Heterostructured Mo2N–Mo2C Nanoparticles Coupled with N‐Doped Carbonized Wood to Accelerate the Hydrogen Evolution Reaction

Mo2C is a promising non‐precious hydrogen evolution reaction (HER) electrocatalyst. However, regulating the strong hydrogen adsorption characteristics of Mo2C and finding suitable support electrodes are essential processes before Mo2C can replace Pt to realize a sustainable hydrogen economy. Herein, the facile synthesis of heterostructured Mo2N–Mo2C nanoparticles on N‐doped carbonized wood (Mo2N–Mo2C/N‐CW) as a self‐supported electrode through carbonization and NH3 plasma treatment is demonstrated. The synergistic effects of heterostructured Mo2N–Mo2C and N‐CW with aligned microchannels provide enhanced catalytic activity, fast charge transfer kinetics, additional active site exposure, and rapid transport of the electrolyte and H2 bubbles, which improves the HER performance. Consequently, the Mo2N–Mo2C/N‐CW electrode exhibits superior HER performance with low overpotentials of only 79 and 311 mV to reach 10 and 500 mA cm−2 in an acidic solution, respectively. It also exhibits long‐term stability for 20 h in the high‐current‐density region (110 mA cm−2). The density functional theory (DFT) calculations at various sites reveal that the heterointerface of Mo2N and Mo2C promoted the catalytic activity by optimizing the adsorption/desorption of hydrogen.


Introduction
Hydrogen is the most promising energy-source candidate to replace fossil fuels and realize a carbon-neutral society; this is because hydrogen is abundant, has a high gravimetric energy density, and is environmentally friendly. [1] Electrochemical hydrogen evolution reaction (HER) by water splitting is the most suitable method for carbon-free, and efficient electrocatalysts are required to lower kinetics barriers. The most effective electrocatalytic materials for the HER are Pt and Pt alloys with small overpotentials and Tafel slopes; [2] however, their high cost and scarcity significantly restrict their largescale practical applications. [3] Therefore, it is critical to explore stable and highly efficient electrocatalysts based on nonprecious and earth-abundant elements to enable economically sustainable hydrogen production.
Among non-precious electrocatalysts, molybdenum carbide (Mo 2 C) is considered one of the most promising electrocatalysts for the HER owing to its high chemical stability, low cost, and d-band structure, which are similar to the properties of Pt group metals. [4] Mo 2 C also exhibits strong adsorption of hydrogen on the surface due to its negative hydrogen-binding energy, which promotes the Volmer step of the HER reaction. However, the desorption of active hydrogen (i.e., the Heyrovsky/Tafel step) is restricted. [5,6] To address this, the introduction of doping with transition metals such as Ni- [7] or Co-dopant [8] can effectively exert electrocatalytic performance, but is limited to inevitable structural damage.
Most electrocatalysts are synthesized in powder form using the hydrothermal method, and for this, the electrode must be coated with a binder such as Nafion for hydrogen production. [9] However, the practical use of a binder is limited owing to disadvantages such as a low catalyst loading range and inhibited diffusion. [10] Manufacturing self-supported electrodes without binders can improve the catalyst loading range and mass transfer, which is promising for maximizing the number of active sites and conductivity for an efficient HER. [11,12] This approach can also prevent the aggregation of electrocatalysts. Indeed, various carbon-based conducting matrices such as carbon nanotubes, [13] graphene, [14] and carbon microflowers [15] have been integrated with Mo 2 C and have demonstrated good HER performance as self-supported electrodes Recently, inspired by nature, carbonized wood (CW) with a noble-metal-free electrocatalyst has emerged as a fascinating platform to replace traditional self-supported electrodes. CW enables rapid mass transport and maximizes the number of active sites in an electrocatalyst owing to its intrinsically anisotropic structure and large porous surface area. [16,17] In particular, low tortuosity and aligned microchannels across the entire electrode can promote rapid electrolyte permeation and quickly release H 2 during the HER. [18,19] Moreover, CW is environmentally friendly and abundant, with a low cost, excellent electrical conductivity, and good chemical and mechanical stabilities, making it a potential candidate as a good supporting electrode for the HER. [20][21][22] Mo 2 C synthesized on a carbon-based substrate is generally obtained through the reaction of molybdenum-based precursor (i.e., ammonium molybdate) and carbon source by annealing. [23] However, in this process, molybdenum oxycarbide (MoO x C y ) is formed together by high carbonization temperature and wetchemical route. [24] The MoO x C y formed on the Mo 2 C surface changes or inhibits the catalytic activity of the pristine Mo 2 C. [25] Therefore, minimizing the unwanted formation of MoO x C y or converting it to other catalysts is an important strategy in Mo 2 C electrocatalyst synthesis to enhance the HER performance.
