Efficient Hydrogen Evolution Reaction in 2H-MoS₂ Basal Planes Enhanced by Surface Electron Accumulation

Abstract

An innovative strategy has been developed to activate the basal planes in molybdenum disulfide (MoS₂) to improve their electrocatalytic activity by controlling surface electron accumulation (SEA) through aging, annealing, and nitrogen-plasma treatments. The optimal hydrogen evolution reaction (HER) performance was obtained on the surface treated with nitrogen-plasma for 120 s. An overpotential of 0.20 V and a Tafel slope of 120 mV dec⁻¹ were achieved for the optimized condition. The angle-resolved photoemission spectroscopy measurement confirmed the HER efficiency enhanced by the SEA conjugated with the sulfur vacancy active sites in the MoS₂ basal planes. This study provides new insight into optimizing MoS₂ catalysts for energy applications.

1. Introduction

Hydrogen (H₂) holds great attention as a green energy carrier for future technologies due to its high energy density and being environment friendly [1,2]. Electrochemical water-splitting is a promising route for sustainable H₂ production due to its advantages of abundant sources and non-pollutants [2,3]. Hydrogen evolution reaction (HER) is one of the most efficient pathways to produce H₂ from water splitting using suitable catalysts. Currently, platinum (Pt) and other noble metals or alloys are the most popular electrocatalysts for HER, while their scarcity and high cost limit their widespread utilization [4,5]. The new challenge for researchers is to develop earth-abundant, low-cost, and high-performance catalysts as an alternative to Pt and other noble metals for electrochemical H₂ production.

Two dimensional (2D) transition metal dichalcogenides (TMDCs) have been extensively investigated as promising alternative catalysts for HERs due to their exceptional physicochemical properties including high mechanical properties, good semiconducting ability, large surface area, and high catalytic activity [6,7,8,9,10,11,12]. Among TMDCs, molybdenum disulfide (MoS₂) has been found to be one of the most economical alternatives to noble metal catalysts for HER due to its high electron mobility, stability, non-toxic nature, low cost, and excellent electrocatalytic properties [13,14,15]. Due to its inert basal plane, the overall activity of the two-layered hexagonal phase MoS₂ (2H-phase MoS₂) catalyzes mainly from the active edge sites, which limits the practical application of MoS₂ for HER [9,16,17,18]. Various techniques have been developed to overcome the limited catalytic activity of the MoS₂ basal plane, such as interface electronic coupling, phase engineering, and introducing active unsaturated defects and strain [19,20,21,22,23,24]. The pioneer strategy to activate the basal plane of MoS₂ that enhances HER activity was the introduction of sulfur (S)-vacancies into the basal plane [22,25,26].

In HER, the optimal H₂ adsorption free energy (ΔG_H) is 0 eV, where H₂ is bound to the catalyst neither too strongly nor too weakly. For the basal plane of pristine 2H-MoS₂, the ΔG_H value is approximately 2 eV, making the basal plane inert to HER performance. The ΔG_H values can be decreased with an increase in S-vacancies on the MoS₂ basal plane. A ΔG_H value between ±0.08 eV for S-vacancies in the range of 9–19% is the most favorable value for HER [22,27,28,29]. To date, a limited number of techniques have been used to create S-vacancies on 2D TMDC materials such as electrochemical reduction, H₂ annealing, and plasma bombardment [29,30,31]. According to our previous work, long term air exposure (aging) spontaneously creates S-vacancies, which cause surface electron accumulation (SEA) on MoS₂ surfaces [32].

In this report, we utilize and control SEA to activate the basal planes of 2H-MoS₂. The SEA was performed through various techniques such as aging, annealing, and nitrogen (N₂)-plasma treatments to create S-vacancies on the basal plane of 2H-MoS₂. The effect of SEA on HER efficiency was investigated. Angle-resolved photoemission spectroscopy (ARPES) measurements were performed to confirm the electronic structures of different MoS₂ surfaces. The results revealed that 2H-MoS₂ was converted from intrinsic (fresh) to degenerate semiconductor (using N₂-plasma treatments) by creating S-vacancies which provided more active sites and charge carriers for HER.

