Rational Design of Phase-Engineered WS2/WSe2 Heterostructures by Low-Temperature Plasma-Assisted Sulfurization and Selenization toward Enhanced HER Performance

Efficient hydrogen generation from water splitting underpins chemistry to realize hydrogen economy. The electrocatalytic activity can be effectively modified by two-dimensional (2D) heterostructures, which offer great flexibility. Furthermore, they are useful in enhancing the exposure of the active sites for the hydrogen evolution reaction. Although the 1T-metallic phase of the transition metal dichalcogenides (TMDs) is important for the hydrogen evolution reaction (HER) catalyst, its practical application has not yet been much utilized because of the lack of stability of the 1T phase. Here, we introduce a novel approach to create a 1T-WS2/1T-WSe2 heterostructure using a low-temperature plasma-assisted chemical vapor reaction (PACVR), namely plasma-assisted sulfurization and plasma-assisted selenization processes. This heterostructure exhibits superior electrocatalytic performance due to the presence of the metallic 1T phase and the beneficial synergistic effect at the interface, which is attributed to the transfer of electrons from the underlying WS2 layer to the overlying WSe2 layer. The WS2/WSe2 heterostructure catalyst demonstrates remarkable performance in the HER as evidenced by its small Tafel slope of 57 mV dec–1 and exceptional durability. The usage of plasma helps in replacing the top S atoms with Se atoms, and this ion bombardment also increases the roughness of the thin film, thus adding another factor to enhance the HER performance. This plasma-synthesized low-temperature metallic-phase heterostructure brings out a novel method for the discovery of other catalysts.


■ INTRODUCTION
−3 However, the major challenge in hydrogen production is the absence of cost-effective and efficient methods.−6 Nevertheless, a suitable catalyst is necessary to enhance the production of H 2 by reducing the thermodynamic barrier and accelerating the overall hydrogen evolution reaction.An ideal electrocatalyst should have a substantial surface area with abundant active sites, facilitating efficient adsorption and desorption.A low or near-zero free-energy adsorption value for hydrogen (ΔG H* ) is considered optimal.While platinum (Pt) is the most effective electrocatalyst, its scarcity and cost present challenges for mass production 7,8 Thus, there is a need to discover an alternative material with high abundance, low cost, and high catalytic performance to serve as a catalyst.
Recently, two-dimensional (2D) materials, specifically transition metal dichalcogenides (TMDs), 9−13 have brought enormous consideration because of their outstanding chemical, mechanical, electrical, and optoelectronic properties. 14,15The formation of large active areas in 2D layered structures leads to wide edges, which are the key active facets for better HER performance. 16,172D TMDs such as WS 2 , MoS 2 , MoSe 2 , WSe 2 , and so on, are recognized for their potential as effective electrocatalyst materials.−21 Recent research highlights 2D WS 2 as a promising platinum (Pt) alternative in electrocatalysis, comparable to MoS 2 .Notably, the high specific area of WS 2 provides abundant active sites, positioning it as a noteworthy candidate for catalytic applications.Its superior electrical conductivity, cost-effectiveness, and corrosion resistance further enhance its potential, suggesting competitiveness with Pt and other TMDs in catalytic efficiency. 22−26 Creating heterostructures is a key method to overcome WS 2 catalyst limitations by achieving electronic transformation, thereby enhancing efficiency in energy conversion and increasing active surface sites. 27Various WS 2 -based heterostructures, including MoS 2 , CoSe 2 , and WSe 2 , have been studied to improve the performance of WS 2 as an electrocatalyst for HER.For instance, Vikraman et al. synthesized the MoS 2 /WS 2 heterostructure using chemical bath deposition and sputtering, achieving an overpotential of 129 mV and a Tafel slope of 72 mV dec −1 . 28Another study by Hussain and co-workers reported the synthesis of a CoSe 2 /WS 2 heterostructure with an onset potential of 95 mV and a Tafel slope of 44 mV dec −1 . 29Recently, Sun et al. reported the one-pot solvothermal formation of WS 2 /WSe 2 heterostructures. 30The fabricated catalyst on a carbon fiber paper (CFP) exhibited good HER performance with a Tafel slope of 74.08 mV dec −1 .Here, WSe 2 was selected for heterostructure formation with WS 2 due to its ease of fabrication and superior electrochemical performance.−30 A recent technique, the Laser planting strategy, involves utilizing laser pulses to generate metal single atoms (SAs) on substrates, particularly for their application as electrocatalysts. 34However, these methods are still inconvenient and time-consuming and also require expensive precursors.CVD has drawbacks, including time-consuming synthesis at high temperatures (exceeding 600 °C), posing a risk of substrate damage.Moreover, the need for film transfer for analysis introduces inefficiencies, leading to a costly fabrication process with issues such as wrinkles and polymer residue on the surface.CVD often yields a stable 2H phase, which is semiconducting.In contrast, the 1T phase of TMDs demonstrates promising catalytic properties for the HER.In a study, Qu et al. investigated the HER activity of different phases of TMDs such as MoS 2 , MoSe 2 , WS 2 , and WSe 2 and found that the 1T phase of these materials possessed significantly higher HER activity than their 2H counterparts. 35his increased activity is attributed to the presence of active sites on the edges and defects of the 1T phase, which are not present in the 2H phase.Maintaining the metastable 1T phase of TMDs under ambient conditions for a prolonged period is challenging, since this phase can easily convert into the stable 2H phase, which is more frequently observed.As a result, the creation of 1T-TMD/1T-TMD heterostructures has not been reported so far.Therefore, there is a need to develop a facile and reliable method for designing an efficient HER electrocatalyst while comprehensively understanding and optimizing the HER activity of the 1T phase of TMDs.