In this study, we synthesized heterostructured Mo 2 N-Mo 2 C nanoparticles (NPs) on N-doped CW (hereafter denoted as Mo 2 N-Mo 2 C/N-CW) as an efficient self-supported electrode. Heterostructured Mo 2 N-Mo 2 C NPs were synthesized by converting MoO x C y into Mo 2 N through a simple NH 3 plasma treatment. The minimization of MoO x C y and formation of Mo 2 N with a metallic electronic structure, downshifted d-band, and good corrosion resistance contributed to the enhanced HER activity and stability. [26] The N doping of CW by molybdenum precursor and NH 3 plasma treatment enhanced the charge transfer, the number of defects acting as active sites, and surface hydrophilicity, which improved the HER. Moreover, the aligned microchannels of the N-CW favored the release of H 2 bubbles. Consequently, Mo 2 N-Mo 2 C/N-CW demonstrated superior HER performance in an acidic solution with overpotentials of À79 and À311 mV at current densities of 10 and 500 mA cm À2 , respectively, as well as robust durability during a 20 h test at a high current density (110 mA cm À2 ). Density functional theory (DFT) calculations further revealed that the H* adsorption Gibbs free energy (ΔG H* ) of the Mo 2 N-Mo 2 C heterointerface approached the optimal zero-energy criterion for the HER process, which shows that the electrical interaction of Mo 2 C and Mo 2 N modulated the strong hydrogen adsorption properties of Mo 2 C. These results present a new strategy for tuning molybdenum-based heterostructures and open up opportunities for the design of transition-metal-based electrocatalysts and applications in other energy-related fields. Figure 1 schematically illustrates the process adopted for fabricating heterostructured Mo 2 N-Mo 2 C NPs on N-doped CW supports (details of the experimental procedures are provided in the Experimental Section). A 2 mm thick slice of balsa wood was immersed in an (NH 4 ) 2 MoO 4 aqueous solution and dried in a vacuum desiccator and dry oven to obtain an (NH 4 ) 2 MoO 4 /wood slice. The slice was pre-carbonized at 260°C for 6 h, after which it was carbonized at 900°C for 6 h in an Ar atmosphere to form the Mo 2 C/N-CW. Different concentrations of (NH 4 ) 2 MoO 4 (5, 10, and 20 wt%) were used to optimize the Mo 2 C content, with the samples labeled as Mo 2 C/N-CW-5, Mo 2 C/N-CW-10, and Mo 2 C/ N-CW-20, respectively. Subsequently, Mo 2 N-Mo 2 C heterostructured NPs fabricated on N-CW (denoted as Mo 2 N-Mo 2 C/N-CW) were obtained via NH 3 plasma treatment of Mo 2 C/N-CW- 20. and MoO x C y , respectively. [27,28] The HRTEM image of Mo 2 N-Mo 2 C/N-CW displayed in Figure 2f presents two adjacent NPs in contact, with interplanar spacings  [29] It can be observed that MoO x C y converts to Mo 2 N owing to the NH 3 plasma treatment. These results are in good agreement with those obtained from the X-ray diffraction (XRD) analysis of the phase composition, which will be discussed in the following.

