Immobilized covalent triazine frameworks films as effective photocatalysts for hydrogen evolution reaction

Covalent triazine frameworks have recently been demonstrated as promising materials for photocatalytic water splitting and are usually used in the form of suspended powder. From a practical point of view, immobilized CTFs materials are more suitable for large-scale water splitting, owing to their convenient separation and recycling potential. However, existing synthetic approaches mainly result in insoluble and unprocessable powders, which make their future device application a formidable challenge. Herein, we report an aliphatic amine-assisted interfacial polymerization method to obtain free-standing, semicrystalline CTFs film with excellent photoelectric performance. The lateral size of the film was up to 250 cm2, and average thickness can be tuned from 30 to 500 nm. The semicrystalline structure was confirmed by high-resolution transmission electron microscope, powder X-ray diffraction, grazing-incidence wide-angle X-ray scattering, and small-angle X-ray scattering analysis. Intrigued by the good light absorption, crystalline structure, and large lateral size of the film, the film immobilized on a glass support exhibited good photocatalytic hydrogen evolution performance (5.4 mmol h−1 m−2) with the presence of co-catalysts i.e., Pt nanoparticles and was easy to recycle.


experiments?
13. There are many grammatical and typographical errors. For example: -'In early studies, it is technically difficult' is a grammatically not correct statement. -The figure captions for all the figures have many errors.
-The statement 'Researchers were faced' is a grammatically not correct statement.
Reviewer #2: Remarks to the Author: The manuscript presented by B. Tan el at. is an excellent report on how to prepare large crystalline films of a COF (in this report a covalent triazine framework (CTF)). As explained by the authors in the manuscript, most of the CTF used as photocatalysts in the hydrogen evolution reaction (HER) or carbon dioxide reduction reaction (CO2RR) are solid powder suspensions, but for their use in photoelectrocatalytic devices, films are a much better solution. In this sense, the material prepared in this manuscript is an excellent step forward in this direction. The synthetic organic approach to obtain the CTF films ground-breaking and it will open many new opportunities to develop this field. In general, the characterization of the material is thoroughly performed, and the results obtained confirm that the materials were obtained without doubts. There are some minor aspects in the characterization that I think that should be improved, but I will describe them at the end of this report. In any case, they do not take away any merit to the manuscript. The experimental details for the preparation of the materials and the characterization seems to be enough detailed to reproduce the preparation of the films. My main concern is related to the photocatalytic tests in the HER reaction. The experiments are acceptable, but I think these comments should be addressed: Page 12: Hydrogen evolution reaction (1 hour) and long-term hydrogen evolution experiments. Authors describe several experiments to determine the applicability of the material as a photocatalyst in HER reaction when assisted with Pt as a co-catalyst. In this regard, I am missing a full characterization of the material after the photocatalytic tests and some further (photo)(electro)chemical experiments: 1) what is the fate of the "molecular" Pt co-catalyst? HR-TEM measurements after photocatalysis could show if Pt NPs were formed during photocatalysis. 2) After some long-term cycles, I would perform the photocatalytic test with a fresh TEOA solution (without co-catalyst): if Pt NPs were formed, the material should still be active. 3) HR-TEM images after catalysis would show if the film was degraded during photocatalysis. 4) Instead of cycles, I would perform a single long-term photocatalytic experiment (+100 hours) to check the stability of the material. Moreover, it is not clear the experimental procedure used in each cycle: new TEOA and Pt catalyst in each cycle? 5) What happens when the reaction is performed without the addition of the Pt co-catalyst in the first cycle (or in a single 1-hour experiment)? 6) Is it possible to check the activity of the "film + Pt" material in an ITO electrode without light or TEOA and applying a potential to check the electrochemical properties of the material? I expect it to be very active: overpotential for HER could be obtained. Moreover, using a Clark electrode, faradaic efficiencies could be obtained, and this would show that the hydrogen evolution reaction is the only reaction taking place. 7) After photocatalytic experiments, and after thoroughly washing the film, ICP analyses of the film will show if undetectable Pt NPs (ultrasmall) or coordination compounds are retained in the film. 8) A more detailed comparison with other similar systems should be added. Please, cite and compare the manuscript results with the results described in: doi.org/10.3390/catal11060754 Minor general points: 1) Page 3: Sentence in Lines 32-34. Unfinished sentence? Or just an "and" missing? 2) Page 4: Sentence in Lines 75-78. Even if I think that the meaning can be understood, it is difficult to read, and I would rephrase the sentence to make it more clear.
3) Page 5. Line 102. "n-Hexane" should be written without capital letters. 4) Page 6. Scheme 1. In the chemical reaction depicted in the scheme, I would add the by-products of the reaction in every step (water?). 5) Page 6. Scheme 1. Line 107. "a) possible reaction mechanism". I would rather describe them as "reaction steps" (or something similar) rather than "reaction mechanism". 6) Page 8. Line 132. Space missing at "be14.55". 7) Page 9. Sentence in Lines 155-157. Again, the sentence seems to be unfinished (or an "and" missing". 8) Page 9. Line 159. XPS description seems unfinished. It seems as if an initial paragraph is missing. 9) Page 9. Lines 163-164. Re-write the text in these lines. 10) Page 12. Line 217. "with Pt as cocatalyst". Even if usually referred just as the cocatalysts, the role of Pt in these photocatalytic reactions is crucial and, at least, the actual Pt source (H2PtCl6) should be detailed here.
All in all, I think that this manuscript should be accepted for publication if these comments are addressed (or refuted).
Reviewer #3: Remarks to the Author: In this paper, Tan et al. developed a smart synthetic strategy to first create a liquid/air interface, and fabricate successfully free-standing, large area, crystalline Covalent triazine frameworks (CTFs) films by this kind of interface. Furthermore, they explored the application of the film as immobilized photocatalysts and showed that the fabricated CTFs film had an excellent performance of photocatalytic hydrogen evolution. From a practical point of view, immobilized CTFs are more suitable for large-scale water splitting applications, however, existing synthetic approaches mainly result in insoluble and non-processable powders, which makes their future device application still a huge challenge. It is indeed highly interesting in this work that an interface is created without a water phase, which provides a rational approach to prepare immobilized photocatalysts, which is a significant landmark for the development of CTF and COF films and their potential integration in future devices for applications in fields as photoelectric, sensing, separation, and energy technologies. After careful evaluation, I would like to recommend this paper for publication in Nature Communications after the authors address the following important points: 1. Page 6, line 104: "Because of the weak polarity of the long carbon chain, the imine precursor was floating at the DMSO layer instead of dissolving in it". Because this process is very important for the creation of liquid/air interface and the preparation of CTFs films, can the authors give further experiment details to describe this process?

