Contents lists available at ScienceDirect

Nano Energy

journal homepage: www.elsevier.com/locate/nanoen

Full paper

Crafting MoC2-doped bimetallic alloy nanoparticles encapsulated within N-doped graphene as roust bifunctional electrocatalysts for overall watersplitting

Qi Hua, Xiufang Liua, Bin Zhua, Liangdong Fana, Xiaoyan Chaia, Qianling Zhanga, Jianhong Liua,Chuanxin Hea,⁎, Zhiqun Linb,⁎

a College of Chemistry and Environmental Engineering, Shenzhen University, Shenzhen, Guangdong 518060, People's Republic of Chinab School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA

A R T I C L E I N F O

Keywords:Hydrogen evolution reactionOxygen evolution reactionPrussian blue analoguesN-doped grapheneNon-precious transition metal

A B S T R A C T

Despite recent vigorous progress in synthesis of monofunctional electrocatalysts for hydrogen evolution reaction(HER) or oxygen evolution reaction (OER), it remains challenging to develop bifunctional electrocatalysts forefficient overall water splitting. Herein, we report the crafting of MoC2-doped NiFe alloy nanoparticles (NPs)encapsulated within a-few-layer-thick N-doped graphene (denoted NG-NiFe@MoC2) via one-step calcination ofhybrid precursors containing polymer-encapsulating binary Prussian blue analogues NPs and Mo6+ cations. Theresulting NG-NiFe@MoC2 nanohybrids were exploited as electrocatalysts and exhibited excellent performanceon either HER or OER separately as a direct consequence of the synergistic effects of unique compositions (i.e.,MoC2 dopants and NiFe alloy NPs; both exerted profound influence on HER and OER) and advantageous ar-chitecture (i.e., a-few-layer-thick N-doped graphene encapsulating shell). Remarkably, an alkaline electrolytecapitalizing on NG-NiFe@MoC2 nanohybrids as bifunctional electrocatalysts achieved overall water-splitting(i.e., concurrent HER and OER) current density of 10mA cm−2 at a low potential of 1.53 V over a period of 10-hoperation, outperforming the precious Pt/C//RuO2 counterpart.

1. Introduction

Water splitting into hydrogen and oxygen is widely recognized as agreen and effective approach to transform and store renewable energy.As such, it has garnered increasing interest for rendering a green energyfuture. Unfortunately, two half reactions in water splitting, that is,anodic oxygen evolution reaction (OER) and cathodic hydrogen evo-lution reaction (HER), require electrocatalysts to reduce overpotentialsand achieve desirable reaction rate. [1] Recently, precious Pt-based andIr-, Ru-based catalysts have shown impressive performance for HER andOER, respectively; [2] however, their high cost and scarcity preventthem from further commercial applications. In this context, efficientelectrocatalysts based on earth-abundant elements, including transitionmetal (TM)-based materials (i.e., carbide, oxides, sulfides, phosphidesand hydroxides) [3–9] and heteroatoms (i.e., N, P and S)-doped carbonmaterials, have been developed for HER or OER [10]. Nevertheless,despite significant advances in synthesis of monofunctional electro-catalysts, effective methods to develop bifunctional electrocatalystspossessing both excellent HER and OER performances for overall water

splitting are comparatively few and limited in scope because of dif-ferent reaction centers required by HER and OER, respectively.

Recently, controlled incorporation of two different metal speciesinto one single nanostructure to accelerate the activation of reactantshas emerged as a powerful strategy for constructing superior electro-catalysts. [11] For example, the bimetallic NiMo catalyst with moderatemetal-hydrogen (M-H) strength displayed a higher HER activity thansingle Ni and Mo, because Mo alone has too strong M-H bindingstrength while Ni alone has too weak M-H binding strength. [12]However, the NiMo catalysts suffer from low activity for OER with anoverpotential of 368mV at 10mA cm−2. [13] On the other hand, it wasfound that incorporation of Fe to Ni-based catalysts could dramaticallyenhance the OER performance due to the stronger affinity of bimetallicNiFe to OH- and the intermediates of OER. [14] In this context, it isplausible that incorporation of Fe to bimetallic NiMo catalyst may im-prove the performance of OER, and the resulting new catalyst couldserve as a bifunctional electrocatalyst for overall water splitting.However, the investigation into the ternary metal electrocatalysts,particularly the relationship between the ternary compositions and

https://doi.org/10.1016/j.nanoen.2018.05.033Received 29 January 2018; Received in revised form 5 May 2018; Accepted 15 May 2018

⁎ Corresponding authors.E-mail addresses: hecx@szu.edu.cn (C. He), zhiqun.lin@mse.gatech.edu (Z. Lin).

Nano Energy 50 (2018) 212–219

Available online 19 May 20182211-2855/ © 2018 Elsevier Ltd. All rights reserved.

