Enhanced hydrogen evolution reaction activity of hydrogen-annealed vertical MoS2 nanosheets

Molybdenum disulfide (MoS2) is a promising electrocatalyst for hydrogen evolution reaction (HER), but only edges and S-vacancies are catalytic active sites for the HER. Therefore, it is crucial to increase edge sites and S-vacancies for enhancing the HER activity of MoS2. Here, we report an enhanced HER activity of MoS2 by combing vertical nanosheets and H2 annealing. Compared to horizontal MoS2 nanosheets, pristine vertical MoS2 nanosheets showed better HER activity due to a larger amount of edges. H2 annealing further enhanced the HER activity of vertical MoS2 nanosheets remarkably. Scanning electron microscopy (SEM), X-ray photoelectron spectra (XPS) and electrochemical impedance spectroscopy (EIS) were used to elucidate the enhanced HER activity by H2 annealing. SEM images showed that H2 annealing roughened the MoS2 edges, leading to more edge sites. XPS data revealed the smaller S : Mo ratio after H2 annealing, meaning more S-vacancies. Meanwhile, EIS measurements showed that charge transfer was accelerated by H2 annealing. These findings elaborated the H2 annealing induced enhancement of the HER activity, which were further confirmed by the subsequent re-sulfurization experiment.


Introduction
Because of its high efficiency, environment-friendliness and renewability, hydrogen is expected to play an important role in superseding the carbon-based fossil fuels. Sustainable, costeffective and efficient production of hydrogen is a prerequisite for realizing hydrogen economy. 1,2 Electrochemical catalytic hydrogen evolution reaction (HER) in acidic media is an efficient method to generate hydrogen from water splitting. 2 Due to the presence of an overpotential, an electrolysis voltage is always needed, which leads to a waste of electric energy. Currently, there are no materials that could compare with platinum (Pt) in terms of the activity and stability. But its commercial application is greatly limited by the high cost and scarcity. 3,4 Therefore, it highly demands to develop other efficient non-noble-metal HER electrocatalysts with high abundance and low cost to make H 2 a competitive alternative energy source. 5,6 A signicant breakthrough was achieved when MoS 2 was introduced as a promising and highly stable electrocatalyst for the HER. [7][8][9][10][11][12][13][14][15][16][17][18][19] For decades, MoS 2 was believed to be inactive of the HER, because the basal plane exhibits a hydrogen adsorption free energy of 1.92 eV. 20,21 Recently, density functional theory (DFT) calculations on the Mo edge of MoS 2 revealed that at 50% hydrogen coverage, it possessed a hydrogen adsorption free energy of 0.08 eV, near the optimal value of 0 eV. 7 Soon aer, it was experimentally conrmed that the edges of MoS 2 are indeed the catalytic active sites for the HER. 9,10 These studies motivated the development of MoS 2 catalysts with a substantial fraction of further exposed edge sites, including hollow spheres, 9 edgeexposed lms, 22 amorphous lms, 23 defect-rich lms, 24 nanodots 25 and vertically aligned nanosheets. 8,24,[26][27][28] Therein, vertical MoS 2 nanosheets, which have vertical channels for ion penetration and intimate contact between the active nanosheets and the random substrate, were shown to remarkably increase the density of edge sites and hence considered as an ideal geometry for improving the HER performance of MoS 2 -based HER catalysts. 9,22,[29][30][31][32][33][34][35][36][37][38][39][40][41] Although extensive efforts have been made to increase the number of edge sites, the overall HER activity is still limited, as generally only a small fraction of edge sites contribute to the reaction rate. 10,42 Therefore, it is necessary and urgent to explore effective methods to increase the HER activity of inert basal plane sites in MoS 2 nanosheets. Fortunately, S-vacancies in the basal plane of MoS 2 nanosheets were recently explored and suggested to have a signicant impact on the HER activity. 10,[43][44][45] By Ar or oxygen plasma exposure, or H 2 treatment, S-vacancies were introduced into MoS 2 nanosheets, and remarkably enhanced the HER activity. 44,46 This method is only effective for at MoS 2 catalysts due to the Ar and oxygen plasma's directionality and thus unsuitable for vertical nanostructures and 3D nanostructures. 47 Meanwhile, conductivity was also identied as a crucial factor for the HER activity, because a high conductivity ensures a fast charge transfer in the HER interface. [48][49][50][51] Therefore, improving the conductivity while increasing active sites is the most promising but challenging task for optimizing the HER activity of MoS 2 nanosheets.
