Overall water splitting by graphdiyne-exfoliated and -sandwiched layered double-hydroxide nanosheet arrays

It is of great urgency to develop efficient, cost-effective, stable and industrially applicable electrocatalysts for renewable energy systems. But there are still few candidate materials. Here we show a bifunctional electrocatalyst, comprising graphdiyne-exfoliated and -sandwiched iron/cobalt layered double-hydroxide nanosheet arrays grown on nickel foam, for the oxygen and hydrogen evolution reactions. Theoretical and experimental data revealed that the charge transport kinetics of the structure were superior to iron/cobalt layered double-hydroxide, a prerequisite for improved electrocatalytic performance. The incorporation with graphdiyne increased the number of catalytically active sites and prevented corrosion, leading to greatly enhanced electrocatalytic activity and stability for oxygen evolution reaction, hydrogen evolution reaction, as well as overall water splitting. Our results suggest that the use of graphdiyne might open up new pathways for the design and fabrication of earth-abundant, efficient, functional, and smart electrode materials with practical applications.

the oxygen and hydrogen evolution reactions and even a bifuctional electrocatalyst. The catalyst presents low overpotential and long-term stability and higher catalytic activity than other reactive catalyst. The manuscript is well organized and solid. I believe this important study will provide a new idea for GDY application. Therefore, I recommend it to publish in nature communication with the following revisions: (1) The authors state that they combine graphdiyne (GDY) with exfoliated iron/cobalt LDH nanosheet arrays grown on nickel foam. The LDH was exfoliated through the intercalating of monomer (HEB). How dose it work?
(2) According to the schematic representation ( Fig. 1), the LDH nanosheets were exfoliated into several thicker nanosheets (from 80 nm to 7 nm). Actually, from SEM images in Fig. 2, it is the same that the number of nanosheet arrays of LDH is not increase after GDY growth. The authors should give the reasons. Is it possible other reasons? Such as the corrosion of pyridine or TMEDA，which makes the LDH layer more thicker.
(3) What are the thickness of GDY film and LDH? (4) In the FT-IR spectra, a sharp peak presents in 2103 cm-1. It corresponds to the stretching of C≡C bonds. According to your previous reported work, the peak at 2103 cm-1 is agree with the terminal alkynes and IR signal of the C≡C bonds in GDY should be as weak as possible because of the molecular perfectly symmetry. How to explain the sharp peak presents in 2103 cm-1 ? (5) According to the electrochemical testing, whatever OER, HER or Overall water splitting, the catalytic activity of e-ICLDH@GDY/NF is better than those of previously reported LDH-related catalysts and nearly close to the Pt. Does it is possible to be used in practical applications? What are the challenges? (6) The authors used two classic synthetic methods (GDY nanowall synthesis and GDY synthesis on arbitrary substrates) for GDY preparation. The related literatures should be cited (J. Am. Chem. Soc. 2015, 137, 7596;Adv. Mater. 2017, 29, 1605308). (7) There are several mistakes about Pt or RuO2 in Fig. 8 and 9 of SI.
Reviewer #3 (Remarks to the Author): The manuscript "Overall water splitting by graphdiyne-exfoliated and -sandwiched layered doublehydroxide nanosheet arrays" demonstrates a highly active catalyst composed of graphdiyne and FeCo LDH for OER and overall water electrolysis. The performance of the electrodes is much higher than that of majority of previous investigations. Here, some points are still not clear enough to understand this work and need further explanation is necessary to help comprehend it.
1. In the experimental section, Cu foils were added during the preparation of e-ICLDH@GDY. What is the effect of Cu foils? Is it possible to dope Cu into e-ICLDH@GDY during this process? More, in supplementary figure 2, the peak at ~8.9keV in EDS spectrum is not indexed to Co, Ni or Fe. However, it is very close to Cu. 2. Considering the reduced thickness of nanosheets from tens of nanometers to several nanometers. The volume expansion should be very large. However, this wasn't observed based on SEM images in Figure 2. Then, is there any loss of the electroactive materials during the growth of GDY? 3. The e-ICLDH@GDY/NF catalyst exhibited good stability for OER, and how is the chemical state of GDY after OER stability test? Is there any possibility of being oxidized for GDY? 4. What are the difference and relationship between figure 6g and supplementary figure 8, which is very confusing? In figure 6, the e-ICLDH@GDY/NF sample performed best for HER, but in figure s8, the activity of e-ICLDH@GDY/NF was very poor and ICLDH/NF performed very well for HER. And in figure 6g/6h and figure s9, the labels of Pt/RuO2 are very confused. Please be careful on those labels, which can significantly mislead the readers. 5. In figure 6k and figure s11, the fitting data are not clear enough with circles, lines may be better and preferred. 6. Are the CV curves in figure 7b iR-corrected or not? If they are iR-corrected, it is better to supply the pristine data without iR corrections. For overall water splitting, polarization curves without iR corrections are more convincing from the view of practical applications. The cell voltage of 0.5V should be double confirmed in figure 7i, which is unbelievable to drive such a high current density for overall water splitting. 7. The authors claimed that "…a greater density of electrons at the electrocatalyst surface should be of more help for the formation of *OOH and thus enhance the OER activity…" (line 230 and 231). However, in figure 5g and 5h, results showed that both Co and Fe suffered great electron transfer to GDY, which is also the result of theoretical calculations. It seems a paradox between the results and analysis. How to understand the effect of electron transfers between metal and GDY on the catalytic activity? More discussions may be given here to clarify this effect for both OER and HER. 8. In this work, the very large catalytic current densities for both HER and OER are driven by small potentials, which is very impressive. In this case, the Faraday efficiencies of OER, HER and overall water splitting should be given to further confirm the performance of electrodes.
Page 13 lines 21-22: Our calculated results (Fig. 5j) further show that the ICLDH@GDY has the smallest Gibbs free energy of 0.34 eV for HER compared with that of pristine GDY (0.46 eV) and ICLDH (1.55 eV), indicating the excellent HER activity for GDY incorporated ICLDH.  Figure S5 and HER results in Figure S8, we can see there are two major factors that contributed to the enhanced activity. One is Fe incorporation into CLDH, and another is GDY introducing. Moreover, from the polarization curves, it is obvious that Fe incorporation contributes more. However, the authors have not mentioned this point at all. More discussions should be added.