T

performance for overall water splitting, is still rare.Because of the low conductivity and stability of TM materials, en-

capsulating TM nanoparticles (NPs) into graphene layers is crucial fordeveloping robust TM-based electrocatalysts. Moreover, optimizing thecompositions (i.e., N-doping) and structure (i.e., thickness) of graphenelayers would further improve the electrocatalytic performance of gra-phene-encapsulated TM NPs. [15] For instance, Ni NPs encapsulated infew-layer N-doped graphene exhibited desirable HER and OER perfor-mance for overall water splitting. [16] On the other hand, Prussian blueanalogues (PBA), as a family of metal-organic frameworks (MOFs),contain transition metals as metallic nods and CN- as linkers, re-presenting attractive precursors in development of efficient electro-catalysts with well-controlled nanostructure. [17,18]

Herein, we report the crafting of MoC2-doped NiFe alloy NPs en-capsulated within a-few-layer-thick N-doped graphene (denoted NG-NiFe@MoC2, where the atomic ratio of Ni:Fe is 0.36:0.64, that is,Ni0.36Fe0.64) by one-step annealing of binary NiFe Prussian blue ana-logues (denoted NiFe-PBA) and Mo6+-grafted polyvinylpyrrolidone(PVP) hybrid precursors (denoted NiFe-PBA/PVP) at high temperature.The unique hybrid precursors composed of PVP-encapsulating NiFe-PBA NPs and grafted Mo6+ enabled the controlled and homogeneouscarbonization reaction, meanwhile the outer PVP overlayers avoidedthe aggregation of the resulting NiFe alloy NPs and MoC2 dopants.Consequently, the advantageous compositions (i.e., MoC2 dopants andNiFe NPs, which synergistically played key role in HER and OER) andunique architecture (i.e., a-few-layer-thick N-doped graphene-en-capsulating shell) rendered the optimized NG-NiFe@MoC2 nanohybridshighly active and stable electrocatalysts for either HER or OER in-dividually. More importantly, they could also be exploited as bothanodic and cathodic electrocatalysts for overall water splitting toachieve a current density of 10mA cm−2 at a voltage of 1.53 V withimpressive durability of 10 h. As such, they emerge as promising bi-functional electrocatalysts to substitute precious metals for efficientoverall water splitting.

2. Experimental section

2.1. Electrocatalysts synthesis

The NiFe-PBA/PVP hybrid precursors were synthesized by a mod-ified co-precipitation method. Solution A: 1.4928 g Ni(CH3COOH)2·4H2O and 12 g PVP were dissolved in 100ml alcohol toform a mixed solution. Solution B: 1.3270 g K3[Fe(CN)6] was dissolvedin 20ml deionized water to form another solution. Solution B wasadded into solution A dropwise under vigorous stirring at ambienttemperature. The resulting aqueous solution was aged for 2 h, and theprecipitate was collected by centrifugation without washing. The pre-cipitate denoted NiFe-PBA/PVP was dried at 80 °C for 24 h. Afterwards,Mo7O24.6NH4·4H2O (0.4 g) was dissolved in 20ml deionized water, towhich NiFe-PBA/PVP (.2 g) was added. The resulting solution was agedat ambient temperature for 6 h, and the precipitate was collected bycentrifugation and dried at 80 °C for 12 h. The obtained sample wascalcined at 800 °C under the atmosphere of Ar for 2 h, yielding NG-NiFe@MoC2 − 2, where 2 refers to as the input mass ratio ofMo7O24.6NH4·4H2O to NiFe-PBA/PVP. The other three samples, that is,NG-NiFe@MoC2 − 0, NiFe@MoC2 − 1 and NG-NiFe@MoC2–3 werealso synthesized through the same process as that for NG-NiFe@MoC2

− 2 except that the input mass ratios of Mo7O24.6NH4·4H2O to NiFe-PBA/PVP were 0:1, 1:1 and 3:1, respectively. For comparison, the NiFe-PBA without PVP overlayers was also prepared using the same proce-dure as NiFe-PBA/PVP but with no PVP added. Afterwards, the NiFe-PBA was calcined as that for NG-NiFe@MoC2 − 2, and the resultingsample denoted NiFe-PBA-C.

Moreover, the NG-MoC2 − 2 sample was prepared by selectivelyetching NiFe alloys away from the NG-NiFe@MoC2 − 2 using 1M hy-drochloric acid solution at 80 °C for 24 h. For the synthesis of NiFe@

MoC2 − 0 without N-doped graphene shell, the NiFe-PBA/PVP wascalcined in air at 500 °C for 6 h, and then reduced in hydrogen at 700 °Cfor 2 h.

2.2. Characterizations

The X-ray diffraction (XRD) measurements on various samples werecollected using a D8ADVANCE diffractometer with graphite-filteredCuKα source (λ=0.15418 nm) over a 2θ range of 5–70°. The elementanalysis was performed through inductively coupled plasma atomicemission spectroscopy (ICP-AES) on a Shimadzu ICPS-7500 spectro-meter. The samples for ICP-AES measurements were first calcined in airat 500 °C for 6 h to remove the encapsulating N-doped graphene, fol-lowed by the treatment with aqua regia at 100 °C for 6 h to ensure thesamples to be completely dissolved. Field emission scanning electronmicroscopy (FESEM) images were obtained on a JEOL JSM-7800F.Transmission electron microscopy (TEM) and high-resolution TEM(HRTEM) were measured by a JEOL 2100 electron microscope at anacceleration voltage of 200 kV. HAADF-STEM-EDS images were col-lected on a JEM-2100F instrument with an EDS system. X-ray photo-electron spectroscopy (XPS) analysis was performed using a Thermo VGESCALAB250 X-ray photoelectron spectrometer with AlKα X-ray ra-diation. The binding energies of different elements were correctedbased on the graphitic C 1 s peak at 284.6 eV.

