Economical and efficient electrocatalysts are crucial to the hydrogen evolution reaction (HER) in water splitting to produce hydrogen. Heterostructured electrocatalysts generally exhibit enhanced HER catalytic activity due to the strong electron coupling effects and synergistic optimization of hydrogen adsorption–desorption. Herein, in-plane heterostructured MoN/Mo2N nanosheets are fabricated as high-efficiency HER electrocatalysts in the alkaline medium from bulk MoS2 by molten salt-assisted synthesis. Density-functional theory calculations and experiments show that the in-plane heterostructured MoN/Mo2N nanosheets facilitate interfacial electron redistribution from Mo2N to MoN, giving rise to more negative H2O adsorption energy and optimal hydrogen adsorption free energy (ΔGH* = −0.017 eV). Consequently, a low overpotential of 126 mV at 10 mA cm−2 and a small Tafel slope of 69.5 mV dec−1 are achieved in the 1M KOH electrolyte, demonstrating excellent HER characteristics. Moreover, the overpotential shows negligible change after operating at 50 mA cm−2 for 12 h, confirming the excellent stability. The results reveal a novel and effective strategy to design highly efficient 2D in-plane heterostructured HER electrocatalysts for water splitting.
The hydrogen evolution reaction (HER) is an environmentally friendly and promising approach to produce clean hydrogen (H2) via water splitting, and high-efficiency and durable electrocatalysts are crucial to reduce the overpotential and promote the HER efficiency.1–3 Although platinum (Pt) shows excellent HER activity, their high cost and scarcity have hampered the use of industrial water splitting to produce hydrogen.4,5 Therefore, it is essential to design and develop alternative non-precious metal-based HER catalysts. In recent years, transition metal compounds, such as metal phosphides,6 carbides,7 nitrides,8 sulfides,9 and borides,10 have been investigated as alternative HER catalysts. In particular, molybdenum-based nitrides, such as MoN and Mo2N, have garnered increasing attention due to their metal-like conductivity, good corrosion resistance, and favorable d-band electronic structure.
Two-dimensional (2D) molybdenum nitrides possess enhanced HER properties due to their unsaturated dangling bonds and abundant active sites.11 However, single-phase 2D MoN or Mo2N delivers unsatisfactory HER performance due to their lopsided hydrogen adsorption–desorption free energy. In this context, proper heterostructures can improve the HER catalytic activity by adjusting the local electronic structure at the interface, optimizing the water dissociation capability, and balancing hydrogen adsorption–desorption. For example, in the work of Sun et al., the authors have reported an Co/MoN heterogeneous nanoflake array with enhanced water dissociation capability in alkaline HER process as a result of promoted electron transport from Co to MoN.12 In the work of Lin et al., a MoN–MnO heterostructure nanosheet catalyst with the interfacial-O configuration has been described, which enables fast hydrogen coupling and desorption on the nitride domains leading to enhanced HER activity in alkaline media.13 However, preparation of MoN nanosheets is usually complicated and the second phases are generally deposited on the MoN surface, resulting in limited active sites. In this context, construction of 2D in-plane nitride heterojunctions with strong electron coupling may deliver better HER performance by taking advantage of the more abundant active sites and better balance between adsorption and desorption of hydrogen. However, controllable and facile synthesis of the 2D heterostructured MoN/Mo2N nanosheets with robust heterointerfaces is still challenging.
Herein, a flower-like 2D heterostructured MoN/Mo2N nanosheets electrocatalyst is designed and demonstrated to be high-efficiency alkaline HER catalysts. The 2D MoN/Mo2N nanosheets are synthesized from bulk MoS2 by Na2CO3-assisted thermal nitriding under NH3 and subsequent annealing in Ar/H2 as schematically illustrated in Fig. 1(a). The flower-like 2D heterostructured MoN/Mo2N nanosheets have several advantages in HER. First, the 2D MoN/Mo2N hybrid structure provides abundant active sites and robust heterointerfaces which facilitate access of electrolytes and release of H2 bubbles due to the 2D structure and flower-like morphology. Second, the 2D in-plane MoN/Mo2N interfaces optimize H adsorption/desorption by coupling strong hydrogen adsorption of MoN and weak hydrogen adsorption of Mo2N, consequently yielding an optimal hydrogen adsorption free energy for MoN/Mo2N (ΔGH* = −0.017 eV). Third, the ratio of MoN/Mo2N can be optimized by controlling the annealing temperature and time in the Ar/H2 ambient. The optimal in-plane MoN/Mo2N heterostructured nanosheets exhibit excellent HER activity as exemplified by a low overpotential of 126 mV at 10 mA cm−2 and a small Tafel slope of 69.5 mV dec−1 in alkaline media, which are superior to those of recently reported MoN- and Mo2N-based electrocatalysts (Table S1).14–22 Moreover, the 2D MoN/Mo2N nanosheets exhibit remarkable stability with only negligible potential changes at a high current density of 50 mA cm−2 after 12 h. The results demonstrate a novel strategy to construct 2D in-plane heterostructured TMNs catalysts for high-performance electrocatalytic water splitting.
