been recognized as attractive potential bifunctional electrocatalysts for both OER and HER, such as metal selenides,[6] sulfide,[7] carbide,[8] metal alloy,[9] borides,[10] and phosphides.[11] To expose more catalytic active sites and facilitate the transport of electrons or protons during the reaction process, an effective way is to decorate metal nanoparticles with conductive carbon, especially, heteroatom-doped (e.g., N, P, S, or B) graphitic carbon or carbon nanotube (CNT) which derived from a heteroatom-containing polymer or metal–organic frameworks (MOFs).[12] Heteroatom-doped conductive graphitic carbon in situ formed, such as N,Bcodoped graphitic carbon, not only serves as a robust support for effectively protecting the metal centers from aggregation, but also alters the electronic configuration of the host metal

1801136 (1 of 9)

© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com

www.afm-journal.de

species via the interplay between heteroatom doped carbon and metal species, which show synergistic effects to improve the catalytic activities.[4c,13] To obtain binary or ternary heteroatom-doped carbon decorated metal nanoparticles, a mixture of MOF and heteroatom-incorporated (e.g., N, P, or B) ligand as the precursor template is often adopted. By contrast, only a few reports are available on the exploration of MOFs as singlesource precursor to construct multi-heteroatom-doped metal nanomaterials by one-step method probably due to the lack of suitable MOFs.[14] Using single MOF as a precursor, it is possible to guarantee that these multiplex heteroatom doped are repeatable. Boron imidazolate frameworks (BIFs) are a new class of zeolite-like metal–organic frameworks consisting of the crosslinking of preassembly boron-imidazolate complex and metal ions.[15] One of the attractive features is that BIFs possess high density of boron and nitrogen components. Additionally, the BIFs having BH group exhibit significant reducing property and can directly reduce encapsulated metal ion into nanoparticles without any external stimulus.[16] The utilization of such order arrangement of zeolite-lite MOF as precursors allows us to obtain uniformly distributed N,B-codoped nanocarbon nanomaterials. Based on these advantages, herein, we report the design and synthesis of cobalt nanoparticles decorated nitrogen and boron codoped graphitic carbon shell (Co/NBC) by direct carbonization of a new cobalt-based boron imidazolate frameworks (BIF-82-Co) without introduction of additional heteroatom-containing ligand. During carbonization, N,Bcontaining ligands are decomposed into N,B-codoped graphitic carbon shell catalyzed by cobalt species.[17] Meanwhile, the carbon source with BH groups can function as the reducing reagent to reduce these cobalt nodes into metallic cobalt nanoparticles under high temperature.[17] Furthermore, controlling the reaction temperature can modulate the surface structure and component to optimize the synergistic effect among different components. As a result, the optimized Co/NBC-900 displays highly efficient and robust electrocatalytic activity toward HER and OER as well as overall water splitting. These findings will promote the development of a new class of boron imidazolate frameworks to fabricate multi-heteroatom-doped nanocarbon hybrid materials for application in OER and HER electrocatalysis.

2. Results and Discussions As illustrated in Scheme 1, porous boron imidazolate framework (BIF-82-Co) was synthesized via a simple solvothermal method through the self-assembly of cobalt nitrate and KBH(2methyl-imidazolate)3 at 80 °C for 72 h. After cooling down to room temperature, the pure purple crystals were obtained. Single-crystal X-ray diffraction study revealed that compound BIF-82-Co crystallized in the orthorhombic space group Pbcn (Table S1, Supporting Information). It features a 2D layered cationic framework. The asymmetric unit of BIF-82-Co consists of one Co(II) ions, one nitrate ion and one BH(mim)3− ligand, as well as a disordered N,N-dimethylformamide (DMF) molecule. The Co(II) is coordinated by one nitrate oxygen atom and three nitrogen atoms from three BH(mim)3− ligands, giving rise to a slightly distorted tetrahedral geometry (Figure S1a, Supporting Information). The BH(mim)3− is tridentate ligand and bridged with three Co ions. The alternating linkage between threeconnected Co(II) ions and the µ3-bridging BH(mim)3− ligand results in the formation of 2D layered structure with a (6, 3) topology, in which the Co···B distances ranging from 5.588 to 5.638 Å (Figure S1b, Supporting Information). These 2D layers are held together along a axis via weak van der Waals interactions to generate a 3D supramolecular structure, with a 1D channel (4.6 × 4.6 Å) along c axis (Figure S1c, Supporting Information). The surface area of BIF-Co was determined by N2 sorption isotherm at 77 K using the Brunauer–Emmett– Teller (BET) calculation (Figure S2a, Supporting Information). The sample displayed Type I isotherm with BET surface area of 201.47 m2 g−1. The powder X-ray diffraction (XRD) pattern indicated that the rod-shaped crystals are pure phase (Figure S2b, Supporting Information). The morphology of BIF-82-Co was first investigated by fieldemission scanning electron microscopy. As shown in Figure 1a, the as-prepared precursor BIF-82-Co consists of some uniform microspheres with a diameter of about 40–45 µm. Further examination reveals that these microspheres are composed of randomly aggregated rod-like nanocrystals with a length of 3–4 µm and an average width of 600 nm (Figure 1b). After the carbonization at 900 °C under N2 atmosphere for 2 h, the resultant sample maintains the rod-like morphology of Co/NBC-900 with abundant cobalt nanoparticles (Figure 1c) and the precursor

Scheme 1. Illustration of the synthesis of Co/NBC.