Herein, we report an enhanced HER activity of MoS 2 by combing vertical nanosheets and H 2 annealing. Vertical MoS 2 nanosheets were grown on glassy carbon by CVD method at 520 C. Pristine vertical MoS 2 nanosheets with a larger amount of edges showed better HER activity than horizontal ones. H 2 annealing further enhanced the HER activity of vertical MoS 2 nanosheets. Scanning electron microscopy (SEM), X-ray photoelectron spectra (XPS) and electrochemical impedance spectroscopy (EIS) were used to elucidate the enhanced HER activity by H 2 annealing. SEM images showed that H 2 annealing roughened the MoS 2 edges, leading to more edge sites. XPS data revealed the smaller S : Mo ratio aer H 2 annealing, meaning more S-vacancies. Meanwhile, EIS measurements showed that charge transfer was accelerated by H 2 annealing. These ndings elaborated the H 2 annealing induced enhancement of the HER activity, which were further conrmed by the subsequent resulfurization experiment.

Growth of vertical MoS 2 nanosheets by CVD method
Vertical MoS 2 nanosheets were synthesized by CVD method as shown in Fig. 1a. The vertical MoS 2 nanosheets samples were grown on glassy carbon inside a tubular furnace equipped with 22 mm diameter quartz tube. A silica boat loaded with MoO 3 powder (purity 99.95%) and glassy carbon substrate were placed one aer another in the center of hot zone inside the tube furnace, and the distance between MoO 3 powder and glassy carbon substrate was $2 cm. At the upstream of the tube, the other boat with sulfur powder (purity 99.5%) was placed outside the hot zone, which was mildly sublimated with heater band at $120 C. Before heating, the tube was pumped down to a base pressure of $0.1 Pa and ushed with Ar to guarantee a favorable growth atmosphere. The furnace was then heated from room temperature to 520 C at 50 C min À1 and kept at 520 C for 15 min with Ar ow of 10 sccm. Aer growth, the furnace was opened directly for rapid cooling to the room temperature with Ar ow of 200 sccm.

Hydrogen annealing of vertical MoS 2 nanosheets
Before the hydrogen annealing experiment, high quality vertical MoS 2 nanosheets sample on glassy carbon was selected by electrochemical measurements. Aer putting selected MoS 2 nanosheets sample into the quartz tube at center of the furnace, the temperature of the furnace was raised up to 780 C at 50 C min À1 with Ar ow of 50 sccm. Aer the temperature reached 780 C, Ar was immediately switched to H 2 with ow of 20 sccm, which is thought as the start point of the H 2 annealing. Aer a desired time (5,10,15,20,25,30 and 35 min, respectively), H 2 was immediately switched to Ar with ow of 50 sccm and the furnace was shut off, leaving to cool down naturally. The annealing was carried out at low pressure of $55 Pa.

Re-sulfurization of H 2 -annealed MoS 2 nanosheets
The re-sulfurization was conducted following a similar process of growth experiment, except for the absence of MoO 3 and longer treatment time. H 2 -annealed MoS 2 sample was placed in the center of the furnace and excessive sulfur powder was put in the center of the heater bend with temperature of $120 C. Before heating, the tube was pumped down to a base pressure of $0.1 Pa and ushed with Ar to guarantee a favorable growth atmosphere. The furnace was then heated from room temperature to 520 C at 50 C min À1 and kept at 520 C for 30 min with Ar ow of 10 sccm. Aer re-sulfurization, the furnace was opened directly for rapid cooling to the room temperature with Ar ow of 200 sccm.

