University of Birmingham Electrochemical sulfidation of WS2 nanoarrays:

The activity of transition metal sul ﬁ des for the hydrogen evolution reaction (HER) can be increased by sulfur- enrichment of active metal-sul ﬁ de sites. In this report, we investigate the electrochemical sul ﬁ dation of atmospherically aged WS 2 nanoarrays with respect to enhancing HER activity. In contrast to MoS 2 , it is found that sul ﬁ dation diminishes HER activity. Electrochemical and XPS experiments suggest the involvement of insoluble tungsten oxides in the altered HER and electron transfer properties. This demonstrates the strong dependence of the transition metal dichalcogenide (TMD) composition with the successful sulfur incorporation and subsequent HER activity.


Introduction
The demand for sustainable sources of electrochemical hydrogen production [1] has triggered the development of the abundant and lowcost TMDs as substitutes to the best performing platinum group metal catalysts for the hydrogen evolution reaction (HER) [2][3][4][5].
To improve their HER activity, research has focussed on the preparation of S-rich TMD structures which surpass the 1:2 M:X stoichiometry found in bulk materials [6][7][8]. Such sulfur enrichment aims to incorporate more bridging S 2 2 − and terminal S 2 − moieties into the TMD structure; both consistently reported as being the active sites involved in proton adsorption and desorption [9,10]. Enhanced HER performance has been reported for S-rich structures such as amorphous MoS 2 + x [11][12][13][14][15] and WS 2.64 electrodeposited thin films [9], as well as on wet chemical synthesis-prepared MoS 2 + x [16,17] or MX 3 /MX 2 physical mixtures [18]. However, some of the proposed structures exhibit diminished HER performances after atmospheric or electrochemically-induced sulfur depletion [14,18] or impurities presence [19]. We report the use of a one step, room temperature electrochemical sulfidation method initially developed for MoS 2 [20], for sulfur-enriching WS 2 . In particular, atmospherically-aged WS 2 nanocone arrays which are of interest due to their enhanced electrocatalytic properties [21]. Changes in the electrocatalytic behaviour are understood via monitoring surface composition, morphology, and electron transfer properties over a one month period by XPS, SEM, and voltammetric experiments.
In short, WS 2 (defect-free, 99.9995% purity, 2D Semiconductors USA) crystals cut into rectangles of approximately 1.5 × 5 mm were affixed to glassy carbon (GC) type 2 stubs (7 mm diameter, 2 mm thick, Alfa Aesar, UK) with carbon tape. A 20 μL mixture of a 216 ± 4 nm diameter polystyrene-latex nanosphere (NS) suspension (3000 Series Nanosphere, 1 wt% in water, Thermo Scientific, UK) with absolute ethanol in a 1:1 vol. ratio was transferred to a silicon wafer (previously cleaned with piranha solution and oxygen plasma) to form a self-assembled, hexagonal close-packed, NS monolayer. The NS monolayer was transferred onto the liquid interface of a water-filled Petri dish containing the TMD-modified GC stubs, and the supernatant extracted with a syringe to promote NS deposition onto the TMD surface.
HER experiments were carried out in a 2 mM HClO 4 (ACS ≥70%, Sigma-Aldrich), 0.1 M NaClO 4 (ACS ≥ 98%, Sigma-Aldrich) solution using a range of voltage scan rates (2-1200 mV s − 1 ). Preconditioning of TMD electrodes prior to HER experiments was via 10 cycles from − 0.045 to − 1.645 V (vs SCE) at a voltage scan rate of 50 mV s − 1 . Additional capacitance (voltage range −0.2 to 0.2 V vs. NHE, scan rates 10-500 mV s − 1 ) and impedance measurements (voltage range 0 to −1.645 V vs. SCE, frequency range 10 − 1 to 10 5 Hz, voltage amplitude 10 mV) were performed alongside all HER measurements to apply roughness factor and iR compensation corrections. HER potentials are referenced versus the normal hydrogen electrode (NHE) by means of Nernstian shift correction (E NHE = 0.242 V + 0.059 pH).