Results and Discussion
The TEM images of all the samples ( Figure S2, Supporting Information) reveal that the amount of Mo 2 C NPs on N-CW increases with the (NH 4 ) 2 MoO 4 concentration, and the 20 wt% Mo 2 C/N-CW sample can be expected to exhibit the highest catalytic activity. In addition, Figure S2c,d, Supporting Information, further confirm that Mo 2 C and Mo 2 N-Mo 2 C NPs are uniformly distributed on N-CW, and the size of nanoparticles is several tens of nm. Figure 2g,h depicts the HRTEM-EDS spectra of Mo 2 C/N-CW-20 and Mo 2 N-Mo 2 C/N-CW, respectively. N detected in Mo 2 C/N-CW-20 indicates that the CW is N-doped owing to the thermal decomposition of (NH 4 ) 2 MoO 4 during the carbonization process. Mo 2 N-Mo 2 C/N-CW (0.67 at%) possesses a higher N content than Mo 2 C/N-CW-20 (0.18 at%). This observation further supports the conclusion that Mo 2 N-Mo 2 C/N-CW is obtained from Mo 2 C/N-CW via the NH 3 plasma treatment. The corresponding elemental mapping images are presented in Figure S3, Supporting Information. Generally, the phase/crystal structure strongly influences the material properties and catalytic performance. The structure of Mo 2 N-Mo 2 C/N-CW was evaluated using Raman spectroscopy, XRD, and X-ray photoelectron spectroscopy (XPS). The Raman spectra of all the samples represented characteristic spectra of graphitic carbon, including the D (%1336 cm À1 ) and G (%1582-1584 cm À1 ) bands ( Figure 3a). [30] The Raman spectra revealed that the D and G bands of Mo 2 N-Mo 2 C/N-CW were red-shifted by 8.28 and 8.5 cm À1 compared with those for the Mo 2 C/N-CW samples, respectively. This shift may be related to the charge transfer between the Mo 2 N-Mo 2 C NPs and N-CW, rather than the doping effect of N atoms in the graphitic lattice. [31] The I D /I G ratio of the Mo 2 C/N-CW samples increased from 1.03 to 1.11 as the (NH 4 ) 2 MoO 4 concentration increased, owing to the insertion of heterogeneous N atoms into the graphitic structure of CW. After the NH 3 plasma treatment of Mo 2 C/N-CW-20, the I D /I G ratio further increased from 1.11 (Mo 2 C/N-CW-20) to 1.15 (Mo 2 N-Mo 2 C/N-CW). This could be attributed to the partial damage caused to the sp 2 hybridized C domains by the NH 3 plasma treatment. NH 3 plasma treatment on CW support can provide abundant defects acting as active sites and good wettability. [32] In other words, N-CW can serve not only as a support for Mo 2 N-Mo 2 C but also as a catalyst, thereby contributing to the improvement in the HER activity owing to its synergistic coupling with Mo 2 N-Mo 2 C.
The XRD patterns of different samples are compared in Figure 3b. All the samples presented two broad peaks at   (002) and (100) crystal planes of graphitic carbon, respectively. [33] For Mo 2 C/N-CW, ten new peaks appeared compared to CW. The two diffraction peaks located at 27.8°and 40.5°corresponded to MoO x C y and metallic Mo impurities (JCPDS 01-1208), respectively. [28] The other eight diffraction peaks that appeared at 34.4°, 37.9°, 39.4°, 51.9°, 61.6°, 69.6°, 73.8°, and 75.6°indicated β-Mo 2 C (JCPDS 35-0787). For Mo 2 N-Mo 2 C/N-CW, the diffraction peaks related to MoO x C y (pink arrow) and metallic Mo (orange arrow) disappeared, indicating that MoO x C y and Mo were successfully converted into γ-Mo 2 N (JCPDS 25-1366) owing to the NH 3 plasma treatment. In addition, the peak positions for β-Mo 2 C did not change significantly, indicating the coexistence of β-Mo 2 C and γ-Mo 2 N, which was consistent with the HRTEM results.