Structural characterization
Surface morphology and structure were characterized by SEM and high resolution transmission electron microscopy (HRTEM, Tecnai TF20) techniques, respectively. SEM characterization was carried out on a Hitachi SU8000 instrument with an accelerating voltage of 15 kV and current of 10 mA. HRTEM samples were prepared by gently rubbing the TEM grid across the face of the MoS 2 thin lm to detach nanosheets and promote their adhesive on to the lacey-carbon TEM grid. Raman spectroscopy was measured using a B&W-Tek confocal Raman microscope with a laser excitation energy of 532 nm (2.33 eV). XPS were recorded on an ESCALAB MKII using an Al Ka excitation source.

Electrochemical measurements
The experiment of HER was carried out by measuring the current correlated with the water dissociation in the ambient conditions. All of the electrochemical measurements were performed in an electrochemical workstation CHI604B with a standard three-electrode conguration. The working electrode was the vertical MoS 2 nanosheets on glassy carbon. A graphite rod and an Ag/AgCl electrode were used as the counter electrode and the reference electrode, respectively. The performance of the HER was measured using linear sweep voltammetry from 0 to À0.6 V (vs. RHE) with scan rate of 5 mV s À1 at room temperature, which was conducted in 0.5 M H 2 SO 4 aqueous solution. All the potentials were referenced to reversible hydrogen electrode (RHE) by the following equation: An iR correction was normally employed to compensate for any potential loss arising from the external resistance of the electrochemical system. The electrolyte resistance and capacitance of the electrocatalysts were characterized by EIS. The ac impedance is measured at overpotential of À0.3 V (vs. RHE) within the frequency range from 0.1 to 10 6 Hz with a perturbation voltage amplitude of 10 mV. Cyclic voltammetry (CV) was conducted from 0 to À0.6 V (vs. RHE) at 100 mV s À1 to investigate the cycling stability.
Results and discussion Fig. 1a shows an illustration of the dual-temperature zone CVD synthesis of vertical MoS 2 nanosheets on glassy carbon. In the dual-temperature approach, heater band is isolated from the tube furnace. Therefore, the temperature of the heater band can be controlled independently from the temperature of the tube furnace. A 1 g portion of S powder was placed upstream of the gas ow in the quartz tube, right in the center of the heater band. 5 mg of MoO 3 powder and glassy carbon was put in the center of the tube furnace. The S powder was heated to 120 C in order to be evaporated by heater band. And the S vapor was carried downstream by Ar. At low pressure, S and MoO 3 vapor reacted and formed MoS 2 on glassy carbon at 520 C, which is much lower than previous study. 22,43 Aer the growth, the whole substrate turned violet from black (shown in Fig. 1b and c), indicating the formation of uniform MoS 2 nanostructures. The morphology and structure of MoS 2 were characterized by SEM and HRTEM, respectively. Fig. 1d shows the SEM image of vertical MoS 2 nanosheets on glassy carbon. It can be seen that the size of MoS 2 nanosheets is about 100-200 nm. From the HRTEM image in Fig. 1e, it can be seen that the nanosheets consisted of 4-5 layers with layer distance of 0.64 nm. Raman measurements were also performed to further conrm the presence of MoS 2 . As shown in Fig. 1f, there are two dominant Raman peaks at 382 and 408 cm À1 , respectively, which originate from the in-plane E 1 2g and out-of-planes A 1g vibrational modes of MoS 2 , respectively.