Response to Reviewer' Comments
Given below are our responses (in BLUE colour) to the reviewer' comments. The changes to the manuscript and supplementary information are marked in RED colour.

Reviewer 1:
The manuscript from Hu et.al. reports the synthesis of large-sized and free-standing thin films of covalent triazine framework (CTF) via an aliphatic amine-assisted interfacial polymerization method. Authors have shown that the CTF films with a lateral size up to 250 cm 2 and the thickness ranging from 30 to 500 nm can be prepared. The large-sized films with good photoelectric performance have been applied as a semiconductor catalyst for photocatalytic hydrogen generation. Although the values reported for photocatalytic hydrogen evolution (5.4 mmol h -1 m -2 ) are good. The crystallinity of the films obtained is very limited.

Response:
We thank the reviewer for understanding and appreciating the significance of manuscript.
We agree that the crystallinity of the films obtained is still limited. We think there are two Here, we show that the film prepared by an irreversible reaction also can possess a certain degree of crystallinity.
In addition, basic characterization such as Grazing-Incidence Wide-Angle X-ray Scattering (GIWAXS), Raman spectroscopy analyses, etc. are missing.
Response: This is a good suggestion. We have now provided the GIWAXS ( Manuscript holds certain novelties about the synthesis of large-sized CTF films and their applications for photocatalytic water splitting, however, it is unclear whether the 2D films are crystalline as the authors trying to claim or it is just random mixture of crystalline and amorphous domains. This is the main reason I would like to withhold my suggestion to recommend publications as it stands. Following are suggestions where the authors could potentially improve and questions regarding the characterizations and properties of these materials: Response: After more careful analysis, we agree that the 2D films are consisted of both the crystalline and amorphous domains. We have updated this information in the revised manuscript and believe that calling it a "semi-crystalline film", will be more appropriate. 1. The basic claim that authors have made is the crystallinity of the CTF films, which is not much clear from SAED, PXRD and SAXS profiles.   Here, we obtained the GIWAXS data carried out on lab X-ray source, which shows clear scattering signal at about q = 0.51 Å -1 ; and also convincing radial integration intensity profiles In the manuscript, we have added the related discussions as follows:  Table 1 and Supplementary Table 2).