Because the HER performance is closely related to the chemical environment of the transition metal catalyst, the chemical composition and surface electronic state of Mo 2 N-Mo 2 C/N-CW were further characterized using XPS (Figure 3c [36] In the N 1s spectrum (Figure 3e), three peaks fitted at 398.2, 399.8, and 401.1 eV were assigned to the pyridinic-N, pyrrolic-N, and graphitic-N species, respectively. After the NH 3 plasma treatment of Mo 2 C/N-CW, the deconvoluted N 1s spectrum of Mo 2 N-Mo 2 C/N-CW revealed a new peak appearing at 395.1, which corresponded to Mo-N bond. [37,38] The presence of Mo-N confirmed its successful conversion into Mo 2 N owing to the NH 3 plasma treatment. Interestingly, the pyridinic-N proportion increased from 17.37% for Mo 2 C/N-CW to 39.97% for Mo 2 N-Mo 2 C/N-CW. Among the N-configuration types in the N-CW matrix, pyridinic N was the most favorable for hydrogen adsorption, which proved beneficial for the HER. [39] In the Mo 3d XPS spectrum of Mo 2 C/N-CW (Figure 3f ), the binding energies of 228.2 and 231.3 eV could be assigned to Mo 3d 5/2 and Mo 3d 3/2 , respectively, and represented Mo 2 C. Furthermore, Mo 2 C/N-CW also presented peaks at 229.3/232.5 eV for Mo 4þ (MoO 2 ) and 231.5/235.7 eV for Mo 6þ (MoO 3 ), which are commonly observed when Mo 2 C is exposed to air. [40] After the NH 3 plasma treatment, the β-Mo 2 C, MoO 2 , and MoO 3 peaks appeared diminished, and new Mo 3d peaks attributed to Mo 2 N (229.1/232 eV) appeared, thus confirming the successful transformation of MoO x C y to Mo 2 N. Additionally, the Mo 3d 5/2 and Mo 3d 3/2 binding energies of Mo 2 C shifted from 228.2 to 227.9 eV and from 232 to 231.1 eV, respectively. This implied a strong surface electronic interaction between Mo 2 N and Mo 2 C, which could induce a charge redistribution at the interface and thus alter the catalytic activity. [41] The aforementioned characterizations confirmed that Mo 2 N-Mo 2 C NPs were fabricated on N-CW by the carbonization and NH 3 plasma treatment.
To understand the HER activity of Mo 2 N-Mo 2 C/N-CW, linear sweep voltammetry (LSV) was performed at a scan rate of 10 mV s À1 in a 0.5 M H 2 SO 4 solution using a three-electrode system. The activities of Mo 2 C/N-CW and CW were measured for comparison. Figure 4a depicts the polarization curves with iR compensation for Mo 2 N-Mo 2 C/N-CW and other samples. CW required a high overpotential (η) of 243 mV to reach a current density of 10 mA cm À2 . After the synthesis of β-Mo 2 C NPs on N-CW, the overpotential was significantly reduced. Mo 2 C/N-CW-20 demonstrated greater HER activity with a low overpotential of 110 mV compared to Mo 2 C/N-CW-10 (159 mV) and Mo 2 C/N-CW-5 (167 mV). As expected, the overpotential decreased as the concentration of the (NH 4 ) 2 MoO 4 precursor increased. After the NH 3 plasma treatment of Mo 2 C/N-CW-20, the catalytic activity of Mo 2 N-Mo 2 C/N-CW substantially improved. Mo 2 N-Mo 2 C/N-CW demonstrated the highest HER activity with the lowest overpotential of 79 mV compared with the other samples. Moreover, Mo 2 N-Mo 2 C/N-CW only required an overpotential of 311 mV to reach a high current density of 500 mA cm À2 , which was lower than that required by Pt (η = 410 mV). [42] The improved catalytic performance of Mo 2 N-Mo 2 C/N-CW at a high current density is expected to be beneficial for large-scale water electrolysis systems.