Responses:
Thank you for your valuable suggestions. Corresponding discussions have been added in the revised text.
Page 16 lines 6-17: From the OER results ( Supplementary Fig. 9), we can see, there are two major factors can make the enhanced activity. One is GDY incorporation, and another is Fe introduction into CLDH.
From the polarization curves, obviously, incorporation of Fe has greater effect on the activity.
Theoretical studies showed the effects of Fe incorporation (Supplementary Fig. 11 and Supplementary Methods). After the configuration optimization, the calculated binding energies (Eb) are -2.01 eV and -2.26 eV for CLDH@GDY and ICLDH@GDY, respectively, which shows that the incorporation of Fe atoms contribute to the binding between GDY and LDH.
Besides, we have also examined the charge transfer between the LDH and GDY. Compared with pure CLDH (0.70 ‫,)|‪e‬|‬ a more electron in total (0.80 ‫)|‪e‬|‬ was transferred from ICLDH to GDY, indicating that the incorporation of Fe can further enhance the electron transfer, which would be helpful for improving the overall catalytic activity.
Page 19 lines 14-16: Similar to that observed in OER ( Supplementary Fig. 9), the HER results ( Supplementary Fig.   13) also showed that both the GDY introducing and Fe incorporation into CLDH contributed to the improvement of the HER activity, and Fe incorporation contributes more.

5.
The authors attribute the improvement of catalytic activity of e-ICLDH@GDY/NF to the introduction of GDY, which significantly facilitates electron transport and increases the number of electrocatalytic active sites. In order to show whether the intrinsic activity is improved or not, TOFs for both HER and OER should be calculated and compared with those of other catalysts.

