In situ interfacial engineering of nickel tungsten carbide janus structures for highly efficient overall water splitting

Journal Pre-proofs Article In situ interfacial engineering of nickel tungsten carbide janus structures for highly efficient overall water splitting Songge Zhang, Guohua Gao, Han Zhu, Lejuan Cai, Xiaodi Jiang, Shuanglong Lu, Fang Duan, Weifu Dong, Yang Chai, Mingliang Du PII: DOI: Reference:

S2095-9273(20)30064-5 https://doi.org/10.1016/j.scib.2020.02.003 SCIB 952

To appear in:

Science Bulletin

Received Date: Revised Date: Accepted Date:

2 December 2019 7 January 2020 4 February 2020

Please cite this article as: S. Zhang, G. Gao, H. Zhu, L. Cai, X. Jiang, S. Lu, F. Duan, W. Dong, Y. Chai, M. Du, In situ interfacial engineering of nickel tungsten carbide janus structures for highly efficient overall water splitting, Science Bulletin (2020), doi: https://doi.org/10.1016/j.scib.2020.02.003

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2020 Science China Press. Published by Elsevier B.V. and Science China Press. All rights reserved.

Article Received 2 December 2019 Received in revised form 7 January 2020 Accepted 4 Feburary 2020

In situ interfacial engineering of nickel tungsten carbide Janus structures for highly efficient overall water splitting Songge Zhang1#, Guohua Gao2#, Han Zhu1*, Lejuan Cai3, Xiaodi Jiang2, Shuanglong Lu1, Fang Duan1, Weifu Dong1, Yang Chai3*, Mingliang Du1* #

These authors contributed equally to this work.

1 Key Laboratory of Synthetic and Biological Colloids, Ministry of Education, School of Chemical and Material Engineering, Jiangnan University, Wuxi 214122, China. 2 Shanghai Key Laboratory of Special Artificial Microstructure Materials and Technology, Key Laboratory of Road and Traffic Engineering of the Ministry of Education, Tongji University, Shanghai 200092, China 3 Department of Applied Physics, The Hong Kong Polytechnic University, Hong Kong 999077, China. Corresponding authors: [email protected]

Email:

[email protected];

[email protected];

Abstract Regulating chemical bonds to balance the adsorption and disassociation of water molecules on catalyst surfaces is crucial for overall water splitting in alkaline solution. Here we report a facile strategy for designing Ni2W4C-W3C Janus structures with abundant Ni–W metallic bonds on surfaces through interfacial engineering. Inserting Ni atoms into the W 3C crystals in reaction progress generates a new Ni2W4C phase, making the inert W atoms in W3C be active sites in Ni2W4C for overall water splitting. The Ni2W4C-W3C/carbon nanofibers (Ni2W4C-W3C/CNFs) require overpotentials of 63 mV to reach 10 mA cm–2 for hydrogen evolution reaction (HER) and 270 mV to reach 30 mA cm–2 for oxygen evolution reaction (OER) in alkaline electrolyte, respectively. When utilized as both cathode and anode in alkaline solution for overall water splitting, cell voltages of 1.55 and 1.87 V are needed to reach 10 and 100 mA cm–2, respectively. Density functional theory (DFT) results indicate that the strong interactions between Ni and W increase the local electronic states of W atoms. The Ni2W4C provides active sites for cleaving H−OH bonds, and the W 3C facilitates the combination of Hads intermediates into H2 molecules. The in situ electrochemicalRaman results demonstrate that the strong absorption ability for hydroxyl and water molecules and further demonstrate that W atoms are the real active sites. Keywords Interfacial engineering; Nickel tungsten carbide; Janus structure; Water splitting 1.

Introduction Rational design of highly efficient, low cost and durable electrocatalysts is fundamentally crucial

for developing new energy technologies [1]. For the sake of sustainable and efficient hydrogen production, electrochemical water splitting is paramount significant due to its inherent advantages,

1

including feasibility of large-scale production and highly pure product [2]. The involved two half reactions, hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) usually perform well at different electrolyte [3]. It is known that even the individual state-of-the-art Pt/C and IrO2 catalyst cannot effectively catalyze HER and OER in overall water splitting cells simultaneously. It is because the Pt/C catalysts are intrinsically highly active for HER and in contrast, the IrO 2 catalysts are only excellent for OER [4]. For industrial condition, the HER and OER involved in alkaline electrolyte always suffer from sluggish reaction kinetics for the multistep proton coupled electron transfer process [5]. Therefore, it is highly desirable to develop the alternative materials featured with earth abundant and high electrocatalytic activity, which is comparable to state-of-art noble metals [6]. To meet the increased demands of overall water splitting, it requires to design bifunctional catalysts to catalyze the HER and OER in the same electrolyte simultaneously. However, transition metal sulfides and nitrides usually exhibit superior activity for HER, while transition metal oxides and hydroxides/oxyhydroxides are active materials for OER [7]. In addition, many OER and HER catalysts are only active in either acidic or alkaline electrolytes [8, 9]. Transition-metal carbides, such as, tungsten carbide featured with platinum-like catalytic behavior, have been extensively investigated in HER due to its low price, high corrosion resistance, and superior electronic conductivity [10]. In order to realize the application in both HER and OER in alkaline, it requires researchers to tackle a few big challenges. Pure tungsten carbide has too strong adsorption for H* and thus exhibits low activity for HER in alkaline media [11]. An ideal HER electrocatalyst should have fast water adsorption and H* releasing ability [12]. Therefore, developing essential strategies to modify the electronic structure of tungsten carbide to promote the activity is highly desirable. The integration of active component with tungsten carbide could be an effectively feasible strategy to improve the electrocatalytic activity. The general reported approaches to improve the HER activity of transition-metal carbide are introducing a second transition metals (Co, Ni, and Fe) as additives [13-15]. However, the doping different components with carbide into one architecture without forming new phases and interfaces have limited effects on the improvement of overall efficient OER and HER in one electrolyte. The integration of different phase’ features with different functionalities can induce remarkable synergistic effects, including electronic regulation, interfacial stabilization, atomic arrangement etc. [16]. One approach is the adoption of Janus structures particle which surface exhibits distinct composition/physical properties and therefore exhibits a distinct interface. Particularly, the abundant interfaces between different phases play crucial roles in binding, transforming, and transporting the surface species such as adsorbents, electrons and intermediates [14-16]. Designing and constructing of novel Janus structures with multi-interfaces could be a fundamental strategy to maximize the advantages of hybrid catalysts towards the overall water splitting [17]. Herein, we provide a facile strategy for designing the unique Ni2W4C-W3C Janus structures with abundant interfaces and superficial Ni–W metallic bonds through in situ interfacial engineering by combining the electrospinning technology and graphitization process. The advantages of this approach would be the new formation of Ni2W4C phase, making inert W atoms in W3C be active sites in Ni2W4C for the overall water splitting. In the Janus structures, the Ni 2W4C provides active

2

sites for cleaving H−OH bonds, and the W3C facilitates the combination of Hads intermediates into H2 molecules. The in situ electrochemical-Raman results demonstrated the strong absorption ability for hydroxyl and water molecules and the in situ Raman spectra for HER and OER process further demonstrating that W atoms is the real active sites. 2.