1801136 (2 of 9)



www.afm-journal.de

Figure 1. a,b) SEM images of BIF-82-Co. c) TEM, d–f) HRTEM images of Co/NBC-900. The inset of (f) is fast Fourier transform (FFT) of the nanocrystal structure.

containing boron and nitrogen decomposes into both boron and nitrogen codoped graphitic carbons, whereas the cobalt species is reduced to metallic cobalt with partial cobalt oxides, thus leading to forming metallic cobalt encapsulated in N,B-codoped graphitic carbon shell (Figure 1d,e). Interestingly, a few CNTs with the lengths of 70–100 nm and diameters of 10–20 nm are also observed on the surface of graphitic carbon (Figure 1d). In general, it is difficult to produce long CNT under N2 atmosphere using simple cobalt-based MOF as precursors, which makes it potential catalyst for electrochemical energy application. High-angle annular dark-field scanning transmission electron microscopy (TEM) and energy dispersive X-ray elemental mapping show the uniform distribution of C, N, B, Co, and O elements throughout the hybrid materials (Figure S3, Supporting Information). Further high-resolution TEM (HRTEM) shows that these cobalt nanoparticles are surrounded by a few layered graphitic carbon shells (thickness of about 10 nm) formed in situ with an interlayer distance of 3.6 Å (Figure 1e). These embedded nanoparticles with the lattice fringes of 0.204 nm correspond to the (111) plane of cubic metallic cobalt (ICSD-53805), in agreement with the d-spacing estimated by fast Fourier transform (FFT) pattern (Figure 1f). This observation implies that the exposed (111) plane dominates in the hybrid. Moreover, Raman spectroscopy analysis reveals that there are two characteristic peaks located at 1350 cm−1 (D band) and 1590 cm−1 (G band), corresponding to disordered and graphitized carbon, respectively (Figure 2a). In comparison to that of NBC-900 (1.42) without cobalt species, Co/NBC-900 displays a decrease in the ID/IG ratio (1.17) (Figure 2a). This observation indicates the formation of more ordered graphitic carbon due to the role of cobalt species,[17] which contributes to the stability and charge transfer of hybrid material. The pore structure information about the Co/NBC-900 was determined by N2 sorption. Figure S4 (Supporting Information) shows a typical IV type isotherm with a significant hysteresis loop, suggesting the mesoporous structure of Co/NBC-900 with a BET surface


area of 73.2 m2 g−1, which is helpful for mass transport and charge transfer during electrocatalysis.[18] The structure of Co/NBC-900 was further identified by XRD technique (Figure 2b). The strong peaks at 44.5°, 51.7°, and 76.1° are consistent with the (111), (200), and (220) planes of face-centered cubic Co (ICSD-53805), respectively. Also, a very weak diffraction peak at about 26° corresponds to a d-spacing of 0.36 nm, which can be assigned to the (002) plane of graphitic carbon, in good agreement with the TEM observations and Raman spectrum analysis. Only a broad and weak signal at about 60° was observed, corresponding to the diffraction of the (220) plane of cubic CoO (JCPDS-431004), signifying that the low crystallinity of the cubic CoO phase in situ formed. The composition and valence state of this sample were explored by X-ray photoelectron spectroscopy (XPS). The survey XPS spectrum of Co/NBC-900 indicated that the sample consists of C, N, B, Co, and O elements (Figure S5a, Supporting Information). The C 1s peak centered at 285.4 eV contributes to the CN group (Figure S5b, Supporting Information). The high resolution XPS spectra of N 1s are shown in Figure 2c. For comparison, metal-free NBC-900 was prepared by direct carbonization of the ligand BH(mim)3− under the same condition without cobalt (Supporting Information). The N 1s spectrum of NBC900 can be deconvoluted into three bands at 397.1, 398.5, and 399.7 eV. They can be assigned to the NB, pyridinic CN, and pyrrolic CN, respectively.[19] By contrast, besides the bands linked with pyridinic CN (398.2 eV, 48%) and pyrrolic CN (398.8 eV, 20%), the Co/NBC-900 sample displays two new bands at 399.1 (17%) and 401.3 eV (14%), which are assigned to the CoNx and graphitic CN structures, respectively. The peak corresponding to the binding energy of NB structure at 397.1 eV disappears probably due to the weak NB bond. The presence of CoNx bonding structure provides evidence of the interaction between cobalt and heteroatom doped graphitic carbon supports.[17b,20] The B 1s band can be deconvoluted into four bands at 190.4, 191.3, and 192.3 eV, corresponding

1801136 (3 of 9)



www.afm-journal.de

Figure 2. a) Raman spectra of the Co/NBC-900 and NBC-900. b) XRD patterns of the Co/NBC samples. High-resolution XPS spectra c) N 1s, d) B 1s, and e) Co 2p of the Co/NBC samples. f) XANES spectra at the Co K-edge of the Co/NBC-900, CoOOH, Co3O4, and Co foil.