II. EXPERIMENTAL SECTION
A. Chemicals and materials
All the chemicals were used without further purification. Molybdenum disulfide (MoS2, 99.5%) was purchased from Aladdin and sodium carbonate (Na2CO3, ≥99.8%) and hydrochloric acid (HCl, 36%–38.0%) were purchased from Sinopharm Chemical Reagent Co, Ltd.
B. Synthesis of 2D MoN/Mo2N nanosheets
The MoN nanosheets were prepared from the bulk MoS2 precursor by Na2CO3-assisted salt synthesis.23 In the typical synthesis, the MoS2 and Na2CO3 powders with a molar ratio of 1:2.5 were mixed in a mortar and annealed in a quartz tube furnace at 750 °C with a heating rate of 10 °C/min for 3 h under NH3. After cooling to room temperature, the product was collected, washed with 1M HCl for 2 h, and rinsed with de-ionized water several times. The MoN nanosheet powder was obtained after freeze drying. The MoN nanosheets were then annealed under Ar/H2 to produce the MoN/Mo2N heterojunction nanosheets at 800 °C with a heating rate of 10 °C/min. The ratio of MoN/Mo2N was controlled via adjusting the annealing time (1.0, 1.5, and 3.0 h and samples denoted as MoN/Mo2N-1, MoN/Mo2N-1.5, and MoN/Mo2N-3). The Mo2N nanosheets were obtained by further annealing MoN/Mo2N-1.5 nanosheets at 800 °C for 8 h under Ar/H2.
C. Characterization and electrochemical assessment
1. Materials characterization
The morphology and microstructure of the materials were characterized by field-emission scanning electron microscopy (FE-SEM, ThermoFisher/Apreo S HiVac) and transmission electron microscopy (TEM, JEM-F200). The crystal structure was analyzed by x-ray diffraction (XRD, Rigaku/SmartLab) and elemental maps were acquired by energy-dispersive x-ray spectroscopy (EDS, OXFORD AZtecLive Ultim Max 100). The chemical states were determined by x-ray photoelectron spectroscopy (XPS, ESCALab250).
2. Electrochemical measurements
The electrochemical measurements were carried out in 1.0M KOH electrolyte using a three-electrode system on electrochemical workstation (Bio-logical VMP 300). To prepare the working electrode, 6 mg of the sample, 20 µl of Nafion, 100 µl of de-ionized water, and 900 µl of isopropanol were mixed, sonicated for 40 min, and introduced onto a copper foam dropwise. The copper foam was heated in an oven at 70 °C for 10 min and 40 µl of diluted tenfold Nafion was used as the protective layer. The modified copper foam electrode was dried at 25 °C in air for 18 h. In the three-electrode system, the electrocatalyst-loaded copper foam (1 × 1.2 × 0.1 cm3) was the working electrode, graphite rod was the counter-electrode, and saturated calomel electrode (SCE) was the reference electrode. All the potentials were iR corrected and calibrated with reference to the reversible hydrogen electrode (RHE) according to Nernst equation (ERHE = ESCE + 0.241 + 0.0592 × pH). The activity of the working electrode was evaluated by linear sweep voltammetry (LSV) at 5 mV s−1. Electrochemical impedance spectroscopy (EIS) was conducted between 100 kHz and 0.01 Hz with an AC perturbation of 5 mV and the electric double-layer capacitance was obtained via cyclic voltammetry (CV) from 10 to 60 mV s−1 in the non-Faradic region between −0.72 and −0.82 V vs SCE. To evaluate the electrochemical stability of MoN/Mo2N, CV was carried out for 5000 cycles at a scanning rate of 100 mV s−1 and a galvanostatic test was conducted at a current density of 50 mA cm−2, and the LSV curves before and after cycling at 5 mV s−1 were recorded.
