Electrochemical imaging correlated to hydrogen evolution reaction on transition metal dichalcogenide, WS₂

Two-dimensional (2D) materials with atom thick layers have been expected as high efficient electrodes for practical applications in electrochemical energy storage/harvesting devices.^1–5 In 2D materials, monolayers or nanosheets of transition metal dichalcogenides (TMDs) such as molybdenum disulfide (MoS₂) and tungsten disulfide (WS₂) exhibit excellent efficiency in water electrolysis, comparable to platinum.^6–15 The origin of their high active sites came from their crystal phase, structures, and chemically modified or strained surface. In particular, their defect and edge sites have been engineered for further improvement. Indeed, for the direct identification of their active sites after being engineered, it is an important issue to confirm and derive optimized molecular structures. The probing techniques have been paid great attention for that purpose, in order to characterize local electrocatalytic reactions in the nanometer scale in recent reports. For example, Jaramillo et al. revealed that there is a clear linear relation between the number of active edge sites in MoS₂ and electrocatalytic hydrogen evolution reaction (HER) in MoS₂, confirmed by scanning tunneling microscopy.¹⁶ Also, Li et al. derived local HER kinetics on sulfur vacancies on strained MoS₂ by scanning electrochemical microscopy.¹⁷ Recently, Bentley et al. visualized the HER process on MoS₂ by scanning electrochemical cell microscopy (SECCM).¹⁸ In particular, the SECCM has more advantages than other characterization techniques for the characterization of the local electrochemical behaviors on 2D TMD materials because their characterization area is confined by a meniscus created by their probe. Similar to the former report, we have developed an SECCM system with a single barrel nanopipet with a hopping mode.¹⁹ This system can evaluate local lithium-ion redox activities on the electrodes of lithium-ion batteries^19–22 and electrochemical HER on 2D materials.^23,24 Notably, WS₂ is also one of promised earth-abundant electrocatalytic electrodes for HER applications because of their high activities with robust chemical stability.²⁵ It is, therefore, required to evaluate WS₂ in detail with a number of layers as well as other 2D materials.

Herein, by the SECCM technique (visualization and local electrochemical evaluation), we reveal the HER activities of the mono/bilayers of WS₂, in particular, their activities on edge and basal structures. The SECCM measurement clearly visualizes that the edge sites show higher electrochemical HER activities than that on the basal plane. Further, the number of layers does not show a linear relation to the current response, and we observe higher activity in the basal plane of a bilayer WS₂ as compared to a monolayer, which could be ascribed to an increased number of defects in the second layer.²⁶ 

We synthesized WS₂ mono/bilayers on cleaved highly oriented pyrolytic graphite (HOPG) substrates via the chemical vapor deposition (CVD) method followed by the recipes reported by our past reports.^27–29 The CVD was conducted with a quartz tube (3 cm in diameter and 100 cm in length). The HOPG substrates, WO₃ powder (Aldrich, 99% purity), and sulfur flakes (Aldrich, 99.99% purity) were placed in the tube. Then, Ar gas was introduced into the tube at a flow rate of 100 cm³/min with the temperature gradually elevated to the sulfurization temperature (900–1100 °C) over 60 min. Once the substrate temperature reached, sulfur was heated at 200 °C for about 30 min to generate sulfur vapor, which was supplied to the substrate via an additional electrical furnace. After sulfurization, the tube was immediately cooled to room temperature, resulting WS₂ films on the substrate. The number of WS₂ layers was characterized by an optical microscope combined with Raman spectroscopy with a 532 nm laser wavelength for excitation (Horiba LabRAM HR-800) at room temperature.

Local HER activities and their electrochemical imaging were examined by the SECCM system as shown in Fig. 1(a). The basic SECCM system configuration with a single barrel nanopipet was described in Refs. 19–24. As a probe of the SECCM system, a 70 nm diameter nanopipet was prepared from a borosilicate glass capillary [Harvard Apparatus (inner diameter = 1.00 mm and outer diameter = 0.78 mm; GC150F-10)] pulled by a typical CO₂ laser puller (P-2000, Sutter Inst.), confirmed by scanning electron microscopy as the inset of Fig. 1(a). 0.5M H₂SO₄ and a palladium wire were used as the aqueous electrolyte and quasi-reference counterelectrode, respectively. As the probe was in the proximity of the sample, a meniscus was created, and the meniscus size was almost equivalent to the pipet diameter.¹⁹ The meniscus was utilized as an electrochemical cell for HER properties such as local linear sweep voltammetry (LSV) and electrochemical imaging with the applied voltage for introducing the HER.