XPS measurements
A Kratos Axis HSi X-ray photoelectron spectrophotometer (Aston University) fitted with a charge neutraliser and operated using a Mg Kα (1253.6 eV) achromatic radiation, was used to record spectra at a pressure of < 1 × 10 − 9 Torr using a spot size of 100 μm. Pass energies used were: 160 eV for survey spectra, and 20 eV for high resolution scans of specific energy regions. Data processing was performed using CASA XPS version 2.3.18PR1.0, with spectral energy corrected to the adventitious C 1 s peak at 284.6 eV. Shirley backgrounds were applied to high resolution peaks before being fitted with individual components. W 4f spectra were fit using a FWHM of 0.98 eV, peak area ratios of 4:3, doublet separations of 2.17 eV and Gaussian-Lorentz (30) lineshape, with W 4f 7/2 WS2 2H (32.7 eV), W 4f 7/2 WS 2 1 T (31.7 eV) and W 4f 7/2 WO 2 (33.1 eV) components, whilst the overlapping W 5p 3/2 feature was fitted with a FWHM of 2 eV, a Gaussian-Lorentz (30) lineshape and a binding energy of 38.1 eV. Sulfur 2p peaks were fitted with a FWHM of 1.03 eV, peak area ratios of 2:1 and a binding energy for S 2p 3/2 WS 2 of 162.3 eV [31,32].

Results and discussion
For reinstating, or improving, the initial HER activity of atmospherically-exposed WS 2 samples, a solution-phase method previously demonstrated for MoS 2 nanoarrays [20] was used. This requires voltage cycling of the TMD samples in a pH 3 solution containing 10 mM Na 2 S 2 O 3 and 0.1 M Na 2 SO 4 whereby S 2 O 3 2 − spontaneously decomposes to form colloidal sulfur [33].
An anodic sweep to fully oxidize the TMD surface, was followed by a cathodic scan to maximize sulfur incorporation onto the TMD surface by reduction For MoS 2 , electrochemically-induced surface oxidation (at E > +1 V vs Ag/AgCl) yields the acid-soluble MoO 4 2 − species [34].

Sulfur incorporation after surface oxidation suggests that MoO 4 2 −
species assist in the overall sulfidation mechanism. For WS 2 , the cyclic voltammogram obtained during the sulfidation treatment is similar to that of MoS 2 [20]. The application of this method to WS 2 was evaluated by monitoring HER performance, oxidation state, and electron transfer properties over a one month period following this sulfidation treatment on previously tested, atmospherically aged WS 2 samples. This provided the following observations: (i) freshly sulfidated samples did not necessarily present enhanced HER performances compared with pre-sulfidated samples, and (ii) the samples' HER peak current, after correction for roughness factor, was inferior after a 3-week environmental exposure compared to the pre-sulfidated, atmospherically-exposed state.
Both phenomena can be understood by changes in oxidation state revealed by XPS. For the 31 ± 1 s (R = 2, Z = 6.4) plasma-etched WS 2 sample, the peak current decays to half of its initial value following sulfidation (Fig. 1a). This is correlated to a decrease in the total S:W ratio (from ca. 2:1 to 1.5:1, see Fig. 2e), and the appearance of WO 2 at the crystal surface up to ca. 24% (W 4f 7/2 /W 4f 5/2 doublet lies at binding energies of ca. 33 and 35.2 eV, respectively; Figs. 1c, 2a) [31]. Previous reports on bulk and chemically-exfoliated WS 2 crystals suggest that incorporation of WO 2 is detrimental for the HER [18,35]. In the case of the 31 ± 1 s sample, this is supported by the changed HER kinetics (Tafel slope increase from 100 to 185 mV dec − 1 , Fig. 3c) and higher onset potentials (| η onset | from 173 to 207 mV).
Conversely, the 61 ± 1 s plasma-etched WS 2 sample presented higher peak currents (Fig. 1b) and kinetics (Tafel slope 130 vs. initial 210 mV dec − 1 , Fig. 3d) following sulfidation, despite the decay in the total S:W ratio (from ca. 2:1 to 1.88:1, see Fig. 2f) and the 14% increase in surface WO 2 content (Figs. 1d and 2c). This initially non-linear trend is found to be linked to the S:W ratio, if calculated solely using the W 4 + XPS components characteristic of WS 2 . Sulfur-rich S:W ratios promote enhanced HER performance and vice versa. Maximum peak currents coincide with the highest sulfur-to-metal ratios for both 31 ± 1 s (S:W = 2.08:1, j p ≈ 9 mA cm − 2 , day 8) and 61 ± 1 s (S:W = 2.18:1, j ≈ 1.6 mA cm − 2 , freshly sulfidated) samples. After these peak values, both post-sulfidated 31 ± 1 s and 61 ± 1 s etched samples exhibited an HER current decrease in subsequent electrochemical testing to values lower or comparable with the freshly sulfidated state, due to lower S:W ratios. This accords with previous investigations which correlated higher sulfur content in TMDs with improved hydrogen turnover frequencies [12,14], and sulfur-depleted W-edge sites of electro-oxidised WS 2 with poor catalytic activity [35]. We hypothesize that the electrochemically-induced restructuring gradually depletes the WO 2 phase, initially exposing underlying WS 2 with high active site densities which are later reconstructed during atmospheric and experimental conditions to a more homogeneous nanostructure (Fig. 2g-h).
The cathodic feature appearing at E ca. − 0.4 V vs NHE in the HER experiments ( Fig. 3a and b) is ascribed to the diffusion decay peak profile of proton reduction catalysed by the WS 2 active sites, characteristic of the fully-supported, low proton concentration electrolyte used [36,37]. Indeed, the resolution of this peak also seems correlated with the S 2 − :W 4 + ratio, and consequently to the active sites present.
With regard to the electron transfer kinetics, both samples exhibit higher k app O values (≈4 × 10 − 5 cms − 1 ) after undergoing the sulfidation treatment (Fig. 1e-f). This agrees with literature reports which found enhanced electrical conductivities of WO x species vs. WS 2 [38], beneficial for mediating in the redox chemistry of surface sensitive species such as Fe(CN) 6 4 − /Fe(CN) 6  These results suggest that this sulfidation method does not incorporate sulfur into the atmospherically aged WS 2 samples. Instead, it promotes the appearance of WO x moieties at the WS x surface which are reduced in the cathodic sweep. We hypothesize that in general, the sulfur incorporation is only effective when the electro-oxidative step of TMDs forms acid-soluble species, as sulfur incorporation into atmospherically-aged MoS 2 crystals was optimal when the cathodic voltage vertex surpassed the reduction potential of the TMD oxidised species (MoO 4 2 − ) [20]. In the case of WS 2 , the oxidised WO x species generated during the electro-oxidative step are insoluble at pH ≤ 3 [39], coinciding with the optimized pH value for the sulfidation electrolyte (pH 3). Consequently, the electroreduced sulfur cannot be incorporated into the WO x structure, and would dissolve under acidic conditions [40]. Hence, we predict that the electrochemical solvent-phase sulfidation method is only suited for MoX 2 (X = S, Se) rather than for WX 2 (X = S, Se).