The electrocatalytic HER activities were investigated in 0.5 M H 2 SO 4 solution by linear sweep voltammetry (LSV) using a threeelectrode setup at a scan rate of 5 mV s À1 . An iR correction was normally employed to compensate for any potential loss arising from the external resistance of the electrochemical system. Fig. 2a shows the cathodic polarization curves of glassy carbon with and without vertical MoS 2 nanosheets, which are indicated by black and orange lines, respectively. It can be seen that vertical MoS 2 nanosheets exhibited great HER activity, compared to bare glassy carbon. A variety of methodologies have been proposed to evaluate the HER activity. Here overpotential at current density of 10 mA cm À2 , 42 onset overpotential and Tafel slope were used. The overpotential for vertical MoS 2 nanosheets at current density of 10 mA cm À2 is 485 mV, which is lower than horizontal MoS 2 nanosheets. 46,52 The improvement is due to more edges in the vertical MoS 2 nanosheets sample, and hence more active sites. 22 Fig. 2b shows the Tafel plot of pristine vertical MoS 2 nanosheets, indicated by black dotted lines. The Tafel slope is determined by tting the linear portion of the Tafel plot to the Tafel equation h ¼ b log |j| + a, where j is the current density, b is the Tafel slope. The value of 118 mV dec À1 was obtained from the pristine sample. The pristine MoS 2 sample exhibited the current density 1 mA cm À2 at the onset overpotential (301 mV), which was determined from Tafel plot as shown in Fig. S1. † To further enhance HER activity, pristine vertical MoS 2 nanosheets were annealed in H 2 atmosphere, which has been demonstrated to be catalytically effective for horizontal MoS 2 nanosheets. 46 But the mechanism was not clear. Before H 2 annealing, the cathodic polarization curves of each vertical MoS 2 nanosheets sample were measured to assure that all pristine samples had similar HER activity (the overpotentials at current density of 10 mA cm À2 are shown by black squares in Fig. 2c). In the H 2 annealing experiment, the selected MoS 2 samples were placed in the center of the tube furnace set at 780 C. For time-dependent H 2 annealing, the different pristine samples were treated for 5, 10, 15, 20, 25, 30 and 35 min, respectively. To precisely control the H 2 annealing time, 20 sccm of H 2 only owed when the temperature reached 780 C, while 50 sccm of Ar owed during the warming-up and coolingdown process. Fig. 2a shows the cathodic polarization curves of pristine and H 2 -annealed vertical MoS 2 nanosheets for 5, 10, 15, 20, 25, 30 and 35 min, respectively. It can be seen that all H 2 -annealed MoS 2 samples had a smaller onset overpotential and a smaller overpotential at current density of 10 mA cm À2 . Meanwhile, the onset potential and the overpotential at current density of 10 mA cm À2 have a relation to the annealing time. So, we plotted the overpotential at current density of 10 mA cm À2 as a function of annealing time (shown by red dots in Fig. 2c). It can be seen that, as the annealing time increased to 20 min, the overpotential decreased to 357 mV from $485 mV for pristine MoS 2 samples, which is reproducible (see Fig. S3 of ESI †). The overpotential was increased when further increasing the annealing time, but it was still smaller than that of pristine MoS 2 samples. The annealing time dependence of cathodic polarization curves also indicates good reproducibility of our method. Fig. 2b shows the Tafel plots of pristine and H 2 -annealed vertical MoS 2 nanosheets for 5, 10, 15, 20, 25, 30 and 35 min, respectively. It can be seen that, when annealing for 5-20 min, all samples exhibited similar Tafel slopes of $75 mV dec À1 , which are much lower than pristine samples with Tafel slopes of $118 mV dec À1 . But when further increasing the annealing time, the Tafel slopes were increased to $102 mV dec À1 . The HER results indicated that the HER activities of vertical MoS 2 nanosheets were enhanced aer H 2 annealing, and was maximized with annealing time of 20 min, which showed better performance than CVD-grown vertical MoS 2 nanosheets and H 2 annealed horizontal MoS 2 in previously reported. 22,46 To gure out the origin of the enhanced HER activity of vertical MoS 2 nanosheets by H 2 annealing, the morphologies of H 2 -annealed samples were characterized by SEM. Fig. 3a-g show the SEM images of H 2 -annealed vertical MoS 2 nanosheets samples for 5,10,15,20,25,30 and 35 min at 780 C, respectively. It can be seen in Fig. 3a that, when the annealing time is 5 min, only a few vertical MoS 2 nanosheets were damaged (indicated by yellow circles), while most ones were intact. When further increasing the annealing time, the damaged MoS 2 nanosheets were fragmentized into smaller nanosheets (indicated by yellow circles in Fig. 3b) and further formed aggregates (shown in Fig. 3c). Aer annealing for 20 min, almost all MoS 2 nanosheets were damaged (shown in Fig. 3d). When the annealing time was increased to 35 min, all nanosheets were melted and the sheet size was greatly reduced (shown in Fig. 3g). It is particularly noted in Fig. 3h that the edge of the nanosheets became rough, indicating that H 2 annealing increased the edges and hence active sites.
