Enhanced Photoelectrochemical Water Splitting at Hematite Photoanodes by Effect of a NiFe-Oxide co-Catalyst

Round 1

Reviewer 1 Report

This paper is a follow-up of the previous one (ref. 7), which describes on the photoelectrochemical water splitting with use of Sn,P-modified hematite as a photoanode (PA), CuO as a photocathode (PC), and a polymer membrane as a gas separator. Here, co-catalyst modification of the PA made improvement of its photocurrent under the application of bias voltage. All materials used here are well-known and this is just the combination study of them.  This is a kind of a quickie joband there are many issues to be added and modified as shown below. Thus, after major revision on the light of the following points, this paper will be re-reviewed.

1. Lack of suitable references on NiFeOx as a WOC in the main text.
2. Compare NiFeOx catalyst with CoPi.
3. Is the loaded amount of the co-catalyst optimized? If not, need optimization.
4. Need confirmation of the composition of the Sn:P-modified anode by EDX and or XPS before the modification of the co-catalysts..
5. Preferably add time-course of evolved O₂/H₂ quantification and Faradaic efficiency.
6. In the previous paper, the polymer anion exchange membrane used was Fumatech FAA-3-20 (20 mm thickness). Here, however, FAA-3-50 (50 mm thickness) is applied. In general, a thicker PEM will give higher resistance and accordingly causes larger potential drops. Why do authors use less efficient materials here?
7. Both the preparation methods of both PA and PC are modified slightly in different methods from the previous one (ref. 7): e.g. for PA, hematite modified with P by spray-pyrolyzed (P: α-Fe₂O₃), though here, Sn-modified hematite is used before P modification. For PC, Fe-modified CuO is used (ref. 7), though unmodified CuO is used here. Need to show the performance differences between the both.
8. The applied light source (AM1.5 vs. AM 1.0) and irradiation power (100 vs. 92 mW/cm²) are subtly different between the previous one and this. In order to make the straight comparison between the two papers possible, the application of the same conditions is recommended.
9. There are no descriptions on the stability of this tandem electrode system after at least 1-hr irradiation in comparison with one without co-catalysts and UV-vis, XDS/SEM and XPS analyses of electrodes before and after the photolysis. Based on the results, describe on the occasion(s) on the photocurrent decay.

Minor points:

1. Ref, 25: J.Phys. Chem. C

 

Author Response

This paper is a follow-up of the previous one (ref. 7), which describes on the photoelectrochemical water splitting with use of Sn,P-modified hematite as a photoanode (PA), CuO as a photocathode (PC), and a polymer membrane as a gas separator. Here, co-catalyst modification of the PA made improvement of its photocurrent under the application of bias voltage. All materials used here are well-known and this is just the combination study of them.  This is a kind of a quickie job and there are many issues to be added and modified as shown below. Thus, after major revision on the light of the following points, this paper will be re-reviewed.

Thank you for your comments. These have been very helpful to improve the manuscript. Regarding the scope and the novelty of this work, our aim is to demonstrate that the promoting activity of the NiFeOx catalyst for water oxidation is also effective in the case of tandem photoelectrolysis cells based on an anion  exchange membrane and ionomer approach. The purpose of the work is also to show that, according to this particular tandem semiconductor configuration based on a transparent membrane, a trade-off  for the catalyst loading at the anode is needed to avoid blocking relevant amount of light transmission to the cathode semiconductor as well as to enhance interface properties.  In our opinion, this study can be also useful when other promoters are investigated for a tandem photoelectrolysis cell configuration approach using a transparent membrane separator.

1. Lack of suitable references on NiFeOx as a WOC in the main text.

 

Thank you very much for this observation. We have now improved the text by providing a large number of references dealing with NiFeOx as a WOC (refs. [48-65] in the revised manuscript)

 

2. Compare NiFeOx catalyst with CoPi.

 

CoPi is an excellent promoter for photoelectrochemical water splitting with excellent catalytic activity demonstrated for both anodic and cathodic processes (ref. [32]). However, the focus of this work was to investigate non-critical raw materials for the EU according to the aims of our EU H2020 project. Cobalt (cobalt phosphate) is unfortunately included in the EU CRM list [36] and CoPi was not thus taken into consideration for an experimental comparison.