Experimental

2.1 Synthesis of Ni2W4C-W3C/CNFs hybrid In a typical procedure, 0.17 g nickel nitrate hexahydrate and 0.43 g ammonium met tungstate were dissolved in 12 mL polyacrylonitrile/dimethylformamide (PAN/DMF) solution with a mass fraction of 13% PAN and stirred by magnetic stirring apparatus to obtain homogeneous solution. Then the mixed solution was transferred to a syringe with a stainless copper needle at the tip. The electrospinning experiment was performed in a home-made electrostatic spinning machine with a flow rate of 0.6 mL h–1 and high voltage of 12 kV. The distance between collector and needle was 13 cm. All experiments were operated at room temperature and ultimately Ni-W-PAN precursor nanofibers were acquired. The directed growth of the Ni 2W4C-W3C/CNFs were carried out in a chemical vapor deposition (CVD) furnace. The as-collected nanofibers mats were placed in a ceramic boat and calcined at the center of heating zone of the furnace. Next, the Ni-W-PAN mats were heated to 230 oC in air for pre-oxidation at a rate of 5 oC min–1 and maintained 3 h for stabilization. After this process, the furnace was heated up to 1000 oC indirectly at a rate of 5 oC min–1, and the temperature was held constant for 3 h under the protection of Ar. After the heat treatment, the furnace was cooled to room temperature in an Ar atmosphere. 2.2 Materials characterizations The field emission scanning electron microscopy (FE-SEM) images were characterized via a JSM-6700F FE-SEM (JEOL, Japan) at an acceleration voltage of 3 kV. Transmission electron microscopy (TEM) images were taken using a JSM-2100 transmission electron microscope (JEOL) at an acceleration voltage of 200 kV. High-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) images and STEM mapping were recorded using a STEM (Tecnai G2 F30S-Twin, Philips-FEI, USA) at an acceleration voltage of 300 kV. X-ray diffraction (XRD) patterns were analyzed using a Bruker AXS D8 DISCOVER X-ray diffractometer with Cu Kα radiation (λ = 1.5406 Å) at a scanning rate of 0.02 (2θ)–1 in the 2θ range of 10°–90°. X-ray photoelectron spectra of the samples were recorded using an X-ray photoelectron spectrometer (Kratos Axis Ultra DLD, Japan) with an Al (mono) Kα source (1486.6 eV). The Al Kα source was operated at 15 kV and 10 mA. Raman spectra were recorded using a Renishaw in Via Raman microscope (LabRAM HR800, French) with a 532 nm laser excitation source. 2.3 Electrochemical characterization The electrochemical measurements were performed with a CHI 660D electrochemical work station (CH Instruments, Inc., China). The conventional three-electrode system was used, which incorporate working electrode, count electrode and reference electrode. The carbon rod was used as

3

count electrode and the saturated calomel electrode (SCE) was utilized as reference electrode. The NiWC-2/CNFs were tailored into a neat square of 1 cm2 and directly used as the working electrode. The potential reported in this study were calibrated and converted to the reversible hydrogen electrode (RHE) by the equation ERHE=ESCE+0.244+0.059×pH. All tests were carried out without the ohmic potential drop (iR) drop compensation. The polarization curves were acquired by linear sweep voltammetry (LSV) at a scan rate of 2 mV s–1 in KOH solution. The HER tests were performed between 0.1 and –0.4 V (vs. RHE) and the OER tests were performed between 1.2 and 1.8 V (vs. RHE). Before LSV tests, cyclic voltammograms were conducted for stability and electrode activation at a scan rate of 100 mV s–1 between 0.5 and 0.5 V (vs. RHE). Tafel plots were obtained by taking the logarithm of current density as X-axis and overpotential as Y-axis according to the Tafel equation η = a + blogj,

(1)

where a is the Tafel constant, b is the Tafel slope, j is the cathodic current density and η is the overpotential. The electrochemical impedance spectroscopy spectras (EISs) were operated in a potentiostatic mode in the frequency ranging from 105 to 0.1 HZ with the amplitude of 15 mV at 1.5 V vs. RHE in the basic media. To evaluate the electrochemical active surface area (ECSA), the electrochemical double layer capacitances (Cdl) were conducted with different scan rates form 20 to 100 mV s–1. The chronopotentiometry tests for stability performance were measured at a constant potential of –0.14 V vs. RHE for HER and 1.5 V vs. RHE for OER for 12 h. The overall water splitting was also carried out in three-electrode system. The voltage ranged from 0 to 0.5 V (vs. RHE). The stability test was conducted with the voltage of 1.5 V. For overall water splitting, the experiment conducted in two-electrode system, in which the samples worked as working electrode and count electrode at the same time. Polarization curves for overall water splitting was obtained at a scan rate of 2 mV s−1. 2.4 In situ electrochemical-Raman tests The Raman test were carried out with a LabRAM HR800 confocal microscope (Horiba, French). A long-distance objective lens with a numerical aperture of 0.1 were used to focus a diode-pumped solid-state laser beam on the sample. The laser wavelength is 785 nm. The Raman signal was collected in a back-scattering geometry. 2.5 First-principles calculations All the calculations in this paper were performed by using the Vienna ab initio simulation package (VASP), the exchange and correlation energy functional was treated by the Perdew–Burke–Ernzerh of variant of the generalized gradient approximation (GGA). Interaction between ion and electrons were described with projector augmented wave pseudo potentials (PAW) approach, and the energy cutoff for the plane wave basis set was set to be 500 eV. The k-point sampling was performed usingcentered 6×6×1 points, which gives the total energy of complex within 10–5 eV. The MoS2 slab was

4

attached on the (010) plane of W3C and Ni2W4C based on experimental data, of which the vacuum slab was set to be 15 Å. 3.

Results and discussion As shown in Fig. 1a, large amounts of small nanoparticles uniformly distribute on the surfaces of

CNFs. The CNFs host provides a stable bracket and reaction zone for the spontaneous formation of the Ni2W4C-W3C nanoparticles and exhibits a unique 3D networks, which are beneficial for the efficient charge separation and fast change transfer [18]. The Ni2W4C-W3C nanoparticles display pea-like morphology and the size ranges from 10 to 80 nm (Figs. 1b and S1 (online)). Fig. 1b and c exhibit the high-angle annular dark field scanning transmission electron microscopy (HAADFSTEM) image and STEM energy dispersive X-ray (STEM-EDX) mapping images of a single Ni2W4C-W3C nanoparticle. As shown in Fig. 1c, the pea-like Ni2W4C-W3C nanoparticle displays the uniform distribution of C, Ni and W elements. It is interesting that the C and W elements both exhibit the same distribution with pea-like morphology. However, the Ni element only locates at the one side of the pea-like morphology. The STEM-EDX mapping results indicate that the pea-like Ni2W4C-W3C nanoparticle exhibits two different crystals phase, Ni2W4C and W3C phases, forming a unique Janus heterostructures. High-resolution TEM image of a single Ni2W4C-W3C nanoparticle (Fig. 1d) suggests that the Ni2W4C-W3C nanoparticle exhibits grain boundary and two different crystal phases. The W3C phase indicates the lattice fringes with the interplanar distances of 2.03 Å, corresponding to the (211) planes (Fig. 1e). The Ni2W4C phase exhibits the lattice fringes of (620) planes with interplanar distances of 1.76 Å (Fig. 1f). The line scan STEM-EDX spectra of a single Ni2W4C-W3C nanoparticle further demonstrates that the Ni element only exists at one side of the Janus nanoparticle while the C and W elements exist throughout the Ni2W4C-W3C nanoparticle (Fig. 1g). The results confirmed the successful formation of Janus-like Ni2W4C-W3C heterostructures with abundant interfaces.