2. Results and Discussion

2.1. Structural Characterization

Figure 1a illustrates the crystal structure of the MoS₂ monolayer formed by a hexagonal plane of Mo and S atoms. These triple planes are stacked together by weak van der Waals forces of attraction to form bulk MoS₂. The van der Waals force allows the preparation of mono and a few layers of MoS₂ flakes through a simple mechanical exfoliation technique using dicing tape (see Figure 1b). The bulk fresh surface was created by the exfoliation of the top layers of pristine MoS₂ and the other pristine MoS₂ crystal was subjected to N₂-plasma treatments. The X-ray diffraction (XRD) pattern of pristine, fresh, and N₂-plasma-treated bulk samples is shown in Figure 2. The diffraction peaks, located at 14.3°, 29.0°, 44.1°, 60.1°, and 77.9° were assigned to (002), (004), (006), (008), and (0010) planes of standard 2H-MoS₂, respectively (JCPDS #872416) [32]. The diffraction peak along the (002) plane was the most prominent compared with the other peaks. Moreover, a broad reflection around 21° was observed for the fresh and pristine MoS₂ samples which may have been due to the glass substrate [33]. A reduction in the intensity of diffraction peaks of fresh and N₂-plasma MoS₂ was observed when compared with pristine MoS₂, and this may have been due to the absence of constructive interference from the bulk crystal planes [34]. The position of diffraction peaks did not show any shift, which indicated that there was no inner structural change in the MoS₂ samples after N₂-plasma treatment.

Figure 3 shows the Raman spectra of pristine, fresh, and N₂-plasma-treated MoS₂. The observed Raman modes at 385 and 411 cm⁻¹ belong to the $E_{2g}^1$ (in-plane) and $A_{1g}$ (out-of-plane) modes of bulk MoS₂, respectively [35,36]. The 2H-MoS₂ phase contains four Raman active vibrations, namely, $E_{2g}^1$, $A_{1g}$,$E_{1g}$ and $E_{2g}^2$, and the absence of the other two modes may have been due to the selection rules of the scattering geometry and limited rejection of Rayleigh scattered radiation [35,37,38]. After the exfoliation and plasma treatments, the positions of the observed two modes did not show any shift. However, the peak intensity of the N₂-plasma-treated sample showed considerable decrement when compared with the pristine MoS₂ sample. This indicated that the lattice distortion on the MoS₂ basal plane was due to the gradual increase in S-vacancies and cracks caused by the N₂-plasma treatment [31].

2.2. Electrochemical HER Efficiency Enhanced by SEA

2.2.1. SEA through Aging

To examine the effects of SEA on a MoS₂ electrocatalyst for a HER, the fresh surface was exposed to air for a different number of days to create SEA on the MoS₂ surface. Figure 4a shows the polarization curves of pristine and fresh MoS₂ surfaces with an overpotential of 0.75 and 0.82 V, respectively. The fresh MoS₂ used a high overpotential of 0.82 V (vs. RHE) to generate a 10 mA/cm² current density. The higher overpotential of fresh MoS₂ was due to the intrinsic (insulating) nature of the cleaved fresh surface, which was confirmed by ARPES and scanning tunneling spectroscopy (STS) measurements in our previous study [32]. The overpotential of the fresh surface was reduced by the aging effect, which created the S-vacancies. The aging effect of the fresh surface was examined over a different number of days, and their polarization curves are shown in Figure 4c. The overpotential reduced gradually with the increase in aging time and reached a minimum value of 0.31 V for a surface aging of 58 days. This reduction in the overpotential was evidence of the HER efficiency enhanced by SEA on the MoS₂ basal plane due to the aging effect. Generally, the exposure of the sulfide surface to air causes desulfurization (escape of sulfur atoms) and adsorption of foreign molecules such as oxygen and water molecules. The ARPES results of an in situ-cleaved surface and the same surface kept for 11 h excluded the possibility of the SEA being induced by the adsorption of foreign molecules, and it was concluded that the SEA was induced by long-term air exposure originating from the S-vacancy surface defects [32]. The S-defects created a donor-like surface state close to the conduction band edge which increased the electron concentration at the MoS₂ surface.

