Hydrovoltaic effect-enhanced photocatalysis by polyacrylic acid/cobaltous oxide–nitrogen doped carbon system for efficient photocatalytic water splitting

Severe carrier recombination and the slow kinetics of water splitting for photocatalysts hamper their efficient application. Herein, we propose a hydrovoltaic effect-enhanced photocatalytic system in which polyacrylic acid (PAA) and cobaltous oxide (CoO)–nitrogen doped carbon (NC) achieve an enhanced hydrovoltaic effect and CoO–NC acts as a photocatalyst to generate H2 and H2O2 products simultaneously. In this system, called PAA/CoO–NC, the Schottky barrier height between CoO and the NC interface decreases by 33% due to the hydrovoltaic effect. Moreover, the hydrovoltaic effect induced by H+ carrier diffusion in the system generates a strong interaction between H+ ions and the reaction centers of PAA/CoO–NC, improving the kinetics of water splitting in electron transport and species reaction. PAA/CoO–NC exhibits excellent photocatalytic performance, with H2 and H2O2 production rates of 48.4 and 20.4 mmol g−1 h−1, respectively, paving a new way for efficient photocatalyst system construction.

The present manuscript reports on Hydrovoltaic effect-enhanced photocatalysis by PAA/CoO-NC system for efficient photocatalytic water splitting. The materials are tested as photocatalysts for H2 evolution.
Due to its optical properties, PAA/CoO-NC system is an interesting material to explore for photocatalytic application. The manuscript in my view presents a novelty to be published in Nature communication. However, some critical issues need to be addressed and manuscript revised prior to acceptance.
For these reasons, I recommend the publication of the present manuscript in Nature communication. Major critical points are outlined in the following: 1. The increased photocatalytic performance upon modification with CoO is certainly not a surprise as it serves as a co-catalyst or a trap center for the material prepared. Moreover, the manuscript omits some relevant information and review on hydrovoltaic effect as it related to photocatalysis check this article and reference it. (Zhong, T., Li, H., Zhao, T., Guan, H., Xing, L., & Xue, X. (2020). Selfpowered/self-cleaned atmosphere monitoring system from combining hydrovoltaic, gas sensing and photocatalytic effects of TiO2 nanoparticles. Journal of Materials Science & Technology. doi:10.1016/j.jmst.2020.11.002) 2. Carbon is known to possess large surface area, however, the work is silent about the effect of CoO on the surface area of the system.
3. The specific wavelength of the light source used in the photocatalytic generation of H2 is not state, even though Figure S16 depicts several wavelengths was used to determine AQY. 4. I am surprised that the effect of calcination at 600 oC on the surface of the CoO-NC was not reported, particularly if one considers that CoO-NC was annealed at 600°C for 1 h, which might cause sintering and a consequent reduction of the surface area. In other words, one expects the bare CoO-NC to feature a higher specific surface area. Fig S14. What is responsible for the difference in rate of H2 evolution in both systems Because the difference is bandgap is not obvious.

If the DRS UV/VIS of both PAA/CoO and CoO-NC shows similar optical properties as presented in
6. The mechanism presented in section 2.5 did not explain the pathway to the formation of H2O2 because in water splitting H2 and O2 are normally the by-product (Theoretically). Figure 1j and S3 did not explain if the CoO is decorated on the surface of the materials or found within the matrix, also the actual amount of the CoO as determined by the XPS is not reported.