Fig. 1. (Color online) FE-SEM image (a) and HAADF-STEM image (b) of the Ni2W4C-W3C/CNFs. STEM-EDX element mapping (c) and HRTEM images (d) of a Ni2W4C-W3C nanoparticle. HRTEM images of the W3C phase (e) and Ni2W4C phase (f), scale bar is 2 nm. (g) Line-scan EDX spectra of the Ni2W4C-W3C nanoparticle. Inset in Fig. 1g is the corresponding HAADF-STEM image. (h) In situ temperature XRD patterns of Ni2W4C-W3C/CNFs. (i) The enlarged range of 2𝜃 from 42° to 48°. (j) Three main peaks of Ni2W4C-W3C/CNFs in XRD patterns obtained at different temperatures.

The in situ temperature XRD is conducted from 50 to 1000 oC with heating rate of 5 oC min–1 to investigate the formation process of the Ni2W4C-W3C Janus structures. As shown in Fig. 1h and i, three peaks locate at 38.6°, 44.7°, 44.8° could be attributed to (200), (211) and (206) plane of WCx (PDF 35-0776). In the heating process, these peaks have tendency of blue shifts, suggesting that the Ni atoms are partially inserted into the crystal lattice of WCx. The interplanar spacings of WCx crystals enlarge during the Ni inserting process, resulting in the decrease of 2𝜃. While in the cooling process from 1000 to 50 oC, these peaks exhibited red shifts due to the formation and perfection of Ni2W4C crystals (Fig. 1j). The XRD pattern of Ni2W4C-W3C (50 oC) exhibited several new peaks

5

which are quite different with the WCx, suggesting that there are new phases formed during the graphitization process in carbon nanofibers. The in situ temperature XRD results strongly revealed the in situ formation process and mechanism of Ni2W4C by inserting the Ni atoms into the crystal lattice of WCx. From the crystal structure of Ni2W4C in Fig. 2a, the density functional theory (DFT) calculation results indicate that with the inserting of Ni into the W 3C unit cell, the Ni2W4C structure generates large amount of Ni–W metallic bonds, which could be the active sites for overall water splitting. As shown in the XRD pattern (Fig. 2b) of the Ni2W4C-W3C/CNFs (50 oC), a broad peak in 22o is ascribed to (002) plane of amorphous C in CNFs [3]. The diffraction peaks located at 41.6°, 51.3°, 71.0° and 72.7° can be well indexed to the (511), (620), (822) and (751) planes of Ni 2W4C phase (JCPDS No. 20-0796). The peaks at 35.6°, 40.0°, 44.1° and 75.3° are in according with the (200), (210), (211) and (400) planes of W 3C phase (JCPDS No. 42-0853). The XRD results indicate that the cubic Ni2W4C and cubic W3C phases both exist, verifying the unique Ni2W4C-W3C Janus structures. Based on the in situ XRD results, the WCx crystals form in CNFs at initial stage (50–400 o

C), and the Ni atoms begin to insert into the crystal of WCx at high temperature stage (500–1000

o

C). After the cooling stage to room temperatures (1000–50 oC), the formed Ni2W4C grown up to

better crystals with the WCx phases transform into W3C phases. As control, the individual W3C/CNFs and Ni3C/CNFs are also prepared at the same synthesis procedure. As shown in Figs. S2 and S3 (online), the XRD patterns of W3C/CNFs and Ni3C/CNFs exhibit the characteristic peaks of pure phase W3C and Ni3C, respectively. Therefore, when the Ni atoms meet W atoms, it will form the unique Ni2W4C-W3C Janus structures.

Fig. 2. (a) Crystal model of the Ni2W4C. Blue: Ni; cyan: W; white: C. (b) XRD pattern of Ni2W4CW3C/CNFs. W 4f (c) and Ni 2p (d) XPS spectra of and Ni2W4C-W3C/CNFs, W3C/CNFs and Ni3C/CNFs.

The XPS is used to investigate the surface electronic state, chemical composition and electron interactions of Ni2W4C-W3C/CNFs. XPS spectra of individual W3C/CNFs and Ni3C/CNFs are also investigated as control. The N 1s region, O 1s region, XPS surveys of W3C/CNFs, Ni3C/CNFs and Ni2W4C-W3C/CNFs are shown in Figs. S4 and S5 (online). As shown in Fig. 2c, the individual W3C/CNFs indicates two spin-orbit doublets with binding energies (BEs) at 32.1, 32.6, 34.2, 35.6 and 37.8 eV, respectively [19]. The spin-orbit doublets with BEs at 32.1 and 34.2 eV correspond to the W 4f7/2 and W 4f5/2 for the attribution of metallic W0. The other spin-orbit doublets at 32.6 and 35.6 eV could be assigned to the W 4f7/2 and W 4f5/2 orbits of W–C bonds. The peak at 37.8 eV could be ascribed to some WOx species due to the exposure in the air [20]. Focusing on the highresolution W 4f spectrum of Ni2W4C-W3C/CNFs, the spin-orbit doublets shift to the lower BEs at 31.0 and 33.2 eV and these correspond to the W 4f7/2 and W 4f5/2 of metallic Ni–W bond in Ni2W4C crystals, suggesting strong electron interaction between the Ni and W atoms. The spin-orbit doublets with BEs at 31.4 and 34.5 eV could be assigned to the W 4f orbits of W–C bonds in W3C crystals.

6

Fig. 2d shows the high-resolution Ni 2p XPS spectra of Ni 2 W4 C-W3 C/CNFs and individual Ni3C/CNFs. In Ni 2p XPS spectra of Ni 3C/CNFs, the characteristic peaks located at 871.8 and 853.9 eV correspond to the Ni 2p 1/2 and Ni 2p3/2 of metallic Ni in Ni3C/CNFs and the BEs of two satellite peaks are located at 880.9 and 862.1 eV. Due to the formation of Ni2 W4C-W3 C Janus structures, the corresponding peaks of Ni 2p 1/2 and Ni 2p3/2 shift to the higher BEs of 874.1 and 856.3 eV, indicating the strong electron transfer from Ni to W atoms. XPS results confirm the strong electron interaction between the Ni and W atoms due to the formation of Janus structures. The inserting of Ni atoms into the W 3 C crystals could enhance the local electronic state of W atoms, making the inert W atoms in W3C be active sites for the water splitting. The control of Ni2 W4C-W3 C Janus structures is the key factor for the rational design of advanced electrocatalysts for electrocatalytic water splitting. Through changing the mole ratio of Ni and W precursor, the structures of binary nickel tungsten carbide evolve from mixed Ni3 C and W3C to Ni2 W4C-W3 C Janus structures. Three different binary nickel tungsten carbide are prepared by adjusting the molar ratio of Ni and W precursor. The various catalysts are denoted as NiWC-1/CNFs (Ni2+/W6+=1/2), NiWC-2/CNFs (Ni2+/W6+=1/3) and NiWC-3/CNFs (Ni2+/W6+=1/4). Fig. 3 shows the morphologies and structures of NiWC1/CNFs and NiWC-3/CNFs and the morphologies of NiWC-2/CNFs are shown in Fig. 1. As shown in Fig. 3a and e, numerous small nanoparticles are evenly immobilized on the NiWC1/CNFs and NiWC-3/CNFs. Fig. 3b and f displays that the size of nanoparticles ranges from 10 to 80 nm.