To fully elaborate the enhanced HER activity of vertical MoS 2 nanosheets by H 2 annealing, we also did the XPS characterization on the pristine and H 2 -annealed samples to explore the chemical states of the vertical MoS 2 nanosheets. Fig. 4 shows the XPS data of Mo 3d region for the pristine and H 2 -annealed vertical MoS 2 nanosheets for 5, 10, 15, 20, 25, 30 and 35 min, respectively. As shown in Fig. 4a, the spectrum of the pristine sample is dominated by a doublet with two sharp peaks, which  are Mo 3d 5/2 with binding energy of 229.7 eV and Mo 3d 3/2 with binding energy of 232.9 eV. The distance between the two peaks was $3.2 eV, which corresponds to the formation of MoS 2 . The peak located at 226.5 eV corresponds to S 2s in MoS 2 . 53 As shown in Fig. 4b, aer H 2 annealing for 5 min, a new doublet appeared in the Mo 3d region located at a lower binding energy to that of the original Mo 4+ doublet of the pristine sample. The two components of the new doublet at 228.1 and 231.5 eV, could be assigned to Mo 3d 5/2 and Mo 3d 3/2 of Mo 0 , respectively, as in elementary molybdenum, which was reduced from the pristine MoS 2 nanosheets by H 2 annealing. The appearance of elementary molybdenum indicates the loss of sulfur and hence the appearance of S-vacancies in H 2 -annealed MoS 2 samples. From the relative intensity between the new Mo 0 doublet and the original Mo 4+ doublet in Fig. 4b, the percentage of S-vacancies was determined to be 22.8%. As the annealing time was increased, the peak intensities of the new Mo 0 doublet were increased, and so did the relative intensities between the new Mo 0 doublet and the original Mo 4+ doublet. It is also determined that the percentage of S-vacancies were 29.2%, 39.2%, 41.5%, 50.0%, 52.4% and 61.2% for H 2 -annealing time of 10, 15, 20, 25, 30 and 35 min, respectively. As shown in Fig. 2, the maximal HER activity was found at H 2 annealing time of 20 min, corresponding to 41.5% of S-vacancies. Recent calculation showed that the optimal hydrogen adsorption free energy for HER occurred for an S-vacancy concentration that is between 12.5 and 15.62% of the surface atoms, 44,47 which is much lower than our experiment result. It is known that besides hydrogen adsorption free energy, conductivity is also a crucial factor in HER, which we think is responsible for the high percentage of Svacancies. Meanwhile, the intensity of the S 2s peak was gradually decreased without changing its shape and binding energy, which also suggested the production of S-vacancies in the H 2annealed MoS 2 sample.
To further conrm that it is the S-vacancy that contributed to the enhanced HER activity of H 2 -annealed vertical MoS 2 nanosheets, we re-sulfurized the H 2 -annealed MoS 2 samples to repair the S-vacancies and measured the catalytic activities aer the resulfurization. The re-sulfurization was achieved in the dualtemperature zone CVD with similar parameters to the synthesis of the vertical MoS 2 nanosheets, except for the absence of MoO 3 and longer treatment time. Fig. 5a shows the cathodic polarization curves of pristine and re-sulfurized MoS 2 samples which were previously H 2 -annealed for 5, 10, 15, 20, 25, 30 and 35 min, respectively. It can be seen that, compared to H 2annealed MoS 2 samples, both onset overpotential and overpotential at current density of 10 mA cm À2 were increased aer re-sulfurization, but smaller than pristine MoS 2 samples, which means HER activity were decreased aer re-sulfurization, but still better than pristine MoS 2 samples. Since the resulfurization lled all S-vacancies induced by H 2 annealing (see XPS data of re-sulfurized vertical MoS 2 nanosheets in Fig. S2 of ESI †), the decrease of the HER activity aer resulfurization could be attributed to the reduction of Svacancies. For the partial recovery of HER activities compared  to pristine MoS 2 samples, we conjecture that H 2 annealing induced rough MoS 2 edges played a role, which also could explain the dependence of the overpotential at current density of 10 mA cm À2 as on annealing time for re-sulfurized MoS 2 samples (indicated by blue triangles in Fig. 5c).