Tafel slope is the linear fitting region in the Tafel plot according to the Tafel equation of η = b log |j| + a, where η is the overpotential, b is the Tafel slope, and a is the current density [39,40]. The Tafel plots of pristine and fresh MoS₂ surfaces are presented in Figure 4b, with corresponding Tafel slopes of 248 and 253 mV dec⁻¹, respectively. The Tafel plots of aging surfaces are illustrated in Figure 4d, and it was noted that the Tafel slope reduced with an increase in aging time. Finally, the lowest Tafel slope of 148 mV dec⁻¹ was obtained for the MoS₂ surface aged for 58 days. Generally, a lower Tafel slope requires less voltage to increase the current density in the HER mechanism by one order of magnitude. Hence, among these, MoS₂ aged for 58 days was the most favorable catalyst for HER. Normally, HER activity depends on the number of active sites on, and electron density of, the catalyst. The major active sites for the electrocatalytic reaction were formed mainly by S-vacancies which were created on the MoS₂ surface due to the spontaneous desulfurization in air [32].

2.2.2. SEA through Annealing

To control the SEA and optimize HER efficiency, the fresh surface was annealed at 80 °C for different times (1, 2, and 3 h). The polarization curves of fresh and annealed MoS₂ surfaces for different times are shown in Figure 5a and corresponding Tafel plots can be seen in Figure 5b. The overpotential of MoS₂ decreased from 0.82 to 0.28 V when the fresh MoS₂ surface was exposed to annealing at 80 °C for 2 h, and the Tafel slope attained the minimum value of 168 mV dec⁻¹. Further, on increasing the heat treatment to 3 h, the overpotential and Tafel slope increased from 0.28 to 0.32 V and 168 to 199 mV dec⁻¹, respectively. This could have been due to the increase in S-vacancies over the optimum when the fresh MoS₂ surface was exposed to heat for more than 2 h. The catalytic activity depends on the density of S-vacancies. Li et al. observed maximal catalytic activity for MoS₂ films when the density of S-vacancies was in the range of 7−10%. Further, an increase in the density of S-vacancies reduced the catalytic activity [41]. Density functional theory studies have revealed that the optimal HER activity of MoS₂ occurs below the S-vacancy concentration of 12.5% [42]. The S-vacancies will disperse homogeneously like point defects up to a vacancy concentration of 12.5%, and with further increases in vacancy concentration (up to 18.75%) some of the vacancies will form agglomerate clusters. At higher vacancy concentrations, the combination of isolated point defects and clustered defects may induce structural defects in MoS₂. The relatively strong hydrogen adsorption (ΔG_H = −0.278 and −0.290 eV) at higher vacancy concentrations (15.63 and 18.75%) shows very high activity for proton adsorption, but desorption of *H to form H₂ would be kinetically more difficult. Hence, the optimal hydrogen adsorption (ΔG_H = ±0.15 eV) occurs below the S-vacancy concentration of 12.5%, which corresponds to the highly active sites for hydrogen evolution.

2.2.3. SEA through N₂-Plasma Treatment

Further, N₂-plasma treatment was conducted on pristine MoS₂ surfaces over different times to enhance HER activity. The polarization curves and corresponding Tafel plots of pristine and N₂-plasma-treated MoS₂ surfaces over different times are shown in Figure 5c and Figure 5d, respectively. During plasma treatment, the pristine surface was kept 1 cm away from the ICP N₂ plasma generator to avoid rapid S-vacancies and surface damage. The overpotential of the MoS₂ surface reduced from 0.75 to 0.20 V on N₂-plasma exposure for 120 s. Further increases in N₂-plasma exposure time resulted in increases in overpotential. Thus, it was concluded that the optimal N₂-plasma treatment time for better HER activity was 120 s. Generally, the Tafel slopes of 120, 40, and 30 mVdec⁻¹ are an indication of the Volmer, Heyrovsky, and Tafel reaction steps in the HER mechanism, respectively [43,44]. The optimal N₂-plasma treatment for 120 s exhibited the Tafel slope value of 120 mV dec⁻¹ (see Figure 5d) and hence, it followed the Volmer mechanism of HER.