Response:
We have now revised the scheme to better understand the reaction.

Scheme 1. Scheme of CTFs film synthesis. a Reaction steps. b
Synthetic procedure for the film on DMSO surface assisted by imine precursor.
6. The elemental mapping for carbon and nitrogen atoms shown in figure 3 give a hint that distribution of carbon atoms is very uniform, however, nitrogen atoms are not present in the proportion. Roughly, the proportion of nitrogen to carbon should be 1:3, this does not seem the case. Authors should provide the elemental analyses.

Response:
The counts per second (CPS) of our previous elemental mapping data are too low, which made the signal of N very weak. Now, we analyzed the samples again by SEM and EDS for better elemental mapping ( Fig. 4d -4f). From Fig. 4d -4f, we can observe the uniform distribution of carbon and nitrogen. Meanwhile, the elemental analyses also show that the ratio of nitrogen and carbon is close to the theoretical value (Supplementary Table 3).   10. Although authors claim that there is a little decrease in the hydrogen evolution performance after the second cycle, from figure 4c, it seems that the performance is decreased by more than 20-25%. What is the reason for such low activity just after 10 hours?

Supplementary
Response: From Fig. 4c, it seems that the performance decreased significantly. To investigate the stability of performance in HER reaction, we performed a single long photocatalytic experiment (100 h), the performance is very stable in the first 60 h, followed with decrease after 60 h, but still active and stable till 150 h by replenishing TEOA. So, the most likely reason for the decrease of performance after 2 cycles may be due to the peeling off of film from the glass support during the cyclic process.
Furthermore, decrease of performance in long-term cycles has also been observed in some film-based inorganic semiconductor photocatalytic systems for overall water splitting (the performance is about 65% of the initial activity after 40 h; Joule, 2018, 2, 2667 -2680). The decrease of the activity in long-term cycles is generally attributed to the backward reactions or the agglomeration/detachment of co-catalysts.

What is apparent quantum yield (AQY) for the films?
Response: The apparent quantum yield for the films has now been calculated and mentioned in the revised manuscript as shown the table below.   13. There are many grammatical and typographical errors.

Supplementary
For example: -'In early studies, it is technically difficult' is a grammatically not correct statement.
-The figure captions for all the figures have many errors.
-The statement 'Researchers were faced' is a grammatically not correct statement.
Response: Thanks for your suggestions; we are sorry for these mistakes, and have tried our best to remove such mistakes in the revised manuscript.
'In early studies, it is technically difficult' is a grammatically not correct statement.
-The figure captions for all the figures have many errors.
-The statement 'Researchers were faced' is a grammatically not correct statement.
Response: All such mistakes/typos have now been removed.

Reviewer #2:
The manuscript presented by B. Tan el at. is an excellent report on how to prepare large crystalline films of a COF (in this report a covalent triazine framework (CTF)). As explained by the authors in the manuscript, most of the CTF used as photocatalysts in the hydrogen evolution reaction (HER) or carbon dioxide reduction reaction (CO2RR) are solid powder suspensions, but for their use in photoelectrocatalytic devices, films are a much better solution. In this sense, the material prepared in this manuscript is an excellent step forward in this direction. The synthetic organic approach to obtain the CTF films ground-breaking and it will open many new opportunities to develop this field.
In general, the characterization of the material is thoroughly performed, and the results obtained confirm that the materials were obtained without doubts. There are some minor aspects in the characterization that I think that should be improved, but I will describe them at the end of this report. In any case, they do not take away any merit to the manuscript. The experimental details for the preparation of the materials and the characterization seems to be enough detailed to reproduce the preparation of the films.

Response:
We thank the reviewer for understanding and appreciating significance of this manuscript. We also appreciate the reviewers' comments and suggestions that helped us to significantly improve the manuscript.
My main concern is related to the photocatalytic tests in the HER reaction. The experiments are acceptable, but I think these comments should be addressed:  2) After some long-term cycles, I would perform the photocatalytic test with a fresh TEOA solution (without co-catalyst): if Pt NPs were formed, the material should still be active.