We present the fabrication of a 1T-WS 2 /1T-WSe 2 heterostructure catalyst through plasma-assisted chemical vapor reactions, specifically plasma-assisted sulfurization and selenization.Inductively coupled plasma (ICP) in these processes lowers the synthesis temperature to 350 °C, forming a 1T-metallic phase.Plasma treatment aids in metal-oxide breakdown, leading to WS 2 creation.Se ion bombardment, assisted by plasma treatment, replaces S atoms in the top layers of WS 2 with Se atoms, forming the WS 2 /WSe 2 heterostructure.Confirmation was achieved via Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), depth-profile XPS, atomic force microscopy (AFM), X-ray reflectivity (XRR), and crosssectional transmission electron microscopy (TEM).Substrate temperature control in plasma sulfurization and selenization systems allows the phase transition from a semiconducting 2H phase to a metallic 1T phase.Dual plasma treatment improves surface roughness, increasing the film surface area.This heterostructure exhibits excellent hydrogen evolution reaction performance with a low Tafel slope of 57 mV dec −1 and remarkable stability, attributed to synergistic effects from bilayer TMD formation, reducing the energy barrier.

■ RESULTS AND DISCUSSION
Figure 1 shows a schematic of the entire growth process from metal-oxide deposition to plasma treatment, resulting in the final TMD heterostructure.The electron gun deposition method was used to deposit a layer of tungsten oxide (WO x ) with a thickness of 10 nm on silicon oxide substrates, followed by plasma sulfurization and plasma selenization processes, leading to the formation of the WS 2 /WSe 2 heterostructure.
The synthesis of the WS 2 /WSe 2 heterostructure and the growth of pure WS 2 and WSe 2 were confirmed using Raman scattering, as shown in Figure 2a.The spectra of WO x obtained by using the electron gun deposition method did not show any characteristic peaks, indicating its amorphous nature.However, Raman spectra of WS 2 and WSe 2 exhibited the characteristic peaks of in-plane W−S phonon (E 2g 1 ) and out-of-plane W−S (A 1g ) at around 352 and 417.2 cm −1 , respectively.Additionally, a low-intensity peak at around 170 cm −1 confirmed the formation of the 1T phase in the WS 2 sample. 36The in-plane vibration E 2g 1 and A 1g modes of W−Se bonds are closely located at 253 and 250 cm −1 , respectively, and are merged into one peak. 37The presence of the characteristic WS 2 and WSe 2 peaks confirms the formation of the WS 2 /WSe 2 heterostructure.A slight shift in the A 1g peak position of WS 2 to a higher wavenumber for the WS 2 /WSe 2 heterostructure strongly supports the presence of WS 2 and WSe 2 interaction. 38he results suggest that the WS 2 /WSe 2 heterostructure synthesized using PACVR retains the distinct Raman peak positions of the individual 2D TMDs rather than exhibiting intermediate peak positions that would be expected for an alloyed WS 2x Se 2(1−x) phase.Therefore, the heterostructure maintains the characteristic properties of the constituent TMDs rather than forming an alloy. 39The peak observed at 300 cm −1 is a distinctive feature of silicon originating from Si/ SiO 2 substrates.The unique characteristics of our process allow for an easy extension of the deposition technique to various substrates.To further demonstrate the versatility of PACVR, we extended the synthesis of heterostructures to other substrates, such as sapphire, as shown in Figure S1.The Raman spectra exhibited a small peak at 235 cm −1 , which is assigned to unreacted selenium (Se) and matches the characteristic Raman shift of crystalline Se (t-Se). 40Additionally, we successfully synthesized the MoS 2 /MoSe 2 heterostructure, and its formation was confirmed by the corresponding Raman spectra in Figure S2.These results highlight the wide applicability of PACVR for synthesizing various 2D heterostructures on different substrates.The characteristic modes A 1g at 235 cm −1 and E 2g 1 at 284 cm −1 originating from MoSe 2 were evident, along with the presence of characteristic peaks for the in-plane Mo−S phonon mode, E 2g 1 , at 380 cm −1 and the out-of-plane Mo−S phonon mode, A 1g , at 405 cm −1 , confirming the MoS 2 formation. 41The presence of Raman peaks for both MoS 2 and MoSe 2 in the heterostructure confirms the formation of the MoS 2 /MoSe 2 heterostructure after following the same growth process as for the WS 2 /WSe 2 heterostructure.However, a similar drift in the A 1g position can be seen, as reported in the WS 2 /WSe 2 heterostructure, confirming that there is an interaction between the two materials synthesized by PACVR.The composition, bonding characteristics, and surface electronic states of the WS 2 /WSe 2 heterostructure were analyzed by using X-ray photoelectron spectroscopy (XPS), as illustrated in Figure 2b−e.The XPS survey spectrum of the WS 2 /WSe 2 heterostructure (Figure 2b) confirms the presence of W, Se, and S elements, which distinguish it from those of WO x , WS 2 , and WSe 2 (Figure S3).The W 4f spectra exhibit two dominant peaks corresponding to W 4f 7/2 and W 4f 5/2 , indicative of W 4+ characteristics. 