Fig. 3. (Color online) FE-SEM image (a), TEM image (b), XRD pattern (c) and STEM-EDX mapping images (d) of NiWC-1/CNFs. FE-SEM image (e), TEM image (f), XRD pattern (g) and STEM-EDX mapping images (h) of NiWC-3/CNFs.

The morphology of NiWC-1/CNFs and NiWC-3/CNFs are similar with NiWC-2/CNFs. However, the XRD results (Fig. 3c and g) show that the main phases have changed to Ni 3C, NiCx, W3C and WCx with the reduced contents of Ni 2W4 C in NiWC-1/CNFs and NiWC3/CNFs, respectively. Figure S6 (online) shows the W 4f and Ni 2p XPS spectra of the NiWC-1/CNFs and NiWC-3/CNFs. The Ni concentration changes from NiWC-1 to NiWC3 would lead to the positive shifts of BEs from 353.2 to 355.3 eV, respectively. Fig. 3d and h show the element distribution of NiWC-1/CNFs and NiWC-3/CNFs in a signal nanoparticle, respectively. Compared with the NiWC-2/CNFs (Fig. 1d–g), the Ni and W distribute in the whole area with no Janus structure forming. The different mole ratios in precursor result in different morphology and structure, thus leading to variant electrocatalytic performance. The Ni/W mole ratio of 1/3 could lead to a better Janus Ni 2 W4C-W3 C without other NiCx and WCx phases. Electrocatalytic water splitting could be divided into two half reactions, HER and OER. The HER performance of NiWC-1/CNFs, NiWC-2/CNFs and NiWC-3/CNFs is conducted in alkaline electrolyze with typical three electrode systems. Fig. S7 (online) shows the

7

electrocatalytic activity of CNFs, indicating that the CNFs is inert to water splitting. As shown in Fig. 4a, the linear sweep voltammogram (LSV) measurement indicate that in alkaline condition, the NiWC-2/CNFs with Janus structures only require an overpotential of 63 mV to reach the current density of 10 mA cm –2. Meanwhile, the NiWC-1/CNFs and NiWC-3/CNFs with larger amounts of mixed Ni 3C and W3 C phases exhibit higher overpotentials of 92 and 126 mV to afford the current density of 10 mA cm –2. At low current density, the commercial Pt/C exhibits the lowest the overpotential of 28 mV (10 mA cm –2). However, when the current density is more than 88 mA cm –2, the commercial Pt/C needs a higher overpotential of 300 mV at 95 mA cm –2, which is higher than that of the prepared Ni2 W4C-W3 C Janus structures (285 mV). Fig. S8 (online) also demonstrates the HER performance of individual Ni3C/CNFs and W3 C/CNFs. The overpotentials of Ni 3C/CNFs and W3C/CNFs are 128 and 107 mV (10 mA cm –2), respectively. Notably, the Ni 2 W4C-W3C Janus structures display superior HER activity in alkaline condition than the mixed Ni 3C and W3 C phases, due to the emerging active sites of Ni-W bimetallic bond for HER [21].

Fig. 4. (Color online) Polarization curves (a), Tafel plots (b), and summarized overpotentials and Tafel slopes (c) at 10 mA cm−2 of NiWC-1/CNFs, NiWC-2/CNFs, NiWC-3/CNFs and commercial Pt/C electrodes for HER in 1.0 mol L–1 KOH solution without IR-correction (scan rate 2 mV s–1). Polarization curves (d), Tafel plots (e), and summarized overpotentials and Tafel slopes (f) at 10 mA cm−2 of NiWC-1/CNFs, NiWC-2/CNFs, NiWC-3/CNFs and commercial Pt/C electrodes for OER in 1.0 mol L –1 KOH solution without IR-correction (scan rate 2 mV s–1). The double-layer capacitance (Cdl) (g) and EIS Nyquist plots (h) of NiWC-1/CNFs, NiWC-2/CNFs and NiWC-3/CNFs. (i) Chronoamperometric responses (i–t) recorded from the NiWC-2/CNFs for HER and OER, respectively.

The Tafel slope is used to characterize the rate-determining step of hydrogen evolution. Electrocatalytic hydrogen generation in alkaline media could be explained by three steps. The first step is the adsorption of hydrogen intermediates (H ad) on the surface of catalysts (H2O(l) ⇄ Had + OH–(aq), Volmer step). Afterwards, by a chemical recoupling of two H ad atoms (2Had⇄ H2(g); the Tafel step) or a second electron transfer (H 2O(l) +Had + e− ⇄ H2(g) + OH−(aq), the Heyrovsky step), molecular hydrogen is formed. As shown in the Figs. 4b and S9 (online), the NiWC-2/CNFs electrode displays a Tafel slope (derived from the current density of 10 mA cm –2) of 112 mV dec–1, which is more superior to that of NiWC-1/CNFs (158 mV dec–1), NiWC-3/CNFs (256 mV dec–1), W3 C (437 mV dec–1) and Ni3C (422 mV dec–1). It is demonstrated that the reaction kinetic on the NiWC-2/CNFs electrocatalyst is a coupled Volmer-Tafel mechanism, while the NiWC-1/CNFs and NiWC-3/CNFs follow the Volmer–Heyrovsky mechanism, implying the remarkable intrinsic HER activity of NiWC 2/CNFs with Ni2 W4 C-W3 C Janus structures. Fig. 4c summarizes the Tafel slopes and overpotentials of various electrocatalysts. The binary nickel tungsten carbide exhibits strong structure-dependent activity and the Ni 2 W4C-W3 C Janus structures show the best electrocatalytic HER performance in alkaline media.