Moreover, the CoPi stability at the interface of strongly alkaline membranes may represent an issue.  CoPi seems to be appropriate for milder pH conditions  [Refs. 37-41 in the revised manuscript].

However, for sake of comparison, some other CRM materials, stable in strong alkaline conditions, were analysed for comparison.

We have now discussed in the text the CoPi properties and mentioned why we have searched for an alternative promoter and relevant references about CoPi have been included (pag.2 lines 24-32 revised manuscript)

 

3. Is the loaded amount of the co-catalyst optimized? If not, need optimization.

 

The loading of the co-catalyst has been optimised to achieve a trade-off between the enhancement of the water oxidation process and the need to assure sufficient transparency in the higher wavelength range to reach the cathode semiconductor through the transparent membrane. The loading optimisation study is now included and properly discussed in the manuscript (pag.7 lines 8-12 and figure 4).

 

4. Need confirmation of the composition of the Sn:P-modified anode by EDX and or XPS before the modification of the co-catalysts.

 

The Sn and P inclusions were confirmed by EDX analysis (pag.8 lines 11-12) and shown in the supplementary information (Fig. S3).

 

5. Preferably add time-course of evolved O₂/H₂quantification and Faradaic efficiency.

 

For these small cells (0.25 cm²) it was not easy to implement specific analytical tools (e.g. gas chromatography) to determine faradaic efficiency.

H₂ evolution during time studies has been monitored for larger cells (4 cm²) of the same tandem configuration using a mass spectrometer. However, a precise estimation of the faradaic efficiency was not possible due to some calibration issues for such low signals. A rough estimation would indicate faradaic efficiency above 90% but without a precise calibration we would avoid mentioning it.

These data are now provided in the supplementary information (Fig. S4). with a short mention in the  manuscript (pag.12 lines 1-3).

 

6. In the previous paper, the polymer anion exchange membrane used was Fumatech FAA-3-20 (20 mm thickness). Here, however, FAA-3-50 (50 mm thickness) is applied. In general, a thicker PEM will give higher resistance and accordingly causes larger potential drops. Why do authors use less efficient materials here?

 

In view of the large area cell fabrication (100 cm²), planned in our project, the 50 µm membrane is more appropriate from the point of view of mechanical handling and compression resistance.

The conductivity of the 50 µm membrane [Ref. 78 in the revised manuscript] is sufficiently high to keep ohmic losses, associated with ion transport in the polymer electrolyte separator, within 1 mV (close to the experimental error) in the range of the recorded current densities (pag.10 lines 28-32 revised manuscript).

 

7. Both the preparation methods of both PA and PC are modified slightly in different methods from the previous one (ref. 7): e.g. for PA, hematite modified with P by spray-pyrolyzed (P: α-Fe₂O₃), though here, Sn-modified hematite is used before P modification. For PC, Fe-modified CuO is used (ref. 7), though unmodified CuO is used here. Need to show the performance differences between the both.

 

A study regarding the modification of CuO by Fe has been reported elsewhere [ref.14]. Regarding the anode, note that Sn doping is a result of employing a F:SnO₂ substrate and a high temperature in the thermal treatment (750ºC), while P doping has been described for a number of years (Energy Environ. Sci., 2015,8, 1231-1236, ref.25). We have employed “standard” electrodes as the focus of this work is regarding the role of the NiFeOX promoter not an investigation of the role of Sn, P doping for hematite or Fe doping for CuO, which, on the other hand, is already described. We have here used a set of electrodes prepared according to a methodology that has been detailed in the text that have shown proper reproducibility characteristics. These have been chosen to assess the role of the NiFeOX promoter.

 

8. The applied light source (AM1.5 vs. AM 1.0) and irradiation power (100 vs. 92 mW/cm²) are subtly different between the previous one and this. In order to make the straight comparison between the two papers possible, the application of the same conditions is recommended.