The voltage was applied at −1600 or −1800 mV versus Pd (equivalent to −800 or −1000 mV versus reversible hydrogen electrode, RHE) between the probe and sample substrates, introducing HER as electrochemical imaging. It is also noted that the conversion in the SECCM system with a quasi-reference counterelectrode inside a single barrel pipet is documented and referred from Ref. 30. The LSV was conducted with a scan rate of 100 mV/s in the applied voltage range from −0.8 to −2.0 V versus Pd. All the measurements have been done under ambient air.

The typical triangular shapes of WS₂ sheets on HOPG substrates were observed by optical microscopy after CVD synthesis in Fig. 1(b). The color of the triangular sheets was mainly two patterns marked with the dashed white line in Fig. 1(b); light pink—“A” and dark pink—“B.” Raman spectroscopy was applied at the centers of the isolated triangular A and B sheets with the corresponding optical microscopy. Figure 1(c) shows Raman spectra of light pink (on the left) and dark pink (on the right) triangle shapes. In both shapes, the peaks of Raman spectra obtained are typical vibration modes of WS₂,³¹ confirming the existence of WS₂. Notably, Raman spectroscopy has been also previously studied by a number of research groups as a powerful technique to determine the number of layers in two-dimensional materials including WS₂.^31,32 In WS₂, each featured peak corresponding to 2LA(M) around 352 cm⁻¹ and A_1g(Г) around 417 cm⁻¹ were used for benchmarks to confirm the number of layers, according to Zeng et al.³³ By using Raman spectra results in Fig. 1(c), their frequency difference [A_1g(Г) − 2LA(M)] and their intensity ratio (I_2LA(M)/I_A1g) were 65.5 cm⁻¹ and 2.2 in light pink WS₂, and 70.5 cm⁻¹ and 1.3 on the right, concluding that the former WS₂ (sheet A) is a monolayer and the latter one (sheet B) is a bilayer.

To investigate their local electrochemical HER activities with the structural configuration, we have applied electrochemical measurements for real-space mapping by SECCM with the applied voltage between the probe and the sample introducing the HER (V_appl. = −1600 mV versus Pd). Figure 2(a) shows typical on-site HER current mapping results on monolayer WS₂. The magnified imaging has been taken to investigate their whole area to cover edge sites, basal sites, and HOPG simultaneously. The higher the absolute value of this responded current, the higher the HER activity. This is because the negative current attributed to the HER from protons to hydrogen due to electrochemical water splitting: 2H⁺ + 2e⁻ → H₂. The HER current mapping showed that the edge region has higher electrochemical activity on HER than other regions (in the order of basal, HOPG regions). The result is consistent with the work in edge rich bulk WS₂ of Sarma et al.³¹ The phenomenon in which the edge region showed higher electrochemical activity compared to the basal region is caused by the atomic structure at the edge sites of WS₂. The atomic structure is presented as isolated vacancies of the tungsten or sulfur atom, recognized that those vacancies are within TMDs, and exhibited heightened energetic favorability with regard to H⁺ adsorption.³⁴ Following the HER mapping measurement, the LSV measurement was conducted by pinning the probe of SECCM at two points, C and D, as shown in Fig. 2(b). The point C indicates the LSV measurements in the edge region and the point D indicates that in the basal region, allowing for their comparison. The LSV measurement results show a rapid negative current increase from below −1.5 V versus Pd. This phenomenon can be also attributed to the electrochemical process of the HER as observed in the current mapping results. Those LST results showed clearly that the edge region exhibits a higher HER reaction than that on the basal region, which is consistent with the electrochemical mapping results. Furthermore, Tafel slopes were obtained for examining their electrocatalytic activity in Fig. 2(c). The Tafel slopes were 165 and 278 mV dec⁻¹ for edge and basal regions, respectively. In Ref. 35, the pristine WS₂ study of bulk systems without hybridization of highly conductive carbon has a Tafel slope value of 187 mV dec⁻¹. The large slope value indicates that the Volmer reaction is a rate-determining step of the HER. Notably, the value was between the slope values in the edge and basal regions as the mixture of the values in both regions. In other words, the SECCM characterization has a clear advantage as localized electrochemical analysis. Additionally, a comparison was made with MoS₂, which has been extensively studied as a representative TMD. In Ref.18, the Tafel slope value of MoS₂ in the edge region was reported to be about 130 mV dec⁻¹. Therefore, as a good candidate for an electrochemical HER catalyst, the electrochemical activities of WS₂ would need further improvement.