Notably, the surface wettability of an electrocatalyst has significant influences on electrolyte wetting, redox electron transfer, and gas release in electrochemical reactions. [43][44][45] Thus, the wettabilities of Mo 2 C/N-CW and Mo 2 N-Mo 2 C/N-CW were tested by dropping water droplets (100 μL) on their surfaces for equal periods of time ( Figure S5, Supporting Information). The initial contact angle of Mo 2 C/N-CW was 72°; in contrast, Mo 2 N-Mo 2 C/N-CW demonstrated improved hydrophilicity, with an initial contact angle smaller than 10°. Subsequently, the water droplets completely spread and entered both samples within 30 s. Consequently, the improved hydrophilicity produced by the NH 3 plasma treatment could enhance the charge transfer rate between the electrocatalyst and electrolyte. [45] In addition, the aligned microchannels with the hydrophilic surface of Mo 2 N-Mo 2 C/N-CW provided a pathway for rapid H 2 bubble release.
To elucidate the HER mechanism of the as-prepared electrocatalysts, the Tafel slopes were obtained by fitting the linear regions of the Tafel plots. The corresponding Tafel plots are displayed in Figure 4B. The Tafel slopes of Mo 2 C/N-CW decreased to 129, 108, and 94 mV dec À1 as the (NH 4 ) 2 MoO 4 precursor concentration increased to 5, 10, and 20 wt%, respectively, which is consistent with the overpotential trend. Mo 2 N-Mo 2 C/N-CW demonstrated a Tafel slope of 86 mV dec À1 , which was lower than that of Mo 2 C/N-CW, indicating that it exhibited the fastest HER kinetics. The catalytic HER of Mo 2 N-Mo 2 C/N-CW followed the Volmer-Heyrovsky mechanism, and the rate-limiting step was the electrochemical desorption of hydrogen. [46] In addition, the exchange current density ( j 0 ) values of all the samples were obtained from the zero-overpotential points of the Tafel plots ( Figure S6, Supporting Information). The j 0 value of Mo 2 N-Mo 2 C/N-CW was 0.309 mA cm À2 , which was higher Note that the double-layer capacitance (C dl ) is proportional to the electrochemical active surface area (ECSA) and is an indicator that directly explains the activity of a catalyst in the HER. [47] To calculate the C dl values of Mo 2 N-Mo 2 C/N-CW and Mo 2 C/N-CW, cyclic voltammetry (CV) analysis was conducted at different scan rates in the non-faradaic region ( Figure S7, Supporting Information). [48] As indicated in Figure 4c, Mo 2 N-Mo 2 C/N-CW exhibited a C dl value of 95.8 mF cm À2 , which was higher than that of Mo 2 C/N-CW. This high C dl value indicated that the Mo 2 N-Mo 2 C heterostructure promoted an increase in the number of exposed catalytic active sites compared to Mo 2 C, contributing to the enhanced HER performance. To gain insights into the reaction kinetics of the as-prepared electrodes, electrochemical impedance spectroscopy (EIS) measurements were performed from 0.01 to 10 6 Hz at η versus reversible hydrogen  electrode (RHE). The resulting Nyquist plots and corresponding equivalent circuit models are depicted in Figure 4d. Here, the diameter of the semicircle reflects the charge transfer resistance (R ct ) at the electrode/electrolyte interface. In Mo 2 C/N-CW, R ct decreased as the precursor concentration increased. Out of all the measured electrodes, Mo 2 N-Mo 2 C/N-CW exhibited the lowest R ct value of 6.5 Ω. This low R ct value suggested that the strong interaction of Mo 2 N with Mo 2 C enhanced the charge transfer ability, thus enabling superior HER activity. Further, electrochemical stability and durability are important criteria for evaluating electrocatalysts to be used in the HER.