30The doublet peaks located at 32.1 and 34.22 eV in WS 2 and 31.1 and 33.2 eV in WSe 2 , as shown in Figure 2c, are consistent with previously reported values for WS 2 and WSe 2 , respectively. 24,37The WS 2 /WSe 2 heterostructure exhibits dominant peaks (W 4+ 4f 7/2 and W 4+ 4f 5/2 ) from the 1T phase at 31.7 and 33.8 eV, respectively.Interestingly, a shift in the binding energies of W 4f 7/2 and W 4f 5/2 for the WS 2 /WSe 2 heterostructure suggests electron transfer from WS 2 to WSe 2 under the unique heterostructure effect, indicating strong electronic interaction between the two materials. 30The Se 3d spectra of WS 2 /WSe 2 and WSe 2 exhibited two well-defined peaks at 53.3 and 54.3 eV for the pure WSe 2 . 37The high valence peak of Se in the heterostructure shifted to a higher binding energy, confirming the strong interaction of the catalyst.Overlapping spectra of the S 2p core level with the Se 3p core level further confirmed the composition of the WS 2 / WSe 2 heterostructure.In addition, XPS was used to confirm the successful formation of the MoS 2 /MoSe 2 heterostructure, as shown in Figure S4.The Mo 3d spectra exhibited two peaks at 229 and 232 eV, corresponding to the binding energies of Mo 3d 5/2 and Mo 3d 3/2 in MoSe 2 .Similarly, the Se 3d spectra showed the corresponding 3d 5/2 (at 54.4 eV) and 3d 3/2 (at 55.3 eV) binding energies for Se in MoSe 2 .These results suggest that WO x and MoO x were effectively transformed into WS 2 /WSe 2 and MoS 2 /MoSe 2 heterostructures through the plasma-assisted sulfurization and selenization processes.Highresolution TEM images of the WS 2 /WSe 2 heterostructure are shown in Figure 2f.The lattice spacings of 0.68 and 0.64 nm corresponding to the (002) facets of WSe 2 and WS 2 were indexed.Also, a similar kind of crystallinity is visible in the TEM images of WS 2 and WSe 2 , suggesting that the lowtemperature synthesis along with plasma could be a reason for the same (Figure S5).An amorphous deposition of WO x is also confirmed by the TEM images of WO x .Figure 2g shows elemental mapping of the WS 2 /WSe 2 heterostructure and reveals the presence of WSe 2 in the top layers.However, WS 2 appears to be near the bottom layers, confirming the conversion of the top layer into WSe 2 .Besides, energydispersive analysis X-ray (EDAX) spectra (Figure S6) show the equivalent presence of S and Se elements in the heterostructure.
Furthermore, atomic force microscopy (AFM) was used to study the thickness and morphology of the WS 2 /WSe 2 heterostructure.Figure 3a−f shows 2D and three-dimensional (3D) images of as-deposited WO x , WS 2 , and WS 2 /WSe 2 heterostructure on SiO 2 /Si substrates.It can be seen from the images that the e-gun-deposited WO x shows a uniform granular surface with a low roughness of about 0.2 nm.However, the plasma-sulfurized WO x film is considerably rougher than the pristine sample and exhibits prominent ridges and valleys.The root-mean-square (rms) surface roughness (σ s ) at an area of 100 × 100 nm 2 of the WS 2 sample increases to ∼0.4 nm.The roughness of the samples increases with each plasma treatment given in the deposition of sulfur and selenium, thus increasing the surface area of the catalyst heterostructure film, as shown in Figure 3g−i.The surface of the WS 2 /WSe 2 heterostructure is observed to be nonhomogeneous, with a mean square roughness of ∼0.63 nm.To evaluate the thickness of the as-grown films, AFM height profile measurements were performed along the edges.The average height of the WS 2 /WSe 2 heterostructure film, measured from the Si substrate surface, is found to be ∼20 nm, which is double the thickness of the initially deposited film.To analyze the interface roughness and the specific thickness of the WS 2 /WSe 2 heterostructure, X-ray reflectivity (XRR) was conducted using an eight-circle Huber diffractometer at the wiggler beamline BL17B at the National Synchrotron Radiation Research Center (NSSRC) in Taiwan.XRR allows the estimation of surface roughness, interface roughness, and layer thickness by analyzing the 2θ dependence of the reflectivity.The oscillation observed in the reflectivity is a result of the interference between X-rays reflected from the surface and interface. 42By analysis of the period of the oscillation, the thickness of different layers can be determined.The amplitude of the oscillation is influenced by the density difference between the film and the substrate with a larger density difference, leading to a higher amplitude.The decay rate of the reflectivity beyond the critical angle of total reflection is also indicative of surface roughness with a larger roughness, causing a higher decay rate.Figure 3j shows the reflectivity from the WS 2 -deposited silicon wafer, where the dotted and solid lines represent the experimental and simulated results, respectively.The difference between the XRR and AFM measurements is that the X-ray beam covers a larger area of 1 × 1 cm 2 on the sample, whereas AFM probes only a local area (around 100 × 100 nm 2 ). 