General heterostructure strategy of photothermal materials for scalable solar-heating hydrogen production without the consumption of artificial energy

Solar-heating catalysis has the potential to realize zero artificial energy consumption, which is restricted by the low ambient solar heating temperatures of photothermal materials. Here, we propose the concept of using heterostructures of black photothermal materials (such as Bi2Te3) and infrared insulating materials (Cu) to elevate solar heating temperatures. Consequently, the heterostructure of Bi2Te3 and Cu (Bi2Te3/Cu) increases the 1 sun-heating temperature of Bi2Te3 from 93 °C to 317 °C by achieving the synergy of 89% solar absorption and 5% infrared radiation. This strategy is applicable for various black photothermal materials to raise the 1 sun-heating temperatures of Ti2O3, Cu2Se, and Cu2S to 295 °C, 271 °C, and 248 °C, respectively. The Bi2Te3/Cu-based device is able to heat CuOx/ZnO/Al2O3 nanosheets to 305 °C under 1 sun irradiation, and this system shows a 1 sun-driven hydrogen production rate of 310 mmol g−1 h−1 from methanol and water, at least 6 times greater than that of all solar-driven systems to date, with 30.1% solar-to-hydrogen efficiency and 20-day operating stability. Furthermore, this system is enlarged to 6 m2 to generate 23.27 m3/day of hydrogen under outdoor sunlight irradiation in the spring, revealing its potential for industrial manufacture.

The work by Li, Wang, Luo, and coworkers demonstrate a novel photothermal material for converting sunlight into non-artificial heat with unprecedented high temperatures reached. The authors demonstrate the use of this material by performing methanol steam reforming (MSR) at 260 degrees Celcius entirely driven by solar heating, which is very impressive and highly interesting in my mind. However, solar photothermal driven thermochemical MSR reactions is not new, as the authors point out themselves. Nevertheless, the presented work does provide an improvement of approximately an order of magnitude compared to the current state-of-the-art, and their device is stable for more than 20 days. I find it confusing, though, that the authors apparently go on to synthesize a new MSR catalyst to confirm the use of their novel photothermal material, instead of demonstrating it with already published thermally driven catalytic reactions. How does their approach validate the superiority of Bi2Te3/Cu over other solar-to-thermal conversion devices? And how do we really compare the results when both the catalyst and the photothermal device are new? If I am to assess the overall activity and stability of the MSR, the work is, in my view, of high significance, but does not warrant publication in Nature Communications in its current form. However, the work should be re-assessed after the following considerations: 1) The concept of "Methanol economy" by Olah ought to be mentioned. 2) If the photothermal device is merely developed to provide breakthrough-level performing solar driven MSR, I suggest to further optimize the results to obtain a more significant improvement over state-ofthe-art. If the MSR is included to merely highlighting the power of the photothermal device, I suggest to test and compare outcome when using known catalysts for MSR.
3) The authors employ MSR as a benchmarking reaction for conducting a thermally driven reaction at a high temperature above 120 degrees Celcius using their novel photothermal device. In my mind, it would be pertinent at least to acknowledge that there indeed exist low-temperature methanol reforming protocols performing well below 120 degrees Celcius (e.g. Beller Nature 2013, 495, 85 as well as Trincado and Grützmacher Nature Chemistry 2013, 5, 342), and to explain why it is pertinent to develop a sun-to-heat device for high-temperature MSR when low-temperature systems have already been developed on for almost a decade. An explanaition becomes even more relevant when considering that PEM fuel cells generates an excess of heat that the low-temperature MSR potentially could exploit. 4) The authors mention that their system produces a low concentration of CO; however, 0.8% of CO (of CO2 and CO) is not a low concentration for PEM fuel cells. Please consider this more carefully, especially in the light of that they mention how this system may be used for powering a vehicle. 5) It would be instrumental with some more pictures that clearly show the function and setup of the device, eg of the outdoor setup.
Reviewer #3 (Remarks to the Author): In this manuscript, the authors propose to combine black photothermal materials and infrared insulating materials, so that an elevated solar-heating temperature can be achieved. As an example, the Bi2Te3/Cu hybrid showed a 1 Sun-heating temperature of 317 °C. In addition, a reaction device based on Bi2Te3/Cu was assembled, which was able to heat a catalyst of CuOx/ZnO/Al2O3 nanosheets to 305 °C and showed a 1 Sun driven hydrogen production rate of 75.9 L/h from methanol and water. These results are impressive, and the following issues need addressed before possible publication of this manuscript.
1 When the thickness of Bi2Te3 layer was 100 nm, the Bi2Te3/Cu exhibited the highest 1 Sun irradiated temperature of 317 °C, higher than those for 15nm and 3 micron thick Bi2Te3. Is this an optimized result? What is the reason behind? 2 What is the fine structure of the Bi2Te3 layer? As seen from the TEM images, there seems to be difference between the structures of the Bi2Te3 layer with different thicknesses. Is the heating performance of the hybrid only thickness-related, or influenced by the structure of Bi2Te3? 3 What is the Bi2Te3/Cu interface structure? How does it affect the Sun-heating performance? 4 An illustration and corresponding discussion clearly showing the working mechanism of this hybrid should be presented, i.e. how is the sun light irradiation transferred to thermal energy efficiently through the Bi2Te3 layer, Cu and their interface? What is the key physics of this hybrid design? First, the authors should clarify the methanol water reforming in this study is a photocatalytic reaction, thermal catalytic reaction or combined reaction.

