Double-shelled hollow rods assembled from nitrogen/sulfur-codoped carbon coated indium oxide nanoparticles as excellent photocatalysts

Excellent catalytic activity, high stability and easy recovery are three key elements for fabricating efficient photocatalysts, while developing a simple method to fabricate such photocatalysts with these three features at the same time is highly challenging. In this study, we successfully synthesized double-shelled hollow rods (DHR) assembled by nitrogen (N) and sulfur (S)-codoped carbon coated indium(III) oxide (In2O3) ultra-small nanoparticles (N,S-C/In2O3 DHR). N,S-C/In2O3 DHR exhibits remarkable photocatalytic activity, high stability and easy recovery for oxidative hydroxylation reaction of arylboronic acid substrates. The catalyst recovery and surface area were well balanced through improved light harvesting, contributed by concurrently enhancing the reflection on the outer porous shell and the diffraction in the inside double-shelled hollow structure, and increased separation rate of photogenerated carriers. Photocatalytic mechanism was investigated to identify the main reactive species in the catalytic reactions. The electron separation and transfer pathway via N,S-codoped graphite/In2O3 interface was revealed by theoretical calculations.

1) According to the gradient variance tendency shown in Fig. 3 by comparing the N,S-codoped In2O3 DHR, N,S-C/In2O3 NP, N-C/In2O3 NP, In2O3 DHR, In2O3 HD and commercial In2O3, it seems that the In2O3-based materials with higher crystal integrities exhibit higher electronic resistances, under the ground or excited states. Also, based upon the statement of photocatalytic mechanism, the oxygen of In2O3 donated an electron to O2 upon photo-irradiation. Thus, it was deduced that, the defects of In2O3, especially that of oxygen sites in In2O3, plays an important role in determining the photoelectronic performances. It was suggested to check the high-resolution XPS spectrum of those In2O3-based materials and compare the different compositions of O1s, for the rationalization of structure-activity relationship.
2) In the photocatalytic section, the author stated that the N,N-diisopropylethylamine was a "cocatalyst", but this is not correct, it should be the hydrogen and electron donor or the so called sacrifice agent.
Reviewer #3 (Remarks to the Author): The manuscript by Sun et. al reports a method to obtain N,S-C/In2O3 photocatalysts with the morphology of double-shelled hollow rods. In general, this is a well-organized paper containing interesting results and the characterizations of the materials are sufficient. The key of this method is adding BIT as the modulator and the N, S sources while synthesis In-based MOF. This seems a very simple and effective method for uniformly doping heteroatoms to MIL-68-In. However, the fabrications of MOF-derived catalytic materials have been well know. It also seems difficult to extend the reported strategy to other systems, raising questions on its universality. I did not find any conceptual breakthrough that would greatly advance the field. I was not convinced this work is of broad interest to the general research community or its meets the high standards of Nature Communications in terms of novelty, significance and impact. Overall, this is an interesting piece of work but a more specialized journal is suggested. A few comments are provided below.
1. The authors might want to emphasize the novelty and significance of the work. Particularly, how good is the catalytic performance by comparing with reported system? 2. The authors mentioned that their catalysts could easily be recovered. However, this means the catalysts have very poor colloidal dispersity. Undoubtably, this should not be called as an advantage! If the catalysts cannot be well dispersed, how good can their catalytic activity be?
3. I was trying to find the challenge that the current work aimed to address. In the abstract, it was claimed "Excellent catalytic activity, high stability and easy recovery are three key elements for fabricating an efficient photocatalyst". However, this claim is too vague for judging the importance and difficulty of the challenge. What is the exact challenge that is addressed in this work? Why is it important? Why is still not addressed in previous studies? 5. Some information is lack of consistency. e.g. "Hence, annealing N,S-codoped MIL-68-In at 500 oC in a vacuum atmosphere with a ramping rate of 2 °C min−1 was sufficient to ensure complete conversion from MIL-68-In to In2O3." (line 109-111 in the main text) and "N,S-C/In2O3 DHR was synthesized via the calcination of the obtained N,S-codoped MIL-68-In at 550 °C in vacuum atmosphere with heating rate of 5 °C min-1 for 45 min." (line 13-14 in the supporting information) What are the exact calcination temperature and heating rate?