 

Due to some recent laboratory work, we have recorded a slight misalignment of our solar simulator. We have planned to have some maintenance service for that; but, unfortunately, our laboratories are closed for the moment. We have properly measured the light irradiation in our photoelectrolysis set up equipped with a solar simulator using a calibrated solar cell and determined the specific irradiation power. All the results have been normalised by the effective incident power. For all the results reported in this work, we have used exactly the same irradiation power.

 This is very close to what we have used in our previous paper and we do not think it may cause any relevant comparison issue. However, the aim of the present work is to focus on the role of the NiFeOx promoter rather than providing an update of our progresses with respect to our previous work.

 

 

9. There are no descriptions on the stability of this tandem electrode system after at least 1-hr irradiation in comparison with one without co-catalysts and UV-vis, XDS/SEM and XPS analyses of electrodes before and after the photolysis. Based on the results, describe on the occasion(s) on the photocurrent decay.

A durability test of several hours is now included in the supplementary information (Fig. S5) with a short mention in the revised text (pag.12 lines 4-6). Due to the short-term operation, a detailed post-operation analysis of the electrode properties was not carried out.  We think that, a proper durability test should last at least some hundred hours to get relevant information about electrode or membrane modifications. This will be the scope of a next work. In any case, now, we do not have access to the laboratories to provide at least some preliminary analysis of the electrodes that have operated for a few hours. In principle, we do not expect relevant changes after this short operation time.  

 Minor points:

1. Ref, 25: J.Phys. Chem. C

 

“C” was added to ref.25 now corresponding to ref.28 in the revised version of the manuscript

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript entitled “Enhanced photoelectrochemical water splitting at hematite photoanodes by effect of a NiFe-oxide surface promoter” presents an interesting proposal as the work carried out tries to improve the behavior of a PEC modifying it with a co-catalyst. However, it has some important weak points that need to be addressed before it can be considered suitable for publication.

• It is important to distinguish and to specify properly what is the role of the added NiFe oxide. In the title the authors name it a surface promoter, later it is named co-catalysts and in the manuscript other terms are used like: promoter, catalytic enhancer, additive. These terms are not synonyms and can not be used indistinctly. The authors must select the most appropriate.
• References must be revised. For example ref 19 deals with doping and refs 30 and 31 are located to refer to P and Sn doping none of them considers P doping, reference 32 does not deal with IrRuOx. These are examples, but the authors should revise citation in general.
• The English expression must be revised. For example, terms and expressions in lines 68, 70, 73, 127, 141, 199
• XRD analysis: regarding data of figure 1a, the authors must mention that there is not only a spinel phase, but a mixture of phases. In fact, the two XRD peaks at about 36 and 37 degree do not have the intensity ratio expected for the spinel. Besides, the authors must indicate which has been the peak selected to calculate the crystal size. XRD of the synthesized Fe2O3 (original and modified) must be also included (at least as supplementary material).
• Table 1 does not include hematite modified with P and Sn. It is not clear if in all cases a modified Fe2O3 is used. It is not clear at all the state (and the role) of P and Sn in the anode.
• Figure 2, left: the figure caption must indicate what is ION; right: If the band gap values of hematite and CuO are diferent, should not they be depicted different?
• If I am not wrong, point 2.2 should be named photoelectrochemical characterization. If not, more mention to the photo process must be done. Also, it is not mentioned the illumination configuration (the experimental section mentions several options) of the system for which the results are presented.
• Figure 5, which is the meaning of the blue arrows in the right image?, how can be appreciate in it the white particles observed in the left image?
• The results and discussion part of the manuscript is poor. It is not clearly explained why does FeNiOx have a positive effect, better than that of the other two co-catalysts (promoters?) used.

Author Response

The manuscript entitled “Enhanced photoelectrochemical water splitting at hematite photoanodes by effect of a NiFe-oxide surface promoter” presents an interesting proposal as the work carried out tries to improve the behavior of a PEC modifying it with a co-catalyst. However, it has some important weak points that need to be addressed before it can be considered suitable for publication.