Response:
The authors thank for the reviewer's constructive comment. The methanol water reforming is thermal catalytic reaction in solar heating catalysis, as the catalysts loaded in the device can not absorb the sunlight, which is depicted in Supplementary Because most of photocatalysis reactions are "Without Consuming Artificial Energy".
The authors "only" compared this study with other sunlight-driven hydrogen production in Fig. 5b and Tab. 1. It is worth noting that, as the Solar Heating on the Photothermal Materials created a high temperature environment, which high enough (above 300 ºC) to trigger a thermal catalysis reaction. Especially the CuZnAl is a highperformance methanol water reforming catalyst. Therefore, I suggest the authors clarifying their catalysis system and setting up a reasonable comparation with another studies.

Response:
The authors thank for the reviewer's comment.
The heat energy used for solar heating MSR is achieved only from solar by the photothermal conversion of Bi2Te3/Cu based device. In comparison, the energy 3 / 27 consumed by traditional thermocatalytic MSR is artificially input electricity or fossil energy, while the energy consumed by photocatalytic MSR is sunlight. Therefore, the comparison between solar heating MSR with photocatalytic MSR is fair. The purpose of this comparison is to show that the path of solar-heat energy-chemicals in our work is more efficient than the path of solar-photogenerated carriers-chemicals in photocatalytic MSR. We can use the path of solar-heat energy-chemicals to drive the low energy barrier reactions (such as MSR) more efficiently.
Furthermore, to make this reaction running, a "Artificial Energy" driven pump is needed.