After the investigation of the edge effect on electrochemical HER of WS₂, we have addressed to examine how the number of layers influence the activity. We have carried out the SECCM measurement on the area where mono/bilayer WS₂ sheets were located closely. The SECCM mapping results are displayed with corresponding optical micrographs in Fig. 3(a)—right. Similar to previous results, the edge sites of both mono/bilayer WS₂ were more active than basal regions. The current distribution of the edge regions was plotted in Fig. 3(b), respectively. Interestingly, the intensity ratio of median values on bilayer and monolayer edge regions was calculated as 1.6. It implies that the HER current response at edges is not in a linear relationship with the number of layers. This tendency was also similarly reported in the case of redox activity on graphene edges of Güell et al.³⁶ The possible mechanism of the activity suppression may relate to the electron scattering between layers after the HER process or interfacial resistance between the sample and substrate. This finding serves as a pivotal guideline for designing catalytic electrode structures of TMDs intended for future use as hydrogen generating electrodes. For instance, if the reactivity remains relatively consistent across different numbers of layers, or if the variation intensifies with increasing layers, it implies that the ideal atomic structure of WS2 should focus on achieving a single atomic layer. Additionally, Fig. 3(a) reveals that the basal region on bilayer WS₂ derives a higher current HER response than that of the basal region on a monolayer. Assumingly, this increase in current response may come from defect sites on the basal plane as was reported in a previous study.¹⁰ 

Electrochemical HER performance on mono/bilayer WS₂ has been studied by SECCM probing and mapping technique. Electrochemical HER mapping and Tafel plots derived from LSV results revealed that the edge sites of monolayer WS₂ show higher HER activity than that of the basal plane. Further, the number of layers on WS₂ was also examined, confirming that an increment in the number of layers will enhance the HER activities at edge sites. On the other hand, as the number of layers increased from the monolayer to bilayers, the current response with respect to HER did not follow a proportional relationship, and the current distribution became wider as the increment of the layer. Additionally, increased HER activity was observed in the basal plane of the bilayer as compared to the monolayer, which may be ascribed to defects. WS₂ has great potential for applications in electrocatalytic electrodes for the electrolysis of water. Particularly, TMDs with edge and defects sites provide a key to introduce electrochemical HER for structure engineering.

This work was supported by JSPS KAKENHI Grant Nos. 19H05814, 22H00278, 22H05459, JP21H05234, and JP21H05232; JST PRESTO Grant No. JPMJPR2274; World Premier International Research Center Initiative (WPI), MEXT, Japan; Advanced Target Project and Fusion research Grants, Tohoku University, Japan; The Asahi Glass Foundation; and The Murata Science Foundation, Toyota Mobility Foundation.

The authors have no conflicts to disclose.

Akichika Kumatani: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Writing – original draft (equal). Hiroto Ogawa: Data curation (equal); Formal analysis (equal). Takahiko Endo: Data curation (equal). Yu Kobayashi: Data curation (equal). Jana Lustikova: Formal analysis (equal); Investigation (equal). Hiroki Ida: Formal analysis (equal); Investigation (equal). Yasufumi Takahashi: Supervision (equal); Writing – review & editing (equal). Tomokazu Matsue: Supervision (equal). Yasumitsu Miyata: Data curation (equal); Supervision (equal); Writing – review & editing (equal). Hitoshi Shiku: Supervision (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.