to the CB, CBN, and OB bonding structures, respectively (Figure 2d),[21] indicating that the boron was successfully doped in the hybrid. The stronger O 1s peak at 529.8 eV in Co/NBC-900 suggested the presence of metal oxides, which is consistent with the XRD result (Figure S5c, Supporting Information). The binding energies of Co 2p3/2 peaks located at about 778.5 and 780.2 eV correspond to the metallic cobalt and cobalt oxides (Figure 2e; Figure S5d, Supporting Information), which is indicative of partial oxidized cobalt species exposed on the surface of metallic cobalt nanoparticles. In addition, the small metallic Co 2p3/2 peak at 778.5 eV is positively shifted compared with the binding energy of metallic Co (778.1 eV), implying the presence of electron transfer from metallic cobalt to heteroatom-doped carbon frameworks. This observation will be further discussed by means of the density functional theory (DFT) below. The content of cobalt in sample was determined by inductively coupled plasma optical emission spectroscopy and found to be 38.24 wt% (Table S2, Supporting Information). To study the effect of carbonization temperature on electrocatalytic performance, the other three samples were also prepared and designed as Co/NBC-X (X = 600, 700, and 800), according to the applied carbonization temperature (600–800 °C), using the same procedure as Co/NBC-900 (Supporting Information). Their morphology is similar to that of Co/NBC-900 (Figure S6 of the Supporting Information and Figure 2). XRD diffraction patterns (Figure 2b) indicate that all the samples possess a similar structure but the XRD diffraction peaks became sharper as increasing the pyrolysis temperature, implying that the high temperature contributes to the crystallization. The contents of carbon and nitrogen were measured by elemental analyzer and


the results were presented in Table S3 (Supporting Information). It has been shown that the content of carbon increases as increasing the pyrolysis temperature from 600 to 900 °C. The weight percentage of carbon increases from 36.38 to 41.36 wt%. This might be due to the decomposition of the DMF molecules and organic ligands in crystal to supply carbon sources, which contributes to the graphitization to form graphitic carbon with increasing the contents of carbon but decreasing the contents of BN species.[19,22] As shown in Figure 2a and Table S4 (Supporting Information), the percentage content of pyridine-like CN and CB increases from 600 to 900 °C, while the contents of the nitrogen of CoNx, NB, and pyrrolic-N decrease with the increasing the carbonization temperature to 900 °C. Accordingly, the intensity ratio of ID/IG decreases from 600 to 900 °C (Figure S5e, Supporting Information), indicating that higher temperature favors the formation of both N and B codoped more ordered graphitic carbon.[19] The existence of graphitic CN and CB bonds also implies the formation of a homogeneous single phase of B and N codoped graphene (h-NBC). They locate at the edge site of the carbon materials, which are prone to changing the distribution of electron density of metal atoms and provide more active sites.[19,23] Furthermore, this type of structure could improve conductivity to facilitate electron transfer during electrocatalysis. Interestingly, it is found that the content of cobalt oxides increases as increasing the carbonization temperature revealed by the high-resolution XPS of Co 2p (Table S4, Supporting Information). The formation of cobalt oxides could be due to the fact that metallic cobalt nanoparticles were protected by a few layers of graphitic carbon shell, while the cobalt was oxidized on the exposed surface of

1801136 (4 of 9)



www.afm-journal.de

Figure 3. a) Polarization curves and b) the corresponding Tafel plots of the Co/NBC samples, NBC-900, and commercial Ir/C. c) Nyquist plots of the Co/NBC samples. d) The current density versus scan rate; the slope of the fitting line allows the estimation of the double-layer capacitance (Cdl). e) The chronopotentiometry (V–t) curve at a constant current density of 10 mA cm−2. f) Fourier transform (FT) spectra of Co K-edge EXAFS of Co/NBC-900 after OER and reference samples.

cobalt nanoparticles, which were not completely encapsulated in the graphitic carbon shell due to the decomposition of BN and pyrrolic-N with the removal of the less graphitized carbon species. These observations demonstrate that high carbonization temperature is helpful for the formation of more surface oxides, pyridine-like CN and CB species, which play an important role as active sites for Co/NBC hybrid in catalyzing OER and HER, as discussed below. In order to further gain more accurate structural information about the coordination model and the oxidation state of cobalt atoms in the Co/NBC hybrid. The X-ray absorption fine structure (X-ray absorption near edge structure, XANES) and extended X-ray absorption fine structure (EXAFS) spectroscopy were performed on the Co/NBC hybrid (Figure 2f; Figure S7, Supporting Information).[24] For comparison, three cobalt-based samples, including Co foil, CoOOH, and Co3O4, were also investigated. As shown Figure 2f and Figure S7 (Supporting Information), the XANS spectra of all the samples are significantly different from those of CoOOH and Co3O4 but exactly match that of the Co foil reference. However, a significant peak above the absorption edge indicates a considerable amount of cobalt oxides presented (Figure 2f). The Fourier transformed EXAFS patterns were used to probe the corresponding coordination environment around cobalt atoms. As shown in Figure S7 (Supporting Information), the Co/NBC samples show one main peak at 2.11 Å, corresponding to the CoCo coordinate shells, which dominates the EXAFS spectra Adv. Funct. Mater. 2018, 1801136

of Co/NBC samples. Furthermore, the intensity increases with increasing the carbonization temperature, indicating that the high temperature is helpful for the crystallization of cobalt nanoparticles. It is noteworthy that there is light backscattered from cobalt in the first coordination sphere (Figure S7b, Supporting Information). The low distance peak located at 1.51 Å without phase correction presumably arises from CoO, CoN, or CoB interactions, implying the existence of the interaction between cobalt and heteroatom doped carbon frameworks, in line with the above observation assessed from XPS above. In brief, all these observations support the successful preparation of metallic cobalt nanoparticles with surface oxides encapsulated in N,B-heteroatom-doped carbon materials by direct carbonization of a cobalt-based boron imidazolate framework. The OER electrocatalytic activity of as-synthesized Co/NBC was first investigated by depositing a uniform catalyst film on glass carbon electrode using typical three electrode systems in 1.0 m KOH solution. The polarization curves were collected by linear sweep voltammetry (LSV) measurements at a scan rate of 5 mV s−1. As shown in Figure 3a, the carbonization temperature significantly affects the OER activity, which increases with the carbonization temperature and reaches a maximum at 900 °C. The hybrid Co/NBC-900 displays the highest OER activity with an onset potential of ≈1.48 V versus RHE and an overpotential of 302 mV required to reach the current density of 10 mA cm−2. Also, the overpotential of Co/NBC-900 is comparable to that of commercial Ir/C (298 mV) at a current density of 10 mA cm−2,