Response to Reviewer #1's Comments:
Comments to the Author: In this manuscript, the authors reported a synthetic method of double-shelled hollow rods assembled by N,S-codoped carbon coated In 2 O 3 ultrafine nanoparticles (N,S-C/In 2 O 3 DHR). The characterization (XRD, SEM, TEM, mapping and XPS) of the prepared NCs was performed clearly, which provided significant scientific insights. A serial of contrast experiments provided a thorough investigation of photoelectric properties of N,S-C/In 2 O 3 DHR. By contrast with different catalysts, N,S-C/In 2 O 3 DHR exhibited excellent photocatalytic activity toward photocatalytic oxidative hydroxylation of various arylboronic acid substrates. Moreover, authors investigated the photocatalytic mechanism by using UV-vis absorption spectra, ESR spectra and theoretical calculation. As we known, it is challenging to synthesize a photocatalyst combining with the functions of high catalytic activity, high stability and easy recovery. In this article, authors achieved the above goals by designing the structure and composition of catalyst. Therefore, it is an interesting and innovative work in the field of the controllable syntheses of metallic oxide photocatalysts. The manuscript seems also well-organized. Therefore, I am glad to recommend the publication of the manuscript in Nature Communications after the revision of following minor issues: Response: We really appreciate the reviewer's useful comments and positive recommendation of publication after revisions. We have carefully revised the manuscript based on the comments.
1. In the photocatalytic oxidative hydroxylation of arylboronic acids, authors used blue LED light irradiation, how about the yield when irradiated with green and yellow LED light? Response: Thanks for the interesting suggestion. According to the advice, the photocatalytic activity of N,S-C/In 2 O 3 DHR toward oxidative hydroxylation of arylboronic acids was evaluated by using green and yellow LED light. On account of weak absorption of N,S-C/In 2 O 3 DHR to green and yellow light, the yield was relatively lower, i.e., 44% under green light and 15% under yellow light for 12 h irradiation ( Figure R1). In order to have more detailed descriptions about the catalytic activity of N,S-C/In 2 O 3 DHR, Figure R1 had been added in the revised supporting file as Figure S15, and corresponding descriptions have been added in the revised manuscript (page 13). 2. Arylboronic acids substrate scope should be further expanded. Response: Thanks for the kind reminder. According to the suggestion, we have further expanded the substrate scope, and corresponding experimental results ( Figure R2) had been added as Figure 4d in the revised manuscript. 4. In the photocatalytic experiment, 1 H NMR was employed to determine the yield. The 1 H NMR spectra of reactant and product should be provided. Response: Thanks for the kind reminder. According to the suggestion, 1 H NMR spectra ( Figure R3) of both reactant and product have been added in the revised supporting file as Figure S14. 5. XPS spectra determined the chemical states of each element in the material. Therefore, the positions of each sub-peaks indicated by XPS spectra should be shown more clearly in other ways. Response: Thanks for the kind suggestion. In order to more clearly show the positions of each sub-peak, we have added corresponding color to each sub-peak in the revised spectra ( Figure  2). 6. In order to measure the recovery of N,S-C/In 2 O 3 DHR, authors compared the precipitation of N,S-C/In 2 O 3 DHR and N,S-C/In 2 O 3 NP after keeping their solutions for 12h, and the corresponding results were shown in Fig. S14. But the sample in Fig. S14 was white, which was not the color for C coated sample. Authors should explain this phenomenon. Response: Thanks for the kind notice. We are sorry that we have described samples in original Figure S14 by the mistake. In fact, original Figure S14 shows the precipitation of N,S-C/In 2 O 3 DHR and commercial In 2 O 3 . The color of commercial In 2 O 3 is white. Original Figure S14 has been changed to Figure S17 ( Figure R4). The legend of Figure S17 and corresponding descriptions have been corrected in the revised manuscript (page 15). 7. The VB potential of In 2 O 3 was calculated to be 2.09 eV. Authors should confirmed this value by VB XPS. Response: Thanks for the kind suggestion. The valence band X-ray photoelectron spectroscopy (VB XPS) of pure In 2 O 3 is shown in Figure R5. It can be seen that the position of the valence band edge of In 2 O 3 is located at about 2.15 eV, which is consistent with the conclusion calculated from the Mulliken electronegativity theory. Figure R5 has been added in the revised supporting file as Figure S20, and corresponding descriptions have been added in the revised manuscript (page 16).

