Two-dimensional metallic tantalum disulfide as a hydrogen evolution catalyst

Two-dimensional metallic transition metal dichalcogenides are emerging as prototypes for uncovering fundamental physical phenomena, such as superconductivity and charge-density waves, as well as for engineering-related applications. However, the batch production of such envisioned transition metal dichalcogenides remains challenging, which has hindered the aforementioned explorations. Herein, we fabricate thickness-tunable tantalum disulfide flakes and centimetre-sized ultrathin films on an electrode material of gold foil via a facile chemical vapour deposition route. Through temperature-dependent Raman characterization, we observe the transition from nearly commensurate to commensurate charge-density wave phases with our ultrathin tantalum disulfide flakes. We have obtained high hydrogen evolution reaction efficiency with the as-grown tantalum disulfide flakes directly synthesized on gold foils comparable to traditional platinum catalysts. This work could promote further efforts for exploring new efficient catalysts in the large materials family of metallic transition metal dichalcogenides, as well as exploiting their applications towards more versatile applications.

1. The authors investigated the effect of temperature on Raman spectra. According to the Figure  4a, there is no significant change in behavior of Raman spectra between 80 to 210 K. It would be better to choose the critical temperature which indicates the effect of heating or cooling, significantly. 2. The DFT calculation is a powerful tool in atomic level calculations. In this wok, the authors tried to justify the electrochemical reaction mechanism by using DFT calculation. They just calculated the Gibbs energy to show and explain the catalyst mechanism. It is not enough for readers. It would be better to bring at least a sentence about the catalytic performance and mechanism or make a reference. 3. What is the effect of temperature on catalytic performance? 4. In this paper, the authors mentioned the electrical transport, but they didn't about the amount of electrical transport. What is your definition from electrical transport? (e.g. mobility, conductivity, sheet resistance, current etc.) In this revised version, we have made great efforts to address the points raised by reviewers by performing more supplementary experiments, especially for Reviewer 2, regarding the possible Pt contamination on the HER efficiency of 2H-TaS 2 /Au foils sample with the Pt working electrode, and regarding the effect of synthesis temperature on the HER performance of TaS 2 raised by Reviewer 3.
Through these careful revisions, we sincerely hope that, our work would be of great interest to the broad readership of Nature Communications in both fundamental researches for the growth of metallic transition metal dichalcogenides and their applications in energy related fields.

Reviewer #1 (Remarks to the Author): The authors developed the synthetic method of thickness-controllable TaS 2 through LPCVD and APCVD. The approach is meaningful because growth of metallic TMDC is very limited. Also, they showed TaS 2 which have extraordinary performance for hydrogen evolution reaction. However, several factors should be mentioned clearly.
Our response: We are very grateful for the reviewer's very positive evaluation on the significance of our work, especially regarding the successful growth of metallic transition metal dichalcogenides (TMDCs) and the extraordinary performance for hydrogen evolution reaction (HER). The issues raised by the reviewer are considered very carefully and addressed point-by-point as follows.
1. The authors explained the flakes are getting smaller, thinner and more disperse after 5000 cycling of HER then the performance is enhanced (Page 11, line 301). They claimed that the thinner effect brought short electron-pathway on basal plane and increasing surface area by improving the accessibility of proton to basal plane so HER performance is enhanced. But, smaller flakes also mean edge sites of TaS 2 would be increased, and the activity of HER can come from more edges than basal plane. I am not sure whether the good performance is from basal plane. The author need to be more careful and clear to explain about the active sites.

Our response:
We are very thankful for the reviewer's very constructive comment. We agree with the reviewer that, the edge sites of TaS 2 will be increased for the smaller flakes after cycling. According to the theoretical calculation data shown in Fig. 5c, the free energies of H adsorption (ΔG H* ) on Ta-edge (-0.04 eV) and S-edge (-0.10 eV) are much closer to thermos-neutral (ΔG H* = 0 eV) than that of the basal-plane (0.15 eV). The edge sites should be more active in the electrocatalytic activity for HER than that of the basal plane. That is to say, the smaller TaS 2 flakes with more edge sites (after cycling) should possess much better HER performance than that of the initial larger TaS 2 domains.
According to the kind suggestion, we have clarified this discussion in page 12 by "In our opinion, the cycling induced morphology change has three beneficial effects on the catalytic activity…, 3) increasing the density of active site at the flake edge of 2H-TaS 2 , considering that the ΔG H* values of both Ta-edge and S-edge are much closer to the thermos-neutral point than that of basal-plane, and the edge sites are catalytically more active than that of the basal plane…"