To determine the stability, the polarization curves at the beginning and after 1000 CV cycles were compared. After 1000 CV cycles, the polarization curves of Mo 2 N-Mo 2 C/N-CW and Mo 2 C/N-CW-20 almost overlapped, with shifts of only 3.7 and 0.6 mV, respectively, demonstrating outstanding stability (inset in Figure 4e). Chronoamperometric tests were performed at a fixed potential that could have a high current density region of 110 mA cm À2 for 20 h in 0.5 M H 2 SO 4 (Figure 4e). The corresponding results for Mo 2 N-Mo 2 C/N-CW and Mo 2 C/N-CW initially decreased slightly and then stabilized over time, primarily owing to the strong adsorption of bubbles blocking the active sites. [49] The morphology, elemental composition, and phase stability of Mo 2 N-Mo 2 C/N-CW were determined via SEM/EDS and XRD after the chronoamperometric test for 20 h ( Figure S8 and S9, Supporting Information). The morphology and structure of Mo 2 N-Mo 2 C/N-CW revealed no apparent changes and maintained a 3D interconnected open-cell architecture and microchannels. The corresponding elemental mapping images indicated that Mo 2 C and Mo 2 N were uniformly well dispersed on N-CW. Moreover, the XRD results showed no difference in peaks after 20 h, and Mo, MoO x C y peaks also were not discovered, indicating the phase stability of Mo 2 N-Mo 2 C/N-CW. Consequently, Mo 2 N-Mo 2 C/N-CW demonstrated high stability and long-term durability as an HER electrode in a highcurrent-density environment. Additionally, the η 10 values of Mo 2 N-Mo 2 C/N-CW and reported CW-or Mo-based electrocatalysts in 0.5 M H 2 SO 4 were compared (Figure 4f and Table S1, Supporting Information). Mo 2 N-Mo 2 C/N-CW presented a relatively low η 10 value compared to the other electrocatalysts, confirming its excellent HER performance. DFT calculations were performed to further elucidate the enhanced catalytic activity of Mo 2 N-Mo 2 C/N-CW. The three structural models depicted in Figure S10, Supporting Information, were γ-Mo 2 N-β-Mo 2 C (denoted as Mo 2 N-Mo 2 C), β-Mo 2 C, and γ-Mo 2 N, which were constructed based on the XRD and HRTEM results. To improve the reliability of the calculation results, various sites where H* could be adsorbed on the surface of the structural models were investigated and grouped into sites 1 and 2 based on two perpendicular planes. Figure 5a depicts the structural model of Mo 2 N-Mo 2 C with H* adsorbable sites at the heterointerface. Figure 5b,c presents side views of heterointerface hydrogen adsorption sites 1 and 2, respectively. The Mo 2 N-Mo 2 C heterointerface was composed of the (110) plane (site 1) and (112) plane (site 2), and the hydrogen adsorption Gibbs free energy (ΔG H* ) values for the Mo, C, and N sites capable of H* adsorption in each plane were calculated ( Figure S11 and S12, Supporting Information). Moreover, the ΔG H* values for the hydrogen adsorption of Mo and C sites in the (111) (site 1) and (112) (site 2) planes of β-Mo 2 C and the Mo and N sites in the (111) (site 1) and (112) (site 2) planes of γ-Mo 2 N were also calculated ( Figure S13 and S14, Supporting Information). Note that ΔG H* is an evaluation factor representing the HER catalytic activity. The optimal catalyst has a ΔG H* value close to zero, which achieves fast HER kinetics by balancing the H* adsorption and desorption on the surface. [50] Figure 5d and Table S2, Supporting Information, indicate the calculated ΔG H* values for H* adsorption at the Mo, C, and N sites in the structural models, and Figure 5e presents the final ΔG H* values for Mo 2 N-Mo 2 C, β-Mo 2 C, and γ-Mo 2 N. The C sites of β-Mo 2 C and N sites of γ-Mo 2 N presented |ΔG H* | values of 0.390 and 0.250 eV, respectively, which were lower than those for the Mo sites. The N site of Mo 2 N-Mo 2 C exhibited the lowest |ΔG H* | value of 0.072 eV compared with those of β-Mo 2 C and γ-Mo 2 N, and the sites at the heterointerface demonstrated the lowest |ΔG H* | values even when compared with the Mo, N, and C sites (Figure 5d). In addition, we discovered that site 2 had a lower |ΔG H* | value than site 1 at all sites except for the Mo site of β-Mo 2 C, indicating that the investigated catalysts demonstrated higher HER activity in the (112) plane. The near-zero ΔG H* value of Mo 2 N-Mo 2 C represented an outstanding HER catalytic performance, indicating that the electronic interaction of Mo 2 N and Mo 2 C at the heterointerface modulated ΔG H* to facilitate both the adsorption and desorption of hydrogen, which was consistent with the electrochemical test results.