The XPS results in
8. The conclusion stated that the hydrovoltaic effect improves the hydrogen binding energy in the formation of active hydrogen during photocatalytic water splitting, resulting in excellent charge separation and transfer and accelerated reaction kinetics. However, the result discussion does not support this claim.
Reviewer #3 (Remarks to the Author): In this manuscript, Xu et al. proposed a hydrovoltaic effect-enhanced photocatalytic system in which an enhanced hydrovoltaic effect was achieved using polyacrylic acid (PAA) and cobaltous oxide (CoO)-nitrogen doped carbon (NC) and CoO-NC acted as a photocatalyst to generate H2 and H2O2 products simultaneously. The hydrovoltaic effect was shown to decrease the Schottky barrier height between CoO and the NC interface and accelerate the kinetics of water splitting, thereby resulting in photocatalytic performance with high H2 and H2O2 production rates. This work realized a synergetic combination of the conventional photocatalysis and the newly emerged hydrovoltaic effect and may open a new sub-direction for sustainable development of clean energy. However, before I can recommend its publication, the authors have to fully address following issues: 1.In the introduction, "Hydrovoltaic technology is a renewable energy harvesting method that uses moisture interactions with nanostructured carbon materials". The definition is not accurate. Hydrovoltaic technology can not only harvest energy from ambient moisture, but also from evaporation, raindrops, waves and so on. It is characterized by energy conversation based on the interaction between water and functional materials. The authors should find in which paper this terminology was first proposed.
2.In section 2.2, the influence of light illumination on the moisture moving velocity was investigated. An increased velocity results in an enhanced voltage generation. The author should clarify whether the enhancement is contributed solely by the increased moving velocity, any additional contribution from other factors, such as the photogenerated carriers or inhomogeneous distribution of the heat induced by light?
3.In the system, CoO was used as the active centers for water splitting and was shown to have a desirable band alignment with NC groups. I'm wondering if such a strategy of materials design can be generalized to a range of metal oxides, notably TiO2 that is mostly used for photocatalysis. The authors are suggested to perform additional experiments to demonstrate the generality.
4.The generated hydrovoltaic electricity is by pumping moisture steam through the materials, which costs extra energy and may result in an overall extremely low efficiency. Actually, hydrovoltaic technology enables electricity harvesting from natural evaporation which does not any extra energy input. In this regard, can the authors employ the evaporation-induced electricity to enhance the photocatalysis? 5.To further verify the photocatalytic effect of PAA/CoO-NC, the authors are suggested to carry out control experiments of photocatalytic H2/H2O2 production and provide in situ Raman spectra of pure PAA membranes loaded with Pt cocatalyst at 1100 ml h-1 Ar/H2O steam injection.
6.There are some misquotes in the introduction, such as "For example, a hydroelectric generator comprising the ionic polymer Nafion and a poly(N-isopropylacrylamide) hydrogel was developed to generate electricity [17,18]". Please check the literature comprehensively. Figure S1, the description on the device fabrication process in figure legend does not match its schematic illustration. For example "Commercial epoxy slurry is used to paint two "L-shaped" with predesigned dimensions on the substrate (Step 1)"; here, the electrodes should be "L-shaped", not epoxy.

7.In
8.In Note S1, "relative humidity by pumping a mixture of dry and wet nitrogen", while in the paper, the humidity is controlled by "Ar/H2O steam injection". 9.In Figure S7 (c) and Figure S8 (b), "The increase in the current density may be attributed to the increased amount of water moving and diffusion as W increases". While in general perception, the current increases with the width of the hydroelectric generator, and the current density is nearly constant. Please explain it.
10.In Figure 2a, the authors need to explain why the voltage at a 1300 ml h-1 Ar/H2O steam injection rate is lower than that at 1100 ml h-1. Does the liquid water cover the PAA/CoO-NC membrane?
11.In 2.5, "The detailed Fermi levels (Ef), conduction band minimum (CBM), and valence band maximum (VBM) are shown in Figure 5c 12.In 2.3, "A 0.1 wt% Pt cocatalyst was loaded on the photocatalyst via photodeposition." The preparation process of Pt cocatalyst in the Supporting Texts should be provided.

Reviewer #1 (Remarks to the Author):
This work uses the hydrovoltaic effect to enhance the performance of photocatalytic water splitting.
The idea is novel. Also, the system has good performance with H2 and H2O2 production rates of 48.4 and 20.4 mmol g -1 h -1 , respectively, and a high AQY of 56.2% at 400 nm, making it among the most efficient photocatalysts. Thus, I recommend the acceptance of this manuscript. However, the experiments and results are not clearly presented, and mechanism is not well explained, major revision is needed before publication, as noted below.
Response: We thank the referee for appreciating the impact and value of our study. We also appreciate the referee's constructive comments, which have helped us improve the quality of our manuscript. We have revised the manuscript accordingly.