TaS 2 with 150 nm shows the best HER results in Figure 5d. However, the initial HER performance of figure 5f doesn't show as good as figure 5d, although they have same thickness (150 nm). The TaS 2 with 150 nm in Figure 5d is same as after 5000 cycling in figure 5f. That point is confusing. Does it have the best result only after cycling? If they have good results only after cycling, TaS 2 doesn't have good catalytic activities.
Our response: We are very grateful for the instructive comment raised by the reviewer. The HER results in Figs. 5d,f are derived from the similar TaS 2 samples with the same thickness. Notably, in consideration that the TaS 2 is a metallic TMDC material, which is unstable under ambient condition and the surface of TaS 2 can be oxidized at the atmospheric condition before the HER measurement. Ohsaka, T. et al., has shown that, these surface oxides (TaO x ) have extra low electrocatalytic activity for HER, featured with much larger overpotential at 0.1 mA/cm −2 (~250 mV) than that of TaS 2 (~1 mV) (Ohsaka, T. et al., J. Mater. Chem. A 3, 16791−16800 (2015)). As a result, at the initial state, the 2H-TaS 2 /Au has relatively low HER activity. Particularly, the surface oxides can be peeled off by hydrogen bubbles during the HER cycling process, and the intrinsic electrocatalytic activities of TaS 2 are presented subsequently (after 500 cycling). Therefore, it is proposed that, the TaS 2 flakes possess extra high electrocatalytic performance after removing the surface oxides through a facile HER cycling process.
We have added this discussion in page 12 by "…Notably, considering that the TaS 2 is a metallic TMDC material, which is unstable under ambient condition and the surface of TaS 2 can be oxidized at the atmospheric condition, and these surface oxides (TaO x ) have extra low electrocatalytic activity for HER 50 . Through a facile HER cycling process, the surface oxides can be peeled off by hydrogen bubbles and the intrinsic electrocatalytic activities of 2H-TaS 2 are presented subsequently…" We have also added this reference in page 12 (in Ref. 50).

Our response:
We are very thankful for the reviewer's very constructive comment. The polarization curves in Fig. 5d have shown that the Au foil has very low electrocatalytic activity.
In order to further exclude the possible penetration of Au atoms into the TaS 2 layers in the CVD growth process, we have performed XPS measurements of transferred TaS 2 on SiO 2 /Si (Fig. R1, also shown in Supplementary Fig. 6 in the revised manuscript). The Au 4f characteristic peaks are not observed, indicating that the Au element is not penetrated into TaS 2 during the CVD growth process.
We have added this discussion in page 6 by "…The XPS measurement of transferred 2H-TaS 2 on SiO 2 /Si was also performed to exclude the possible Au penetration into the 2H-TaS 2 layers during the CVD growth process ( Supplementary Fig. 6

)…"
More details were provided in Supplementary Information (Supplementary Fig. 6) by "The XPS measurement of transferred 2H-TaS 2 flakes on SiO 2 /Si was also performed to exclude the possible mixing of Au in the 2H-TaS 2 layers during the CVD growth process ( Supplementary Fig. 6). The Au 4f characteristic peaks are not observed, strongly indicating that Au is not penetrated into TaS 2 layers during the growth process." Figure R1 (also shown in Supplementary Fig. 6

in the revised manuscript) | XPS spectrum of transferred 2H-TaS 2 on SiO 2 /Si acquired over a wide range of binding energies (0-1100 eV). Inset is the zoon-in scan from
the Au 4f range.

Our response:
We are very thankful for the reviewer's very constructive comment. We have added the deconvolution peaks from Ta and S in Fig. R2 (also shown in Fig. 1b and Fig. 2b). We have also supplied a sentence in page 4 by "…Notably, additional peaks at 28.4 and 26.9 eV are attributed to Ta 5+ , in consideration of the oxidation susceptibility of metallic TaS 2 …" Figure R2 (also shown in Fig. 1b and Fig. 2b

Reviewer #2 (Remarks to the Author): The paper by Shi et al reports the CVD growth of two-dimensional 2H TaS 2 on gold substrate and unusually high catalytic activities towards hydrogen evolution reactions, together with carefully structure characterizations using AFM, XPS, TEM and Raman. The paper reads very well and the results are new and interesting. In principle I agree the paper for acceptance of publication after minor revision.
Our response: We are very grateful for the reviewer's very positive evaluation on the novelty of our work. The issues raised by the reviewer are considered very carefully and addressed point-by-point as follows.