43To determine the density of the WS 2 /WSe 2 film grown on a Si wafer, XRR measurements were performed in the 2θ range 0−10°.The experimental data were fitted, revealing that the WSe 2 film had a density of ∼4.6 g/cm 3 and a thickness of 8.02 nm, whereas WS 2 had a lower density of 2.4 g/cm 3 with a thickness of 12 nm as shown in Figure 3k, resulting in a total thickness of the WS 2 /WSe 2 heterostructure film, agreeing roughly with the result obtained by AFM.The interface roughness plays an important role in determining the adhesion between the layers.After analyzing the XRR results, the roughness of the WSe 2 layer was determined to be 0.63 nm and the interface roughness of WS 2 was 0.19 nm.As can be seen from the tables in the insets of Figure 3j,k, we can see that the roughness of the WS 2 sample is lower as compared to the heterostructure, which is due to the double exposure of the sample to plasma, resulting in hills and valley formations, thus increasing the roughness of the surface.The observation that the surface roughness is inversely proportional to the density of the film suggests that the film with a lower density has a higher surface roughness, as reported in previous studies. 44However, this relationship might not directly apply to all interfaces or conditions, as factors like deposition process, material properties, or heterogeneity could alter this relationship at the interface of the films. 45e intensity of X-ray reflectivity for a rough surface is reduced by a factor of exp −σd 2 q z d 2 ,where σ is the root-mean-square roughness and q z is the wavevector perpendicular to the sample surface.This indicates that the surface roughness affects the slope of the X-ray reflectivity curve.As shown in Figure 3j,k, the intensity decreases more rapidly with the increase in roughness.The thickness oscillations, known as "Kiessig fringes", are clearly visible. 46The period of oscillations is directly related to the film's thickness (D), given by = t D 2 , for which t and D represent the period of oscillations and film thickness, respectively.The XRR analysis may be affected by systematic errors due to even minor variations in the thickness of the film as the beam in this experimental configuration completely covers the sample.Owing to the fundamental difference in the interaction between the AFM tip and the surface compared to X-rays with the sample, the information obtained about surface roughness from AFM and XRR can be different.The surface irregularities and type of roughness play crucial roles in determining the results of AFM and XRR measurements.Typically, XRR and AFM yield similar roughness parameters for very flat surfaces with root-meansquare (rms) roughness within the range of several angstroms.However, for surfaces with high irregularities and complex height distributions, significant differences in the roughness parameters obtained by AFM and XRR can be observed. 47urface roughness, determined by XRR, provides a broader view, assessing the roughness across the entire film surface of the catalyst.The results show an increase in roughness from 0.12 nm in WS 2 to 0.63 nm in the WS 2 /WSe 2 heterostructure, indicating the impact of dual plasma on the same surface.Additionally, interface roughness elucidates electron transfer mechanisms and adhesion, confirming the conversion and the presence of an interface, significantly influencing the HER performance.
To investigate interfacial effects in WS 2 /WSe 2 heterostructures and to study the depth of penetration of Se ions into the WS 2 film to form the WS 2 /WSe 2 heterostructure, we analyzed the XPS depth profiles.Initially, the intensity of W 4f increases as we move deeper into the layers, and a similar trend is followed for the Se 3d peak.However, the peak of Se diminished after 8 nm, which indicates that the conversion of WS 2 to WSe 2 is possible only in the top layers (represented by the orange graph as shown in Figure 4, while the bottom 8−10 nm still corresponds to WS 2 (depicted by the blue graph)).The S 2p spectra show an overlapping spectrum with the Se 3p core level in the upper 10 nm-thick WS 2 /WSe 2 heterostructure, exhibiting a picture of the Se atoms knocking out the S atoms in the top layers.Quantifying selenium in the presence of sulfur using XPS presented challenges due to spectral overlap between the S 2p and Se 3p peaks (Se 3p 3/2 and Se 3p 1/2 ), along with the S 2s and Se 3s peaks.Considering the Se 3p 3/2 −3p 1/2 doublet separation (5.8 eV) and constraining the 3p 3/2 −3p 1/2 peaks to a 2:1 ratio, 48 the remaining structure corresponded to the S 2p peak(s), which appeared as a minor peak visible in the fitted spectra of the WS 2 /WSe 2 heterostructure (Figure 2d).This interface provides an electron transfer from WS 2 to WSe 2 , thus explaining the enhanced performance of the heterostructure compared to the bare WS 2 and WSe 2 catalysts.