Response:
The questions 1 and 2 can be answered together. We have defined the direction of the water steam transport vertically through the PAA/CoO-NC film. The water steam is exposed on the surface of PAA/CoO-NC and then diffused along the nanochannel to the bottom. Figure R1a shows the photo of PAA/CoO-NC hydroelectric generator. The PAA/CoO-NC is placed in a closed reactor, in which the wetted Ar gas (Ar/H2O) is brought, and a certain ambient humidity is formed in the reactor.
The direction of steam flow and produced hydrovoltage effect: Figure R1b shows the schematic of water steam diffusion on the PAA/CoO-NC film. When PAA/CoO-NC film exposed on water steam environment, the water steam enables H + ions to diffuse from the surface to the bottom of PAA/CoO-NC due to its porous structure, which is similar to previous literature (Adv. Mater. 2018, 30, 1705925). Thus, the water steam flow diffused along the nanochannel to the bottom. The hydrovoltaic voltage is generated because of a gradient in the concentration of water molecules along the nanochannels from the top surface to the bottom of the film. The water steam diffusion on the PAA/CoO-NC film of a nanochannel based on the electrokinetic effect is illustrated in Figure R1c, d.
The PAA/CoO-NC generates a pressure-driven flow carriers counter-ions of H + to form an electric current in the flow, eventually reaching an equilibrium of H + ions diffusion, resulting in a constant voltage output.
How hydrovoltaic enhanced photocatalytic reaction: The photocatalytic reaction has been enhanced by generated hydrovoltaic effect based on two sides: (1) The hydrovoltaic effect induces an electric field in photocatalyst of CoO-NC and generates a strong interaction between the H + carriers and reaction centers of the nanostructure, thereby further improving the kinetics of water splitting ( Figure R1d).
(2) The height of the Schottky barrier between CoO and the NC interface decreases due to the hydrovoltaic effect, which significantly promotes carrier separation and transfer on the photocatalyst (Figure 5f). These features collectively enhance the photocatalytic performance of the proposed system. The hydrovoltaic-enhanced photocatalytic mechanism has already been discussed in the page 11 of main text.  In the revised version, Figure R1 is added as new Figure S13 in the supporting information. Figure R2 is added as new Figure S1 in the supporting information. The corresponding photo and schematic of PAA/CoO-NC have been updated in Figure 1b, c. The relative discussion has been added in the page 7 of main text, page 18 of supporting information, and page 6 of supporting information and copied below: "The PAA/CoO-NC hydroelectric generator has been successfully constructed based on a moisture electrokinetic effect in a nanochannel of PAA/CoO-NC. [15,30] The water steam is exposed on the surface of PAA/CoO-NC and then diffused along the nanochannel to the bottom due to its porous structure (the detail mechanism and water steam diffusion path seen in Figure S13). The PAA/CoO-NC generates a pressure-driven flow carriers counter-ions of H + to form an electric current in the flow, eventually reaching an equilibrium of H + ions diffusion, resulting in a constant voltage output. [15,30] "(Page 7 of main text) " Figure S13a shows the photo of PAA/CoO-NC hydroelectric generator. Figure S13b shows the schematic of water steam diffusion on PAA/CoO-NC film. When water steam is exposed on the surface of PAA/CoO-NC film, the water molecules enable mobility of H + ions diffused from surface to the bottom of PAA/CoO-NC along with its porous structure. The diffusion direction of water steam is perpendicular to the membrane downward, which is similar to previous literature [25] . The hydrovoltaic voltage is generated because of a gradient in the concentration of water molecules along the nanochannels from the top surface to the bottom of film as illustrated in Figure S13c