 Thank you very much for these observations. These have been very useful. We have revised the manuscript accordingly.

1. It is important to distinguish and to specify properly what is the role of the added NiFe oxide. In the title, the authors name it a surface promoter, later it is named co-catalysts and in the manuscript other terms are used like: promoter, catalytic enhancer, additive. These terms are not synonyms and can not be used indistinctly. The authors must select the most appropriate.

The terms, surface-promoter, co-catalyst and catalytic enhancer are indeed considered synonyms in this field (see for example refs [37-41] in the revised manuscript).

A heterogeneous catalyst is an agent promoting the reaction at the interface with an enhancement of the reaction rate. Thus, all these terms essentially refer to the same species.

However, following the reviewer’s suggestion we have reduced the number of possible synonyms.  

 

2. References must be revised. For example ref 19 deals with doping and refs 30 and 31 are located to refer to P and Sn doping none of them considers P doping, reference 32 does not deal with IrRuOx. These are examples, but the authors should revise citation in general.

These references have been corrected in the revised text

3. The English expression must be revised. For example, terms and expressions in lines 68, 70, 73, 127, 141, 199.

The language was improved.  

 

4. XRD analysis: regarding data of figure 1a, the authors must mention that there is not only a spinel phase, but a mixture of phases. In fact, the two XRD peaks at about 36 and 37 degree do not have the intensity ratio expected for the spinel. Besides, the authors must indicate which has been the peak selected to calculate the crystal size. XRD of the synthesized Fe2O3 (original and modified) must be also included (at least as supplementary material).

We have now clarified these aspects. The following sentence has been added to the text (pag.3 lines 9-13):

The NiFeO_x co-catalyst (Ni:Fe = 1:1 wt%) showed mainly the crystalline structure of the NiFe₂O₄ spinel phase (JCPDS card n° 10-0325). However, also a small presence of Fe₂O₃ and NiO phases (JCPDS cards n° 24-0072 and 22-1189, respectively) was identified. The crystallite size was determined from the  average of the peak broadening of the four main reflections. This is now specified in the text (pag.3 lines 9-13). The XRD pattern of Fe₂O₃ is now reported in the supplementary information (Figure S1).

5. Table 1 does not include hematite modified with P and Sn. It is not clear if in all cases a modified Fe2O3 is used. It is not clear at all the state (and the role) of P and Sn in the anode.

We apologize for this inconvenience. A P and Sn doped hematite has been used in all experiments. This is now clearly reported in the text (Table 1 revised manuscript).

Regarding the anode, note that Sn doping is a result of employing a F:SnO₂ substrate and a high temperature in the thermal treatment (750ºC), while P doping has been described for a number of years know (Energy Environ. Sci., 2015,8, 1231-1236; ref.25). We have employed “standard” electrodes as the focus of this work is regarding the role of the NiFeOx co-catalyst rather than an investigation of the role of Sn and P doping for hematite. We have here used a set of electrodes prepared according to a methodology detailed in the text that has shown proper reproducibility characteristics. These have been chosen to assess the role of the NiFeOX promoter.

 

6. Figure 2, left: the figure caption must indicate what is ION; right: If the band gap values of hematite and CuO are diferent, should not they be depicted different?

ION is now replaced with “ionomer” in the figure and the widths of Fe₂O₃ and CuO band gaps in the energy diagram are now reflecting their difference in energy gap.

 

7. If I am not wrong, point 2.2 should be named photoelectrochemical characterization. If not, more mention to the photo process must be done. Also, it is not mentioned the illumination configuration (the experimental section mentions several options) of the system for which the results are presented.

The point 2.2 is now named photoelectrochemical characterisation and the employed illumination configuration is now described in the experimental section 3.2 (pag.11 lines 8-10). The cell is placed horizontally and illuminated by the solar simulator from the FTO photoanode back contact.

 

8. Figure 5, which is the meaning of the blue arrows in the right image?, how can be appreciate in it the white particles observed in the left image?