Responses:
Assuming that all metal ions in the nanosheets were active, the activity of samples was further estimated by their apparent turnover frequencies (TOFs). This provides a crude upper bound for TOF values. The TOF value was calculated according to the following equation: where j is the current density at an overpotential; A is the area of LDHs electrode; n is the stoichiometric number of electrons consumed in the electrode reaction; F is the faraday constant; N is molar number of metal ions on the electrode. The metal content of the catalysts was calculated from the ICP-AES data. TOF values are 12.77 s −1 at η = 100 mV for HER and 28.16 s −1 at η = 250 mV for OER, respectively. These values are larger than most of recently reported catalysts, as summarized in Table R1.
These have been added in the revised manuscript.
Page 21 lines 3-7: Turnover frequencies (TOFs) have also been used to assess the information about the intrinsic activity. Assuming that all sites were active, the e-ICLDH@GDY/NF. TOF values are 12.77 s −1 at 100 mV for HER and 28.16 s −1 at 250 mV for OER, respectively. These values are larger than most of recently reported catalysts (Supplementary Table 4).

Responses:
-It is a very good question. Under ideal conditions, when the cathode surface is the same to the anode surface, the voltage for overall water splitting would be equal to the voltage difference (ΔV) between HER and OER at the same current density. However, in practical operations, we can find some differences from the ideal state owing to the surface complexity. This phenomenon can also be found in some literatures (Table R1), such as Nat.
Commun ICLDH@GDY/NF. This oxidation peak corresponds to the oxidation of Co 2+ to Co 3+ . Table R1. Summary of the potentials for HER, OER, ∆VOER-HER and OWS at 10 mA cm -2 , respectively. Catalysts Ref.  This has been corrected in the revised paper. Figures 7g to 7i have been discussed in the revised paper.

Reviewer #2:
The study of green energy has become an increasingly important issue due to the change of climate and the exhaustion of fossil fuels. Splitting of water into hydrogen and oxygen is a sustainable and very important technology for energy conversion and storage. Oxygen and hydrogen evolution reactions are the basic half reaction for water splitting. Developing efficiency, inexpensive and sustainable catalyst plays a key role for water splitting reaction. In this work, the authors developed a comprising graphdiyne and iron/cobalt layered doublehydroxide nanosheet arrays and applied as the oxygen and hydrogen evolution reactions and even a bifunctional electrocatalyst. The catalyst presents low overpotential and long-term stability and higher catalytic activity than other reactive catalyst. The manuscript is well organized and solid. I believe this important study will provide a new idea for GDY application.
Therefore, I recommend it to publish in Nature Communication with the following revisions: Adv. Mater. 2017, 29, 1605308). The GDY films could be uniformly grown on LDH surfaces.
Owing to the flexibility nature of both GDY and LDH, the stress/deformation caused by the intimate contact in-between will further enlarge the layer spacing, leading to the complete exfoliation of b-LDHs into thicker LDH nanosheets (e-LDH). The resulted product is denoted here as e-LDH@GDY/NF.
Corresponding discussions have been added in the revised manuscript.
Page 6 lines 1-9: In general, the anions located in the interlayer regions can be easily replaced. And the large interlayer distance in LDH make it easy for the entrance of hexaethynylbenzene (HEB) monomers into the LDHs gallery. The acetylenic hydrogen of HEB could then form hydrogen bonds with the hydroxide layer, which can lead to the formation of HEB film on the LDH surfaces. A Glaser−Hay reaction would then take place efficiently with the aid of TMEDA. The GDY films could be uniformly grown on LDH surfaces. Owing to the flexibility nature of both GDY and LDH, the stress/deformation caused by the intimate contact in-between will further enlarge the layer spacing, leading to the complete exfoliation of b-LDHs into thicker LDH nanosheets (e-LDH). The resulted product is denoted here as e-LDH@GDY/NF.