8

OER involves a four-electron pathway and exhibits slow dynamic process, which is the bottleneck of overall water splitting [22]. The OER performance of the various electrocatalysts are shown in Figs. 4d–f and S10, S11 (online). Due to the high redox peak appearing in some electrocatalysts, many researchers also used the overpotentials at 20, 50 or 100 mA cm–2 to evaluate the performance [23]. As shown in Fig. 4d, the polarization curve of NiWC-2/CNFs exhibits superior electrocatalytic performance with lowest overpotentials of 0 mV at 30 mA cm–2. These values are much smaller than NiWC-1/CNFs (310 mV), NiWC-3/CNFs (380 mV) and commercial IrO 2 (310 mV). At high current density of 50 and 100 mA cm–2, the NiWC-2/CNFs also exhibits the lowest overpotentials of 350 and 410 mV, which are much smaller than NiWC-1/CNFs (390 and 450 mV) and NiWC-3/CNFs (410 and 480 mV), and is very closer to that of commercial IrO 2 (340 and 400mV). In addition, the NiWC-2/CNFs electrode indicates small Tafel slope of 52 mV dec –1, which is smaller than the commercial IrO2 (74 mV dec–1) (Fig. 4e). Similar to HER activity as expected, the NiWC2/CNFs electrode with Ni 2W4 C-W3 C Janus structures obtains the enhanced OER activity when compared with other catalysts and commercial IrO 2/C (Fig. 4f). HER and OER overpotentials of Ni2W4C-W3C Janus structure have been tested in 0.5 mol L–1 H2SO4 and 0.1 mol L–1 PBS solution. As shown in Fig. S12 (online), for HER, the Ni2W4C-W3C/CNFs requires overpotentials of 82 and 177 mV to reach 10 mA cm–2 in acidic and neutral media, respectively. For OER, the Ni2W4C-W3C/CNFs requires overpotential of 280 and 490 mV to reach 10 mA cm–2 in acidic and neutral media. Notably, the Ni2 W4C-W3 C/CNFs electrode shows great activity in acidic and netural media. In order to investigate the electrochemical active surface area (ECSA), the double-layer capacitance (Cdl) is calculated via cyclic voltammetry (CV) tests at various scan rates ranging from 20 to 100 mV s–1 [24]. As illustrated in Figs. 4g and S13–S15 (online), the Cdl of NiWC2/CNFs is 160 mF cm–2, which is much higher than these of NiWC-3/CNFs (152 mF cm–2), NiWC-1/CNFs (143 mF cm–2) and W3C (Fig. S16 (online), 87.0 mF cm–2). The higher electrochemical active surface area indicates that the Ni 2W4C-W3 C Janus structures exhibit abundant active sites [25]. The electrochemical impedance spectroscopy (EIS) spectras of as-prepared electrocatalysts are shown in Fig. 4h. The charge transfer resistance (Rct) reflected the electrochemical kinetics could be acquired through the semicircle in the high frequency range. According to the equivalent circuit (Fig. S17 online), the semicircle Nyquist plots in Figs. 4h and S18 (online) illustrate that the charge transfer process is a control step of electrode process. The NiWC-2/CNFs electrode exhibits smallest Rct of 35.8 Ω at the potential of 50 mV vs. RHE when compared with the NiWC-1/CNFs (91.6 Ω) and NiWC3/CNFs (251.6 Ω). Significantly, the electrocatalytic durability is another crucial factor for the industrial application. As shown in Fig. 4i, the long-term cycling performance of NiWC2/CNFs is evaluated via chronopotentiometry for 12 h. The current density of NiWC-2/CNFs under continuous OER process keep stable around 30 mA cm -2 while the NiWC-2/CNFs used as HER electrode also exhibit slightly degradation in the current density of about 15 mA cm–2. The stability tests demonstrate that the Ni 2 W4C-W3 C/CNFs electrocatalysts both display superior OER and HER stability in alkaline solution. XPS spectra is utilized to

9

indicate the chemical state change of NiWC-2/CNFs after HER and OER (Fig. S19 online). Fig. S19 (online) shows the W 4f, Ni 2p and O 1s XPS spectra of Ni2 W4C-W3 C/CNFs after HER and OER stability test for 12 h. For W 4f region, the binding energies (BEs) of W–O, W–C and metallic W did not change when compared with the original Ni 2 W4C-W3 C/CNFs (Fig. 2d). As for Ni 2p region, the peaks at 874.1 and 856.3 eV could be ascribed to the Ni 2p1/2 and 2p3/2, also demonstrating the negligible changes after HER and OER tests. Fig. S18c (online) shows the O 1s region of NiWC-2/CNFs. The peaks at 533.3 and 531.7 eV correspond to C=O and C–O bond and no metal–O bond emerges, indicating the excellent stability of Ni2 W4C-W3 C. Authors also provide the ECSA results after OER test (Fig. S20 online). The Cdl of Ni2 W4 C-W3 C/CNFs is 159 mF cm–2 after OER test, which is similar with the original Ni2 W4C-W3 C/CNFs (160 mF cm–2), demonstrating that there are no additional active sites formed during the OER test. The intrinsic HER and OER activity based on the ECSA were also investigated. The ECSA of the NiW-1/CNFs, NiW-2/CNFs and NiW-3/CNFs are 2383, 2667 and 2533 cm 2, respectively. As shown in Figure S21 (online), the NiW-2/CNFs exhibits the lowest overpotentials of 164 mV at 0.01 mA cm -2 for HER and 470 mV at 0.05 mA cm -2 for OER than those of the NiW-1/CNFs (193 mV for HER and 520 mV for OER) and NiW-3/CNFs (236 mV for HER and 590 mV for OER), suggesting the excellent intrinsic HER and OER activity. The geometric and intrinsic activity of samples for HER and OER exhibit the same trends. Encouraged by the excellent HER and OER performance in the same alkaline media, the NiWC-2/CNFs served as self-supported electrode for overall water splitting is investigated (Fig. 5a). Fig. 5b shows the polarization curve of NiWC-2/CNFs|| NiWC-2/CNFs cell in 1.0 mol L–1 KOH solution. The measured overpotentials are only 270 mV for HER and 450 mV for OER at the current density of 100 mA cm –2, which is much lower than these reported electrocatalysts. The overall water splitting results demonstrate that the NiWC-2/CNFs electrodes possess superior electrochemical activities towards HER and OER in 1.0 mol L–1 KOH. Fig. 5c describes the polarization curves of NiWC-1/CNFs, NiWC-2/CNFs and NiWC-3/CNFs for overall water splitting in 1.0 mol L–1 KOH at the scan rate of 2 mV s–1. The NiWC-2/CNFs displays the best electrocatalytic for overall water splitting by obtaining a cell voltage of 1.55 V (10 mA cm –2), which is much lower than those of NiWC-1/CNFs (1.72 V) and NiWC-3/CNFs (1.71 V) under the same condition. As shown in Fig. S22 (online), with a cell voltage of 2.0 V, the NiWC-2/CNFs obtains the current density of 200 mA cm–2 in 1 mol L–1 KOH, and it is 4.3 and 4.2 times higher than those of NiWC-1/CNFs (48 mA cm–2) and NiWC-3/CNFs (46 mA cm–2), respectively. To examine the durability of this cell device, the CV measurement and chronopotentiometry for overall water splitting are investigated. As shown in Fig. 5d and e, the polarization curve of NiWC-2/CNFs before and after 5000 cycles only indicate slightly changes in the current density, suggesting the excellent durability of NiWC-2/CNFs for overall water splitting. Moreover, the obvious hydrogen and oxygen bubbles could be seen in specific photograph (Fig. 5f) of NiWC2/CNFs during the continuous overall water splitting test. Compared with the recently

10

reported new electrocatalysts for overall water splitting, the Ni 2 W4 C-W3 C/CNFs obtains a relatively small overpotential at 10 mA cm –2 comparing to reported electrocatalysts. (Fig. 5g) [26–48].

Fig. 5. (Color online) (a) The schematic shows the home-built electrolyser using Ni 2 W4CW3 C/CNFs as both anode and cathode. (b) Polarization curves for overall water splitting with the NiWC-2/CNFs electrode as both anode and cathode at a scan rate of 2 mV s −1. (c) LSV curves of water electrolysis for NiWC-1/CNFs, NiWC-2/CNFs, NiWC-3/CNFs in a twoelectrode configuration with a scan rate of 2 mV s −1 without iR correction. (d) LSV curves of the NiWC-2/CNFs after 5000 cycles. (e) Stability test of NiWC-2/CNFs in the potential of 1.72 V vs. RHE. (f) Optical photograph showing the generation of H 2 and O2 bubbles on the electrodes. (g) Comparison of the overpotentials required at 10 mA cm−2 among our catalyst and available reported electrocatalysts [26–48].