Our response:
We are very thankful for the reviewer's very constructive comment.
Three results have been presented to rule out the possible Pt contamination on the HER performance of 2H-TaS 2 : (1) The HER performances of 2H-TaS 2 /Au samples with different thicknesses have been re-measured by using the carbon rod as the counter electrode, with the results shown in Figs. R3a-c (also shown in Supplementary Figs. 14a-c, and Supplementary Table 1 in the revised manuscript). At the cathodic current density (j) of 10 mA/cm 2 , the overpotentials (η) are falling in 100-190 mV, the Tafel slopes are calculated to be 50-55 mV/dec, and the exchange current density (j 0 ) are measured to be 85-143.2 µA/cm 2 . Such three parameters are very close to the results achieved by using Pt as the counter electrode, preliminary indicating that the Pt counter electrode has very small effect on the HER performance of 2H-TaS 2 . Notably, in consideration of the relatively low conductivity of carbon rod, the Tafel slope values are slightly larger than that of the results obtained from the HER measurement using Pt counter electrode; (2) The Nafion proton exchange membrane assisted electrochemical measurements are also performed, as shown in Figs. R3d-f (also shown in Supplementary Figs. 14d-f, and Supplementary Table 1 in the revised manuscript). The overpotentials at j = 10 mA/cm 2 are 80-160 mV, the Tafel slopes are 45-49 mV/dec, and the j 0 values are 90-149.5 µA/cm 2 , such results are very close to those of Pt as the counter electrode, further suggestive the negligible effect of Pt on the HER performance of TaS 2 /Au; (3) The XPS characterizations of 2H-TaS 2 after HER measurements were also performed (with the data shown in Supplementary Fig. 13). The Pt characteristic peaks are not observed, indicating that Pt is not covered on or mixed in TaS 2 during the HER process.
The aforementioned three results suggest that, the Pt counter electrode has negligible effect on the HER performance of 2H-TaS 2 .
We have added this discussion in page 12 by "…In order to rule out the effect of the possible Pt contamination on the HER performance of 2H-TaS 2 , we have re-measured the HER performance of 2H-TaS 2 /Au by using the carbon rod as the counter electrode, and performed the Nafion proton exchange membrane assisted electrochemical measurement, respectively. Similar catalytic results have been achieved among the different methods, highly indicative the extra high elextrocatalytic performance of 2H-TaS 2 /Au foils (Supplementary Fig. 14 More results were provided in Supplementary Information (Supplementary Fig. 14) by "In order to rule out the possible effect of Pt contamination, we have re-measured the HER performance of various 2H-TaS 2 samples on Au foils by using the carbon rod as the counter electrode. The overpotentials at the cathodic current density of 10 mA/cm 2 are falling in the range of 100-190 mV, and the Tafel slope values are in the range of 50-55 mV/dec. Such results approach to those obtained from the Pt counter electrode based HER measurements. Additionally, by applying extrapolation method to the Tafel plots, the exchange current densities (j 0 ) are also achieved in the range of 85-143.2 µA/cm 2 , similar with those of using the Pt counter electrode and Supplementary Table 1). Notably, such three parameters are very close to the results derived by using Pt as the counter electrode. Briefly, the Pt counter electrode has very small effect on the HER performance of 2H-TaS 2 . And the slightly larger Tafel slope value than that of using the Pt counter electrode based HER measurement may be induced by the relatively low conductivity of the carbon rod.
We also performed the Nafion proton exchange membrane assisted electrochemical measurements by using Pt counter electrode. This is because the proton exchange membrane only permits the transfer of protons but impedes the other species. The overpotentials for the different samples at j = 10 mA/cm 2 are in the range of 80-160 mV, the Tafel slopes are 45-49 mV/dec, and the j 0 values are 90-149.5 µA/cm 2 , such values are very close to the results only by using Pt as the counter electrode and Supplementary Table 1 Supplementary Fig. 14

Electrochemical impedance spectroscopy should offer experimental evidence to support the authors' claims on the effect of gold substrates on charge transfer and the kinetic origins of the high HER activities of 2H TaS 2 . Suggest the authors add those data in the paper.
Our response: We are very thankful for the reviewer's very kind suggestion. The electrochemical impedance spectroscopies (ESI) of 2H-TaS 2 samples and Au foils have been added in Fig. R4 (also shown in Figs. 5g,h in the revised manuscript). We have also added some discussion in page 13 by "…The extra low charge-transfer resistance (5-11 Ω)

The authors reported that a 2H-TaS 2 /Au foil could be used as catalyst in high hydrogen evolution reaction. The manuscript is clearly written. But, there are somethings to be revised for readers.
Our response: We are very grateful for the reviewer's very positive evaluation on our work. The issues raised by the reviewer are considered very carefully and addressed point-by-point as follows.