2.3. Evidence of Enhanced SEA Observed by ARPES

ARPES measurements were performed for the in situ-cleaved, annealed for 2 h, and N₂-plasma treatment for 120 s -MoS₂ surfaces to confirm the existence of SEA as shown in Figure 6a. The magnified normal spectra at Γ point can be seen in Figure 6b. The valence band edge (E_V) of the in situ-cleaved surface was at a binding energy of −0.80 eV which is the indication of intrinsic Fermi level (E_F) according to the bulk MoS₂ band gap of $>$1.3 eV at 85 K [32]. The difference between E_F and E_V (E_F − E_V$=$0.80 eV) of the in situ-cleaved surface was in good agreement with the reported value of 0.80 eV for in situ exfoliated MoS₂ under UHV conditions [45]. The SEA of aging surfaces was confirmed by ARPES in our previous work [32]. The E_V of the MoS₂ surface annealed for 2 h resulted in red shift and moved to a binding energy of −1.16 eV. This moved E_F to near the conduction band edge (E_C) (E_C − E_F$=$0.19 eV). The carrier concentration (n) of the annealing surface can be calculated using the following formula [46]:

$$n=N_C\exp \left[\frac{-(E_C-E_F)}{kT}\right]$$

(1)

where $N_{C }$ is the effective density of states function in the conduction band and is given by $N_C=2{\left(2\pi m_e^{*}kT/h^2\right)}^{\raisebox{1ex}{$3$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}$. Here, $m_e^{*}$ is the effective mass of an electron, k is Boltzmann’s constant, T is the room temperature, and h is Planck’s constant. The values of $m_e^{*}$ for bulk MoS₂ are in the range of 0.45−0.73 m₀; here m₀ is the free electron rest mass [47]. The calculated n of annealing surfaces is 3.4 $\times$ 10¹⁶–7.0 $\times$ 10¹⁶ cm⁻³.

The E_V showed a remarkable red shift and moved to −1.56 eV for the N₂-plasma-treated MoS₂ surface. This moved the E_F to beyond E_C (E_F − E_C$=$ 0.26 eV), which showed the n-type degenerate semiconductor nature. n can be calculated using the following formula [48]:

$$E_F-E_c=kT\left[ln\left(\frac{n}{N_c}\right)+2^{-3/2}\left(\frac{n}{N_c}\right)\right]$$

(2)

The calculated n value was in the range of 1.50 $\times$ 10²⁰–3.10 $\times$ 10²⁰ cm⁻³. This value was around five orders of magnitude higher than the MoS₂ bulk value (n $=$ 2 $\times$ 10¹⁵ cm⁻³) [49]. The higher n on the N₂-plasma-treated MoS₂ surface denotes the highly conductive nature of the basal plane. This result was consistent with the electrochemical HER results in which the lowest overpotential was obtained for the N₂-plasma treatment for 120 s on the MoS₂ surface. Hence, it was confirmed that HER activity can be enhanced by producing conjugated active sites and SEA in the 2H-MoS₂ basal plane.

The best HER activity of N₂-plasma-treated MoS₂ with an overpotential of 0.20 V was compared with other MoS₂-based catalysts including nanoflakes [50], thin films [51,52,53], hierarchical hollow architectures [54], homostructures [55], and hybrid nanostructures [56,57], and is tabulated in

Table 1. The comparison of MoS₂ bulk overpotential with reported MoS₂ nanostructures.