Conclusions
In contrast to MoS 2 , the application of the solution-phase, roomtemperature electrochemical sulfidation method to obtain S-rich structures did not lead to S-rich WS x but to S-deficient WS x structures with high WO 2 surface content. The inferior HER performances but improved electron transfer properties are in agreement with the detrimental effect reported after WO 2 incorporation into WS 2 for the HER catalysis. The unsuccessful incorporation of electroreduced sulfide in the WS x structure is suspected to arise from the nature of the sulfidation mechanism: redeposition of acid-soluble MoO 4 2 − species for MoX 2 improves S 2 − incorporation onto the surface, which is not possible in the case of WX 2 as the WO x compounds formed are acid insoluble. This demonstrates the key role of the nature of the TMDs in the successful electrochemical incorporation of sulfur in their structure, and reveals that an electrochemistry-based sulfidation method universally applicable for any TMDs remains to be developed. Stacked high-resolution XPS spectra of W 4f and S 2p for a)-b) 31 ± 1 s (R = 2, Z = 6.4) and c)-d) 61 ± 1 s atmospherically aged, sulfidation treated, plasma-etched WS 2 samples over a three-week ambient exposure period. e)-f) Comparison of total S:W XPS atomic photoemission ratios. Representative SEM micrographs g) before and h) after solution phase-sulfidation. Fig. 3. Left column: Linear sweep voltammograms in the 0 to − 1.2 V voltage range of a) 31 ± 1 s (R = 2, Z = 6.4) and b) 61 ± 1 s atmospherically aged, sulfidation treated plasmaetched WS 2 samples over a three week ambient exposure period. Right column: Tafel plots (η vs. log |j geom |) of c) 31 ± 1 s (R = 2, Z = 6.4) and d) 61 ± 1 s atmospherically aged, sulfidation treated plasma-etched WS 2 samples over a three week ambient exposure period. Labels: pre-sulfidated (black), post-sulfidated (red), 8-day atmosphere exposed (green), 15-day atmosphere exposed (blue) and 22-day atmosphere exposed (magenta). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)