For HER performance, a conducting surface is required.For this reason, the as-synthesized WS 2 /WSe 2 heterostructure films were transferred onto a gold-coated silicon substrate as shown in Figure S7, where the synthesized film was separated from the Si surface using a PMMA layer as support.The PMMA/2D material film was then removed from the pristine substrate by scraping its edges with tweezers, and then it was transferred to a diluted ammonia solution (NH 4 OH/deionized (DI) water = 1:5) for separation and finally peeled off in DI water for cleaning before being transferred onto the conducting Au/Si substrate.The HER performance of the WS 2 /WSe 2 heterostructure was evaluated using a three-electrode setup in a 0.5 M H 2 SO 4 solution as shown in Figure S8.To facilitate comparison, the linear sweep voltammetry (LSV) curves of the as-synthesized WS 2 /WSe 2 , WS 2 , and WSe 2 , as well as a bare Au-coated silicon substrate and commercial Pt/C powders, were collected and plotted together as shown in Figure 5a.A comparison of the results showed that the Au substrate had a negligible impact on the catalytic performance, thus confirming that the HER activity observed was solely due to the assynthesized 2D films.The ohmic drop is eliminated by performing IR correction using the resistivity (R s ) value obtained from EIS curves (Table S1). Figure 5b shows a bar graph presenting a comparative study of overpotentials at 2 mA cm −2 and 10 mV cm −2 , respectively, and it can be seen that the WS 2 /WSe 2 catalyst exhibits the best performance and the lowest overpotential of 237 and 291 mV, respectively.However, bare WS 2 and bare WSe 2 show higher overpotentials of approximately 392 and 364 mV, respectively, under a current density of 10 mA cm −2 .The efficient electron−hole separation resulting from the introduction of heterointerfaces leads to a significant decrease in the overpotential. 49dditionally, 1T phase materials exhibit a lower overpotential compared to 2H materials due to their metallic properties and lower ΔG H at the basal plane and edge sites, which results in denser active sites than 2H phase materials 49 It is believed that the 1T-WS 2 /1T-WSe 2 catalyst exhibits superior electrocatalytic performance in the HER due to the diffusion of Se atoms into the top layer, replacing the S atoms.This diffusion leads to low energy barriers for orbital overlap in H* migration, known as the Volmer reaction, and H 2 formation, resulting in better catalytic performance than those of pristine WS 2 and WSe 2 materials. 49This observation is shown by elemental mapping (Figure 2f), which reveals the presence of diffused Se atoms in upper layers; however, as we follow the spectra further, only S atoms are present.The rate-limiting stages and electrocatalyst behavior can be further understood through the Tafel plot derivative.In the case of hydrogen evolution, the reaction kinetics can be derived from a series of reactions.First, during the discharge step, hydrogen ions are adsorbed on the active sites, which is termed the Volmer reaction, as shown in eq 1.This is followed by either an electrochemical desorption step, known as the Heyrovsky reaction, as shown in eq 2, or a recombination step, known as the Tafel reaction, as shown in eq 3 (2) 2H H ads 2 (3) Note that the Tafel slope is a crucial parameter used to evaluate catalyst performance and determine the rate-limiting reaction of the HER.The Volmer, Heyrovsky, and Tafel reactions can each limit the HER when the Tafel slope is around 120, 40, and 30 mV dec −1 , respectively.To determine the rate-determining step of the HER process on the heterostructure catalyst, Tafel plot analysis was conducted, as shown in Figure 5c.The Tafel plots were obtained from the corresponding LSV curves using the Tafel equation, η = b log(j/j 0 ), where b, j 0 , and j represent the Tafel slope, the exchange current density, and the current density, respectively. 50A Tafel slope of 31 mV dec −1 was confirmed for the Pt/C powders, which is consistent with the reported values.
The extracted Tafel slope of WS 2 /WSe 2 is 57 mV dec −1 , which is smaller than those of WS 2 (139 mV dec −1 ) and WSe 2 (102 mV dec −1 ) (Figure 5c).The obtained results indicate that the HER process on the heterostructure catalyst occurs through Volmer−Tafel and/or Volmer−Heyrovsky reaction steps.Moreover, the low Tafel slope value suggests that the WS 2 / WSe 2 heterostructure catalyst can generate a fast response current at a lower overpotential, indicating its superior hydrogen evolution efficiency. 30Compared with the recently reported values presented in Table 1, the Tafel slope of the WS 2 /WSe 2 heterostructure as the catalyst is notably better.Thus, WS 2 /WSe 2 heterostructures are flourishing electro- catalysts for HER.The intrinsic activity of the catalyst materials was examined by analyzing the turnover frequency (TOF).Figure S9 shows the TOF performance of the WS 2 / WSe 2 heterostructure at different overpotentials.The heterostructure exhibits a better TOF of 2.7 s −1 than pure WS 2 and WSe 2 whose calculated values are 0.4 and 0.89 s −1 , respectively, at an overpotential of 300 mV.Furthermore, the stability and durability of the WS 2 /WSe 2 heterostructure as the catalyst were assessed through polarization curve measurements and Raman spectroscopy analysis after 5000 cycles.The results, as shown in Figure 5d, demonstrate that there is almost no degradation of the current density, indicating the robustness of the WS 2 /WSe 2 heterostructure.This is attributed to the compositional and structural stability of the WS 2 /WSe 2 heterostructure.The Raman spectra of the initial and cycled samples, as shown in Figure S10, also indicate similar peak positions, further confirming the durability of the catalyst.Additionally, Figure 5e shows that the current density remains highly stable after 24 h, supporting the stability of the WS 2 /WSe 2 heterostructure as the catalyst.The HER kinetics were investigated using Nyquist plots obtained from electrochemical impedance spectroscopy (EIS) measurements, and each plot exhibits a similar semicircle profile without Warburg impedance in the low-frequency  range, indicating rapid mass transport and a kinetically controlled reaction as shown in Figure 5f.When analyzing EIS at different overpotentials, it was found that the impedance properties were similar, suggesting that the same electrochemical processes occurred in 0.5 M H 2 SO 4 during all tests and the equivalent circuit depicted in Figure S11 supports this finding, for which the electrochemical kinetics between the electrocatalysts and electrolytes were represented by the charge transfer resistance (R ct ).Analysis in Table S1 confirms that the R ct of the WS 2 /WSe 2 heterostructure is significantly lower than those of WS 2 and WSe 2 , indicating that the WS 2 /WSe 2 heterostructure possessed a higher catalytic performance because of the faster electron exchange.The electronic structure of the heterostructure was modulated by the heterointerface between WS 2 and WSe 2 , which boosted the charge transference and improved the catalytic activity of the WS 2 /WSe 2 heterostructure as the catalyst.At the heterointerface between the metallic 1T materials, an overlap of d-orbitals between the WS 2 and WSe 2 heterostructure forms a lower energy barrier, which enhances the electrochemical activity. 