1801136 (5 of 9)



www.afm-journal.de

Figure 4. a) HER polarization curves and b) the corresponding Tafel plots of the Co/NBC samples, NBC-900, and commercial Pt/C. c) HER polarization curves of the Co/NBC-900 before and after 5000 cycles of CV in 1.0 m KOH solution. d) Polarization curves of the Co/NBC-900||Co/NBC-900 electrode and pure Ni foam||Ni foam in 1.0 m KOH solution without iR compensation in a two-electrode system. e) Chronopotentiometric (V–t) curve of overall water splitting in 1.0 m KOH solution. f) Calculated free energy diagram of the HER on Co/NBC-900 and surface oxidation Co/NBC-900, respectively.

despite the smaller onset potential of Ir/C (1.45 V vs RHE). In addition, the catalytic current of Co/NBC-900 at high potentials (>1.56 V vs RHE) also exceeds that of commercial Ir/C. The overpotential of Co/NBC-900 is also comparable to those of other cobalt-based OER catalysts (Tables S5 and S6, Supporting Information). The catalytic OER kinetics of the Co/NBC samples was analyzed by linear fitting the Tafel plots (log j–η) derived from LSV curves according to the Tafel equation (η = blog j + a, where η is the overpotential, b corresponds to the Tafel slope, j is the current density, and a is a constant) (Figure 3b). The Tafel slope of Co/NBC-900 is 70 mV dec−1, which is significantly smaller than those of the samples obtained at different temperatures. The lower Tafel slope for Co/NBC-900 indicates its beneficial kinetics during OER in alkaline media. It is worth noting that Co/NBC-900 displays the best OER activity, although the BET surface area of Co/NBC-900 (73.2 m2 g−1) is lower than that of Co/NBC-800 (131 m2 g−1) and Co/NBC-600 (78.1 m2 g−1) (Figure S4, Supporting Information). It signifies that the OER performance of the Co/NBC catalysts is attributed to the intrinsic catalytic activity of active sites rather than only to the enhanced specific surface area, implying that the intrinsic activity of Co/NBC-900 is significantly higher than that of the other three samples. This can be ascribed to the higher content of the conductive pyridinic CN and CB structures as well as surface oxides on the outer layer of metallic cobalt of Co/NBC-900 compared with the other samples, which contributes to enhance electron transport capability and provides


more exposed catalytic active sites allowing the easily access of electrolyte, and thus facilitates the OER process. To evaluate electron transfer capacity, electrochemical impedance spectrum was collected. Figure 3c indicates that the Nyquist plot of Co/ NBC-900 presents a smaller semicircle than those of the other Co/NBC samples, suggesting a faster electron transfer during the catalytic process. We further estimated the electrochemical double-layer capacitance (Cdl), which can indirectly reflect the electrochemically active surface areas (ECSA) of the catalysts, by collecting cyclic voltammograms in a non-Faradaic region of 1.1–1.2 V versus RHE. Figure 3d demonstrates that with the increase of carbonization temperature, the Cdl increased from 19 mF for Co/NBC-600 to 134 mF for Co/NBC-900. Co/NBC900 possesses the highest ECSA, suggesting that it allows more effective accessibility of its active sites. Meanwhile, chronopotentiometry (V–t) response test at a constant anodic current density of 10 mA cm−2 exhibits negligible increase of overpotential after at least 10 h operating in alkaline solution, confirming the long-term stability of Co/NBC-900 (Figure 3e). Previous studies show that the true catalytically active species of metal nonoxide catalysts for OER reaction could be metal oxide/hydroxides.[25] To explore the surface structure of Co/NBC-900 after OER reaction, we performed a series of post-OER measurements. The XRD of post-OER sample shows cobalt oxides are converted into CoOOH besides the presence of metallic cobalt (Figure S8, Supporting Information). TEM measurements show that the morphology and structure of the

1801136 (6 of 9)