In order to evaluate the kinetics of charge transfer in these samples, EIS measurements were carried out in a traditional three-electrode system. The diameter of the semicircle is considered to be related to charge transfer at the interface of the HER. A smaller diameter corresponds to more efficient charge transfer and smaller conductivity. Fig. 6a shows the Nyquist plots of pristine, H 2 -annealed and re-sulfurized MoS 2 samples, respectively, which were collected by scanning from 0.1 to 10 6 Hz with an overpotential of 0.3 V. H 2 -annealed MoS 2 sample had a smaller semicircle diameter compared to pristine sample, indicating more efficient charge transfer aer H 2 annealing. This also contributes to the enhanced HER activity of H 2annealed MoS 2 samples. The efficient charge transfer could be attributed to the elementary Mo which has better conductivity than MoS 2 , and reduced size of H 2 -annealed MoS 2 nanosheets. This could explain the higher percentage of S-vacancies than previous calculation, at which the optimum HER activity was achieved. 44,47 Aer re-sulfurization, the semicircle diameter of MoS 2 samples became larger, meaning slower charge transfer than H 2 -annealed MoS 2 sample. This is also considered to be responsible for the worse HER activity than H 2 -annealed MoS 2 samples (shown in Fig. 5c). It is also found in Fig. 6a that the semicircle diameter of re-sulfurized MoS 2 sample was not completely recovered to that of pristine sample, indicating faster charge transfer than pristine sample, which resulted in the partial recovery of HER activity (shown in Fig. 5c). The faster charge transfer could be attributed to the reduced size of resulfurized MoS 2 nanosheets and hence the smaller resistance.
Long-term durability, which demonstrates the thermodynamic stability, is also an important factor to evaluate the overall performance of an electrocatalyst. We investigated the cycling stability of H 2 -annealed vertical MoS 2 nanosheets sample for 20 min at 780 C, which had the best electrocatalytic performance. The LSV scanning was performed before and aer repeating cycling voltammetry treatment for 10 000 cycles in an acidic environment with a fast scan rate of 100 mV s À1 in order to simulate the practical working conditions of water-splitting devices. Fig. 6b shows the cathodic polarization curves of the sample before and aer the cycling treatment. It can be seen that there is a negligible change in the cathodic current. Meanwhile, the time-dependent electrochemical measurement for the sample (inset of Fig. 6b) suggests that such electrocatalyst maintained its current density for at least 8 h. Both results indicated the excellent durability of the H 2 -annealed vertical MoS 2 nanosheets as an electrocatalyst.

Conclusions
We demonstrated the enhanced HER activity of MoS 2 nanosheets with vertical conguration and H 2 annealing. Vertical conguration offered more edge sites compared to horizontal MoS 2 nanosheets, and hence improved the HER activity. In addition, H 2 annealing produced a synergistic enhancement in two aspects. First, H 2 annealing roughened the edges and produced S-vacancies in the basal plane of vertical MoS 2 nanosheets, which increased the number of active sites. Second, H 2 annealing leads to the enhancement of the conductivity due to the production of elementary Mo and the decreased sheet size. Moreover, prominent electrochemical durability is also achieved. The H 2 -annealed vertical MoS 2 nanosheets reported here represents a novel approach for producing highly active nanostructured MoS 2 as HER catalysts, and can be extended to other applications, such as supercapacitor and lithium-ion battery.

Conflicts of interest
The authors declare no competing nancial interest.