49dditionally, the 1T phase WS 2 /WSe 2 heterostructure can be catalytically active at both the basal plane and edge sites, leading to a high HER performance.Compared with a single WS 2 or WSe 2 , the WS 2 /WSe 2 heterostructure exhibits enhanced HER performance. 49he electrochemical double-layer capacitance (C dl ) was used to assess the electrocatalytically active surface area (ECSA).The value of ECSA is directly proportional to the value of C dl , which can be determined from cyclic voltammetry (CV) curves.The C dl values for the WS 2 , WSe 2 , and WS 2 /WSe 2 heterostructure electrocatalysts were evaluated in the nonfaradaic region by conducting CV studies at scan rates ranging from 10 to 200 mV s −1 .The results of these studies are shown in Figure 6a−c.ECSA was evaluated using the following relation: ECSA = C dl /C s , where C dl is the double-layer capacitance and C s is the specific capacitance.C s = 0.035 mF cm −2 for 0.5 M H 2 SO 4 . 1 Table 2 includes the electrocatalytic parameters that were extracted from the plotting of different electrochemical graphs.The WS 2 /WSe 2 heterostructure catalyst exhibited a linear slope of 5.29 mF cm −2 , which is higher than the slopes of WS 2 (2.03 mF cm −2 ) and WSe 2 (3.3 mF cm −2 ), indicating that the WS 2 /WSe 2 heterostructure has a higher exposure to efficient active sites, which contributes to its excellent HER performance as shown in Figure 6d.To account for the impact of surface topology on catalyst performance, the LSV data were adjusted by using ECSA normalization, aiming  to ascertain the inherent HER activity for each catalyst (Figure S12).Normalizing the current using ECSA (Figure 5a) resulted in a shift in the order of HER activity compared to the current normalized based on the geometric surface area.The WS 2 /WSe 2 heterostructure catalyst still exhibits the lowest onset potential compared to the pure WS 2 and WSe 2 catalysts.
To study the performance of the WS 2 /WSe 2 heterostructure as the electrocatalyst under alkaline conditions, the WS 2 /WSe 2 heterostructure was tested in a 0.5 M KOH solution.However, it was not as impressive as its performance under a 0.5 M H 2 SO 4 acidic solution, as shown in Figure S13.Nevertheless, it still outperformed WS 2 and WSe 2 , which supports the successful combination of WS 2 and WSe 2 .To elucidate the mechanism and assess the impact of specific sulfur-to-selenium ratios in the WS 2 /WSe 2 heterostructure, experiments were conducted, and the corresponding results are shown in Figure S14.The WS 2 to WSe 2 ratio in the heterostructure was manipulated by varying the growth time, with the conversion of WS 2 to WSe 2 in the top layers increasing as the growth time extended from 15 to 60 min.Figure S14a illustrates the variation in WS 2 /WSe 2 ratios with the growth time at 350 °C.The conversion ratio was confirmed by XPS, as shown in Figure S14b.In Figure S14b, the W 4f spectra were deconvoluted into specific peaks corresponding to WSe 2 and WS 2 in the WS 2 /WSe 2 heterostructure.The intensity ratio of the peaks for WS 2 (represented by yellow and magenta colors for W 4f 7/2 and W 4f 5/2 , respectively) and WSe 2 (represented by blue and green colors for W 4f 7/2 and W 4f 5/2 , respectively) changes with the concentration of the two components in the heterostructure.To quantify the conversion ratio of WS 2 to WSe 2 in the top layers, the area under the plasma-assisted W 4f 5/2 and W 4f 7/2 bands (A W 4fd 5/2 + A W 4fd 7/2 for WSe 2 ) is compared to the total area of the two components in the spectra, including WSe 2 and WS 2 , (A W 4fd 5/2 + A W 4fd 7/2 ) WSed 2 + (A W 4fd 5/2 + A W 4fd 7/2 ) WSd 2 , irrespective of the presence of oxidation states of tungsten (WOx).The conversion increased from 27% for a growth time of 15 min to 73% for a growth time of 60 min.Subsequently, the impact of each component on the HER performance was evaluated using the obtained WS 2 /WSe 2 (%) ratios.Figure S14c demonstrates that the optimal performance can be achieved with a WS 2 /WSe 2 ratio of 53−47%, followed by the presence of 73% WSe 2 in the heterostructure.These findings provide valuable insights into the correlation between the compositions of the heterostructure and its catalytic performance in the HER.The results indicate that the improved electrocatalytic properties of the WS 2 /WSe 2 heterostructure can be attributed to an increase in active facets, good electron transfer, and large surface area.The outstanding HER performance of the WS 2 /WSe 2 heterostructure can be attributed to several factors.First, the similar crystal structure and lattice constants of WS 2 and WSe 2 provide a foundation for forming the heterostructure at a low temperature, leading to metallic 1T phase formation.Second, heterojunction engineering enables the heterointerface to accelerate charge deployment and promote interface charge transfer.Third, the integration of WS 2 and WSe 2 leads to the redistribution of charge density at the heterointerface, which facilitates efficient electrocatalytic activity.The plasmaenhanced roughness of the WS 2 /WSe 2 heterostructure has a fourth beneficial effect, enabling exposure of more active sites, which in turn promotes electron transfer between the electrocatalysts and electrolytes.Finally, the connection of WS 2 with WSe 2 creates a synergistic effect that incorporates the intrinsic characteristics of both components, shortening the electron transfer channels and enhancing the HER performance and stability.