The authors investigated the effect of temperature on Raman spectra. According to the Figure 4a, there is no significant change in behavior of Raman spectra between 80 to 210 K. It would be better to choose the critical temperature which indicates the effect of heating or cooling, significantly.
Our response: We are very thankful for the reviewer's kind comment. For the cooling process, a broad Raman peak is observed below 100 cm −1 at >150 K. In contrast, the fine peak (79.8 cm −1 , indicated by a red arrow in Fig. 4a) is visible at <150 K, suggesting that a NCCDW/CCDW phase transition takes place at ~150 K. Notably, a similar change tendency is also recorded for the heating process, namely, a broad Raman peak is visible below 100 cm −1 at >210 K and the fine peak is noticeable at <210 K, indicating that the NCCDW/CCDW phase transition takes place at ~210 K for the heating process.
In order to precisely determine the critical temperature of NCCDW/CCDW phase transition, we have re-measured the Raman spectra of 2H-TaS 2 flakes from 80 to 210 K in Fig. 4a and added this discussion in page 10 by "…Therefore, the critical temperature of NCCDW/CCDW phase transition is identified as ~150 and ~210 K for the cooling and heating processes, respectively..."

The DFT calculation is a powerful tool in atomic level calculations. In this wok, the authors tried to justify the electrochemical reaction mechanism by using DFT calculation. They just calculated the Gibbs energy to show and explain the catalyst mechanism. It is not enough for readers. It would be better to bring at least a sentence about the catalytic performance and mechanism or make a reference.
Our response: We are very grateful for the constructive comment raised by the reviewer. We agree with the reviewer that, the DFT calculation is a powerful tool to understand and explain the catalyst mechanism.
We have supplemented some discussion about the catalytic performance and the internal mechanism of 2H-TaS 2 according to the DFT calculation results from the published references in page 11 by "…In Fig. 5c,

What is the effect of temperature on catalytic performance?
Our response: We are very thankful for the reviewer's constructive comment. We have measured the HER performance of TaS 2 flakes with the similar coverage and domain size but synthesized at different temperature (650, 700, 750, 800, and 850 ºC) in Figs. R5a,b (also shown in Supplementary Figs. 16a,b in the revised manuscript). Interestingly, the HER performances of TaS 2 flakes synthesized at 650, 700, and 750 ºC are very similar, featured with the similar overpotential values at the cathodic current density of 10 mA/cm 2 of 120-125 mV, and the Tafel slope values of 45-48 mV/dec. This phenomenon indicates that, the growth temperature (from 650 to 750 ºC) has negligible effect on the HER performance of TaS 2 . However, for the TaS 2 flakes synthesized at higher temperature (800 and 850 ºC), the overpotential (at the cathodic current density of 10 mA/cm 2 ) and Tafel slope values are raised to 152-171 mV and 58-62 mV/dec, respectively. This indicates that the high synthesis temperature reduces the HER performance of the derived TaS 2 flakes. Briefly, the low synthesis temperature (<800 ºC) has very little effect on the HER performance, and the catalytic activity of TaS 2 can be reduced at higher growth temperature (>800 ºC). This difference is mainly mediated by their different phases of 2H-TaS 2 and 1T-TaS 2 . Supplementary Fig. 16  We have supplied this details in Supplementary Information (Supplementary Fig. 16) by "The HER performances of TaS 2 samples synthesized at different temperatures were also measured to detect the effect of synthesis temperature on the catalytic activity (Supplementary Figs. 16a,b We have also added a short discussion in page 13 by "…The effect of CVD synthesis temperature on the HER performance of TaS 2 was also presented in Supplementary Fig. 16, where low synthesis temperature (<800 ºC) has negligible effect on the HER performance of TaS 2 , and the high synthesis temperature (>800 ºC) reduces the catalytic activity of TaS 2 . This can be explained from the generation of different phases..."

In this paper, the authors mentioned the electrical transport, but they didn't about the amount of electrical transport. What is your definition from electrical transport? (e.g. mobility, conductivity, sheet resistance, current etc.)
Our response: We are very grateful for the reviewer's very constructive comment. The electrical transport in the revised manuscript has been defined in page 6 by "…Recent electrical transport measurements (temperature-dependence resistance) have revealed that…"