Why H2O2 is generated instead of O2?
Response: The photocatalytic product in our system is H2O2 rather than O2, which is proved through the DFT-based calculations. As shown in Figure R3a, the photocatalyst CoO-NC shows a much lower free energy barrier of 0.37 eV for the H2O2 production process in the hydrovoltaic electric field (U=0.4 V) than the intrinsic barrier of 2.05 eV, indicating the hydrovoltaic effect is beneficial for the H2O2 reaction in PAA/CoO-NC system. For O2 production path, the photocatalyst CoO-NC shows free energy barriers of 2.07 and 2.64 eV with or without hydrovoltaic effect, as shown in Figure R3b, much higher than the limiting step barrier for H2O2 generation. As a consequence, the DFT-based calculations indicate that the CoO-NC photocatalyst prefers to generate H2O2 rather than O2 evolution. Figure R3. (a) DFT calculation of free energy diagram for the three-step H2O2 production process on CoO-NC with (U=0.4 V) or without (U=0 V) hydrovoltaic electric field and the corresponding adsorption geometries structures on CoO-NC; (b) DFT calculation of free energy diagram for O2 production process on CoO-NC with (U=0.4 V) or without (U=0 V) hydrovoltaic electric field and the corresponding adsorption geometries structures on CoO-NC.
In the revised version, Figure R3a and 3b are added as new Figure S27 in the supporting information, and the relative discussion has been added in the page 11 of main text, and page 38 of supporting information and copied below: "Density functional theory (DFT) based calculations further indicate that the CoO-NC photocatalyst prefers to generate H2O2 rather than O2 evolution in the system ( Figure S27)." (Page 11 of main text, detail see DFT calculations of supporting information) Fig. 5 is based on water, but most experiments were performed in steam, will the mechanism stay the same?

The schematic in
Response: I am sorry for the mistake. The word "water" in schematic of Figure 5h should be "steam".
we have checked and corrected the relevant annotations of water to Ar/H2O steam in Figure 5h of page 20 in main text.

5.
What is the condition of steam? Temperature? Is there any water droplets? According to the phase diagram (J. Phase Equilib., 24 [3] 212-227 (2003)), if the temperature is high, CoO will convert to Co3O4. Thus temperature is important.
Response: Thanks very much for the valuable comments. The condition of steam is Ar/H2O, the temperature of Ar/H2O steam in the hydrovoltaic system is about 40 °C, and there are no water droplets ( Figure R4), which is added as Figure S10a in the revised version. The CoO will not convert to Co3O4 because the temperature in hydrovoltaic system is lower than the CoO phase transformation temperature (2103 K from Ref. J. Phase Equilib., 24 [3] 212-227 (2003)). Figure R4. The detected temperature on PAA/CoO-NC surface over time with light illumination (light intensity: AM 1.5G, 100 mW cm -2 ) and at Ar/H2O injection rate of 1100 ml h -1 .

What determines the flow velocity of steam or water? If it is determined by external pressure, then it should not depend on the illumination as in section 2.2, the authors sated "light illumination can increase the moving velocity".
Response: Thanks very much for the valuable comments. The moisture of Ar/H2O, acting as an external force, affects the diffusion of water, which in turn creates the relative motion of water molecules and nanochannel of PAA/CoO-NC. The light illumination induces a higher temperature on the surface of PAA/CoO-NC as shown in Figure R4, bringing an increased moving velocity of water steam in the nanochannel of PAA/CoO-NC.
The relation between elevated temperature induced by light illumination and water steam velocity can be mathematically described as the following equations of diffusion coefficient equations R1. (Small 2018, 14, 1704473;Angew. Chem. Int. Ed. 2016, 55, 8003): where D, D0, Ea, e, R and T represent the diffusion coefficient of protons, the maximum diffusion coefficient at infinite temperature, activated energy, unit electric charge, gas constant and temperature, respectively. As the temperature increases of the surface by light irradiation, the thermal motion of molecules and the collisions between molecules are intensified, and the energy Ea is greatly increased, thus increasing the diffusion coefficient D according to equation R1.
In the revised version, the equation of R1 is added as new equation 9 in Note S2 on page 13 of supporting information; Figure R4 is added as new Figure S10a. The relative discussion has been added in the page 6 of main text, and page 13 of supporting information and copied below: Response: We investigated photocatalytic performance for a longer reaction period of 80 h. The performance retained 92% of its initial activity for PAA/CoO-NC with a slight degradation after a longer reaction period of 80 h, as shown in Figure R5. In the revised version, Figure R5 is added as new Figure S19 in the page 28 of supporting information, and the relative discussion has been added in the page 8 of main text Figure R5. Cycling of photocatalytic hydrogen evolution over PAA/CoO-NC hydrovoltaic generator for a reaction period of 80 h. Response: We thank the referee for appreciating the impact and value of our study. We also appreciate the referee's constructive comments, which have helped us improve the quality of our manuscript. We have revised the manuscript accordingly. Regarding question (a), the novelty of the work is that we proposed a hydrovoltaic effectenhanced photocatalytic system. Generally, the CoO nanomaterial serves as a co-catalyst or a photogenerated holes trap center for enhanced photocatalytic performance. However, CoO itself can also act as a main photocatalyst for photocatalytic water splitting as previous literatures reported (Nat. Nanotechnol. 2014, 9, 69;Nat. Commun. 2021, 12, 1343. In our constructed hydrovoltaic generator system of PAA/CoO-NC, the CoO-NC forms an intimate heterostructure, in which CoO was used as a main photocatalyst and NC as a cocatalyst based on the matching energy band structure, achieving an efficient hydrovoltaic effect enhanced photocatalytic system. Other photocatalyst, such as TiO2 can also be used in the hydrovoltaic effect-enhanced photocatalytic system, demonstrating the universality of the system, which has been discussed in Question 3 of Reviewer 3.