www.afm-journal.de

post-OER sample and the cobalt species are still encapsulated in the graphitic carbon shell (Figure S8, Supporting Information), which was also confirmed by XPS measurements (Figure S8, Supporting Information). To further give insight into the chemical structure, X-ray absorption spectroscopy (XAS) measurements were performed (Figure S9, Supporting Information). From the Co K-edge XANES spectra of post-OER Co/NBC-900, it can be found that the Co K-edge of Co/NBC-900 is different from that of Co foil but close to that of Co3O4, indicating that the oxidation state of Co is higher than +2. However, the preedge region at 7709 eV is in disagreement with that of Co3O4 which displays a strong peak, suggesting that the post-Co/NBC900 could contain metallic cobalt and cobalt oxide/hydroxide. Fourier transformed k3-weight extend EXAFS spectrum shows peak I and peak II at 1.47 and 2.27 Å, corresponding to the first coordination sphere CoO and nearest second shell CoCo, respectively. Their distance and relative intensity ratio (I < II) are in good agreement with the CoOOH phase but differ from that of Co3O4 (Figure 3f; Figure S9, Supporting Information). Of note, there is noticeable difference between post-Co/ NBC-900 and CoOOH. Compared with the CoOOH, the postCo/NBC-900 sample shows a relative intense peak at higher R value (region III) besides shorter CoCo distance (region II) (Figure 3f). These characteristics bear resemblance to that of the metallic cobalt phase, implying that the metallic cobalt phase preserved in the Co/NBC-900 after reaction, in good accordance with the XRD and TEM results. The shorter CoCo distance confirms a chemical interaction presents between metallic cobalt nanoparticles and oxide species. Thus, together with the XPS, XRD, and XAS, the post-OER catalyst consists of cobalt oxide/hydroxides at the surface of metallic cobalt, indicating that the cobalt hydroxides dominate on the surface of the catalyst. The observation is consistent with previous studies that the cobalt oxide layer on the surface of 3d transition metal nanoparticles served as catalytic species toward OER in basic solution.[10,25,26] The electrocatalytic HER performance of the Co/NBC catalysts was also assessed using a three-electrode system containing graphite rod as reference electrode in 1.0 m KOH electrolyte. The polarization curves of all Co/NBC samples are shown in Figure 4. All the Co/NBC samples show HER activity, and their activities depend on the carbonization temperature of the samples. As shown in Figure 4, Co/NBC-600 shows poor HER activity and demands an overpotential of 292 mV to achieve the current density of −10 mA cm−2. Comparatively, Co/NBC-900 affords a smaller onset overpotential of 15 mV and requires an overpotential of 117 mV to drive the current density of −10 mA cm−2. The activity is comparable with that of many reported precious-metal-free HER catalysts in alkaline media including some transition metal (Co or Ni) phosphides (Tables S5 and S7, Supporting Information), but the activity of Co/NBC900 is inferior to that of Pt/C with the overpotential of only 24 mV at −10 mA cm−2 (Figure 4a). The Tafel slope of Co/NBC900 was calculated to be 146 mV dec−1 and is lower than those of other Co/NBC samples (Figure 4b). Also, the Co/NBC-900 exhibits high durability during the HER. After being subjected to 5000 CVs between −0.2 and −0.5 V versus RHE in 1.0 m KOH solution, the Co/NBC-900 experienced slight catalytic degradation described in polarization curve with an overpotential of


−117 mV (initial) and −142 mV (final) at a current density of −10 mA cm−2 (Figure 4c). The XPS spectrum of the post-HER sample shows that the main composition remains unchanged although the amount of surface cobalt oxides slightly increases (Figure S10, Supporting Information), which is due to the surface oxidation of cobalt in strong alkaline condition. It is noteworthy that there are few reports about B,N-codoped cobalt-based catalysts possessing such a favorable HER activity in basic media. Inspired by the promising half-reaction performance in OER and HER, we investigated its overall water splitting activity with a two-electrode configuration by depositing Co/NBC-900 on Ni foam as both anode and cathode at room temperature in 1 m KOH solution (see the Supporting Information). Figure 4d indicates high performance, reaching a water-splitting current density of 10 mA cm−2 with the requirement of a cell voltage of 1.68 V in 1.0 m KOH. Impressively, the overall splitting performance of Co/NBC-900 is comparable to other non-noble metal bifunctional electrocatalysts (Table S8, Supporting Information). Furthermore, the long-term stability of the bifunctional catalyst was also studied. As shown in Figure 4e, the Co/NBC900 shows excellent stability and slightly increased voltage at fixed current of 10 mA cm−2 after 6 h of testing, demonstrating its superior stability. These results confirm that as-prepared Co/NBC-900 material is a promising electrocatalyst for overall water-splitting. The above experimental results indicate that the Co/NBC900 can be established as a bifunctional active electrocatalyst for OER and HER. In order to understand the effect of cobalt species on the catalytic activity, metal free NBC-900 was also investigated. It is found that a slightly active B,N-codoped carbon catalyst seems highly active by incorporation of cobalt, indicating the crucial role of the cobalt species. To gain further insight into the excellent performance of Co/NBC-900, the DFT calculations were performed to investigate electronic structure, adsorption energies, H* adsorption free energy, and the variation of Gibbs free energy in those elementary steps of watersplitting reaction on Co/NBC and surface oxidation Co/NBC models. A good catalyst should have a moderate free energy for H adsorption |ΔGH*|, close to zero, to compromise the reaction barriers of the adsorption and desorption steps. In our DFT simulations, the hollow (H) site on the h-Co/NBC and surface oxidation Co/NBC (referred to as h-CoO@Co/NBC) models shows the high free energy for H* adsorption (Figure 4f). When H is at the title (T) site, t-Co/NBC and t-CoO@Co/NBC present small value of Gibbs free energy (Figure 4f). In addition, the |ΔGH*| of oxygen atoms bonded with metallic cobalt decrease from 0.18 to 0.02 eV, resulting in a more favorable H* adsorption–desorption property. The introduction of CoO optimizes not only H* adsorption property but also the adsorption capacity of CoO@Co/NBC to O*, OH*, and OOH*, as shown in Table S9 and Figure S10 (Supporting Information). The adsorption energies and overpotential in the OER process were listed in Table S9 (Supporting Information). For the two models, the formation of OOH* turns out to be the highest and becomes the rate-limiting step for water oxidation reaction. Obviously, according to overpotential value, the OER activity of CoO@Co/NBC is better than Co/NBC. This is probably due to the suitable interactions with OH*, O*, and OOH*. Moreover,