S. No	System	Overpotential at 10 mA/cm2 (V)	Reference
1.	MoS2 bulk (N2-plasma-treated)	0.20	This work
2.	1T-MoS2 ultra-thin flakes	0.25	[50]
3.	MoS2 thin films	0.38	[51]
4.	MoS2 thin films (ozone treated for 10 min)	0.36	[52]
5.	MoS2 thin films	0.45	[53]
6.	h-rGO@MoS2	0.23	[54]
7.	MoS2 nanosheet/MoS2 nanoflake	0.26	[55]
8.	MoS2/GO	0.21	[56]
9.	MoS2/rGO	0.30	[57]

. It was noticed that MoS₂ bulks performed better than MoS₂-based nanostructures and, hence, control of SEA can be a promising strategy to enhance HER activity in 2D TMDCs. In addition, the obtained overpotential of 0.20 V was higher than the overpotential of Ru doped 2H-MoS₂ (0.16 V at 10 mAcm⁻² in 0.5 M H₂SO₄) [58], and the MoS₂ nanosheet (0.07 V 10 mAcm⁻² in 0.5 M H₂SO₄ under back-gate voltage of 3V) [59]. This suggests that HER activity in MoS₂ basal planes can be improved further by a combination of SEA with doping or Fermi level modulation.

2.4. Discussion

The schematic of SEA and its enhancement on HER in 2H-MoS₂ is shown in Figure 7. The SEA was produced by the S-vacancies which act as donor-like surface states. The accumulated free electrons were injected from the donor-like surface states and leave the positively charged state on the surface which produced the downward band bending. This moved the E_C below E_F according to the ARPES measurement of the N₂-plasma-treated MoS₂ surface due to its n-type degenerate nature. S-vacancies are the origin of high surface electron concentration and, meanwhile, provide abundant active sites. The conjugated SEA and HER active sites provide the required conditions to enhance the electrochemical activity of HER.

Usually, in MoS₂, the metallic 1T-phase basal plane is considered for the HER due to the possible source of active sites [9,23]. The SEA in 2H-MoS₂ enhances HER activity and this result may be comparable with the excellent HER activity of the metallic 1T-phase MoS₂ [7,23,60]. In our previous reports, we observed SEA formation in MoS₂ and MoSe₂ due to the S- and Se-vacancies, respectively [32,48]. The MoS₂ catalyst has been widely investigated for HER but, unexpectedly, the effect of SEA on HER efficiency has been ignored in previous reports. This investigation gives a new insight into the tuning of basal plane active sites through S-vacancies by keeping the edge defects that can enhance HER activity. Moreover, the catalytic vacancy sites and abundant free electrons at the surface are the major areas responsible for efficient HER activity.

3. Experimental Section

3.1. Preparation of MoS₂ Layer Crystals

The chemical vapor transport method was used to synthesize MoS₂ single crystals. Fine powders of sulfur (99.99%) and molybdenum (99.99%) were used as source materials. Bromine (Br₂) was used as the transport agent and enabled an effective and faster vapor transport to produce MoS₂ single crystals. The source materials including sulfur and molybdenum powders together with Br₂ were sealed in a quartz ampoule of a length of 30 cm at a vacuum of 10⁻⁵ Torr. The inner and outer diameters of the quartz ampoule were 1.3 and 1.6 cm, respectively. The preheated material temperatures of growth MoS₂ single crystals were kept at 950 and 1000 °C, respectively. The controlling temperatures of the source and crystallization ends were maintained at 1050 and 960 °C, respectively. After ten days of processing, the as-synthesized MoS₂ layer crystals area was in the range of a few square millimeters to centimeters. The thickness of the crystals ranged from a few nanometers to hundreds of nanometers. Typical mechanical exfoliation was used for obtaining fresh surface MoS₂ from the pristine layer crystals using dicing tape.

3.2. Characterization of MoS₂ Bulks

An X-ray diffraction (XRD-Bruker D₂ phaser diffractometer) system with Cu K_α1 radiation (λ = 1.54056 Å) and Raman spectroscopy (Jobin-Yvon LabRAM HR800) with a 633 nm He-Ne laser as the excitation source were used to examine the structural characterization of the MoS₂ bulks. The pristine surface (non-fresh) meant that the surface of the as-synthesized crystal was exposed to air for a prolonged period. The fresh surface was obtained by stripping the top layers of the bulk crystal with dicing tape. The N₂-plasma treatment was performed on a different pristine MoS₂ surface.