Response:
We first did long-term cycles (100 h); after this cycle, fresh TEOA solution (without co-catalyst) was added, the materials were still active and showed stable HER performance for another 50 h, which confirmed the formation of Pt NPs. Response: This is a good suggestion. We first did long-term photocatalytic experiment (100 h) as shown in figure 6b, the performance was very stable for 60 h, decreases after 60 h, but still active under reaction conditions till 100 h. And after adding some fresh TEOA into the reactor, and purging the reactor with N2, then another 50 h photocatalytic experiment was carried out. It was still active and stable for 50 more hours; thus the film is active under reaction conditions as long as 150 h. Response: We did a single 5-hour experiment without the addition of the Pt co-catalyst and the catalytic performance is shown in Fig. 6a. The performance without Pt is significantly lower than that of the film catalysts with Pt, and the HER rate of the film catalysts with Pt is about 32 times higher than that of the catalysts without Pt in the first 5 h. The overpotential was about 120 mV vs RHE (Fig. 6c), and the Faradaic efficiency was about 92 % when using a constant voltage (-0.35V vs. RHE), which confirmed the good selectivity for H2 evolution (Supplementary Table 7). However, the current density was low because of the low Pt loading (2.0 wt % by ICP).  All in all, I think that this manuscript should be accepted for publication if these comments are addressed (or refuted).

Response:
We thank the reviewer for understanding and appreciating the significance of this manuscript again. We also appreciate the reviewers' comments and suggestions that helped us to significantly improve the manuscript.
Reviewer #3: In this paper, Tan et al. developed a smart synthetic strategy to first create a liquid/air interface, and fabricate successfully free-standing, large area, crystalline Covalent triazine frameworks (CTFs) films by this kind of interface. Furthermore, they explored the application of the film as immobilized photocatalysts and showed that the fabricated CTFs film had an excellent performance of photocatalytic hydrogen evolution. From a practical point of view, immobilized CTFs are more suitable for large-scale water splitting applications, however, existing synthetic approaches mainly result in insoluble and non-processable powders, which makes their future device application still a huge challenge. It is indeed highly interesting in this work that an interface is created without a water phase, which provides a rational approach to prepare immobilized photocatalysts, which is a significant landmark for the development of CTF and COF films and their potential integration in future devices for applications in fields as photoelectric, sensing, separation, and energy technologies. After careful evaluation, I would like to recommend this paper for publication in Nature Communications after the authors address the following important points: Response: We thank the reviewer for understanding and appreciating the significance of this manuscript.   Figure 3).
We have added the corresponding discussion as follows: Because of the weak polarity of the long carbon chain, imine precursor float at the surface of DMSO layer to generate an interface (Supplementary Fig. 1). The structure of precursor was confirmed by 1 H-NMR (Supplementary Fig. 2) and Fourier transform infrared (FT-IR) spectroscopy (Supplementary Fig. 3). Peak around 7.8 ppm can be assigned to C-H of benzene ring; peak around 8.3 ppm to C-H of imine bond and the peaks from 3.7 ppm to 0.9 ppm can be assigned to C-H of aliphatic chain ( Supplementary Fig. 2). In FT-IR spectrum, peak of the amino group (3300 cm -1 ) in hexylamine was disappeared, and the appearance of a new peak at 1645 cm -1 confirmed the formation of imine bond (Supplementary Fig. 3).  3. The XPS data (figure 2d) should be clarified further, the peak at 403 eV is not fitted in figure 2d, the authors should give more explanation.
Response: Thanks for your good suggestion. We have re-fitted the XPS data, and we added explanation in the revised manuscript. 4. Did the authors do some special treatment of the films and glass support in the photocatalytic experiments, such as using some adhesive? The decrease of HER performance may be caused by the peel-off of film from the support.

Response:
We did not do any special treatment of the films and glass support in the photocatalytic experiments. The films were directly loaded on the glass for HER experiments. We also proposed that the decrease of performance in the four cycles may be caused by the peeling-off of film, which would make weight loss and active area loss. And in a long photocatalytic experiment, the performance is stable for 60 h.