Arrows indicate co-catalyst clusters. These correspond to the white particles at low magnification in the left image. This is now specified in the caption.

 

9. The results and discussion part of the manuscript is poor. It is not clearly explained why does FeNiOx have a positive effect, better than that of the other two co-catalysts (promoters?) used.

 

As shown in the recent literature [Refs. 48-58], NiFe-oxide catalysts show outstanding properties for the oxygen evolution reaction in the alkaline environment because of the possibility of a specific tuning of the electronic properties for the adsorption of oxygen species (e.g. hydroxyl species) and their oxidation [Refs. 59-65]. As well known, it is important in electrocatalysis to modulate the strength of the bonds between the catalyst surface and the adsorbed species. This must not be either strong or too labile. In alkaline solutions, the NiFeOx co-catalyst shows a better compromise with respect to IrRuOx. IrRuOx forms strong bonds with hydroxyl species in the alkaline environment whereas this provides optimum characteristics in the acidic environment where the main interaction is with water molecules [43, 66]. LSFCO has promising surface characteristics, but being almost an insulator at low temperature [45, 67], electron transfer is somewhat impeded. This discussion is now added to the manuscript (pag. 6 lines 7-16).

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper reported the preparation of NiFeOx catalyst on Fe₂O₃[P, Sn doped] nanostructured photocathode for Tandem photoelectrochemical cells applications. The content of NiFeOx was prepared by wet chemical synthesis (oxalate method) and compared with LSFCO and IrRuOx OER catalysts. The results presented in the paper are quite interesting, but some of the discussions are incomplete and needs to be modified. The authors have to revise the manuscript along these points mentioned below;

 

1. Ni based OER catalysts such as NiFeOx require an activation step (NiO -> NiOOH) using an electrochemical redox reaction. Please explain the activation conditions to form OER active Ni_1–xFe_xOOH. *Applied Surface Science 493 (2019) 1150–1158, Int. J. Hydrogen Energy 44 (2019) 26118–26127, J. Am. Chem. Soc. 2015, 137, 3, 1305-1313

2. Further analysis after applying Pt catalyst to CuO is requested. Low HER performance of CuO could act as a rate limiting step in this tandem structure. * Int. J. Hydrogen Energy, 44(48), 26148-26157 (2019)., Int. J. Hydrogen Energy 39(15), 7686–7696 (2014)., Scientific Reports 6, 35158 (2016).

3. For figure 3, authors should provide the full scale of J_light, J_dark each. The detailed description of lower PEC properties of LSFCO and IrRuOx from -0.5V is needed.
4. For figure 5, EDS elemental mapping result is needed to investigate the components of the NiFeOx at surface.
5. Experimental results such as UV-Vis, Mott-schottky need to be provided for the band structure of photoelectrode. *Catalysts 2020, 10(4), 358, ACS Appl. Energy Mater. 2020, 3, 1, 666-674, Nanoscale Horiz., 2016,1, 243-267, Scientific Reports 6, 35158 (2016).

Author Response

This paper reported the preparation of NiFeOx catalyst on Fe₂O₃[P, Sn doped] nanostructured photocathode for Tandem photoelectrochemical cells applications. The content of NiFeOx was prepared by wet chemical synthesis (oxalate method) and compared with LSFCO and IrRuOx OER catalysts. The results presented in the paper are quite interesting, but some of the discussions are incomplete and needs to be modified. The authors have to revise the manuscript along these points mentioned below;

 Thank you very much for these comments; we have revised the manuscript accordingly.

1. Ni based OER catalysts such as NiFeOx require an activation step (NiO -> NiOOH) using an electrochemical redox reaction. Please explain the activation conditions to form OER active Ni_1–xFe_xOOH. *Applied Surface Science 493 (2019) 1150–1158, Int. J. Hydrogen Energy 44 (2019) 26118–26127, J. Am. Chem. Soc. 2015, 137, 3, 1305-1313.