1801136 (7 of 9)



www.afm-journal.de

the OOH* is decomposed into O* and OH* in Co/NBC, which causes bigger energy barrier (1.43 eV). In addition, Figure S11 (Supporting Information) also suggested the effective electron transferring from N,B-codoped carbon to cobalt cluster as revealed by charge density difference, indicating the presence of coupled interactions between Co/NBC with OOH, O, and OH. The above results suggest that the Co/NBC-900 catalyst is excellent bifunctional electrocatalyst for water splitting. On the basis of experimental and DFT studies, the high performance could be ascribed to be the following features: i) XRD, XPS, and XAS results indicate that the cobalt species in the Co/NBC samples consist of metallic cobalt and surface cobalt oxides. The surface oxides can coordinate and discharge the H2O/OH− species at the surface of metal nanoparticles, which contributes to the formation of adsorbed OOH species revealed by postmortem of XPS, XRD, and XAS characterizations. Meanwhile, the cobalt oxides not only contribute to catalyzing the OER but also play a role in stabilizing the metallic cobalt, whereas the metallic cobalt core serves as the conductive support for the surface oxide/hydroxide layer, which accelerated the whole OER reaction.[26b] On the other hand, the surface oxidation metallic cobalt cores possess moderate reversible adsorption and desorption energy barriers for hydrogen on the catalyst surface, favoring HER process, as revealed by DFT. ii) The abundance of pyridine-like CN and CB could supply more accessible active sites for the electrocatalytic process. High content of these structures favors high activity. Thus, it is reasonable that the activity of Co/NBC-900 is better than the other three samples. iii) Furthermore, the pore structure allows the contact between the active sites and electrolyte, facilitating fast charge and mass transport during the electrocatalysis process. iv) The cobalt species encapsulated in heteroatom doped carbon supports, and the intimate interactions between these components via the CoNx protect the metal nanoparticles from aggregation which facilitate the charge transfer, which is of help for improving the conductivity and stability. v) DFT calculations reveal that the surface oxidation metallic cobalt species have a rational energy barrier for H adsorption, whereas the electron transfer from N,B-codoped carbon could increase the occupancy of the antibonding orbital of cobalt, which contributes to improve the catalytic performance.[27] Overall, all of these observations above confirmed that the excellent activity of Co/NBC hybrids is attributed to the synergistic effect between surface oxidation metallic cobalt species, N,B-codoped interconnected graphene and carbon nanotube, and the interactions between these components.

3. Conclusion In summary, we have successfully fabricated N,B-codoped graphitic carbon decorated cobalt nanomaterials (Co/NBC) by direct carbonization of a new porous boron imidazolate framework. The optimized Co/NBC-900 hybrid displayed efficient bifunctional electrocatalytic activity and stability toward both OER and HER in 1.0 m KOH solution. It required overpotential of 302 mV to reach the current density of 10 mA cm−2 for the OER, whereas for the HER the overpotential is 117 mV. When it acts as an overall water splitting catalyst, it approaches


−10 mA cm−2 at a cell voltage of 1.68 V with long-term stability. Both experiments and theoretical calculations showed that the superior activity and stability of Co/NBC-900 benefited from the strong synergistic effects between metallic cobalt with surface oxides and N,B-codoped interconnected graphitic carbon and carbon nanotube in alkaline media. This work demonstrates the potential applications of a new class of boron imidazolate frameworks and paves a new way to develop highly active and non-noble metal-based catalysts.

Supporting Information Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements The authors acknowledge the financial support from the National Natural Science Foundation of China (grant nos. 21425102, 21673238, 21773242 and 51772291) and the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB20000000). The authors also thank Dr. L. Zheng from Beijing Synchrotron Radiation Facility, Institute of High Energy Physics, Chinese Academy of Sciences, for the XANES and EXAFS experiments.

Conflict of Interest The authors declare no conflict of interest.

Keywords electrocatalysis, hydrogen evolution, metal–organic framework, oxygen evolution, water splitting Received: February 11, 2018 Revised: March 16, 2018 Published online:

[1] J. A. Turner, Science 2004, 305, 972. [2] M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. Mi, E. A. Santori, N. S. Lewis, Chem. Rev. 2010, 110, 6446. [3] T. R. Cook, D. K. Dogutan, S. Y. Reece, Y. Surendranath, T. S. Teets, D. G. Nocera, Chem. Rev. 2010, 110, 6474. [4] a) J. Nai, H. Yin, T. You, L. Zheng, J. Zhang, P. Wang, Z. Jin, Y. Tian, J. Liu, Z. Tang, L. Guo, Adv. Energy Mater. 2015, 5, 1401880; b) Z. Peng, D. Jia, A. M. Al-Enizi, A. A. Elzatahry, G. Zheng, Adv. Energy Mater. 2015, 5, 1402031; c) H. Tabassum, W. Guo, W. Meng, A. Mahmood, R. Zhao, Q. Wang, R. Zou, Adv. Energy Mater. 2017, 7, 1601671; d) S. Wan, J. Qi, W. Zhang, W. Wang, S. Zhang, K. Liu, H. Zheng, J. Sun, S. Wang, R. Cao, Adv. Mater. 2017, 29, 1700286; e) H. Wang, S. Zhuo, Y. Liang, X. Han, B. Zhang, Angew. Chem., Int. Ed. 2016, 55, 9055; f) Y. Zhu, W. Zhou, J. Sunarso, Y. Zhong, Z. Shao, Adv. Funct. Mater. 2016, 26, 5862; g) Q. Yin, J. M. Tan, C. Besson, Y. V. Geletii, D. G. Musaev, A. E. Kuznetsov, Z. Luo, K. I. Hardcastle, C. L. Hill, Science 2010, 328, 342. [5] a) J. Wang, L. Ji, S. Zuo, Z. Chen, Adv. Energy Mater. 2017, 7, 1700107; b) Y. Zhao, C. Chang, F. Teng, Y. Zhao, G. Chen, R. Shi, G. I. N. Waterhouse, W. Huang, T. Zhang, Adv. Energy Mater. 2017,