2.3. Electrochemical characterizations

The electrochemical performance of as-synthesized electrocatalyststowards HER and OER was studied on an electrochemical workstation(CHI 760 E) in 1M KOH electrolyte. All electrochemical measurementswere conducted on a three-electrode system with a carbon electrode ascounter electrode, saturated calomel as reference electrode and elec-trocatalyst modified glassy carbon electrode (geometric surface area:0.07 cm2) as working electrode, respectively. The electrocatalyst inkswere prepared by dispersing as-synthesized electrocatalysts and Nafion(acting as adhesive) in absolute ethanol and sonicating for 0.5 h. Theresulting electrocatalyst inks were loaded on glassy carbon electrodewith the electrocatalyst loading of 0.2mg cm−2. The electrochemicalimpedance spectroscopy (EIS) was tested from 1MHz to 1 Hz at po-tentials of − 0.3 V for HER and 1.53 V for OER in 1M KOH solution.Cyclic voltammogram (CV) was performed on NG-NiFe@MoC2 − 2electrocatalyst at a scan rate of 100mV s−1 for 10,000 cycles in 1MKOH to investigate the cycling durability for HER and OER, respec-tively.

As for the two-electrode system in overall water splitting, the NG-NiFe@MoC2 − 2 electrocatalyst was dispersed in the mixed solution ofethanol and Nafion, the resulting solution was dropped on nickel foam(1×1 cm) with electrocatalysts loading of 1mg cm−2. The stability ofNG-NiFe@MoC2 − 2 for overall water splitting was investigated by thetime-dependent current density curves in 1M KOH. As for the pre-paration of precious Pt/C and RuO2 electrodes, the commercial Pt/C orRuO2 catalyst was dispersed in the mixed solution of ethanol andNafion, the resulting solution was dropped on nickel foam (1×1 cm)with catalysts loading of 1mg cm−2.

2.4. Electrical double layered capacitance measurement

It is well-known that the electrochemically active surface area(ECSA) can be measured by double layer capacitance (Cdl). The Cdl wasdetermined by the cyclic voltammograms (CVs) of various catalysts atdifferent scan rates (0.04, 0.06, 0.08, 0.10, 0.12 and 0.14 V/s) in a non-Faradic potential range of 1.23–1.33 V vs RHE. Afterwards, the corre-sponding capacitive current was plotted at 1.28 V at different scan ratesof CV, and the slope of the fitted line was the twice Cdl. The ECSA canthen be calculated from the following equation:

Q. Hu et al. Nano Energy 50 (2018) 212–219

213

=ECSA CC

dl

s

where the common specific capacitance of Cs = 0.04 mF/cm2 was

applied in our calculation. [19]

3. Results and discussion

The crafting of NG-NiFe@Mo nanohybrids is illustrated in Scheme1. First, spherical PVP-encasuplating NiFe-PBA NPs (NiFe-PBA/PVP)(1st panel in Scheme 1) were synthesized via a facile co-precipitationmethod. Then, the Mo6+ cations were grafted on as-prepared NiFe-PBA/PVP by coordination interaction between pyridyl groups of PVPand Mo6+ (2nd panel in Scheme 1). After calcination at 800 °C under Aratmosphere, the metal ions (Ni2+, Fe3+ and Mo6+) transformed intoMoC2-doped NiFe alloy NPs due to the inter-diffusion of those metalions at such a high temperature (3rd panel in Scheme 1). Meanwhile,both CN- linkers from PBA and PVP overlayers acted as carbon andnitrogen sources, forming in-situ a-few-layer-thick N-doped graphenesurrounding the alloy NPs. Moreover, the outer PVP layer could alsorestrain the aggregation of alloy NPs and MoC2 dopants, thus uniformlydoping MoC2 onto the alloy NiFe NPs. Subsequently, in order to opti-mize the structure and electrocatalytic performance, a series of NiFe

Scheme 1. One-step calcination of hybrid precursors containing binary NiFePrussian blue analogues (NiFe-PBA), polymers (polyvinylpyrrolidone; PVP) andMo6+ cations (second panel), yielding MoC2-doped NiFe alloy NPs en-capsulated within N-doped graphene (denoted NG-NiFe@MoC2; where theatomic ratio of Ni:Fe is 0.36:0.64 (i.e., Ni0.36Fe0.64) as revealed by XRD). ThePVP overlayers and CN- linkers from PBA were converted into N-doped gra-phene.

Fig. 1. SEM images of (a) NiFe-PBA/PVP hybrid precursors (prior to coordination with Mo6+ cations and calcination) and (b) NG-NiFe@MoC2 − 2 (after calcinationof Mo6+-grafted NiFe-PBA/PVP). (c) TEM and (d, e, f) HRTEM images of NG-NiFe@MoC2 − 2. (g) HADDF-STEM image of NG-NiFe@MoC2 − 2 with (h) EDXmapping and (i) EDX line mapping along the white line in image (g).

Q. Hu et al. Nano Energy 50 (2018) 212–219

214

alloy NPs encapsulated within N-doped graphene nanohybrids, here-after referred to as NG-NiFe@MoC2 − 0, NG-NiFe@MoC2 − 1, NG-NiFe@MoC2 − 2, and NG-NiFe@MoC2–3, respectively, where 0, 1, 2and 3 denote the mass ratios of introduced dopant precursor ofMo7O24.6NH4·4H2O to NiFe-PBA/PVP (see Experimental Section inSupporting Information), were prepared.