Carbon is known to possess large surface area; however, the work is silent about the effect of CoO
on the surface area of the system.

Response:
Thanks very much for the valuable comments. From the N2 adsorption-desorption isotherm plots, the specific surface area of NC ( Figure R6) is 61 m 2 g -1 . After the CoO was introduced, the specific surface area of CoO-NC is 73 m 2 g -1 , showing a slight difference. And the pore size distribution of both samples features typical mesoporous characteristics. In the revised version, Figure R6 is added as new Figure S6 in the page 9 of supporting information. 3. The specific wavelength of the light source used in the photocatalytic generation of H2 is not state, even though Figure S16 depicts several wavelengths was used to determine AQY.

Response:
The H2 generation tests were conducted in the PAA/CoO-NC generator using a 300 W Xenon lamp irradiation with λ > 300 nm (PLS-SXE300, Beijing Perfectlight Technology Co., Ltd, 300 mW cm −2 ). The dependence of different wavelengths for H2 generation in PAA/CoO-NC system was also estimated by various band-pass filters with different wavelengths of 400, 420, 425, 450, 500, 550, and 650 nm as shown in Table R1. In the revised version, the relative discussion has been added in the page 2 of supporting information and copied below:  [2,3] ." (Page 2 of supporting information)

If the DRS UV/VIS of both PAA/CoO and CoO-NC shows similar optical properties as presented in
Fig S14. What is responsible for the difference in rate of H2 evolution in both systems Because the difference is bandgap is not obvious.

Response:
The photocatalytic H2 evolution activity for PAA/CoO-NC and CoO-NC showed a big difference due to the different hydrovoltaic electric field induced in the systems. The generated hydrovoltaic electric field in PAA/CoO-NC is much stronger than that in the CoO-NC, resulting in a promoted carrier separation and transfer on the photocatalyst and higher kinetics of photocatalytic H2 evolution as proved in the main text of Figure 2.
6. The mechanism presented in section 2.5 did not explain the pathway to the formation of H2O2 because in water splitting H2 and O2 are normally the by-product (Theoretically).
Response: Thanks very much for the valuable comments. We added the pathway to the formation of H2O2 and stated the reason why the by-product is H2 and H2O2, not H2 and O2, by DFT calculation. As shown in Figure R7a, the photocatalyst CoO-NC shows a much lower free energy barrier of 0.37 eV for the H2O2 production process in the hydrovoltaic electric field (U=0.4 V) than the intrinsic barrier of 2.05 eV, indicating the hydrovoltaic effect is beneficial for the H2O2 reaction in PAA/CoO-NC system. The photocatalyst CoO-NC shows free energy barriers of 2.07 and 2.64 eV for O2 production with or without hydrovoltaic effect as shown in Figure R7b, much higher than the limiting step barrier for H2O2 generation. As a consequence, the photocatalyst of CoO-NC prefers to generate H2O2 rather than O2 evolution.
In the revised version, Figure R7a and 7b are added as new Figure S27 in the supporting information, and the relative discussion has been added in the page 11 of main text and page 38 of supporting information and copied below: "Density functional theory (DFT) based calculations further indicate that the CoO-NC photocatalyst prefers to generate H2O2 rather than O2 evolution in the system. (Figure S27)." (Page 11 of main text, detail see DFT calculations in supporting information) Figure 1j and S3 did not explain if the CoO is decorated on the surface of the materials or found within the matrix, also the actual amount of the CoO as determined by the XPS is not reported. Figure 1 show that the CoO-NC nanoparticles are wrapped by the cross-linked PAA chains and distributed in the entire matrix. The XPS fits of Co 2p, N 1s, and O 1s in CoO-NC and PAA/CoO-NC are listed in Table   R2, R3 and R4, respectively. The fitted CoO content percentage in CoO-NC and PAA/CoO-NC is about 41% and 23%, respectively. In the revised version, Table R2, R3 and R4 are added as new Table   S1, S2 and S3 in the supporting information, respectively, and the fitted results have been added in the revised version. Response: We thank the referee for appreciating the impact and value of our study. We also appreciate the referee's constructive comments, which have helped us improve the quality of our manuscript. We have revised the manuscript accordingly.