3. Theoretical calculation
III. RESULTS AND DISCUSSION
The 2D in-plane heterostructured MoN–Mo2N nanosheets are prepared by two steps [Fig. 1(a)]. The flower-like MoN assembled by 2D nanosheets is prepared from bulk MoS2 by Na2CO3-assisted thermal nitriding under NH3 at 750 °C for 3 h.23 The bulk MoS2 precursor with a layered structure has an average size of 50 µm and thickness of 10 µm [Figs. 1(b) and S1(a)]. The sharp (002) diffraction peak observed from bulk MoS2 (PDF# 37–1492) [Fig. S1(b)] discloses high crystallinity.28 After Na2CO3-assisted thermal nitriding reaction, the layered MoS2 power is fully converted into flower-like morphology assembled by 2D nanosheets with a thickness of 6.2 nm [Figs. 1(c) and S2]. XRD patterns of the product is ascribed to hexagonal MoN (PDF# 25–1367) and no MoS2 peaks are observed, suggesting that the MoS2 is fully converted into 2D MoN nanosheets. The 2D heterostructured MoN/Mo2N nanosheets are prepared by further annealing the MoN nanosheets under Ar/H2. The MoN/Mo2N structure retains a similar flower-like morphology with an average particle size of 5 µm [Fig. 1(d)].
The XRD patterns of MoN/Mo2N-1, MoN/Mo2N-1.5, and MoN/Mo2N-3 are depicted in Fig. S3. The characteristic peaks at 37.4°, 43.5°, 63.2°, and 75.7° correspond to the (111), (200), (220), and (311) facets of cubic Mo2N (PDF# 25–1366) [Fig. 2(a)]. Moreover, the amount of Mo2N in the MoN/Mo2N nanosheets increases with annealing time (Fig. S3). After further annealing MoN/Mo2N for 8 h under Ar/H2, only Mo2N phase is observed, suggesting formation of single-phase Mo2N nanosheets [Fig. 2(a)]. The MoN, MoN/Mo2N, and Mo2N have the similar morphology, suggesting topochemical chemical conversion from MoN to MoN/Mo2N and Mo2N (Fig. S4). The chemical states of MoN/Mo2N-1.5 are determined by XPS. XPS full-scan survey reveals the existence of Mo and N elements in MoN/Mo2N-1.5 [Fig. S5(a)]. High-resolution Mo 3d XPS spectra can be resolved to four pairs of 3d5/2/3d3/2 doublets: for Mo2+ (228.7/231.8 eV), Mo3+ (228.9/232.1 eV), Mo4+ (229.7/232.6 eV), Mo6+ (233.1/235.9 eV) [Fig. 2(b)].29–32 It has been suggested the dominance of Mo2+ in Mo2N and Mo3+ in MoN and higher valence states of Mo4+ and Mo6+ are due to partial surface oxidation.33–35 Notably, the Mo3+ peak red shifts (∼0.2 eV), whereas the Mo2+ peak blue shifts (∼0.15 eV), suggesting electron transfer from Mo2N to MoN in MoN/Mo2N. The N 1 s spectrum of MoN/Mo2N-1.5 in Fig. S5(b) shows two deconvoluted peaks at 397.9, and 399.3 eV attributing to Mo–N bond, and N–H bond, respectively.36–40 Similarly, there are two O 1 s peaks [Fig. S5(c)] at 530.9 and 532.1 eV arising from N–Mo-Ox and N–Mo-(OH)x in MoN/Mo2N due to surface oxidation during XPS test.33–35,41,42
TEM and HR-TEM images are depicted in Figs. 2(c) and 2(d). Distinct interfaces between MoN and Mo2N phases are observed. The (200) crystal plane of MoN forms an interface with the (111) crystal plane of Mo2N with interplanar distances of ∼0.25 and ∼0.24 nm, respectively.43 To further investigate the formation process of the in-plane MoN/Mo2N heterojunctions, MoN/Mo2N-3 with a higher Mo2N concentration is immersed in 6M NaOH solution at 80 °C for 5 min and the composition and structure before and after immersion are evaluated by XRD and TEM. After alkali washing, the intensity of the XRD peaks of Mo2N of MoN/Mo2N-3 (Fig. S6) decreases, indicating that the Mo2N phase has been partially removed during this alkali etching process. The TEM images of MoN/Mo2N-3 after alkali immersion (Fig. S7) show a large number of pores formed as a result of washing away Mo2N. Hence, it is conjectured that the MoN phase is partially reduced to form Mo2N by annealing MoN under Ar/H2, resulting in the formation of in-plane MoN/Mo2N heterojunction interface (Fig. S8). The elemental maps (Fig. S9) reveal homogeneous distributions of Mo and N in the MoN/Mo2N nanosheets.