1801136 (8 of 9)



www.afm-journal.de

7, 1700005; c) L. Yu, H. Zhou, J. Sun, F. Qin, F. Yu, J. Bao, Y. Yu, S. Chen, Z. Ren, Energy Environ. Sci. 2017, 10, 1820; d) J. Duan, S. Chen, C. Zhao, Nat. Commun. 2017, 8, 15341; e) N. Jiang, B. You, M. Sheng, Y. Sun, Angew. Chem., Int. Ed. 2015, 54, 6251; f) J. Li, Y. Wang, T. Zhou, H. Zhang, X. Sun, J. Tang, L. Zhang, A. M. Al-Enizi, Z. Yang, G. Zheng, J. Am. Chem. Soc. 2015, 137, 14305; g) Y. Zhu, W. Zhou, Y. Zhong, Y. Bu, X. Chen, Q. Zhong, M. Liu, Z. Shao, Adv. Energy Mater. 2017, 7, 1602122; h) J. Liu, D. Zhu, C. Guo, A. Vasileff, S. Z. Qiao, Adv. Energy Mater. 2017, 7, 1700518. [6] a) C. Panda, P. W. Menezes, C. Walter, S. Yao, M. E. Miehlich, V. Gutkin, K. Meyer, M. Driess, Angew. Chem., Int. Ed. 2017, 56, 10506; b) Y. Hou, M. R. Lohe, J. Zhang, S. Liu, X. Zhuang, X. Feng, Energy Environ. Sci. 2016, 9, 478. [7] a) K. Jayaramulu, J. Masa, O. Tomanec, D. Peeters, V. Ranc, A. Schneemann, R. Zboril, W. Schuhmann, R. A. Fischer, Adv. Funct. Mater. 2017, 27, 1700451; b) P. Chen, T. Zhou, M. Zhang, Y. Tong, C. Zhong, N. Zhang, L. Zhang, C. Wu, Y. Xie, Adv. Mater. 2017, 29, 1701584; c) Y. Liu, Q. Li, R. Si, G. D. Li, W. Li, D. P. Liu, D. Wang, L. Sun, Y. Zhang, X. Zou, Adv. Mater. 2017, 29, 1606200. [8] Z.-Y. Yu, Y. Duan, M.-R. Gao, C.-C. Lang, Y.-R. Zheng, S.-H. Yu, Chem. Sci. 2017, 8, 968. [9] a) Y. Zhang, X. Xia, X. Cao, B. Zhang, N. H. Tiep, H. He, S. Chen, Y. Huang, H. J. Fan, Adv. Energy Mater. 2017, 7, 1700220; b) Y. Pi, Q. Shao, P. Wang, J. Guo, X. Huang, Adv. Funct. Mater. 2017, 27, 1700886. [10] J. Masa, P. Weide, D. Peeters, I. Sinev, W. Xia, Z. Sun, C. Somsen, M. Muhler, W. Schuhmann, Adv. Energy Mater. 2016, 6, 1502313. [11] a) Z. H. Xue, H. Su, Q. Y. Yu, B. Zhang, H. H. Wang, X. H. Li, J. S. Chen, Adv. Energy Mater. 2017, 7, 1602355; b) L. Yan, L. Cao, P. Dai, X. Gu, D. Liu, L. Li, Y. Wang, X. Zhao, Adv. Funct. Mater. 2017, 27, 1703455; c) J. Masa, S. Barwe, C. Andronescu, I. Sinev, A. Ruff, K. Jayaramulu, K. Elumeeva, B. Konkena, B. Roldan Cuenya, W. Schuhmann, ACS Energy Lett. 2016, 1, 1192; d) J. Song, C. Zhu, B. Z. Xu, S. Fu, M. H. Engelhard, R. Ye, D. Du, S. P. Beckman, Y. Lin, Adv. Energy Mater. 2016, 7, 1601555; e) M. Ledendecker, S. Krick Calderón, C. Papp, H.-P. Steinrück, M. Antonietti, M. Shalom, Angew. Chem., Int. Ed. 2015, 54, 12361. [12] a) X. F. Lu, L. F. Gu, J. W. Wang, J. X. Wu, P. Q. Liao, G. R. Li, Adv. Mater. 2017, 29, 1604437; b) J. Zhou, Y. Dou, A. Zhou, R. M. Guo, M. J. Zhao, J. R. Li, Adv. Energy Mater. 2017, 7, 1602643; c) Y. Li, W. Zhou, J. Dong, Y. Luo, P. An, J. Liu, X. Wu, G. Xu, H. Zhang, J. Zhang, Nanoscale 2018, 10, 2649. [13] a) K. Liu, H. Zhong, F. Meng, X. Zhang, J. Yan, Q. Jiang, Mater. Chem. Front. 2017, 1, 2155; b) Y. Tong, P. Chen, T. Zhou, K. Xu, W. Chu, C. Wu, Y. Xie, Angew. Chem., Int. Ed. 2017, 56, 7121; c) H. Tabassum, R. Zou, A. Mahmood, Z. Liang, S. Guo, J. Mater. Chem. A 2016, 4, 16469; d) Z. Jiang, Z.-J. Jiang, T. Maiyalagan, A. Manthiram, J. Mater. Chem. A 2016, 4, 5877; e) H. Zhang,