DFT calculation is applied to understand the intrinsic influence of Ni 2 W4C-W3 C Janus structures on the HER activity. Firstly, surface possibilities of W 3 C and Ni2 W4 C are examined. There are three configurations of surfaces for W 3C, W3C (Fig. S23 online), W3 (Fig. S24 online), and W3C3 (Fig. S25 online). Based on the structural symmetric, we chose the [222] plane as the basic model. There are three for [222] plane, and the optimized structures and the corresponding free energies are shown in Fig. S26 and Table S1 (online), respectively. The configuration of W3C@W3 (layer 2) with three W atoms outside of the layer is the most energetically stable. For Ni 2 W4C, there are also three kinds of surfaces, Ni 3 W3C (Fig. 6a), Ni3 W3 (Fig. 6b), and W3C3 (Fig. 6c). However, the layer 3(Ni 2 W4 C@W3 C3) with three tungsten and three carbon atoms outside the layer is the most possible one (Fig. S27 online), which means the Ni2 W4C@Ni3 W3 surface cannot be formed spontaneously. Thus, the formation of W3C@W3 makes it possible that doping Ni atoms grow along the surface of W3 C@W3 to form Ni2 W4C@Ni3 W3 with Ni–W metallic bonding all over it. Thus, the effective area of the active surface has doubled. The density of states for the Ni 2 W4C@Ni3 W3 surfaces are also calculated, as shown in Figs. 6d and S28 (online). The main electron density near the Fermi level comes from the W, which means the W is still the active point after Ni inserting. This is in good agreement with the XPS examination.

Fig. 6. (a–c) Three configurations of surfaces for Ni 2W4 C. (d) The density of state (DOS) plot of W atoms in W3C and Ni2 W4C-W3 C. (e) HER free-energy diagram for [222] plane of W3 C and Ni2 W4C and [111] plane of Pt. (f) Electron density differences between H and NiW alloys. (g) Raman spectra of pure Ni 2 W4C-W3 C (blank) and Ni2 W4C-W3 C put in the in situ electrochemical-Raman cell in 1.0 mol L–1 KOH solution (red). In situ electrochemicalRaman spectra during OER (h) and HER (i) process.

11

The HER activity is based on the H trapping and H 2 desorption, which requires a Gibbs free energy (∆GH* based on Pt [111]) to be zero. Next, we examined the H binding energy on the surfaces of W3C@W3, Ni2 W4C@W3 C3 and Pt [111]. The adsorption energy is derived according to Eq. (1). 1

∆𝐸H = (𝐸(surf + 𝑛H) − 𝐸(surf) − 𝑛𝐸(H)) , 𝑛

(2)

where n is the number of adsorbed hydrogen atoms, E(surf) is the total energy of the surface without H adsorption, and E(surf+nH) is the total energy of the surface with n adsorbed atoms, and the results are shown in Fig. 6e. H binding energy of W3C@W3 surface is nearly –0.92 eV, which is 0.42 eV larger than that of Pt [111] surface. The H binding energy of Ni2 W4C@Ni3 W3 is only 0.15 eV higher than Pt [111], indicating a high reactivity for HER performance. In the Janus structures, the Ni 2 W4C provides active sites for cleaving H−OH bonds, and the W3 C facilitates the combination of H ads intermediates into H2 molecules. The formation of Ni2 W4C is a secondary growth of Ni on W 3 C, where Ni3 alloys are formed between W3 substitute. A d band center calculation was performed, where d band centers of W of W3C and Ni2 W4 C are both –2.98 eV, indicating that there is few electron transfers between Ni and W. A separate charge density difference (Δρ(r)) method is used to identify the charge transfer between H and Ni-W alloys, as shown in Fig. 6f, where yellow density means positively charged and blue density means negatively charged. The charge transfer between the Ni-W makes Ni with low charge density, leading to a low binding energy of H ions. The real active sites and structural changes of the Ni 2W4 C-W3 C during HER and OER process are monitored by using in situ electrochemical-Raman spectroscopy. The electrochemical-Raman spectras are conducted in 1.0 mol L–1 KOH with the sample acting as working electrode, Ag/AgCl acting as reference electrode and Pt wire acting as count electrode. As shown in the Fig. 6g, the Raman peaks located at 148, 215, 703, 806, 886, 966 cm–1 could be attributed to Ni 2 W4C-W3 C while the peaks at 1340 and 1572 cm –1 are ascribed to the D band and G band of CNFs (blank line). However, when the sample is put in the electrolyte solution (red line), many new peaks aroused at 337, 409, 495, 608, 733, 843, 909, 1058 cm–1, respectively. These peaks changes suggested the formation of new chemical bonds among the Ni2 W4 C-W3 C, hydroxyl, protons and water molecules in 1.0 mol L–1 KOH solution, indicating the strong absorption ability for ions and molecules. For the OER process (Fig. 6h), when applying bias potential of 0.5 V, the emerged peaks at 337 and 909 cm –1 correspond to the νW-OH and νW-OOH bands of Ni2 W4C-W3C [49, 50]. With the increased potentials in the range of 0.5–1.1 V vs SHE, the intensities of these bands of Ni 2 W4C-W3C enhanced gradually, probably due to the fact that OH - in alkaline media are driven to adsorb on the W atoms and the formation of the OOH* intermediates. For the HER process (Fig. 6i), with the increased potentials in the range of –1.2– –1.5 V vs SHE, there is no νW-OOH bands at 337 cm–1 can be observed. At high potential of –1.5 V, there is a very slight peak for νW-OH at 909 cm–1 emerged, suggesting the adsorbed water molecules dissociate into H ads species and OH− ions during cathodic polarization. Designing the Ni 2 W4C-W3 C Janus

12

structures through the in situ interfacial engineering could generate the Ni–W metallic bond all over the surface of Ni 2W4 C, leading to the doubled effective area of the active surface. The strong interaction between Ni and W increases the local electronic state of W atoms, decreasing the hydrogen adsorption energy for protons on W atoms and increasing the intrinsic activity of catalysts. 4.

Conclusions

In summary, the unique Ni 2 W4C-W3 C Janus structures with abundant interfaces and superficial Ni–W metallic bonds have been demonstrated through the in situ interfacial engineering by combining the electrospinning technology and graphitization process. The in situ temperature XRD results strongly reveal the in situ formation process of Ni 2 W4 C by inserting the Ni atoms into the crystal lattice of WCx. The Ni2 W4C-W3 C/CNFs showed outstanding electrocatalytic performance, requiring overpotentials of 63 mV to reach 10 mA cm–2 for HER and 270 mV to reach 30 mA cm –2 for OER in alkaline, respectively. The Ni2 W4C-W3 C/CNFs steadily drives alkaline electrolyzer at cell voltage of 1.55 V (10 mA cm−2) and 1.87 V (100 mA cm–2), which is much lower than the values for Pt/C–IrO2/C couple and the reported catalysts in literatures. DFT results indicat that the generated Ni-W metallic bond all over the surface of Ni 2 W4C leads to the doubled increased effective area of the active surface. The strong interaction between Ni and W increases the local electronic state of W atoms, decreasing the hydrogen adsorption energy for protons on W atoms and increasing the intrinsic activity of catalysts. The in situ electrochemical-Raman results demonstrate the strong absorption ability for hydroxyl, protons and water molecules and the in situ Raman spectra for HER and OER process further demonstrate that W atoms are the real active sites. Conflict of interest The authors declare that they have no conflict of interest.