 

Formation of active Ni_1–xFe_xOOH surface states possibly occurs in our opinion during the preliminary electrode treatment in KOH and during polarisation (this should in principle occur between 0.4 and 0.5 V vs. RHE [refs. 63, 68-69]). If the polarisation of the CuO electrode is not very large upon bias application, the anode experiences potentials between 0 and 1.2 V vs RHE. We have recorded an increase of the activation effect at high bias potentials (at more negative cell voltages in the photodiode representation). This evidence points to the formation of the active Ni_1–xFe_xOOH surface states as promoting species. However, it is also important to point out that a small promoting effect is also appearing at 0 V bias. Such aspects are now discussed  in the manuscript (pag.6 lines 24-31) and the references indicated above have been cited (refs. 63, 68-69).  

 

2. Further analysis after applying Pt catalyst to CuO is requested. Low HER performance of CuO could act as a rate limiting step in this tandem structure. * Int. J. Hydrogen Energy, 44(48), 26148-26157 (2019)., Int. J. Hydrogen Energy 39(15), 7686–7696 (2014)., Scientific Reports 6, 35158 (2016).

 

We do not apply any Pt catalyst to CuO. CuO is coated by the transparent ionomer. The hematite photoanode, the NiFeOx cocatalyst dispersion over the hematite and the transparent membrane have been designed to allow that the high wavelength range of the light irradiation can reach the photocathode, thus avoiding that the photocathode limits the behaviour of the device. This is confirmed by the results in Fig. 3, obtained with the same photocathode and different photoanodes (different co-catalysts). If the photocathode would limit the device performance, the photocurrent-voltage curves would not be sensitive to changes in the photoanode. This is now discussed in the revised manuscript (pag.7 lines 1-7) and the references indicated are cited in the text (refs.70-72).

 

3. For figure 3, authors should provide the full scale of J_light, J_dark  The detailed description of lower PEC properties of LSFCO and IrRuOx from -0.5V is needed.

 

Polarisation curves under illumination and in the dark for each cell reported in Fig 3 are now provided in the supplementary information (figure S2) and further description for the behaviour of co-catalyst properties was implemented in the revised version of the manuscript (pag.6 lines 5-16).

 

The lower photoelectrochemical properties of LSFCO and IrRuOx compared the bare cell at potentials more negative than -0.5 V, in the photodiode representation mode, may be probably related to some strong adsorption of hydroxyl species at these potentials. A strong adsorption of hydroxyls may hinder the release of oxygen molecules. As an example, in conventional electrolysis, due to these strong adsorption characteristics, IrRuOx is less effective in alkaline media compared to acidic environments. In the present case, such characteristics may cause some interference to the photoelectrochemical conversion. (pag.6 lines 17-23).  

 

4. For figure 5, EDS elemental mapping result is needed to investigate the components of the NiFeOx at surface.

 

The distribution of NiFeOx on the surface is clear from the charging contrast. Conductive particles appear brighter than the semiconducting underlying hematite layer under these specific analysis conditions. Unfortunately, no access to our labs is possible at the moment to collect EDS elemental mapping.

EDX analysis of the outer NiFeOx-coated hematite photoanode surface is now reported in the supplementary information (figure S3).

 

5. Experimental results such as UV-Vis, Mott-schottky need to be provided for the band structure of photoelectrode. *Catalysts2020, 10(4), 358, ACS Appl. Energy Mater. 2020, 3, 1, 666-674, Nanoscale Horiz., 2016,1, 243-267, Scientific Reports 6, 35158 (2016).

 

A reference is reported (ref. 73). Unfortunately, no access to our labs is possible at the moment. More detailed analysis can be provided in a next paper.