Response:
The authors thank for the reviewer's comment.
Firstly, what we call the demand for artificial energy mainly refers to the catalysis (MSR) itself, which does not involve peripheral pumping and other auxiliary processes. We have changed the expression in the revised manuscript.
"Promoting industrial catalysis without artificial energy input is of great significance for human beings.
Therefore, constructing artificial-energy-input-free catalysis is the key to human sustainable development." Secondly, the efficiency of our solar heating MSR is much higher than that of photocatalysis, and the consumption rate of methanol aqueous solution is too fast, which requires continuous pumping to provide methanol aqueous solution. If photocatalysis consumes all methanol, photocatalysis also requires a pumping process to add methanol.  as the substrate to grow CuZnAl nanosheets. As we all know, using AAO/Al to synthesize catalyst cannot produce catalyst on large scale, that is, the kg scale production of catalyst cannot be carried out. However, our method can mass-produce CuZnAl nanosheets in kg level directly, which is the biggest highlight of our method for catalysts synthesis.
We tested the density of both CuZnAl NS and C-CuZnAl, and the density of CuZnAl NS powder is about 0.12g cm -3 and the one of C-CuZnAl powder is about 0.795g cm -3 .
We have added the data in the revised manuscript.
"We tested the density of both CuZnAl NS and C-CuZnAl, the density of CuZnAl NS powder was ~0.12g cm -3 and C-CuZnAl powder was ~0.795g cm -3 ." Third, in this study, the H2 production rates were depending on the Photothermal Materials, hybrid of Bi2Te3 and Cu showing the best performance. As show in the references listed in this manuscript, the Bi2Te3 as photothermal materials have been heavily study. And novelty of hybrid of Bi2Te3 and Cu preparation can not reach the level of Nat. Comm.

Response:
The authors thank for the reviewer's comment. As the reviewer said, the Bi2Te3 has been heavily studied as photothermal material, but the 1 Sun illuminated temperature of our Bi2Te3 structure is 317 °C, far higher than that of any reported Bi2Te3 structures (Nano Energy 77 (2020) 105102) and photothermal materials (Nature Energy 6 (2021) 807-814) under 1 Sun illumination. More importantly, this manuscript reveals that the heterostructure of narrow-band gap materials (e. g. Bi2Te3) and polished metals (e. g. Cu) can generally improve the sunlight illuminated temperature of narrow-band gap materials, which can provide high-quality heat energy for many fields, such as catalysis, power generation, etc. It is of great innovation and application significance.
Here some comments on the details: 1) In the introduction section: "But, hydrogen generation from methanol and water by 2) Table 1, the authors should list reaction conditions such as reaction temperature.
Response: Thanks for the reviewer's advice. We have added the reaction temperature in Table 1.

Tab. 1
The sunlight driven hydrogen generation from MSR of CuZnAl NS loaded in Bi2Te3/Cu based device, in comparison with the reported advanced sunlight driven hydrogen generation systems. (T: test temperature).
3) The authors should list the methanol conversion.
Response: Thanks for the reviewer's good suggestion. We have added the methanol conversion in Supplementary Fig.14
Response: Thanks for the reviewer's question. We are sorry for our confusing diagram.
To avoid confusing, we have deleted the Figure   " Supplementary Fig. 13 showed that our synthesized sample could full fill a 40 L bottle, revealing the ability of scalable preparation." "As a result, the hydrogen production rate of CuZnAl NS was 1.02 mol g -1 h -1 at 260 °C, quintupling the 0.2 mol g -1 h -1 of C-CuZnAl at 260 °C (Fig. 4g), and the methanol conversion rate was 5.28 % (Supplementary Fig.14)."

/ 27
Reviewer #2 (Remarks to the Author): The work by Li Supplementary Fig. 11 The hydrogen production rates from two MSRs of C-CuZnAl loaded in the Bi2Te3/Cu based device and C-CuZnAl not in the device but directly under different sunlight irradiations.
"we added the commercial CuZnAl (C-CuZnAl, Supplementary Fig. 10) into Bi2Te3/Cu based device to test the sunlight driven MSR performance. As shown in Supplementary   Fig. 11, the 1 Sun driven MSR H2 generation rate through C-CuZnAl was 79.3 mmol g -1 h -1 , which is much higher than that reported record photocatalytic MSR value (46.6 mmol g -1 h -1 ), 1 thus highlighting the importance of the Bi2Te3/Cu based device. To achieve higher sunlight driven MSR performance, the more efficient catalysts for MSR need to be developed."

/ 27
1) The concept of "Methanol economy" by Olah ought to be mentioned.