2.
According to the schematic representation ( Fig. 1), the LDH nanosheets were exfoliated into several thicker nanosheets (from 80 nm to 7 nm). Actually, from SEM images in Fig. 2, it is the same that the number of nanosheet arrays of LDH is not increase after GDY growth. The authors should give the reasons. Is it possible other reasons? Such as the corrosion of pyridine or TMEDA, which makes the LDH layer more thicker.
Responses: As can be seen from Fig. 2 in the manuscript, it is clear that the number of e-ICLDH@GDY nanosheets is much larger than that of pristine ICLDH nanosheets. The density of e-ICLDH@GDY nanosheets was roughly calculated to be ~70 nanosheet per μm 2 , which is much larger than that of pristine ICLDH nanosheets (~18 nanosheet per μm 2 ). In addition, pristine ICLDH nanosheets exhibit a thickness of approximately 40-80 nm and a smooth surface. After the cross-coupling reaction, nanosheets with thinner thickness (around 7.0 nm) and more winkled morphology were observed.
In order to determine whether the pyridine or TMEDA can damage the morphology of the LDHs, pristine LDHs were put into the reactor containing pyridine or TMEDA only (without HEB). Other experimental conditions were the same as that used for synthesizing e-LDH@GDY/NF. SEM images exhibited almost no changes in morphology before and after the treatments ( Fig. R4 and R5). These results further indicated the GDY-induced exfoliation of the bulk LDHs. Based on these facts, we employed the schematic diagram for illustrating the e-LDH@GDY/NF synthetic process. These discussions have also added in the revised manuscript as presented here.
In order to confirm whether the morphologies of the LDHs in pyridine and TMEDA can be damaged, pristine LDHs were put into the reactor containing pyridine or TMEDA only (without HEB). Other experimental conditions were the same as that used for synthesizing e-ICLDH@GDY/NF. SEM images exhibited no any changes in morphology before and after the treatments by pyridine or TMEDA (Supplementary Figs. 1 and 2).

What are the thickness of GDY film and LDH?
Responses: Figure R6 shows HRTEM image of the lateral standing e-ICLDH@GDY nanosheet. Because of the imaging angle, the GDY layer coated on the other side was not observed. The thickness of GDY (or LDH) is estimated from the number of fringes corresponding to the distance between two consecutive carbon (or LDH) fringes. Therefore, the thickness of GDY and LDH are determined to be about 1.2 nm and 5.1~6.5 nm, respectively.
Corresponding contents were also added in the revised text as follows.
Page 10 lines 11-14: Supplementary Fig. 3 shows the HRTEM image of the lateral standing e-ICLDH@GDY nanosheet. The thickness of GDY and LDH are determined to be about 1.2 nm and 5.1-6.5 nm, respectively, which is in accordance with the SEM (Fig. 2i) and AFM imaging (Figs. 2o and 2p) results. Figure R6. HRTEM image of the lateral standing e-ICLDH@GDY nanosheet. The thickness of GDY and LDH are determined to be about 1.2 nm and 5.1~6.5 nm, respectively.

4.
In the FT-IR spectra, a sharp peak presents in 2103 cm -1 . It corresponds to the stretching of C≡C bonds. According to your previous reported work, the peak at 2103 cm -1 is agree with the terminal alkynes and IR signal of the C≡C bonds in GDY should be as weak as possible because of the molecular perfectly symmetry. How to explain the sharp peak presents in 2103 cm -1 ?

Responses:
The peak at 2103 cm −1 corresponds to the internal alkyne stretch. Our previous reported work (Chem. Commun., 2010, 46, 3256-3258) also indicated that the band appeared here corresponded to the typical C≡C stretching (not the terminal alkynes).
In fact, the signal indicative of the presence of C≡C in the FT-IR spectra is really very weak.
As shown in Fig. 4a (the original spectra of our samples), it is very hard to see signals in the region of 2050 to 2150 cm -1 . In order to make the readers more clearly see, we magnified this region as shown in Fig. 4c. But it is sorry that the over-magnification leads to a misunderstanding. The updated figure with proper magnification was shown in Fig. R7c.