Author Response File: Author Response.pdf

Reviewer 4 Report

This work from Lo Vecchio et al. describes the preparation and characterization of complete photoelectrochemical cells based on hematite photoanodes and CuO photocathodes for water splitting. The effects of deposited co-catalysts are also evaluated. Globally, this study could be of interest for a broad community of researchers working in the field of sunlight-driven water splitting systems. Nevertheless, the described findings show that the photoelectrochemical performance of cocatalyst-modified cells is somewhat similar to that of cocatalyst-free cells : only an 0.15 mA cm-2 increase at a -0.6 V voltage is measured (see Figure 3). In other words, the interest of their systems for efficient PE cells is not really demonstrated. Moreover, the materials used in this work have not been sufficiently characterized. For example, SEM images of the photoanode before polarization must be shown to assess the post-electrochemistry changes; the EDX data are missing,... Consequently, I cannot recommend the publication of this manuscript in Catalysts

Author Response

This work from Lo Vecchio et al. describes the preparation and characterization of complete photoelectrochemical cells based on hematite photoanodes and CuO photocathodes for water splitting. The effects of deposited co-catalysts are also evaluated. Globally, this study could be of interest for a broad community of researchers working in the field of sunlight-driven water splitting systems. Nevertheless, the described findings show that the photoelectrochemical performance of cocatalyst-modified cells is somewhat similar to that of cocatalyst-free cells : only an 0.15 mA cm-2 increase at a -0.6 V voltage is measured (see Figure 3). In other words, the interest of their systems for efficient PE cells is not really demonstrated. Moreover, the materials used in this work have not been sufficiently characterized. For example, SEM images of the photoanode before polarization must be shown to assess the post-electrochemistry changes; the EDX data are missing,... Consequently, I cannot recommend the publication of this manuscript in Catalysts

We thank the reviewer for the constructive criticism.

However, we have to say that the photocurrent improvement at a bias of 0.6 V is about 30% whereas it is 50% at 0.4 V bias. This is not a minor variation, but a substantial one. 

Of course, the photocurrent values are still low compared to the best results reported in the literature. However, it is also evident that very often in the literature the reported photocurrents refer to an applied bias of 1.23 V (reversible potential). It is important to mention that our cell configuration does not include any critical raw material and it employs a polymer electrolyte membrane separator rather than a liquid electrolyte. This is a practical advantage to avoid recombination of the evolved gases.

In our opinion, the promoting effect should be considered in relation to the level of applied bias. Our results show that under mild bias conditions of 0.4 V, we have 50% enhancement (pag.5 lines 28-29 revised manuscript).

We know that this configuration needs to be significantly improved but the concept appears very promising for practical applications and with further optimisation of co-catalyst and interface significant progresses can be achieved. 

We have reported SEM analysis after operation to show that the co-catalyst is well present on the surface.

In the revised version, EDX analysis of the outer NiFeOx-coated hematite photoanode surface is reported in the supplementary information (figure S3) of the revised manuscript.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors have given satisfactory answers to most of the reviewer queries.

 

There are still, however, minor details to address. For example the authors explain: Regarding the anode, note that Sn doping is a result of employing a F:SnO2 substrate. 

It is not clear to me what is F.

 

I am sorry but I still do not see the purpose of the arrows included in the image of Figure 6. 

Author Response

1. The authors have given satisfactory answers to most of the reviewer queries.

 Thank you for your comments. These have been helpful to improve the manuscript.

2. There are still, however, minor details to address. For example the authors explain: Regarding the anode, note that Sn doping is a result of employing a F:SnO2 substrate. 

It is not clear to me what is F.

2. A fluorine-doped SnO2 substrate was used for the anode. This is now clarified.

 

3. I am sorry but I still do not see the purpose of the arrows included in the image of Figure 6. 

The arrows indicate co-catalyst clusters corresponding to bright particles in the low magnification image; this has been indicated in the caption. We would prefer keeping the arrows in the figure.

 

4. The language has been improved.

Author Response File: Author Response.pdf

Reviewer 3 Report

Accept as is. 

Authors properly revised the manuscript according to the reviewer's comment. 

Author Response

 Thank you for your comments. These have been helpful to improve the manuscript.

1. The language has been improved.

Author Response File: Author Response.pdf

Reviewer 4 Report

Extensive revision has been made by the authors and most issues/comments raised by the reviewers have been carefully addressed. Consequently, the present form is now publishable in Catalysts.

Author Response

Thank you for your comments. These have been helpful to improve the manuscript.

 

1. The language has been improved.

Author Response File: Author Response.pdf