The XRD pattern of NiFe-PBA/PVP hybrid precursors displayed aseries of diffractions and can be well indexed into (200), (220), (400),(422), (440), (660) and (620) planes for PBAs (JCPDS no. 20–0915)(Supporting Information, Fig. S1). After calcination at 800 °C, the NG-NiFe@MoC2 − 0 sample (i.e., without the introduction of MoC2)showed diffractions at 43.9°, 51.1° and 75.1°, corresponding to (111),(200) and (220) planes for the Ni0.36Fe0.64 alloy phase at the Ni/Fe ratioof 0.36/0.64 (JCPDS no. 47–1405) (Fig. S2). Notably, the atomic ratiosof Ni/Fe in alloys for all the four calcined samples noted above wereapproximately 0.4/0.6 as measured by inductively coupled plasmaatomic emission spectroscopy (ICP-AES) (Table S1), correlating wellwith the XRD study (0.36/0.64). With the addition ofMo7O24.6NH4·4H2O, the intensity of diffractions for the alloy phaseobtained after calcinations was markedly reduced, and new diffractionpeaks that can be assigned to MoC2 (JCPDS no. 35–0787) were clearlyevident on NG-NiFe@MoC2 − 2 and NG-NiFe@MoC2–3 samples. Theseresults indicated that MoC2 species was integrated into NiFe (i.e.,Ni0.36Fe0.64) alloy phase, thus restraining the crystallization of the alloyNPs. The NG-MoC2 − 2 (i.e., selectively dissolving Ni0.36Fe0.64 alloys)and NiFe@MoC2 − 0 (i.e., without MoC2 dopants and N-doped gra-phene shell) and MoC2 samples were also synthesized and used ascontrol (see Experimental Section, Figs. S3-S4).

The representative SEM image of NiFe-PBA/PVP hybrid precursorsshowed spherical NiFe-PBA NPs with a diameter around 20 nm en-capsulated by several PVP layers (Fig. 1a). The encapsulating PVPoverlayers can also be seen in the TEM and HRTEM images of NiFe-PBA/PVP (Fig. S5). After grafting Mo6+ and calcination, the irregularparticle-like morphology was yielded (Fig. 1b), and the correspondingTEM image indicated that it was composed of small NPs with anaverage diameter of 30 ± 5 nm (Fig. 1c). A further scrutiny by high-resolution TEM revealed that most of NPs were encapsulated within3–11 layers of graphene shell (Fig. 1d-e). Fig. 1f displayed clear latticefringes with a spacing of 0.21, 0.23 and 0.34 nm, corresponding to(111) plane of Ni0.36Fe0.64 alloy, (101) plane of MoC2 and (002) planeof graphitic carbon, respectively. It is worth noting that the interfacebetween Ni0.36Fe0.64 alloy and MoC2 species was difficult to be identi-fied, signifying that MoC2 was in close contact with Ni0.36Fe0.64 alloy.Moreover, both EDS element mapping (Fig. 1h) and line curves of arandom NP (Fig. 1g) revealed that Ni, Fe and Mo elements were uni-formly distributed and intimately contacted with each other at nano-scopic level (Fig. 1h–i), suggesting that MoC2 was homogeneouslydoped onto Ni0.36Fe0.64 alloy. For comparison, the morphology of NiFe-PBA (control sample; see Experimental Section) without the addition ofPVP and the resulting calcined product (i.e., NiFe-PBA-C) were alsoinvestigated by SEM (Fig. S6). The NiFe-PBA displayed uniform PBANPs with an average size of 20 nm (Fig. S6a). In contrast, after calci-nation at 800 °C, the resulting calcined sample (i.e., NiFe-PBA-C)showed an agglomerated morphology with the presence of some largeparticles of 200 nm (Fig. S6b). These observations suggested that thePVP overlayers could effectively prevent the agglomeration of the re-sulting Ni0.36Fe0.64 alloy NPs during high-temperature pyrolysis.

The XPS studies were conducted on NG-NiFe@MoC2 − 0 (i.e.,without the incorporation of MoC2) and NG-NiFe@MoC2 − 2 (i.e., withMoC2) to scrutinize the electronic interaction (Fig. S7). The Ni 2p andFe 2p spectra for the NG-NiFe@MoC2 − 0 sample indicated the pre-sence of metallic Ni° (852.9, 870.1 eV) and Fe° (706.8, 720.1 eV), aswell as surface oxidized Ni2+ (855.9, 873.4 eV) and Fe2+(711.4,724.5 eV) (Figs. S7a–S7b). [20] Interestingly, for the NG-NiFe@MoC2

− 2 sample, with the introduction of MoC2, the binding energies of Ni°shifted to 853.4 and 870.6 eV, and the binding energies of Fe° shifted to

707.2 and 720.5 eV. The shift to higher binding energies signified thecharge transfer from Ni0.36Fe0.64 alloys to MoC2 dopants. [13,21]Moreover, the Mo 3d spectrum for NG-NiFe@MoC2 − 2 revealed thepresence of Mo° (227.0, 230.3 eV), Mo3+ (227.9, 231.0 eV) and Mo4+

(228.5, 231.6 eV) species (Fig. S7c). Notably, the binding energies ofMo° were lower than that reported in literature (228.2, 231.8 eV), [22]offering further evidence of the charge transfer from Ni0.36Fe0.64 alloysto MoC2 dopants. The high-resolution of N 1 s spectrum for NG-NiFe@MoC2 − 2 electrocatalysts can be deconvoluted into two different Nspecies: pyridinic-N (398.4 eV) and pyrrolic-N (399.8, 401.1 eV) (Fig.S7d). Intriguingly, there were two different binding energies of pyr-rolic-N species. The higher binding energy (401.1 eV) is owing to theinteraction between the part of pyrrolic-N and metal atoms, thus in-ducing charge transfer between them. [23] Moreover, the C 1 s peak ofNG-NiFe@MoC2 − 2 electrocatalyst can be divided into two peakscentered at 284.6 and 285.5 eV, corresponding to C-C and C-N bonds,respectively (Fig. S8). The presence of C-N bonds suggested the N-doping into the carbon matrix.