5.
According to the electrochemical testing, whatever OER, HER or Overall water splitting, the catalytic activity of e-ICLDH@GDY/NF is better than those of previously reported LDHrelated catalysts and nearly close to the Pt. Does it is possible to be used in practical applications?
What are the challenges?
Responses: It is very good question. The material of e-ICLDH@GDY/NF is easy to prepare, easy to treatment, and shows excellent catalytic activity, which is one of the most promising electrocatalysts for practical applications. The combined material not only shows excellent catalytic activities but also exhibit robust, long-term stability. These superior properties are rarely found together, and their combined effect make it very possible to be used in practical applications.
According to our results and previously reported literatures, the combination of monolayer LDHs and monolayer GDY should have a great potential to further enhance the catalytic activity. This is still a big challenge. Related research work is now underway.

Responses:
The related literature have been cited in the revised text.

7.
There are several mistakes about Pt or RuO2 in Fig. 8 and 9 of SI.

Responses:
Corrections have been made in the revised manuscript.

Reviewer #3:
The manuscript "Overall water splitting by graphdiyne-exfoliated and -sandwiched layered double-hydroxide nanosheet arrays" demonstrates a highly active catalyst composed of graphdiyne and FeCo LDH for OER and overall water electrolysis. The performance of the electrodes is much higher than that of majority of previous investigations. Here, some points are still not clear enough to understand this work and need further explanation is necessary to help comprehend it.
1. In the experimental section, Cu foils were added during the preparation of e-ICLDH@GDY.
What is the effect of Cu foils? Is it possible to dope Cu into e-ICLDH@GDY during this process?
More, in supplementary figure 2, the peak at ~8.9 keV in EDS spectrum is not indexed to Co, Ni or Fe. However, it is very close to Cu.

Responses:
-In the experiment, Cu foils can easy release Cu ions in the presence of a catalytic amount of base to form Cu-based catalysts for the Glaser−Hay reaction.
-Cu species cannot be doped into the e-ICLDH@GDY structure. In order to avoid the copper contamination, all e-ICLDH@GDY/NF products were washed with 2 M HCl, 2 M NaOH, acetone, DMF and deionized water subsequently before use (please see Methods section in the manuscript for details). Moreover, XPS (Fig. 4), HRTEM (Fig. 3) and EDS (Supplementary Figure 4) results also demonstrated that there is no Cu species present in the e-ICLDH@GDY nanosheets.
-It needs to be mentioned that the EDS spectrum shown in Supplementary Figure 2 (Supplementary Figure 4 in the revised manuscript) was measured on a molybdenum (Mo) grid. We have discussed with the engineer regarding the peak located at ~8.9 keV. She said that this peak was ascribed from to the TEM system, for example, the sample rod. In order to clarify this issue, a control measurement on pure molybdenum grid was conducted. The obtained EDS result is shown in Fig. R8.   3. The e-ICLDH@GDY/NF catalyst exhibited good stability for OER, and how is the chemical state of GDY after OER stability test? Is there any possibility of being oxidized for GDY?

Responses:
The chemical state of the sample after OER durability tests has been examined by using XPS measurements (Fig. R11). In the C 1s spectrum, compared with pristine GDY, there is a slightly increasing in the relative content of O=C-OH from 9.9 % to 15.3 % (located at 288.6 eV), indicating the slight oxidization of GDY after stability test. Please be careful on those labels, which can significantly mislead the readers. 5. In figure 6k and figure s11, the fitting data are not clear enough with circles, lines may be better and preferred.

Responses:
These have been changed in the revised text.
6. Are the CV curves in figure 7b iR-corrected or not? If they are iR-corrected, it is better to supply the pristine data without iR corrections. For overall water splitting, polarization curves without iR corrections are more convincing from the view of practical applications. The cell voltage of 0.5V should be double confirmed in figure 7i, which is unbelievable to drive such a high current density for overall water splitting.