Acknowledgements This work was supported by the National Natural Science Foundation of China (51803077, 51872204),

the

National

Key

Research

and

Development

Program

of

China

(2017YFA0204600), the Natural Science Foundation of Jiangsu Province (BK20180627), Postdoctoral Science Foundation of China (2018M630517, 2019T120389), the Ministry of Education (MOE) and the State Administration for Foreign Expert Affairs (SAFEA), 111 Project (B13025), the national first-class discipline program of Light Industry Technology and Engineering (LITE2018-19), and the Fundamental Research Funds for the Central Universities. Author contributions Han Zhu, Mingliang Du and Yang Chai supervised the research. Han Zhu and Songge Zhang designed and engineered this work. Songge Zhang performed the experiments with support

13

from Lejuan Cai, Xiaodi Jiang, Shuanglong Lu, Fang Duan and Weifu Dong. Guohua Gao carried out the DFT calculations. Songge Zhang wrote the manuscript and Guohua Gao, Han Zhu, Mingliang Du and Yang Chai revised the manuscript. All authors contributed to the general discussion.

References [1] Yamaguchi A, Inuzuka R, Takashima T, et al. Regulating proton-coupled electron transfer for efficient water splitting by manganese oxides at neutral pH. Nat Commun 2014; 5: 4256 [2] Zhang B, Zheng X, Voznyy O, et al. Homogeneously dispersed multimetal oxygenevolving catalysts. Science 2016; 352: 333-337. [3] Zhu H, Zhang J, Yanzhang R, et al. When cubic cobalt sulfide meets layered molybdenum disulfide: A core–shell system toward synergetic electrocatalytic water splitting. Adv Mater 2015; 27: 4752-4759. [4] Wu H, Lu X, Zheng G, et al. Topotactic engineering of ultrathin 2D nonlayered nickel selenides for full water electrolysis. Adv. Energy Mater. 2018; 8: 1702704. [5] You B, Liu X, Liu X, et al. Efficient H2 evolution coupled with oxidative refining of alcohols via a hierarchically porous nickel bifunctional electrocatalyst. ACS Catal, 2017; 7: 4564-4570. [6] Yo B, Sun Y. Hierarchically porous nickel sulfide multifunctional superstructures. Adv Energy Mater 2016; 6: 1502333. [7] Chen G, Wang T, Zhang J, et al. Accelerated hydrogen evolution kinetics on NiFe-layered double hydroxide electrocatalysts by tailoring water dissociation active sites. Adv Mater 2018; 30: 1706279. [8] Yilmaz G, Yang T, Du Y, et al. Stimulated electrocatalytic hydrogen evolution activity of MOF-derived

MoS2

basal

domains

via

charge

injection

through

surface

functionalization and heteroatom doping. Adv Sci 2019; 6: 1900140. [9] Lim KJH, Yilmaz G, Lim YF, et al. Multi-compositional hierarchical nanostructured Ni3S2 @MoSx/NiO electrodes for enhanced electrocatalytic hydrogen generation and energy storage. J Mater Chem A 2018; 6: 20491-20499. [10]

Fan X, Peng Z, Ye R, et al. M3C (M: Fe, Co, Ni) nanocrystals encased in graphene nanoribbons: an active and stable bifunctional electrocatalyst for oxygen reduction and hydrogen evolution reactions. ACS Nano 2015; 9: 7407-7418.

[11]

Jin H, Chen J, Mao S, et al. Transition metal induced the contraction of tungsten carbide lattice as superior hydrogen evolution reaction catalyst. ACS Appl Mater Interfaces 2018; 10: 22094-22101.

[12]

Meng FL, Yilmaz G, Ding TP, et al. A hybrid solar absorber-electrocatalytic N-doped carbon/alloy/semiconductor electrode for localized photothermic electrocatalysis. Adv Mater 2019; 31: 1903605.

14

[13]

Xu B, Chen Z, Yang X, et al. Electronic modulation of carbon-encapsulated NiSe composites via Fe doping for synergistic oxygen evolution. Chem Commun 2018; 54: 9075-9078.

[14]

Xiong K, Li L, Zhang L, et al. Ni-doped Mo2C nanowires supported on Ni foam as a binder-free electrode for enhancing the hydrogen evolution performance. J Mater Chem A 2015; 3: 1863-1867.

[15]

Zhou Y, Sun S, Song J, et al. Enlarged Co-O covalency in octahedral sites leading to highly efficient spinel oxides for oxygen evolution reaction. Adv Mater 2018; 30: 1802912.

[16]

Wang D, Li Q, Han C, et al. When NiO@Ni meets WS2 nanosheet array: a highly efficient and ultrastable electrocatalyst for overall water splitting. ACS Cent Sci 2017; 4: 112-119.

[17]

Yilmaz G, Peh SB, Zhao D, et al. Atomic- and molecular-level design of functional metal–organic frameworks (MOFs) and derivatives for energy and environmental applications. Adv Sci 2019; 6; 1901129.

[18]

Wang Y, Wang S, Lou X. Dispersed nickel cobalt oxyphosphide nanoparticles confined in multichannel hollow carbon fibers for photocatalytic CO2 reduction. Angew Chem Int Ed 2018; 48: 17236-17240.

[19]

Chen J D, Yu D N, Liao W S, et al. WO3–x nanoplates grown on carbon nanofibers for an efficient electrocatalytic hydrogen evolution reaction. ACS Appl Mater Interfaces 2016; 8: 18132-18139.

[20]

Li Y, Zhang H, Jiang M, et al. 3D self-supported Fe-doped Ni2P nanosheet arrays as bifunctional catalysts for overall water splitting. Adv Funct Mater 2017; 27: 1702513.

[21]

Xue Z H, Su H, Yu Q Y, et al. Janus Co/CoP nanoparticles as efficient Mott–Schottky electrocatalysts for overall water splitting in wide pH range. Adv Energy Mater 2017; 7: 1602355.

[22]

You, B, Zhang, Y, Yin, P et al. Universal molecular-confined synthesis of interconnected porous metal oxides-NC frameworks for electrocatalytic water splitting. Nano Energy 2018; 48: 600-606.

[23]

Li J, Xu W, Luo J, et al. Synthesis of 3D hexagram-like cobalt–manganese sulfides nanosheets grown on nickel foam: a bifunctional electrocatalyst for overall water splitting. Nano-Micro Lett 2018; 10: 6.

[24]

You B, Jiang N, Sheng M, et al. Hierarchically porous urchin-like Ni2P superstructures supported on nickel foam as efficient bifunctional electrocatalysts for overall water splitting. ACS Catal 2015; 6: 714-721.

[25]

Liu, X, Dong, J, You, B et al. Competent overall water-splitting electrocatalysts derived from ZIF-67 grown on carbon cloth. RSC Adv 2016; 6: 73336-73342.

[26]

Sun Y, Xu K, Wei Z, et al. Strong electronic interaction in dual-cation-incorporated NiSe2 nanosheets with lattice distortion for highly efficient overall water splitting. Adv Mater 2018; 30: 1802121.

15

[27]

Wang Z, Liu H, Ge R, et al. Phosphorus-doped Co3O4 nanowire array: a highly efficient bifunctional electrocatalyst for overall water splitting. ACS Catal 2018; 8: 2236-2241.