The HER electrocatalytic characteristics of MoN, MoN/Mo2N, and Mo2N samples are evaluated in 1M KOH electrolyte. The electrocatalysts are drop-coated on the copper foam that has negligible HER activity. The commercial 20 wt. % Pt/C with the same mass loading dispersed on copper foam is also studied for comparison.44 As shown in Fig. 3(a), the in-plane heterostructured MoN/Mo2N electrode shows enhanced HER activity compared with MoN and Mo2N nanosheets. To reach current densities of 10 and 100 mA cm−2, MoN/Mo2N-1.5 only requires low overpotentials of 126 and 220 mV, respectively, which are lower than those of MoN (249 and 360 mV), Mo2N (221 and 344 mV) and the other Mo-based catalysts [Figs. 3(a) and S10, Table S1]. Figure 3(b) shows that the Tafel slope of MoN/Mo2N-1.5 is 69.5 mV dec−1 and smaller than those of the MoN nanosheets (98.0 mV dec−1) and Mo2N (81.3 mV dec−1), suggesting the Volmer–Heyrovsky mechanism.45 The EIS results in Fig. 3(c) show that the charge transfer resistance of MoN/Mo2N-1.5 is 4.0 Ω and smaller than those of MoN (5.4 Ω) and Mo2N (5.2 Ω), indicative of fast interfacial charge transfer kinetics.46 The electrochemical active surface area (ECSA) of MoN, MoN/Mo2N-1.5, and Mo2N is calculated by double-layer capacitance (Cdl) measurement experiments (Fig. S11). As shown in Fig. 3(d), the Cdl value of MoN/Mo2N-1.5 is 40.02 mF cm−2, which is larger than those of MoN (20.75 mF cm−2) and Mo2N (24.61 mF cm−2), suggesting that MoN/Mo2N-1.5 has more active sites. Since MoN, Mo2N, and MoN/Mo2N-1.5 have similar morphology, the morphology effect on HER activity is negligible. The intrinsic activities of MoN, Mo2N, and MoN/Mo2N-1.5 are analyzed by ECSA normalization (Fig. S12) and MoN/Mo2N-1.5 shows the highest catalytic activity for HER. Continuous CV and chronopotentiometry tests are conducted to assess the stability of the MoN/Mo2N-1.5 catalyst. As shown in Fig. 3(e), the polarization curves of MoN/Mo2N-1.5 almost overlap with the initial one after 5000 CV cycles and the overpotentials show negligible change for 12 h at 50 mA cm−2. SEM images reveal the flower-like morphology assembled by nanosheets are well-retained and XRD patterns have no change (Fig. S13). Moreover, the overpotential shows negligible change after operating at 50 mA cm−2 for 12 h, demonstrating good electrochemical stability of MoN/Mo2N-1.5 in the alkaline electrolyte [Fig. 3(e)]. Figure 3(f) and Table. S1 compare the HER performance of MoN/Mo2N-1.5 with other reported molybdenum-based nitride electrocatalysts, revealing MoN/Mo2N-1.5 has excellent HER characteristics.14–22
To elucidate the mechanism of enhanced HER performance of the in-plane of MoN/Mo2N, density-functional theory (DFT) calculation is carried out on the (200) plane of MoN and (111) plane of Mo2N. Figures S14 and 4(a) show the geometric structures and charge density difference map of the MoN/Mo2N heterojunction and the cyan and yellow regions in the charge density difference map represent loss and aggregation, respectively. The electron density of Mo2N at the MoN/Mo2N interface decreases and charges aggregate at the Mo site of adjacent MoN in line with the XPS results. The density of state (DOS) of the Mo atoms on MoN, Mo2N, and MoN/Mo2N is calculated to analyze the interactions between the adsorbate and metal sites.47 As shown in Fig. 4(b), the MoN/Mo2N heterojunction has higher DOS than MoN and Mo2N near the Fermi level (EF) at the Mo site, indicating that electron rearrangement at the MoN/Mo2N interface and more electrons on the Mo site promote H2O adsorption and optimize H adsorption and desorption.48,49 Based on the water adsorption model in Fig. S15, the calculated H2O adsorption energy of the MoN/Mo2N interface is −1.537 eV, which is less than that of MoN (−1.034 eV) and Mo2N (−1.051 eV), indicating stronger adsorption capacity of water at the heterogeneous interface [Fig. 4(c)]. The hydrogen adsorption free energy (ΔGH*) is an important parameter for the HER activity and the closer ΔGH* is to zero, the more suitable the production of H2.50,51 Figures. 