The HER performance of as-synthesized NG-NiFe@MoC2 nanohy-brids and commercial RuO2 catalysts was then evaluated using a typicalthree-electrode cell in 1M KOH aqueous electrolyte. Among all theelectrocatalysts, the NG-NiFe@MoC2 − 2 nanohybrid displayed thehighest inherent HER activity with the highest current density (j) acrossthe whole potential windows (Fig. 2a). The HER performance of dif-ferent electrocatalysts was further assessed by overpotential (η) at j of10mA cm−2, which is the j expected for solar water-splitting device. Asshown in Table S2, the η for NG-NiFe@MoC2 − 2 electrocatalyst toachieve j of 10mA cm−2 was only 150mV, which is much lower thanthat of NG-NiFe@MoC2 − 0 (413mV) and NG-NiFe@MoC2 − 1(317mV), suggesting the significance of the incorporated MoC2 as ac-tive sites. However, with further increasing the amount of MoC2, the ηto reach j of 10mA cm−2 inversely increased to 199mV for the NG-NiFe@MoC2–3 electrocatalyst. Furthermore, the NG-NiFe@MoC2 − 2electrocatalyst yielded a high j of 104.4 mA cm−2 at η of 300mV, whichwas 73.6-fold higher than that of NG-NiFe@MoC2 − 0. Among all thefour electrocatalysts (i.e. NG-NiFe@Mo-0, NG-NiFe@Mo-1, and NG-NiFe@Mo-3), NG-NiFe@MoC2 − 2 displayed the smallest Tafel slope(92mV/dec) (Fig. 2b) and charge transfer resistance (22Ω) (Fig. 2c; seeFig. S10 for the calculation details), in accordance with the highest HERactivity demonstrated in the polarization curves (Fig. 2a). It is notablethat the polarization curve of NG-NiFe@MoC2 − 2 after cycling 10000times was almost the same as that at the initial test, only with increasedoverpotential of 1 mV at 100mA cm−2 (Fig. 2d). Interestingly, the a-few-layer-thick graphene-encapsulating shell was well-retained aftercycling 10000 times (Figs. S11a-S11b) as revealed by TEM, furthersubstantiating the outstanding stability of NG-NiFe@MoC2 − 2.Clearly, NG-NiFe@MoC2 − 2 stands out as an excellent HER electro-catalyst with a small Tafel slope, a small charge transfer resistance anda high stability.

The electrochemically active surface area (ECSA) of various NG-NiFe@MoC2 electrocatalysts was assessed on the basis of double-layercapacitance (Cdl) (see Experimental Section, Fig. S12 and Fig. S13a).Clearly, with the addition of MoC2, the ESCA improved from 43.5 cm2

on NG-NiFe@MoC2 − 0–143.3 cm2 on NG-NiFe@MoC2 − 2. (Table S2)However, with a further increase in MoC2, the ECSA of NG-NiFe@MoC2–3 decreased to 97.5 cm2. This may be due to the part of activecenters on the Ni0.36Fe0.64 alloy NPs blocked by excess MoC2 dopants.Typically, the ECSAs of electrocatalysts directly determined their cat-alytic performance. [24] Thus, with higher ESCA, the NG-NiFe@MoC2

− 2 exhibited a higher HER activity than the other three electro-catalysts. Moreover, the ECSAs of different electrocatalysts were alsoapplied to yield the current density per ECSA (jECSA), and the corre-sponding polarization curves were depicted in Fig. S13b. Clearly,among all the electrocatalysts, the NG-NiFe@MoC2 − 2 displayed thehighest jECSA at each overpotential, suggesting a higher HER activity peractive area. It has been reported that bimetallic NiMo and FeMo

Q. Hu et al. Nano Energy 50 (2018) 212–219

215

electrocatalysts with moderate M-H strength displayed the higher HERactivity than single Mo. [12,25] In our study, the EDS element mappingand XPS results suggested that there was a strong interaction betweenclosely contacted Ni0.36Fe0.64 alloy NPs and MoC2 dopants. On the basisof the higher activity per active area for the NG-NiFe@MoC2 − 2sample, a possible synergy between Ni0.36Fe0.64 alloy NPs and MoC2

dopants could thus be proposed, that is, MoC2 dopants facilitated theoptimization of the M-H strength of Ni0.36Fe0.64 alloy NPs. It is well-known that MoC2 possesses a strong M-H strength. [12,26] As both Niand Fe have the weak M-H strength, their alloy may likely exhibit aweak M-H strength as well. Thus, the incorporation of MoC2 dopantsenhanced the M-H strength of Ni0.36Fe0.64 alloy NPs, thereby favoringthe adsorption of hydrogen.

The very low HER activity of NG-NiFe@MoC2 − 0 electrocatalystreflected that the a-few-layer-thick N-doped graphene shell (denotedNG) may not be active for HER. However, the high conductivity andanti-corrosion capacity of NG are noteworthy. The pure NiFe@MoC2

− 0 electrocatalyst without a NG shell exhibited a larger Rct (404Ω)than that of NG-NiFe@MoC2 − 0 with a NG shell (Fig. S14a), indicatingthat the intrinsic high-conductivity of NG could promote the fast chargetransfer during HER. This was further substantiated by the low Rct ofNG-NiFe@MoC2 − 2 electrocatalyst (22Ω). Moreover, the cycling testfor 10,000 times showed that the NG-NiFe@MoC2 − 2 possessed anoutstanding stability for HER, which may be resulted from the goodstability of NG shell-encapsulating nanostructure. Therefore, taken to-gether, the excellent HER performance of NG-NiFe@MoC2 − 2 may bea direct consequence of the presence of highly conductive NG, theelectrocatalytic synergy between Ni0.36Fe0.64 alloy NPs and MoC2 do-pants, and the stabilizing effect of NG shell.