3.3. N₂-Plasma Treatment of Pristine MoS₂

An inductively coupled plasma (ICP) system with a commercial 13.56 MHz RF source was used to perform the N₂-plasma treatment on the MoS₂/glassy carbon electrode. In the quartz tube, the distance between the samples and the center of the coil was maintained at 1 cm and the samples were placed on the downstream side. After that, the vacuum in the chamber was maintained at 10 mTorr. High purity N₂ (99.999%) was allowed into the quartz tube at a flow rate of 400 sccm. The N₂-plasma was generated under a pressure of 130 mTorr due to excited N₂ ions when powered by 100 W at ambient temperature. Finally, the pristine MoS₂ surfaces were exposed to a N₂-plasma atmosphere for fixed different time intervals.

3.4. Electrochemical Measurements

An electrochemical test system (Biologic Bi-stat) was used to perform the HER activity measurements using a standard three electrode cell at ambient temperature. The working electrode had a rotating disk electrode (RDE, PINE AFE5T050GC) with a glassy carbon (GC) disk (diameter 5 mm) and a PTFE shroud (diameter 15 mm). The platinum foil and Ag/AgCl were utilized as counter and reference electrodes, respectively. The measured potential values were transferred to a reversible hydrogen electrode (RHE), where E_RHE = E_Ag/AgCl + 0.059 pH + E°_Ag/AgCl. The HER polarization curves were recorded in a 0.5 M H₂SO₄ (pH ≈ 0.3) electrolyte at a scan rate of 5 mVs⁻¹. The high purity N₂ gas was passed through the electrolyte to remove other gases. As-prepared MoS₂ catalyst was loaded onto the surface of the GC electrode and thread sealant (Loctite^®) was used to cover the uncovered GC area beside the MoS₂ catalyst. For the pristine surface, the bulk MoS₂ crystal was placed on the surface of the GC electrode and fixed. For the fresh surface, first we fixed the pristine bulk on the GC electrode and removed the top surface using dicing tape; then for the annealing process, the electrode with the fresh surface was exposed to heating. For the N₂-plasma-treated samples, first we fixed the pristine bulk on the GC electrode and then exposed it to plasma. Cyclic voltammetry (CV) was performed in the potential range of −1.0 to +0.2 V (vs. RHE) and a scan rate of 50 mVs⁻¹ until the curves reached a stable state. Linear sweep voltammetry (LSV) was executed in the same potential range as CV with a scan rate of 5 mVs⁻¹. The electrochemical measurements were repeated three times to obtain concordant data.

3.5. Angle-Resolved Photoelectron Spectroscopy (ARPES) Measurement

The ARPES experiment was performed at the National Synchrotron Radiation Research Center (NSRRC) in Hsinchu, Taiwan, at a BL21B1 U9-CGM beamline. The photoemission spectra of MoS₂ annealed and N₂-plasma-treated surfaces were measured in a UHV chamber equipped with a hemispherical analyzer (Scienta R4000) with a collecting angle of $\pm$15$°$ at a base pressure of 5.6 $\times$ 10⁻¹¹ Torr. The polarization was invariably in the angular dispersive plane. The spectra of all MoS₂ surfaces were recorded with a photon energy of 42 eV at 85 K. The energy resolution was better than 23 meV and the angular resolution was 0.2$°$. After loading samples into a load–lock vacuum chamber, the samples were left in a vacuum until the pressure and temperature reached the required steady state.

4. Conclusions

We improved the electrocatalytic activity of 2H-MoS₂ by activating its basal planes through SEA using aging, annealing, and nitrogen-plasma treatments. The SEA caused by the S-vacancies was catalytically active, and a high electron concentration in the order of ~10²⁰ cm⁻³ was obtained for the N₂-plasma-treated MoS₂ surface. The optimal HER performance was obtained for the surface which underwent N₂-plasma treatment for 120 s, with an overpotential of 0.20 V vs RHE at 10 mA cm⁻². The ARPES measurements confirmed that the HER efficiency enhanced by the SEA conjugated with the S-vacancy active sites in the 2H-MoS₂ basal planes. This work provides an efficient and low cost MoS₂-based HER catalyst and also opens up comprehensive insights into basal planes of 2D TMDCs through SEA for energy related applications.