4(d) and S16 show the ΔGH* values of MoN, Mo2N, and MoN/Mo2N. ΔGH* of the Mo site at the interface of MoN (200)/Mo2N (111) is −0.017 eV, which is lower than those of the Mo site on MoN (−0.525 eV) and Mo site on Mo2N (−0.346 eV). These results suggest excellent HER activity on the heterogeneous MoN/Mo2N catalyst in the alkaline solution, which is more conducive to water adsorption and hydrogen desorption, consequently boosting the enhanced HER activity and kinetics.
In-plane heterostructured MoN/Mo2N nanosheets are designed and demonstrated to be high-performance HER electrocatalysts in the alkaline medium. The in-plane heterostructured MoN/Mo2N nanosheets exhibit a low overpotential of 126 mV at a current density of 10 mA cm−2 and a small Tafel slope of 69.5 mV dec−1 in the 1M KOH electrolyte, which are superior to those of recently reported Mo-based nitride electrocatalysts. DFT calculations and experiments show that electron transfer in the 2D in-plane interface between MoN and Mo2N produces a more negative H2O adsorption energy and moderate hydrogen adsorption free energy (ΔGH* = −0.017 eV). Moreover, the flower-like morphology of MoN/Mo2N facilitates access of electrolytes and release of H2 bubbles and the unique 2D structure increases exposure of active sites. This work is expected to pave a new way for new and rational design of highly efficient 2D in-plane heterostructured electrocatalysts for HER.
See the supplementary material for detailed experimental information and structural characterization [Figs. S1–S9 and S13], electrochemical performance tests [Figs. S10–S12], the atomic models of computation [Figs. S14–S16], and comparison table of electrochemical performance (Table S1).
This work was financially supported by National Natural Science Foundation of China (Grant Nos. U2004210, 21875080, 51572100, and 52002297), the Basic Research Program of Shenzhen Municipal Science and Technology Innovation Committee (Grant No. JCYJ20210324141613032), Application Foundation Frontier Project of Wuhan Science and Technology (Grant No. 2020010601012199), Hubei Province Natural Science Foundation Innovation Group Project (Grant No. 2019CFA020), City University of Hong Kong Strategic Research Grant (SRG), Hong Kong, China (Grant No. 7005505), and City University of Hong Kong Donation Research Grant, Hong Kong, China (Grant No. DON-RMG 9229021). The authors are grateful to the facility support provided by the Analytical and Testing Center of HUST.
Conflict of Interest
The authors have no conflicts to disclose.
X.L. and H.S. contributed equally to this work.
Xiuen Luo: Formal analysis (equal); Investigation (equal); Validation (equal); Writing – original draft (equal). Hao Song: Formal analysis (equal); Investigation (equal); Validation (equal); Writing – original draft (equal). Yulei Ren: Data curation (equal); Investigation (equal). Xuming Zhang: Resources (equal); Writing – review & editing (equal). Kaifu Huo: Conceptualization (lead); Funding acquisition (lead); Project administration (equal); Supervision (lead); Writing – review & editing (lead). Paul K. Chu: Formal analysis (equal); Writing – review & editing (equal).
The data that support the findings of this study are available within the article and its supplementary material.