Responses:
The CV curves in Figure 7b were iR-corrected. According to the review's suggestion, polarization curves without iR corrections are provided (Fig. R12).
The cell voltage in Figure 7i is 1.56 V versus RHE. This has been corrected in the revised text.
These results have been updated in the revised manuscript. Figure R12. CV curves (without iR-correction) for overall water splitting.
7. The authors claimed that "…a greater density of electrons at the electrocatalyst surface should be of more help for the formation of *OOH and thus enhance the OER activity…" (line 230 and 231). However, in figure 5g and 5h, results showed that both Co and Fe suffered great electron transfer to GDY, which is also the result of theoretical calculations. It seems a paradox between the results and analysis. How to understand the effect of electron transfers between metal and GDY on the catalytic activity? More discussions may be given here to clarify this effect for both OER and HER.
Responses: It seems that there was a misunderstanding about the chemical structure of our reported electrocatalyst (e-ICLDH@GDY). As detailed described in the manuscript, the synthesized e-ICLDH@GDY comprises of ICLDH core and GDY shell. Our experimental and theoretical results showed that the great electron transfer from Fe and Co to GDY could result in the increase of the electron density in GDY structure (the surface structure of our electrocatalyst). For OER, e-ICLDH@GDY has a smaller free energy of 0.54 eV for the formation of *OOH species than pure GDY (0.93 eV, Fig. R13), which indicates an enhanced OER activity for e-ICLDH@GDY. For HER, e-ICLDH@GDY possesses the smallest Gibbs free energy (0.34 eV) compared to pure GDY (0.46 eV) and ICLDH (1.55 eV), indicating the best HER activity for GDY modified ICLDH (Fig. R14).
Corresponding discussions have been added in the revised manuscript as follows: Page 15 lines 3-5: the obvious electron transfer from LDH to GDY in e-ICLDH@GDY should, therefore, favor the formation of *OOH species with a smaller free energy of 0.54 eV than that of pure GDY (0.93 eV, Fig. 5i) and benefit the OER activity.
Page 15 lines 8-10: Our calculated results (Fig. 5j) further show that the ICLDH@GDY has the smallest Gibbs free energy of 0.34 eV for HER compared with that of pristine GDY (0.46 eV) and ICLDH (1.55 eV), indicating the excellent HER activity for GDY incorporated ICLDH.  8. In this work, the very large catalytic current densities for both HER and OER are driven by small potentials, which is very impressive. In this case, the Faraday efficiencies of OER, HER and overall water splitting should be given to further confirm the performance of electrodes.  Thank you again for your kind consideration.

Responses
Best wishes,

Yuliang Li
Dear Reviewers, Thank you for the sincere advice and comments on our manuscript titled "Overall water splitting by graphdiyne-exfoliated and -sandwiched layered double-hydroxide nanosheet arrays" for Nature Communications. Based on your comments, we have made the corresponding revisions on the manuscript and the following point-by-point responses to the comments.

Reviewer #1:
The authors addressed some of my concerns, and the manuscript is much improved. However, some questions are still not clearly illustrated. A major revision is still required.
Response: Thank you for your time and detailed comments on our manuscript. We have carefully addressed all your comments and make necessary revisions on our manuscript to address your concerns. The point-by-point responses are as follows.
Question 1. HER in alkaline media is more complex and sluggish than that in acid due to an additional water activation step, which is normally the rate-limiting step. Therefore, it is not accurate to evaluate the HER activity just by ∆G. The authors can refer to the following paper and carry out similar computations (Nature Communications (2017) 8, 15437; Nano Energy (2018) 53, 492).