Conclusions
In summary, we developed a scalable synthesis method for heterostructured Mo 2 N-Mo 2 C NPs on N-CW, as a self-supported electrode, based on earth-abundant elements. Mo 2 N-Mo 2 C/N-CW exhibited not only an outstanding HER performance with low overpotentials of 79 and 311 mV to achieve 10 and 500 mA cm À2 , respectively, but also robust stability and durability at a high current density. DFT calculations of hydrogen adsorption at various sites revealed that the heterointerface of Mo 2 N and Mo 2 C contributed to the enhanced HER performance and exhibited a ΔG H* value that was close to zero. The impressive HER performance of Mo 2 N-Mo 2 C/N-CW could be attributed to the following three aspects. 1) The coupling of Mo 2 N and Mo 2 C enhanced the catalytic activity compared to a single catalyst, enabled fast charge transfer kinetics, and created several heterointerfaces that acted as active sites.
2) The N-doped CW not only exhibited better conductivity and charge transfer but also acted as a secondary active site.
3) Hydrophilic-aligned microchannels could facilitate rapid electrolyte permeation and H 2 bubble release during the HER. These strategies and results could be widely used to design eco-friendly, earth-abundant functional electrodes for electrochemical energy conversion or storage-related applications.

Experimental Section
Materials: Ammonium molybdate ((NH 4 ) 2 MoO 4 , 99.98%) and sulfuric acid (H 2 SO 4 ) were obtained from Sigma-Aldrich. Balsa wood was obtained from Midwest Products (USA). The balsa wood was cut into 20 Â 10 Â 2 mm 3 pieces. All the materials were used as received without further purification. The water used in all the experiments was purified using a Millipore system.
Synthesis of Mo 2 C/N-CW: Balsa wood slices with a thickness of 2 mm along the growth direction were immersed in aqueous solutions of (NH 4 ) 2 MoO 4 at different concentrations (5, 10, and 20 wt%). The concentration of the aqueous solution was expressed as the weight ratio of sliced wood and (NH 4 ) 2 MoO 4 , and 5, 10, and 20 wt% of samples used 1, 2, and 4 mg mL À1 of (NH 4 ) 2 MoO 4 solutions, respectively. After gradual stirring, the wood slices were placed in a vacuum desiccator at room temperature for 1 h. The resulting (NH 4 ) 2 MoO 4 /wood slice samples were dried overnight in an oven (80°C). The wood slices were pre-carbonized in an Ar atmosphere at 260°C for 6 h and further carbonized at 900°C for 6 h under an Ar atmosphere with a heating rate of 5°C·min À1 in a quartz tube furnace to form both Mo 2 C and N-CW. The corresponding sample labels for different concentrations of 5, 10, and 20 wt% were Mo 2 C/N-CW-5, Mo 2 C/ N-CW-10, and Mo 2 C/N-CW-20, respectively. Synthesis of Mo 2 N-Mo 2 C/N-CW: The as-obtained Mo 2 C/N-CW-20 was placed inside a plasma-enhanced chemical vapor deposition system (SCS-5000) manufactured by Snteck Co., Ltd and equipped with a hybrid plasma source and magnetron radio frequency (RF) sputtering system. In the experiment, 100 sccm Ar was used as the carrier gas to achieve a stable supply of NH 3 gas. The RF power, gas pressure, and substrate temperature in the chamber were 300 W, 950 mTorr, and 300°C, respectively. Immediately after plasma ignition, Mo 2 C/N-CW was irradiated by RF plasma under NH 3 flow (200 sccm), as a reactive gas, for 10 min. After the NH 3 plasma treatment, the sample was allowed to cool to room temperature naturally in the chamber. Finally, the opposite side of the plasma-treated Mo 2 C/N-CW was also treated with NH 3 plasma in the same manner to obtain the Mo 2 N-Mo 2 C/N-CW sample used for the experiment.