Response:
We thanks for the reviewer's good suggestion. Therefore, we have added the "Methanol economy" by Olah in the revised manuscript.
"Owing to those storage limitations of hydrogen such as high pressure, leakage, and extensive safety precautions, Olah has proposed methanol economy as the methanol can be act as the hydrogen carrier in the future hydrogen energy system, 12

/ 27
Supplementary Fig. 11 The hydrogen production rates from two MSRs of C-CuZnAl loaded in the Bi2Te3/Cu based device and C-CuZnAl not in the device but directly under different sunlight irradiations.
"we added the commercial CuZnAl (C-CuZnAl, Supplementary Fig. 10) into Bi2Te3/Cu based device to test the sunlight driven MSR performance. As shown in Supplementary   Fig. 11, the 1 Sun driven MSR H2 generation rate through C-CuZnAl was 79.3 mmol g -1 h -1 , which is much higher than that reported record photocatalytic MSR value (46.6 mmol g -1 h -1 ), 1 thus highlighting the importance of the Bi2Te3/Cu based device. To achieve higher sunlight driven MSR performance, the more efficient catalysts for MSR need to be developed." 3) The authors employ MSR as a benchmarking reaction for conducting a thermally driven reaction at a high temperature above 120 degrees Celcius using their novel photothermal device. 4) The authors mention that their system produces a low concentration of CO; however, 0.8% of CO (of CO2 and CO) is not a low concentration for PEM fuel cells. Please consider this more carefully, especially in the light of that they mention how this system may be used for powering a vehicle.
Response: Thanks for the reviewer's comment. The CO concentration that can be used for PEM fuel cells is less than 10 ppm (Beller Nature 2013, 495, 85), the hydrogen produced by our system requires purification for a vehicle that we will do in the future.

/ 27
Therefore, we have deleted the expression in the manuscript.
5) It would be instrumental with some more pictures that clearly show the function and setup of the device, eg of the outdoor setup.
Response: Thanks for the reviewer's comment. We provide a video (Supplementary Movie 1) for the outdoor setup to facilitate reviewers' reading.

/ 27
Reviewer #3 (Remarks to the Author): In this manuscript, the authors propose to combine black photothermal materials and infrared insulating materials, so that an elevated solar-heating temperature can be achieved. As an example, the Bi2Te3/Cu hybrid showed a 1 Sun-heating temperature of 317 °C. In addition, a reaction device based on Bi2Te3/Cu was assembled, which was able to heat a catalyst of CuOx/ZnO/Al2O3 nanosheets to 305 °C and showed a 1 Sun driven hydrogen production rate of 75.9 L/h from methanol and water. These results are impressive, and the following issues need addressed before possible publication of this manuscript.

Response:
The authors are grateful for the reviewer's thorough comment. 2 What is the fine structure of the Bi2Te3 layer? As seen from the TEM images, there seems to be difference between the structures of the Bi2Te3 layer with different thicknesses. Is the heating performance of the hybrid only thickness-related, or influenced by the structure of Bi2Te3?
Response: Thanks for the reviewer's question. We are also very confused about this phenomenon. After discussion with experts, we think that the morphology difference of samples might be caused by the sample preparation. To check this speculation, TEM samples of Bi2Te3/Cu films with 100 nm and 15 nm thickness of Bi2Te3 were prepared by ion thinning, and then it was found that the cross-section TEM images of the two samples were similar (Fig. 1d, e). In comparison, the SEM sample of Bi2Te3/Cu film with 3 μm thickness of Bi2Te3 (Fig. 1c) is synthesized not by ion thinning but by direct shear. Therefore, we can see that the cross-section SEM image is rough, different from  3 What is the Bi2Te3/Cu interface structure? How does it affect the Sun-heating performance?
Response: The authors thank for the reviewer's question. We provided the interface structure of Bi2Te3/Cu in Supplementary Fig. 2. The description of interface structure and effect on solar heating catalysis is added in Supplementary Information, which is also shown below here for the convenience of the reviewer.