■ CONCLUSIONS
In summary, a novel synthesis of the WS 2 /WSe 2 heterostructure was demonstrated by using a PACVR system.A metallic WS 2 /WSe 2 heterostructure was synthesized at 350 °C, whereas a semiconducting heterostructure can be obtained at a higher temperature.Systematic characterizations revealed that the diffusion of Se atoms into the WS 2 film leads to low reaction energy barriers by facilitating orbital overlap in the HER mechanisms.Interfacial electron transfer was observed in the WS 2 /WSe 2 heterostructure, leading to the reduction of hydrogen atoms and an increase in the absorption capacity of the electrode surface.This resulted in a significant improvement in the kinetics of the HER.The WS 2 /WSe 2 heterostructure fabricated by using this methodology exhibited excellent catalytic activity, with a low Tafel slope of 57 mV dec −1 and an overpotential of 291 mV at 10 mA cm −2 .The electrochemically active surface area of the heterostructure confirmed its ability to improve electrocatalytic behavior.The surface electron conductivity was significantly improved due to the highly interacting interfaces in the heterostructure.The findings of this study offer a potential solution that is highly sought-after in the field of sustainable energy conversion technologies.
■ EXPERIMENTAL SECTION Synthesis of the WS 2 /WSe 2 Heterostructure Film.To prepare the WS 2 /WSe 2 heterostructure, a WO x seed layer was initially deposited on silicon wafers with a 300 nm thick SiO 2 buffer layer under high vacuum (>10 −5 Torr) conditions, through an E-gun evaporator system (a deposition rate of 0.1 Å s −1 ).This was followed by a two-step plasma-assisted chemical vapor reaction with the optimized parameters.First, the WO x -deposited wafers were transferred into the PACVR for the step of sulfurization.During the reaction, the sulfur precursor was vaporized at 180 °C, and the vapor was carried by a N 2 /H 2 forming gas to reach the WO x -deposited substrate on a stage at a fixed elevated temperature of 350 °C through an ICP coil with a plasma of 150 W. Both the reduction of WO x by hydrogen radicals and the chemical reaction between S radicals and W atoms were activated simultaneously, resulting in WS 2 films.Subsequently, the sulfurized sample was quickly transferred to the selenization PACVR furnace for the conversion of the top layers of WS 2 into WSe 2 .The synthesis process of the WS 2 /WSe 2 heterostructure involved heating selenium pellets to 300 °C while maintaining the substrate at a low temperature of 350 °C in the presence of a N 2 /H 2 atmosphere.This process was carried out for 1 h.
Material Characterization.Various techniques were employed to characterize the synthesized films, as the presence of atom mixing, diffusion across interfaces, and precipitation of nanoparticles, which could affect the hydrogen uptake and release cycling, as well as the electronic and thermodynamic properties of the films. 54Raman spectroscopy using a 532 nm laser was used to investigate the uniformity and structure of the films, with the laser power set at 20 mW and the Si peak at 520 cm −1 serving as the standard peak.X-ray photoelectron spectroscopy (XPS) was utilized to analyze the binding energies and chemical compositions of the heterostructures using monochromatized Al Kα X-rays (hν = 1486.6eV) with an ECSA ULVAC-PHI 5000 Versaprobe II.The atomic geometry of the films was determined by cross-sectional transmission electron microscopy (TEM) using a JEOL JEM-200F with an acceleration voltage of 200 keV, and high-resolution TEM (HR-TEM) specimens were prepared using a focused ion beam instrument (FIB) (SII nanotech SMI3050) with the lift-out technique.The thickness, density, and roughness of each layer, as well as the heterostructure, were analyzed using X-ray reflectivity (XRR) on an eight-circle Huber diffractometer at the wiggler beamline BL17B at the National Synchrotron Radiation Research Center (NSSRC) in Taiwan, with GenX3 software used for data analysis.To study the depth of Se diffusion and interface characteristics, depth-spectrum XPS was conducted.Atomic force microscopy (AFM) (Bruker, dimension Icon) was used to examine the surface morphology and thickness of the films.