Z. Ma, J. Duan, H. Liu, G. Liu, T. Wang, K. Chang, M. Li, L. Shi, X. Meng, K. Wu, J. Ye, ACS Nano 2016, 10, 684; f) H. Tabassum, A. Mahmood, Q. Wang, W. Xia, Z. Liang, B. Qiu, R. zhao, R. Zou, Sci. Rep. 2017, 7, 43084; g) S. Liu, Z. Wang, S. Zhou, F. Yu, M. Yu, C. Y. Chiang, W. Zhou, J. Zhao, J. Qiu, Adv. Mater. 2017, 29, 1700874; h) B. You, N. Jiang, M. Sheng, S. Gul, J. Yano, Y. Sun, Chem. Mater. 2015, 27, 7636. [14] a) Y. Qian, Z. Hu, X. Ge, S. Yang, Y. Peng, Z. Kang, Z. Liu, J. Y. Lee, D. Zhao, Carbon 2017, 111, 641; b) R. Wang, X. Y. Dong, J. Du, J. Y. Zhao, S. Q. Zang, Adv. Mater. 2018, 30, 1703711. [15] a) H.-X. Zhang, M. Liu, T. Wen, J. Zhang, Coord. Chem. Rev. 2016, 307, 255; b) H.-X. Zhang, H.-R. Fu, H.-Y. Li, J. Zhang, X. Bu, Chem. Eur. J. 2013, 19, 11527; c) J. Zhang, T. Wu, C. Zhou, S. Chen, P. Feng, X. Bu, Angew. Chem., Int. Ed. 2009, 48, 2542. [16] H.-X. Zhang, M. Liu, X. Bu, J. Zhang, Sci. Rep. 2014, 4, 3923. [17] a) T. Zhou, Y. Du, S. Yin, X. Tian, H. Yang, X. Wang, B. Liu, H. Zheng, S. Qiao, R. Xu, Energy Environ. Sci. 2016, 9, 2563; b) T. Zhou, Y. Du, D. Wang, S. Yin, W. Tu, Z. Chen, A. Borgna, R. Xu, ACS Catal. 2017, 7, 6000; c) Y.-Z. Chen, C. Wang, Z.-Y. Wu, Y. Xiong, Q. Xu, S.-H. Yu, H.-L. Jiang, Adv. Mater. 2015, 27, 5010; d) G. Wu, K. L. More, C. M. Johnston, P. Zelenay, Science 2011, 332, 443. [18] H. Jin, J. Wang, D. Su, Z. Wei, Z. Pang, Y. Wang, J. Am. Chem. Soc. 2015, 137, 2688. [19] J. Jin, F. Pan, L. Jiang, X. Fu, A. Liang, Z. Wei, J. Zhang, G. Sun, ACS Nano 2014, 8, 3313. [20] K. Artyushkova, S. Pylypenko, T. S. Olson, J. E. Fulghum, P. Atanassov, Langmuir 2008, 24, 9082. [21] a) S. Chen, J. Duan, W. Han, S. Z. Qiao, Chem. Commun. 2014, 50, 207; b) L. Ci, L. Song, C. Jin, D. Jariwala, D. Wu, Y. Li, A. Srivastava, Z. F. Wang, K. Storr, L. Balicas, F. Liu, P. M. Ajayan, Nat. Mater. 2010, 9, 430. [22] Y. Xu, W. Tu, B. Zhang, S. Yin, Y. Huang, M. Kraft, R. Xu, Adv. Mater. 2017, 29, 1605957. [23] a) S. Wang, L. Zhang, Z. Xia, A. Roy, D. W. Chang, J. B. Baek, L. Dai, Angew. Chem., Int. Ed. 2012, 51, 4209; b) Z. Lu, J. Wang, S. Huang, Y. Hou, Y. Li, Y. Zhao, S. Mu, J. Zhang, Y. Zhao, Nano Energy 2017, 42, 334. [24] Y. Du, J.-o. Wang, L. Jiang, L. S. Borgna, Y. Wang, Y. Zheng, T. Hu, J. Synchrotron Radiat. 2014, 21, 756. [25] S. Jin, ACS Energy Lett. 2017, 2, 1937. [26] a) J. Masa, I. Sinev, H. Mistry, E. Ventosa, M. d. l. Mata, J. Arbiol, M. Muhler, B. R. Cuenya, W. Schuhmann, Adv. Energy Mater. 2017, 7, 1700381; b) L. Wu, Q. Li, C. H. Wu, H. Zhu, A. Mendoza-Garcia, B. Shen, J. Guo, S. Sun, J. Am. Chem. Soc. 2015, 137, 7071. [27] a) Y. Zheng, Y. Jiao, M. Jaroniec, S. Z. Qiao, Angew. Chem., Int. Ed. 2015, 54, 52; b) J. Suntivich, K. J. May, H. A. Gasteiger, J. B. Goodenough, Y. Shao-Horn, Science 2011, 334, 1383; c) J. O. M. Bockris, T. Otagawa, J. Electrochem. Soc. 1984, 131, 290.

1801136 (9 of 9)