/ 27
Supplementary Fig. 2 The HRTEM image of the interface of Bi2Te3 and Cu. Response: The authors thank for the reviewer's comment and we have added the description in Supplementary Fig. 8, which is also shown below here for the convenience of the reviewer.
Supplementary Fig. 8 The picture of catalysts loaded in Bi2Te3/Cu based device.
"When sunlight directly illuminated the solar heating device, the layer of Bi2Te3 could efficiently convert solar energy to thermal energy and radiate little infrared light; the inner layer of Cu could stop infrared emission of catalysts; while the outer vacuum layer can block the thermal conduction from the inner device. In this way, we could create a high temperature to drive MSR under solar illumination ( Supplementary Fig. 8).
The key physics of this hybrid is that the infrared radiation is directly proportional to the amount of sunlight absorber (Bi2Te3). Therefore, the thinner the thickness of Bi2Te3, the less its heat radiation will be. At the same time, if the thickness of Bi2Te3 film is less 26 / 27 than 100 nm, its ability to absorb sunlight will be greatly reduced. Therefore, controlling the thickness of Bi2Te3 to about 100 nm can achieve the balance of high sunlight absorption and low infrared radiation to produce high sunlight irradiation temperature." Dear Editor and authors, I carefully read through the revised manuscript and response letter. The authors made a perfect revision, all my concern have been addressed.
In the first round of review, the only reason I did not recommend publishing is that this study is engineering, or device orientated. I found some similar studies have been published in the Nature Communications previously. More importantly, the authors convinced me with their careful and perfect revision. Therefore, I recommend publishing it in Nature Communications.
Reviewer #2 (Remarks to the Author): In this resubmission by Li et al, the authors fully address my concerns regarding their initial submission. However, in my mind, the outcome of this effort does not lead to a satisfactory improvement of the overall work for warranting publication in Nature Communications. More specifically, the reported catalytic performance of the MSR is indeed significantly superior to literature precedence, but they are not radical improvements. As such, I recommend publishing in a more specialized journal.
Reviewer #3 (Remarks to the Author): The authors answered my questions, but not completely. My main concern is how the Bi3Te2/Cu system achieves a superior solar-to-thermal conversion performance, which is a main novelty of this work. Is it related to the thickness and structure of Bi3Te2 layer, the Bi3Te2/Cu interface, or some new physics of this device? However, the authors did not give a clear answer and description on this point. As a result, there is insufficient deepened discussion and understanding on the achieved device and phenomena. At the current stage, this manuscript is not suitable for publication in Nature Commun.

REVIEWER COMMENTS
Reviewer #1 (Remarks to the Author): Dear Editor and authors, I carefully read through the revised manuscript and response letter. The authors made a perfect revision, all my concern have been addressed.
In the first round of review, the only reason I did not recommend publishing is that this study is engineering, or device orientated. I found some similar studies have been published in the Nature Communications previously. More importantly, the authors convinced me with their careful and perfect revision.
Therefore, I recommend publishing it in Nature Communications.

Response:
The authors very thank indeed for your professional and kind support. indeed significantly superior to literature precedence, but they are not radical improvements. As such, I recommend publishing in a more specialized journal.