In the introduction, "Hydrovoltaic technology is a renewable energy harvesting method that uses moisture interactions with nanostructured carbon materials". The definition is not accurate.
Hydrovoltaic technology can not only harvest energy from ambient moisture, but also from evaporation,

raindrops, waves and so on. It is characterized by energy conversation based on the interaction
between water and functional materials. The authors should find in which paper this terminology was first proposed.
Response: Thanks very much for the valuable comments. In the revised version, the relevant literatures have been supplemented as Refs. [18-22] (Annu. Rev. Fluid Mech. 2004, 36, 381;Nat. Commun. 2014, 5, 3582;Nat. Nanotechnol. 2014, 9, 378;Phys. Rev. Lett. 2001, 86, 131;Nat. Nanotechnol., 12, 2017, 317); and the definition has been revised and copied below:  In the revised version, Figure R8 has been added as new Figure S10b and the relative discussion has been added in the page 6 of main text, and page 13 of supporting information of Note S2 and copied below: 3. In the system, CoO was used as the active centers for water splitting and was shown to have a desirable band alignment with NC groups. I'm wondering if such a strategy of materials design can be generalized to a range of metal oxides, notably TiO2 that is mostly used for photocatalysis. The authors are suggested to perform additional experiments to demonstrate the generality.
Response: Thanks very much for the valuable comments. We introduced TiO2 into our system to demonstrate the generality. We prepared PAA/TiO2-NC hybrid (Figure R9a-c). The energy band structure of PAA/TiO2-NC based on UV-vis DRS and UPS analysis is obtained (Figure R9d-f). We investigated the electricity generation of PAA/TiO2-NC generator and its enhanced photocatalytic performance. As shown in Figure R9g, PAA/TiO2-NC reaches the highest voltage (~345 mV) at a 1100 ml h -1 Ar/H2O steam injection rate with light illumination, confirming the generation of hydrovoltaic effect in PAA/TiO2-NC generator. The H2 evolution amount of PAA/TiO2-NC nanogenerator is 1.2 mmol at the Ar/H2O steam injection rate of 1100 ml h -1 , and O2 gas is detected to be 0.5 μmol in the first cycle as shown in Figure R9h, higher than the performance of PAA/TiO2-NC in bulk water (Figure R9i), suggesting enhanced photocatalytic water splitting performance in the system of PAA/TiO2-NC induced by the hydrovoltaic effect. ; 100 mW cm -2 ); (h) Time-dependent photocatalytic H2 and O2 production of PAA/TiO2-NC at different Ar/H2O rate at 1100 ml h -1 , and submerged in water; Pt cocatalyst is loaded using a photodeposition method; The light source is a solar simulator at AM 1.5G illumination (100 mW cm -2 ); (i) The corresponding time-dependent photocatalytic H2 and O2 production submerged in water.
In the revised version, Figure R9 is added as new Figure S21 in the supporting information, and the relative discussion has been added in the page 8 of the main text and in page 30 of the supporting information and copied below: "The PAA/TiO2-NC photocatalyst has also prepared to further demonstrate the generality of hydrovoltaic effect enhanced photocatalysis as shown in Figure S21." (Page 8 of main text) "...The PAA/TiO2-NC reaches a voltage (~345 mV) at a 1100 ml h -1 Ar/H2O steam injection rate with light illumination, confirming the generation of hydrovoltaic effect in PAA/TiO2-NC generator (Figure S21g). The H2 evolution amount of PAA/TiO2-NC nanogenerator is 1.2 mmol at the Ar/H2O steam injection rate of 1100 ml h -1 , and O2 gas is detected to be 0.5 mmol in the first cycle as shown in Figure S21h, higher than the performance of PAA/TiO2-NC in bulk water (Figure S21i) Response: Thanks very much for the valuable comments. The experimental for natural waterevaporation-induced hydrovoltaic effect in a PAA/CoO-NC sample has been conducted as illustrated in Figure R10, and it indeed has effect on photocatalysis. Figure R10. (a) Photo of the PAA/CoO-NC hydroelectric generator for measuring water evaporation-induced voltage under light illumination (AM 1.5G, 100 mW cm -2 ); (b) Measured output voltage for PAA/CoO-NC over time with or without light illumination (light intensity: AM 1.5G; 100 mW cm -2 ); (c) Time-dependent photocatalytic H2 and H2O2 production of PAA/CoO-NC; Pt cocatalyst is loaded using a photodeposition method; The light source is a solar simulator at AM 1.5G illumination (100 mW cm -2 ).