We now turn our attention to evaluate the OER performance of NG-NiFe@MoC2 nanohybrids and commercial RuO2 electrocatalysts via atypical three-electrode cell in 1M KOH aqueous electrolyte. As shownin Fig. 3a, the NG-NiFe@MoC2 − 2 electrocatalyst exhibited a loweronset potential and a higher current density (j) at each overpotential (η)than other three electrocatalysts, suggesting a higher inherent OERactivity. It is not surprising that a small η of 320mV was required for

NG-NiFe@MoC2 − 2 electrocatalyst to achieve j of 10mA cm−2, whichwas lower than that of NG-NiFe@MoC2 − 1 (350mV) and NG-NiFe@MoC2–3 (330mV). However, without the introduction of MoC2, a largeη of 360mV was needed for NG-NiFe@MoC2 − 0 to reach j of 10mAcm−2, corroborating that MoC2 dopants are crucial for OER. Moreover,the Tafel slope of NG-NiFe@MoC2 − 2 (31mV/dec) was much smallerthan that of NG-NiFe@MoC2 − 0 (47mV/dec), NG-NiFe@MoC2 − 1(51mV/dec) and NG-NiFe@MoC2–3 (78mV/dec) (Fig. 3b), which wasin good agreement with the higher OER activity displayed in the po-larization curves (Fig. 3a). It is important to note that the small η at10mA cm−2 and Tafel slope for NG-NiFe@MoC2 − 2 were markedlysmaller than those of commercial RuO2.

In sharp contrast, the NG-MoC2 − 2 where Ni0.36Fe0.64 alloys wereselectively etched from NG-NiFe@MoC2 − 2 displayed almost negli-gible OER responses (Fig. S9b), indicating that Ni0.36Fe0.64 alloys werethe primary active sites for OER while MoC2 dopants were indifferent.However, on the basis of the XPS results noted above, because of theinteraction with MoC2, the metallic Ni° and Fe° were positively chargedon the Ni0.36Fe0.64 alloy NPs (Fig. S7). According to the previous re-ports, positively charged metal species (i.e., NiOx, CoOx) would easilyadsorb OH-, thus facilitating the electrons withdrawing from OH- andpromoting OER in the alkaline electrolyte. [27,28] Thus, the high OERactivity of NG-NiFe@MoC2 − 2 eletrocatalyst may be ascribed to thepositively charged Ni and Fe species, resulting from electron transferNi0.36Fe0.64 alloy NPs to MoC2 dopants. Similarly, the small chargetransfer resistance (Rct) (24Ω) (Fig. S15a) and outstanding stability ofNG-NiFe@MoC2 − 2 electrocatalyst for OER (Fig. S15b) signified thatthe N-doped graphene can not only accelerate the charge transfer butonly improve the stability during OER.

Intrigued by the outstanding HER and OER performance of NG-NiFe@MoC2 − 2, we capitalized on the NG-NiFe@MoC2 − 2 as bi-functional cathodic and anodic electrocatalysts for overall water split-ting in 1M KOH electrolyte (concurrent HER and OER; seeExperimental Section; Fig. 4). The water-splitting current density of10mA cm−2 can readily be achieved by applying a cell voltage of1.53 V (Fig. 4a), which was smaller than that of other non-precious

Fig. 2. (a) Steady-state polarization curves,and (b) the corresponding Tafel plots of NG-NiFe@MoC2 − 0, NG-NiFe@MoC2 − 1, NG-NiFe@MoC2 − 2, NG-NiFe@MoC2–3 and Pt/Celectrocatalysts for HER. (c) Electrochemicalimpedance spectroscopy (EIS) measurementson these samples noted above. (d) HER polar-ization of NG-NiFe@MoC2 − 2 before andafter cycling for 10000 times.

Q. Hu et al. Nano Energy 50 (2018) 212–219

216

metal-based electrocatalysts (Table S3) and even smaller than that ofprecious Pt/C//RuO2 electrolyser (1.61 V). The H2 and O2 bubbles wereclearly observed on both electrodes at 1.53 V (Fig. 4b), unambiguouslysuggesting the presence of HER and OER at the same time. The long-time durability of the overall water splitting was further explored bycontinuous operation at 1.53 V for 10 h. As demonstrated in the inset ofFig. 4a, the NG-NiFe@MoC2 − 2 electrocatalyst exhibited an out-standing durability with a small current density loss of 0.5 mA cm−2

over a 10-h operation. Subsequently, XPS spectra were performed toexamine NG-NiFe@MoC2 − 2 electrocatalyst after the stability tests ofHER and OER. As shown in Fig. S16, after the stability test of HER for

10 h, the electronic states of Ni, Fe, Mo and N elements remained nearlyunchanged compared to those in as-prepared NG-NiFe@MoC2 − 2sample (Fig. S7). However, the stability test of OER led to the surfaceoxidation of Ni0.36Fe0.64 alloys and MoC2 species in NG-NiFe@MoC2

− 2 (Fig. S17). It has been demonstrated that the in-situ formed oxi-dized species (i.e., Ni2+, Fe2+ and Mo3+) can also serve as active sitesfor OER [16] Such excellent performance and durability render NG-NiFe@MoC2 − 2 electrocatalyst a promising alternative to preciouselectrocatalysts for water splitting in real applications.