Response:
We thank the reviewer for the valuable comments. Here we further gained some insightful interpretations on the high alkaline HER performance on the HER by DFT calculations (Figure R1). The projected partial density of states (PDOSs) show the related pand d-band distributions for the nearest interfacing GDY and LDH layers. The overall 3d-band center of ICLDH is high next to the Fermi level (EF), which is responsible for activating the initial adsorptions of O-species from alkaline condition. The partial intrinsic overlaps of p-and d-bands demonstrate a weak inter-layer bonding. This indicates a rather long-ranged interacting interfacing layer with large distance for H2O molecule easily passing through, without obviously impacting the catalyst interface ( Figure R1a).
Details on the 3d-bands of Co and Fe sites from LDH interfacing layer show the contrast in 3dband centers, which potentially denotes a dynamically self-activated electronic activity of redox.
The Co-3d bands with majority-spin is pinned at the EF and locates 1.4 eV higher than one of Fe-3d band. This trends show Fe-3d acts as electron-rich center while the Co-3d band plays as electron-depletion channel. However, their d-d overlapping is rather weak with few overlaps in PDOS. This indicates the highly effective electron-transfer occurs between the inter-layers instead of intra-layer ionic sites. The O-2p band from H2O exhibits a large overlapping across the Fe-3d band confirming the Fe-site acts as dominant role for H2O initial adsorption within interlayer regions (Figure R1b).
We move onto the energetic pathway of the alkaline HER for this interface system. Overall, the e-ICLDH@GDY system performs the most energetic favorable path with lowest adsorption (-0.56 eV) and preferred exothermic reaction heat for H2 (-1.40 eV). In the contrast, the GDY exhibits the highest barrier for initial H2O adsorption (1.92 eV) and an endothermic reaction (0.23 eV) for HER. The capabilities of water-splitting demonstrate the preference as e-ICLDH@GDY > ICLDH > GDY, respectively ( Figure R1c).
Meanwhile, the chemisorption energy of HER has been also illustrated. Among of them, the e-ICLDH@GDY indicates even more energetic favorable trend for HER compared with pristine-GDY, while the LDH meets the unflavored uphill for H-adsorption ( Figure R1e). From the local structures (Figure R1f), the stable adsorbed H is bonding with C2 site at the GDY layer and the H2O location will pass through the GDY locating near Fe-site on the ICLDH surface.
The splitting H2O process occurs between the C2 site and Fe-site from LDH layer. The two closely adsorbed H exhibit a favorable trend for combination in structural configuration for potentially efficient H2 generation. With assistance of C2 site, the leaving group (OH -) can be stably located on the surface of GDY.
We have added the above results and the suggested References in the revised manuscript (page  we carried out similar calculation method. The number of active sites was estimated from the roughness factor (eq. 1).
The roughness factor (Rf) can be determined by the electrochemically double-layer capacitance (Cdl). The specific capacitance can be converted into an electrochemical active surface area (ECSA) using the specific capacitance value for a flat standard with 1 cm 2 of real surface area.
According to previous reports (Energy Environ. Sci. 2015, 8, 3022;Nano Energy 2018, 53, 492), we assume 60 µF cm -2 for a flat electrode and the surface sites of 2 × 10 15 for the flat standard electrode. As a result, the number of surface active sites for e-ICLDH@GDY is calculated to be 0.067 × 10 18 surface sites/cm 2 .
The TOF values can then be obtained according to the following formulas: The number of total hydrogen turn overs is calculated from the current density extracted from the LSV curves: TOF per site for e-ICLDH@GDY at different overpotentials (η) is calculated as follows: = 3.12 × 10 15 0.067 × 10 18 × = 0.047 × (4) j corresponds to the current density at different overpotentials.  Table 5).
3. The authors' response to my sixth question in the previous review comments is still not clear.
Why will the surface complexity lower the cell voltage for two-electrode water splitting?
Response: It is a very good question.
For overall water splitting (OWS), the two-electrode cell system was employed to conduct corresponding measurement. In the two-electrode cell system, two electrodes are the same e-ICLDH@GDY/NF samples, one is used as the working electrode and the other is used as counter/reference electrode, to measure the potential across the complete cell. It should be mentioned that, compared to the SCE electrode (a standard electrode), the prepared e-ICLDH@GDY/NF electrodes have different sizes, shapes, and morphologies. The distances between the WE and the RE are different, which would result in different solution resistances.
Accordingly, the cell voltages in OWS system is different from that for OER and HER.