[28]

You, B Sun, Y. Innovative strategies for electrocatalytic water splitting. Acc Chem Res 2018; 51: 1571-1580.

[29]

Zhang H, Li X, Hähnel A, et al. Bifunctional heterostructure assembly of NiFe LDH nanosheets on NiCoP nanowires for highly efficient and stable overall water splitting. Adv Funct Mater 2018; 28: 1706847.

[30]

Zhang G, Feng Y S, Lu W T, et al. Enhanced catalysis of electrochemical overall water splitting in alkaline media by Fe doping in Ni3S2 nanosheet arrays. ACS Catal 2018; 8: 5431-5441.

[31]

Pan Y, Sun K, Liu S, et al. Core–shell ZIF-8@ZIF-67-derived CoP nanoparticleembedded N-doped carbon nanotube hollow polyhedron for efficient overall water splitting. J Am Chem Soc 2018; 140: 2610-2618.

[32]

Fang L, Li W, Guan Y, et al. Tuning unique peapod-like Co(SxSe1–x)2 nanoparticles for efficient overall water splitting. Adv Funct Mater 2017; 27: 1701008.

[33]

Liang H, Gandi A N, Anjum D H, et al. Plasma-assisted synthesis of NiCoP for efficient overall water splitting. Nano Lett 2016; 16: 7718-7725.

[34]

Li H, Wen P, Li Q, et al. Earth-abundant iron diboride (FeB2) nanoparticles as highly active bifunctional electrocatalysts for overall water splitting. Adv Energy Mater 2017; 7: 1700513.

[35]

Duan J, Chen S, Zhao C. Ultrathin metal-organic framework array for efficient electrocatalytic water splitting. Nat Commun 2017; 8: 15341.

[36]

Yang L, Guo Z, Huang J, et al. Vertical growth of 2D amorphous FePO4 nanosheet on Ni foam: outer and inner structural design for superior water splitting. Adv Mater 2017; 29: 1704574.

[37]

Hou J, Sun Y, Wu Y, et al. Promoting active sites in core–shell nanowire array as Mott–Schottky electrocatalysts for efficient and stable overall water splitting. Adv Funct Mater 2018; 28: 1704447.

[38]

Tang T, Jiang W J, Niu S, et al. Kinetically controlled coprecipitation for general fast synthesis of sandwiched metal hydroxide nanosheets/graphene composites toward efficient water splitting. Adv Funct Mater 2018; 28: 1704594.

[39]

Zou X, Wu Y, Liu Y, et al. In situ generation of bifunctional, efficient Fe-based catalysts from mackinawite iron sulfide for water splitting. Chem 2018; 4: 1139-1152.

[40]

Li W, Gao X, Xiong D, et al. Hydrothermal synthesis of monolithic Co3Se4 nanowire electrodes for oxygen evolution and overall water splitting with high efficiency and extraordinary catalytic stability. Adv Energy Mater 2017; 7: 1602579.

[41]

Yan L, Cao L, Dai P, et al. Metal-organic frameworks derived nanotube of nickelcobalt bimetal phosphides as highly efficient electrocatalysts for overall water splitting. Adv Funct Mater 2017; 27: 1703455.

16

[42]

Yang Y, Zhang K, Lin H, et al. MoS2–Ni3S2 heteronanorods as efficient and stable bifunctional electrocatalysts for overall water splitting. ACS Catal 2017; 7: 2357-2366.

[43]

Menezes P W, Indra A, Das C, et al. Uncovering the nature of active species of nickel phosphide catalysts in high-performance electrochemical overall water splitting. ACS Catal 2016; 7: 103-109.

[44]

Zhang Y, Xia X, Cao X, et al. Ultrafine metal nanoparticles/N-doped porous carbon hybrids coated on carbon fibers as flexible and binder-free water splitting catalysts. Adv Energy Mater 2017; 7: 1700220.

[45]

Du C, Yang L, Yang F, et al. Nest-like NiCoP for highly efficient overall water splitting. ACS Catal 2017; 7: 4131-4137

[46]

Huang S, Meng Y, He S, et al. N-, O-, and S-tridoped carbon-encapsulated Co9S8 nanomaterials: efficient bifunctional electrocatalysts for overall water splitting. Adv Funct Mater 2017; 27: 1606585.

[47]

Zhang B, Xiao C, Xie S, et al. Iron–nickel nitride nanostructures in situ grown on surface-redox-etching nickel foam: efficient and ultrasustainable electrocatalysts for overall water splitting. Chem Mater 2016; 28: 6934-3941.

[48]

Liu Y, Li Q, Si R, et al. Coupling sub-nanometric copper clusters with quasiamorphous cobalt sulfide yields efficient and robust electrocatalysts for water splitting reaction. Adv Mater 2017; 29: 1606200.

[49]

Daniel M, Desbat B, Lassegues J, et al. Infrared and Raman study of WO3 tungsten trioxides and WO3, xH2O tungsten trioxide tydrates. J Solid State Chem 1987; 67: 235247.

[50]

Jiang J, Sun F, Zhou S, et al. Atomic-level insight into super-efficient electrocatalytic oxygen evolution on iron and vanadium Co-doped nickel (oxy) hydroxide. Nat Commun 2018; 9: 2885.

Songge Zhang received his bachelor degree from Zhejiang SciTech University in 2018 under the supervision of Prof. Mingliang Du and Dr. Han Zhu. In 2019, he joined Prof. Mingliang Du’s lab in Jiangnan University as a research assistant. His research interests mainly focus on the design of high-entropy alloy nanoparticles and electrospun nanomaterials for electrocatalytic CO2 reduction.

Guohua Gao received his Ph.D. degree from Tongji University in 2010. He was promoted to Associate Professor of Physics in 2014 at Tongji University. His current research mainly engages in the study of effect of ionic diffusion, material lattice distortion, band

17

structure and phonon regulation on 2D and porous materials by theoretical and experimental methods for new energy applications.

Han Zhu received his Ph.D. degree from Zhejiang Sci-Tech University in 2016 under the supervision of Prof. Mingliang Du. In 2017, he joined the School of Chemical and Materials Engineering in Jiangnan University as an Associate Professor. His current research interests mainly focus on design and synthesis of electrospun nanofiber based novel nanostructured materials applied in electrocatalysis, heterogeneous catalysis, electrochemical sensors and environmental catalysis.

Yang Chai is an Associate Professor at the Hong Kong Polytechnic University. He is a member of Hong Kong Young Academy of Sciences, and the vice president of Physical Society of Hong Kong. His current research interest includes low-dimensional materials for electron devices and energy applications.

Mingliang Du received his Ph.D. degree in 2007 at South China University of Technology and he also worked there as a Postdoctoral Research Fellow. In 2009, he joined the Zhejiang Sci-Tech University in Hangzhou, as Associate Professor, and was promoted to Professor in 2014. In 2017, he joined the Department of Chemistry and Materials Engineering, Jiangnan University. His current research focuses on synthesis of nanomaterials and carbon materials and their applications in clean energy, electrochemical biosensing, etc.; electrospinning and its applications.

18

19

20

21

22

23

24

The Ni2W4C-W3C Janus structures with abundant superficial Ni-W metallic bonds have been demonstrated. The inserting of Ni atoms into the W3C crystals generate a new Ni2W4C phase, making the inert W atoms in W3C be active sites in Ni2W4C for the overall water splitting.

25