4. Conclusions

In summary, we have demonstrated a facile and cost-effective routeto judiciously crafting MoC2-doped Ni0.36Fe0.64 alloy NPs encapsulatedwithin a-few-layer-thick N-doped graphene. The resulting NG-NiFe@MoC2 − 2 electrocatalyst exhibited remarkable performance for eitherHER or OER as well as concurrent HER and OER. Such excellent per-formance can be attributed collectively to the high conductivity of N-doped graphene, the electrocatalytic synergy of MoC2 dopants andNi0.36Fe0.64 alloy NPs, and the good stability of N-doped graphene-en-capsulating nanostructures. Impressively, when the NG-NiFe@MoC2

− 2 electrocatalyst was employed as bifunctional electrocatalysts foroverall water splitting, only a low voltage of 1.53 V was required toreach a current density of 10mA cm−2, greatly outperforming mostnon-precious metal-based electrolysers and precious Pt/C//RuO2 elec-trolyser (1.61 V). By extension, this viable yet effective strategy mayrender the development of a rich variety of robust TM-based electro-catalysts for highly efficient and low-cost water splitting.

Acknowledgements

The financial support of the National Natural Science Foundation(NNSF) of China (21574084, 51602199 and 21571131), PostdoctoralScience Foundation of China (2017M612745), the Natural ScienceFoundation of Guangdong (2015A030313554 and 2017A040405066),and Shenzhen Government's Plan of Science and Technology(JCYJ20160308104704791 and JCYJ20170818091657056) are grate-fully acknowledged.

Appendix A. Supporting information

Supplementary data associated with this article can be found in theonline version at http://dx.doi.org/10.1016/j.nanoen.2018.05.033.

References

[1] G. Ou, P. Fan, H. Zhang, K. Huang, C. Yang, W. Yu, H. Wei, M. Zhong, H. Wu, Y. Li,Nano Energy 35 (2017) 207–214.

[2] R. Subbaraman, D. Tripkovic, D. Strmcnik, K.-C. Chang, M. Uchimura,

Fig. 3. (a) Steady-state polarization curves, and (b) the corresponding Tafel plots of NG-NiFe@MoC2 − 0, NG-NiFe@MoC2 − 1, NG-NiFe@MoC2 − 2, NG-NiFe@MoC2–3 and RuO2 catalysts for OER.

Fig. 4. (a) LSV of water electrolysis based on non-precious NG-NiFe@MoC2 − 2and precious combination of Pt/C//RuO2 in 1M KOH. The inset depicts thedurability test of electrolyser at 10mA cm−2. (b) Digital image showing thegeneration of hydrogen and oxygen bubbles on Ni foam at the same time.

Q. Hu et al. Nano Energy 50 (2018) 212–219

217

A.P. Paulikas, V. Stamenkovic, N.M. Markovic, Science 334 (2011) 1256–1260.[3] W. Xu, F. Lyu, Y. Bai, A. Gao, J. Feng, Z. Cai, Y. Yin, Nano Energy 43 (2018)

110–116.[4] Y. Cheng, J. Guo, Y. Huang, Z. Liao, Z. Xiang, Nano Energy 35 (2017) 115–120.[5] Q. Ma, C. Hu, K. Liu, S.-F. Hung, D. Ou, H.M. Chen, G. Fu, N. Zheng, Nano Energy

41 (2017) 148–153.[6] L. Tao, C.-Y. Lin, S. Dou, S. Feng, D. Chen, D. Liu, J. Huo, Z. Xia, S. Wang, Nano

Energy 41 (2017) 417–425.[7] Y. Li, D. Yan, Y. Zou, C. Xie, Y. Wang, Y. Zhang, S. Wang, J. Mater. Chem. A 5

(2017) 25494–25500.[8] S. Dou, C.-L. Dong, Z. Hu, Y.-C. Huang, J.-l. Chen, L. Tao, D. Yan, D. Chen, S. Shen,

S. Chou, S. Wang, Adv. Funct. Mater. (2017), http://dx.doi.org/10.1002/adfm.201702546.

[9] Y. Wang, Y. Zhang, Z. Liu, C. Xie, S. Feng, D. Liu, M. Shao, S. Wang, Angew. Chem.Int. Ed. 129 (2017) 5961–5967.

[10] K. Gong, F. Du, Z. Xia, M. Durstock, L. Dai, Science 323 (2009) 760–764.[11] L. Yu, B.Y. Xia, X. Wang, X.W. Lou, Adv. Mater. 28 (2016) 92–97.[12] W.-F. Chen, K. Sasaki, C. Ma, A.I. Frenkel, N. Marinkovic, J.T. Muckerman, Y. Zhu,

R.R. Adzic, Angew. Chem. Int. Ed. 51 (2012) 6131–6135.[13] Z.-Y. Yu, Y. Duan, M.-R. Gao, C.-C. Lang, Y.-R. Zheng, S.-H. Yu, Chem. Sci. 8 (2017)

968–973.[14] X. Long, J. Li, S. Xiao, K. Yan, Z. Wang, H. Chen, S. Yang, Angew. Chem. Int. Ed. 126

(2014) 7714–7718.[15] X. Cui, P. Ren, D. Deng, J. Deng, X. Bao, Energy Environ. Sci. 9 (2016) 123–129.[16] Y. Xu, W. Tu, B. Zhang, S. Yin, Y. Huang, M. Kraft, R. Xu, Adv. Mater. 29 (2017)

1605957.[17] Y. Yang, Z. Lun, G. Xia, F. Zheng, M. He, Q. Chen, Energy Environ. Sci. 8 (2015)

3563–3571.[18] L. Han, X.-Y. Yu, X.W. Lou, Adv. Mater. 28 (2016) 4601–4605.[19] A. Dutta, A.K. Samantara, S.K. Dutta, B.K. Jena, N. Pradhan, ACS Energy Lett. 1

(2016) 169–174.[20] Y. Yang, Z. Lin, S. Gao, J. Su, Z. Lun, G. Xia, J. Chen, R. Zhang, Q. Chen, ACS Catal.