Response:
The authors thank for your helpful comment on our catalytic performance of the MSR that is indeed significantly superior to literature precedence. We also believe that the performance is indeed a radical improvement. This is because, firstly,  Bi2Te3 thin film on Cu support (Bi2Te3/Cu).
"For achieving high solar irradiated temperature, in addition to superior solar-tothermal conversion, it is also necessary to localize the sunlight converted heat energy in Bi2Te3, that is reducing the heat dissipation of Bi2Te3. Despite a vacuum protection was applied to cut off the heat conduction loss of pure Bi2Te3 film, the 1 Sun (1 kW m -2 ) illuminated temperature of pure Bi2Te3 film was only 93 °C ( Supplementary Fig. 1d).
As a blackbody material (Fig. 1a), 8 the heat dissipation of pure Bi2Te3 film includes not only the heat conduction loss but also importantly the violent heat radiation loss caused by the infrared light (IR) radiation (0.91 of IR emissivity shown in Supplementary   Information). 9 Therefore, minimizing the IR radiation of Bi2Te3 is the key for increasing its solar irradiated temperature. The IR lights radiated by Bi2Te3 are produced by lattice vibrations and the lattice vibrations are proportional to the number of atoms in Bi2Te3. 10 From the physical principle, reducing the number of atoms of Bi2Te3 structure can weaken the IR radiation, so that our strategy is synthesizing Bi2Te3 thin film to minimize the number of atoms to minimize the IR radiation as shown in Fig. 1b. In order to achieve low IR radiation of Bi2Te3 thin film structure, the supports used for depositing Bi2Te3 thin film need to have the property of low IR radiation too. However, the supports used for depositing Bi2Te3 thin film are usually silicon, which is also a typical blackbody material with strong IR radiation and cannot be used as the support for reducing the IR radiation of whole Bi2Te3 thin film structure. 8 Different from blackbody materials, highly conductive metal of Cu contains a large number of near free electrons that can prevent the spillover of IR lights, 11 and makes Cu have near zero 8 / 12 IR radiation (~3 % of IR emissivity, Supplementary Fig. 2). 12,13 Therefore, Cu film is selected as the support for synthesizing Bi2Te3 thin film to make the hybrid have the merits of superior solar-to-thermal conversion from Bi2Te3 and low IR radiation from Cu. 14 "

/ 12
as 317 °C through controlling the thickness of the Bi2Te3 thin film in Bi2Te3/Cu, which is not only 224 °C higher than that of pure Bi2Te3 (93 °C) but also 197 °C higher than the reported highest temperature of photothermal materials (120 °C) under 1 Sun irradiation (Adv. Funct. Mater. 31, 7, 2021). 1 The mechanism of the Bi2Te3 thickness control is to balance the solar absorption and the IR radiation of Bi2Te3/Cu. We have described the detail in the revised manuscript and also place the description below for your convenience. Further, in order to show more clearly the thickness effect of the Bi2Te3/Cu heterostructure on the IR radiation, we have added the IR radiation intensities of all samples at 93 °C in Fig. 2d, e, f and Supplementary Fig. 1e. 89 %, 43 %, respectively. Bi2Te3 has a narrow bandgap of ˂ 0.2 eV, 4,5 thus, the sunlight has enough energy to excite electrons transition in Bi2Te3. 15,16 But, the film thickness of Bi2Te3 must be ≥ 100 nm to ensure more than 89% of solar spectrum absorption.
Whereas, the IR region absorption was 4 %, 5 % as the thickness of Bi2Te3 thin film in Bi2Te3/Cu was 15, 100 nm, respectively (Fig. 2b, c), and it increased to 60 % when the thickness of Bi2Te3 thin film extended to 3 μm (Fig. 2a). As the absorptivity of light is equal to the emissivity of corresponding light, 11 the 60 % IR absorption depicted that the IR emissivity of Bi2Te3/Cu with 3 μm of Bi2Te3 thin film is 60 %, at least 10 times higher than the Bi2Te3/Cu with 100 nm (5 %), 15 nm (4 %) thickness of Bi2Te3 thin 11 / 12 film. For a more intuitive embodiment, we directly tested the IR radiation intensity (4 μm-20 μm) of these samples heated to 93 °C. As shown in Fig. 2d, e, f, the IR radiation intensity in the range of 4 μm-20 μm is 248 W m -2 , 20.7 W m -2 , 16.6 W m -2 for the Bi2Te3/Cu with 3 μm, 100 nm, 15 nm thicknesses of Bi2Te3 thin film, respectively, significantly lower than the corresponding IR radiation of pure Bi2Te3 film of 377 W m -2 ( Supplementary Fig. 1e)