To simulate the natural evaporation, we designed a PAA/CoO-NC generator by inserting it into a 100 ml beaker with deionized water covering the bottom electrode under light illumination as shown in Figure R10a. An open-circuit voltage between the two electrodes is generated and gradually rises to 267 mV when the capillary water reaches its maximum height along the PAA/CoO-NC sheet in about 30 mins. The voltage output increases to 318 mV with light illumination at AM 1.5G illumination (100 mW cm -2 ) as shown in Figure R10b. We conducted H2 production experiments to investigate water-evaporation-enhanced photocatalytic water splitting reactions. As shown in Figure R10c, the H2 evolution amount of PAA/CoO-NC nanogenerator is 8.3 mmol, and the oxidation product of H2O2 is measured to be 3.7 mmol, higher than the generalized powder photocatalyst system as shown in Figure 3, suggesting enhanced photocatalytic performance by the water-evaporation-induced hydrovoltaic effect.
In the revised version, Figure R10 is added as new Figure S20 in the supporting information, and the relative discussion has been added in the page 8 of main text and in page 29 of supporting information and copied below: "In addition, the hydrovoltaic effect generation and its enhanced photocatalysis are also demonstrated in a natural water-evaporation-induced hydrovoltaic system ( Figure S20)." (Page 8 of main text) " Figure S20a shows the photo of PAA/CoO-NC film for water-evaporation hydrovoltaic generation. As shown in Figure S20b, an open-circuit voltage between the two electrodes is generated and gradually rises to 267 mV. The voltage output increases to 318 mV with light illumination at AM 1.5G illumination (100 mW cm -2 ). As shown in Figure S20c, the H2 evolution amount of PAA/CoO-NC nanogenerator is 8.3 mmol, and the oxidation product of H2O2 is measured to be 3.7 mmol, suggesting an enhanced photocatalytic performance by the water-evaporationinduced hydrovoltaic effect." (Page 29 of supporting information) 5.To further verify the photocatalytic effect of PAA/CoO-NC, the authors are suggested to carry out control experiments of photocatalytic H2/H2O2 production and provide in situ Raman spectra of pure PAA membranes loaded with Pt cocatalyst at 1100 ml h -1 Ar/H2O steam injection.
Response: Thanks very much for the valuable comments. We conducted the in situ Raman spectra and photocatalytic performance on PAA/Pt at the 1100 ml h -1 Ar/H2O steam injection under light illumination to eliminate the catalytic activity of the substrate. As shown in Figure R11a, a peak centered at 2100 cm −1 assigned to Pt-H vibration, which gradually increases in intensity over time.
The redshift of the Pt-H peak is from 2097 to 2104 cm −1 , indicating a weak hydrovoltaic electric field.
The O-H stretching observed in PAA/Pt evolves at approximately 3200 and 3400 cm −1 , which increases slowly in intensity over time, indicating a weak interaction between PAA/Pt and water molecules. The H2 evolution amount of PAA/Pt membrane is 7.8 μmol, and a trace amount of H2O2 of 0.3 μmol is detected due to the introduced weak hydrovoltaic effect of PAA and intrinsic hydrogen production characteristic of Pt, indicating a slight influence of substrate for the system (Figure R11b). Figure R11. (a) In situ Raman spectra of the PAA/Pt surface over time with hydrovoltaic effect under light irradiation; (b) Time-dependent photocatalytic H2 and H2O2 production of PAA/Pt; Pt cocatalyst is loaded using a photodeposition method; The light source is a solar simulator at AM 1.5G illumination (100 mW cm -2 ).
In the revised version, Figure R11 is added as Figure S23 in the supporting information, and the relative discussion has been added in the page 10 of the main text and page 33 of the supporting information and copied below: 6.There are some misquotes in the introduction, such as "For example, a hydroelectric generator comprising the ionic polymer Nafion and a poly(N-isopropylacrylamide) hydrogel was developed to generate electricity [17,18]". Please check the literature comprehensively.
Response: Thanks very much for the valuable comments. We have checked the literature comprehensively, and the cited reference [17,18] has been corrected to reference [33] Energy Environ. Sci. 15, 2489Sci. 15, -2498Sci. 15, (2022. Figure S1, the description on the device fabrication process in figure legend does not match its schematic illustration. For example "Commercial epoxy slurry is used to paint two "L-shaped" with predesigned dimensions on the substrate (Step 1)"; here, the electrodes should be "L-shaped", not Response: Thanks very much for the valuable comments. The schematic of device electrode has been revised and more clearly represented as shown in Figure R12. In the revised version, it is added as new Figure S1 in page 6 of Supporting Information. Figure R12. Device schematic and configuration used for electricity measurements.