7 (2017) 469–479.[21] Q. Hu, L. Yang, G. Fan, F. Li, ChemNanoMat 2 (2016) 888–896.[22] S. Wang, J. Wang, M. Zhu, X. Bao, B. Xiao, D. Su, H. Li, Y. Wang, J. Am. Chem. Soc.

137 (2015) 15753–15759.[23] K. Artyushkova, B. Kiefer, B. Halevi, A. Knop-Gericke, R. Schlogl, P. Atanassov,

Chem. Commun. 49 (2013) 2539–2541.[24] X. Cui, P. Xiao, J. Wang, M. Zhou, W. Guo, Y. Yang, Y. He, Z. Wang, Y. Yang,

Y. Zhang, Z. Lin, Angew. Chem. Int. Ed. 56 (2017) 4488–4493.[25] Y. Zheng, Y. Jiao, M. Jaroniec, S.Z. Qiao, Angew. Chem. Int. Ed. 54 (2015) 52–65.[26] M.M. Jaksic, Int. J. Hydrog. Energy 26 (2001) 559–578.[27] C.C.L. McCrory, S. Jung, J.C. Peters, T.F. Jaramillo, J. Am. Chem. Soc. 135 (2013)

16977–16987.[28] K. Xu, H. Cheng, L. Liu, H. Lv, X. Wu, C. Wu, Y. Xie, Nano Lett. 17 (2017) 578–583.

Qi Hu received the Ph. D. degree from Beijing University ofChemical Technology, China in 2016. He is currently apostdoc in the research group of Prof. Chuanxin He atShenzhen University, China. His research interests focus ondesign and synthesis of nanocomposite materials for effi-cient electrocatalysis and heterogeneous catalysis.

Xiufang Liu received the B.S. degree in Hubei University ofScience and Technology, China in 2016. She is currently aMaster student in the research group of Prof. Chuanxin Heat Shenzhen University, China. Her research interests focuson the synthesis of hierarchical nanomaterials for electro-chemical hydrogen evolution.

Bin Zhu received the B.S. degree in Henan University,China in 2016. He is currently a Master student in the re-search group of Prof. Chuanxin He at Shenzhen University,China. His research interests focus on the design of one-dimensional nanomaterials for electrochemical water oxi-dation.

Liangdong Fan is currently a principal investigator forgroup of fuel cell and electrocatalysis at ShenzhenUniversity, China. He received double Ph.D degrees one inEnergy Technology from Royal Institute of Technology andanother in Chemical Technology from Tianjin University in2014. He has published more than 55 journal articles withan H index of 22 (by Google Scholar). His research groupfocuses on design and synthesis of functional nanostructurematerials for highly efficient energy conversion and storageapplication, such as solid oxide fuel cell, water electrolysisand supercapacitor.

Xiaoyan Chai is an Assistant Professor of Chemistry andEnvironmental Engineering, Shenzhen University. She re-ceived her PhD from the South China University ofTechnology in 2015. Her research interests focus on elec-trospinning synthesis of carbon nanofibers for high elec-trocatalysis of oxygen reduction.

Qianling Zhang is a Professor at Shenzhen University,China. She received M.S. from Changchun Institute ofApplied Chemistry Chinese Academy of Sciences and Ph.D.from Sun Yat-Sen University. Her research mainly focuseson design and synthesis of ruthenium complexes for bioi-maging and cancer phototherapy.

Jianhong Liu is a Professor at Shenzhen University, China.He received M.S. from University of Science andTechnology of Chengdu in 1986 and Ph.D. from FudanUniversity in 1989. His research interests mainly focus onsynthesis and application of carbon materials for lithiumion batteries.

Q. Hu et al. Nano Energy 50 (2018) 212–219

218

Chuanxin He received his PhD in polymer chemistry andphysics from University of Science and Technology of Chinain 2010. He is currently the Assistant Dean of College ofChemistry and Environmental Engineering, ShenzhenUniversity. He published more than 60 journal papers, in-cluding Advanced Materials, Carbon, Journal of MaterialsChemistry A/B, Journal of Power Sources, and so on. Hiscurrent research interest lies in design and synthesis ofnovel nanostructured materials for fuel cells, water splittingand electrochemical reduction of carbon dioxide.

Zhiqun Lin received his Ph.D. in Polymer Science andEngineering from University of Massachusetts, Amherst in2002. He is currently Professor of Materials Science andEngineering at the Georgia Institute of Technology. Hisresearch interests include solar cells, photocatalysis, elec-trocatalysis, batteries, semiconductor organic–inorganicnanocomposites, conjugated polymers, block copolymers,hierarchical structure formation and assembly, surface andinterfacial properties, multifunctional nanocrystals, andJanus nanostructures.

Q. Hu et al. Nano Energy 50 (2018) 212–219

219