Electrochemical Measurements.A conducting surface is required for electrochemical measurements.The as-synthesized WS 2 /WSe 2 heterostructure films were transferred onto a gold-coated silicon substrate.To prepare the conducting substrate, a 300 nm SiO 2 /Si substrate was cleaned with acetone, isopropyl alcohol (IPA), and DI water, followed by the deposition of a 50 nm gold (Au) layer using an E-gun evaporation system under a high vacuum at a rate of 0.2−0.3Å s −1 .The synthesized film was separated from the Si surface using a PMMA layer as the support.The PMMA/2D material film was then removed from the pristine substrate by scraping its edges with tweezers.Between the PMMA/2D material film and the substrate, a space was created, and then it was transferred to a diluted ammonia solution (NH 4 OH/DI water = 1:5) for separation and finally peeled off in DI water for cleaning before being transferred onto the conducting Au/Si substrate.The whole transferred film Au/ Si substrate was kept in an acetone solution, followed by cleaning with IPA and DI water to dissolve the PMMA.The electrochemical characteristics of the samples were assessed by using a three-electrode configuration and a Bio-Logic VSP potentiostat inside a Teflon cylinder cell equipped with an O-ring at the base.The reference and counter electrodes used were Ag/AgCl saturated with 3 M NaCl and a Pt wire, respectively.The electrolyte solution used for all electrochemical measurements was 0.5 M H 2 SO 4 .The WS 2 /WSe 2 heterostructure film was transferred to a 50 nm thick Au-deposited Si substrate and used as a catalyst for the hydrogen evolution reaction, with Cu tape (3 M) for connection.The measurements were recorded after ten cycles without magnetically stirring the solution.Polarization curves were obtained by conducting linear sweep voltammetry (LSV) at a scan rate of 5 mV/s.All potentials were calibrated to potentials versus a reversible hydrogen electrode (RHE) with reference to Ag/ AgCl using the Nernst equation given by E RHE = E Ag/AgCl + 0.0591pH + E°A g/AgCl .The correction for internal resistance (IR) was performed in order to adjust for the potential drop losses that occur due to the resistance of the solution.This correction is carried out using the equation given by E corrected = E RHE − iR S .The Tafel plots, which show the relationship between the overpotential and the logarithm of current density, were analyzed by fitting them to a linear equation.The linear slope was then determined as the Tafel slope using the provided formula, η = a + b log j, for which η is the overpotential, j is the current density, b is the Tafel slope, and a is the intercept.The long-term stability was obtained at a 6 mV constant potential from the chronoamperometry test by recording the current versus time.In order to calculate the electrocatalytically active surface area (ECSA), cyclic voltammetry (CV) was performed in the non-faradaic reaction potential range at different scan rates.The double-layer capacitance was derived from the CV data and used in the following equation given by ECSA = C dl /C s , for which C s is the specific capacitance.The Nyquist plot was obtained by conducting electrochemical impedance spectroscopy (EIS) measurements in the frequency range of 0.1 Hz to 100 kHz at an amplitude of 10 mV.

* sı Supporting Information
The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsami.4c03513.Schematic process flows of the WS 2 /WSe 2 film transfer on conducting the Au substrate for device formation; three-electrode system for electrochemical measurements using a Bio-Logic VSP potentiostat in a cylindrical cell made of Teflon with an O-ring at the bottom; Raman spectra for WS 2 and WS 2 /WSe 2 growth on a sapphire substrate; Raman spectra of the MoS 2 /MoSe 2 heterostructure with MoS 2 and MoSe 2 ; XPS spectra: (a) survey of WO x , (b) survey spectra of WS 2 , (c) survey spectra of WSe 2 , (d) O 1s binding energy for the WS 2 / WSe 2 heterostructure, WS 2 , and WO x ; XPS spectrum analysis of the MoS 2 /MoSe 2 heterostructure; TEM analysis: WO x and WS 2 ; high-resolution TEM images; energy-dispersive spectroscopy (EDS) elemental mapping of each element, respectively; Raman spectra of the WS 2 /WSe 2 heterostructure after HER; equivalent circuit model for EIS analysis of the catalysts; ECSAnormalized LSV curves; electrocatalytic performance of different catalysts for HER in 0.5 M KOH; turnover frequency (TOF) calculations, and effect of component ratios on HER performance (PDF) ■

Figure 6 .
Figure 6.Cyclic voltammetry curves at different scan rates for (a) WS 2 , (b) WSe 2 , and (c) the WS 2 /WSe 2 heterostructure.(d) Linear relation between the scan rate and the current density difference for the WS 2 /WSe 2 heterostructure.

Table 1 .
Comparison of the Electrochemical Performance of TMD-Based HER Catalysts a CC: carbon cloth, FTO: fluorine-doped tin oxide, GCE: glassy carbon electrode, and CFP: carbon fiber paper.

AUTHOR INFORMATION Corresponding Author Yu
-Lun Chueh − Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan; College of Semiconductor Research, National Tsing Hua University, Hsinchu 30013, Taiwan; Department of Physics, National Sun Yat-Sen University, Kaohsiung 80424, Taiwan; Department of Materials Science and Engineering, Korea University, Seoul 02841, Republic of Korea; orcid.org/0000-0002-0155-9987;Email: ylchueh@ mx.nthu.edu.twDepartment of Materials Science and Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan; College of Semiconductor Research, National Tsing Hua University, Hsinchu 30013, Taiwan; Department of Physics, National Sun Yat-Sen University, Kaohsiung 80424, Taiwan Yu-Ren Peng − Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan; College of Semiconductor Research, National Tsing Hua University, Hsinchu 30013, Taiwan; Department of Physics, National Sun Yat-Sen University, Kaohsiung 80424, Taiwan Mayur Chaudhary − Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan; College of Semiconductor Research, National Tsing Hua University, Hsinchu 30013, Taiwan; Department of Physics, National Sun Yat-Sen University, Kaohsiung 80424, Taiwan Feng-Chuan Chuang − Department of Materials Science and Engineering, Korea University, Seoul 02841, Republic of Korea Complete contact information is available at: https://pubs.acs.org/10.1021/acsami.4c03513