7.In
8.In Note S1, "relative humidity by pumping a mixture of dry and wet nitrogen", while in the paper, the humidity is controlled by "Ar/H2O steam injection".
Response: Thanks very much for the valuable comments. We have revised the annotation to "argon gas" in Note S1 in page 10 of supporting information. Figure S7 (c) and Figure S8 (b), "The increase in the current density may be attributed to the increased amount of water moving and diffusion as W increases". While in general perception, the current increases with the width of the hydroelectric generator, and the current density is nearly constant. Please explain it.

9.In
Response: Thanks very much for the valuable comments. We re-tested the voltage and current density for PAA/CoO-NC and CoO-NC devices by taking the surface area and thickness as the major influencing factors. The device surface area is selected as 15 cm 2 and the thickness of film is selected to be 300 μm.
The PAA/CoO-NC and CoO-NC devices with sandwiched electrodes for electricity tests are shown in Figure R13a. The surface area and thickness are important factors for electricity generation.
As the surface area of the device increases from 1 cm 2 to 18 cm 2 , the voltage and current density for PAA/CoO-NC and CoO-NC all showed a slight decrease, which is due to the additional defects introduced into the devices as the surface area increases. The voltages are measured to be ≈280 mV and ≈100 mV for PAA/CoO-NC and CoO-NC devices, respectively, while the current density showed a basically stable value around ≈12 μA cm -2 and ≈4 μA cm -2 for PAA/CoO-NC and CoO-NC device, respectively (Figure R13b, c). These results indicate that the surface area of devices has little effect on the electricity generation for the devices. Considering a larger surface area of the device is beneficial for photocatalysis due to complete exposure to light. We selected the device surface area of 15 cm 2 for the generation of electricity and its photocatalytic reactions.
For the effect of thickness exploration, we prepared different devices with thicknesses of 100, 300, 500 and 700 μm for electricity generation. The devices showed higher electricity generation at 300 μm for PAA/CoO-NC and CoO-NC ( Figure S8).