Material Characterizations: The morphologies and microstructures of all the samples were characterized using field-emission SEM (JEOL, JSM-7100 F) and HRTEM (Thermo Fisher Scientific, Talos F200X). The different sp 2 structures of CW were recorded using a WITec Alpha 300 Series equipped with an Nd:YVO 4 green laser with a wavelength of 532 nm. XRD measurements were performed using an X-ray diffractometer (SMARTLAB, Rigaku) equipped with a conventional Cu Kα X-ray radiation source. The XPS analysis was performed on a PHI5000 VersaProbe II (ULVAC-PHI) instrument using Al-Kα radiation. The water contact angles of all the samples were measured using a Data Physics Model OCA 15EC. The water contact angle measurements were averaged over five different positions on the surface.
Electrochemical Measurements: All the electrochemical experiments were conducted using a standard three-electrode system on a VersaSTAT3 electrochemical workstation (AMETECK, Princeton Applied Research, USA) with 0.5 M H 2 SO 4 as the electrolyte at room temperature. A graphite rod and saturated calomel electrode (SCE) were used as the counter and reference electrodes, respectively. The prepared Mo 2 N-Mo 2 C/N-CW electrode was used as the working electrode. All the potentials were calibrated to an RHE according to the Nernst equation: E (RHE) = E (SCE)þ 0.241 V þ 0.059 pH. Polarization curves were obtained using LSV at a scan rate of 10 mV s À1 with iR compensation. The C dl of each sample was derived by an integration of the current density using the simple CV method in a potential range of 0.2-0.4 V vs the RHE in the non-Faradaic region under scan rates from 20 to 100 mV s À1 . The EIS analysis was conducted from 10 À2 to 10 6 Hz at an overpotential (vs RHE) for the HER with an amplitude of 5 mV. The stability was determined by accelerated degradation testing using continuous CV sweeps from 0.0 to À0.4 V (vs RHE, in 0.5 M H 2 SO 4 ) at a scan rate of 100 mV s À1 . An electrode durability test was performed using chronoamperometry at a fixed potential.
DFT Computation Details: DFT calculations were performed using the generalized gradient approximation (GGA) with Perdew-Burke-Ernzerhof (PBE) parameterization and Quantum espresso (QE) from Materials square (a web-based DFT calculation platform). For bulk calculations of β-Mo 2 C and γ-Mo 2 N, the primitive unit cell was used. 7 Â 7 Â 7 k-point grids and a 60 Ry energy cutoff of the wave function were used, which ensured electronic and ionic convergence. For the structural calculation of the Mo 2 N-Mo 2 C interface, the interfacial model (144 atoms) which of the distance between periodic interfaces was more than 10 Å was made. This interfacial model was relaxed on 3 Â 2 Â 1 k-point grids and a 60 Ry energy cutoff of the wave function, which ensured electronic and ionic convergence. The convergence criteria of the structural relaxation and electronic self-consistency for energy and forces set were chosen as 10 À8 Ry and 0.000038 Ry bohr À1 , respectively.
For calculating the absorption of hydrogen, an 20 Å thick vacuum layer was placed within the periodic cells repeated in the z-axis to obviate interactions. 2 Â 2 Â 1 k-points mesh was used for β-Mo 2 C and γ-Mo 2 N and 1 Â 1 Â 1 k-points mesh for Mo 2 N-Mo 2 C interface, and a 60 Ry energy cutoff of the wave function. The van der Waals interactions using the DFT-D3 (BJ) method was considered. All calculations were spin-polarized. The convergence criteria of the structural relaxation and electronic self-consistency for energy and forces set were chosen as 10 À6 Ry and 0.00038 Ry bohr À1 , respectively.
The adsorption energy of H atom (ΔE H ) was calculated by where ΔE ab H is the adsorbing energy of the H atom on surfaces, ΔE slab is the energy of the clean surface, and ΔE H 2 is the energy of the gas-phase hydrogen. Those parameters were calculated using DFT calculations. The Gibbs free energy of H adsorption was obtained by where ΔE ZPE is the zero-point energy and ΔS H is the entropy. ΔE ZPE À TΔS H is 0.24 eV under the standard conditions at room temperature. [51] Therefore, the ΔG H was calculated by

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