Response to Reviewer #2's Comments:
Comments to the Author: Han, Zhao and coworkers developed the double-shelled hollow rods assembled from N,S-codoped carbon coated In 2 O 3 nanoparticles (N,S-C/In 2 O 3 DHR) and their application as photocatalysts, which is of fundamental importance in photocatalysis and will abstract broad interests. The photoelectronic performances of N,S-C/In 2 O 3 DHR are well characterized. Thus, we recommended to publish this manuscript in Nature Communications after minor revisions. The comments and arguments are listed as shown below: Response: We really appreciate the reviewer's useful comments and positive recommendation of publication after minor revisions. We have carefully revised the manuscript based on the comments.
1) According to the gradient variance tendency shown in Fig. 3 by comparing the N,S-codoped In 2 O 3 DHR, N,S-C/In 2 O 3 NP, N-C/In 2 O 3 NP, In 2 O 3 DHR, In 2 O 3 HD and commercial In 2 O 3 , it seems that the In 2 O 3 -based materials with higher crystal integrities exhibit higher electronic resistances, under the ground or excited states. Also, based upon the statement of photocatalytic mechanism, the oxygen of In 2 O 3 donated an electron to O 2 upon photo-irradiation. Thus, it was deduced that, the defects of In 2 O 3 , especially that of oxygen sites in In 2 O 3 , plays an important role in determining the photoelectronic performances. It was suggested to check the high-resolution XPS spectrum of those In 2 O 3 -based materials and compare the different compositions of O1s, for the rationalization of structure-activity relationship. Response: Thanks very much for the useful suggestion. According to the suggestion, we have checked the high-resolution O 1s spectra of these In 2 O 3 -based materials. As shown in Figure  R6, high resolution O1s XPS peaks of these samples had some slight differences. The O1s core level peaks could be resolved into three centered Gaussian components (Appl. Surf. Sci. As shown in Table R1, the relative proportion of the O V component in different samples is N,S-C/In 2 O 3 DHR > N,S-C/In 2 O 3 NP > N-C/In 2 O 3 NP > In 2 O 3 DHR > In 2 O 3 HD > commercial In 2 O 3 , which is well consistent with the photocatalytic activity. The results indicated that the surface structure (especially oxygen deficiency sites of materials) was indeed an important factor affecting the catalytic performance of corresponding materials. This is also a very important topic to understand how surface oxygen defects could affect photocatalytic performance. Figure R6 and Table R1 have been added in the revised supporting file as Figure S26 and 2) In the photocatalytic section, the author stated that the N,N-diisopropylethylamine was a "co-catalyst", but this is not correct, it should be the hydrogen and electron donor or the so called sacrifice agent. Response: Thanks for the kind reminder. The description about "co-catalyst" has been changed to "sacrifice agent" in the revised manuscript (page 13).

Response to Reviewer #3's Comments:
Comments to the Author: The manuscript by Sun et. al reports a method to obtain N,S-C/In 2 O 3 photocatalysts with the morphology of double-shelled hollow rods. In general, this is a well-organized paper containing interesting results and the characterizations of the materials are sufficient. The key of this method is adding BIT as the modulator and the N, S sources while synthesis In-based MOF. This seems a very simple and effective method for uniformly doping heteroatoms to MIL-68-In. However, the fabrications of MOF-derived catalytic materials have been well know. It also seems difficult to extend the reported strategy to other systems, raising questions on its universality. I did not find any conceptual breakthrough that would greatly advance the field. I was not convinced this work is of broad interest to the general research community or its meets the high standards of Nature Communications in terms of novelty, significance and impact. Overall, this is an interesting piece of work but a more specialized journal is suggested. A few comments are provided below. Response: Thanks for reviewing our manuscript. We really appreciate the reviewer's comments that the main point of this article is to develop a simple and effective method for uniformly doping heteroatoms to MIL-68-In by adding 1,2-benzisothiazolin-3-one (BIT) as the modulator and N,S sources. The novel double-shelled hollow rods were then assembled from N,S-codoped carbon coated In 2 O 3 nanoparticles produced by calcining the N,S-codoped MIL-68-In. Although the fabrications of MOF-derived catalytic materials have been reported, using MOFs for the preparation of heteroatom-codoped carbon coated metal oxide hollow rods as excellent photocatalysts has not been well studied.
[Redacted] Excellent photocatalysts would have three of following qualities at the same time: the enhanced optical absorption abilities to increase the number of photo-generated charge carriers, the improved separation of photo-generated charge carriers to prolong the lifetime of carriers, and the large surface area to provide more reactive sites for photocatalytic reactions. One effective approach to achieve these qualities is to prepare metal oxide nanoparticles coated by a carbon layer as efficient photocatalysts. On account of excellent mobility of charge carriers, the carbon layer can act as photo-generated-electron acceptor to improve the separation of photo-generated electrons and holes. Increasing the specific surface area of photocatalysts by decreasing their sizes could provide more reactive sites. There are two bottlenecks that restrict the application of this approach in fabricating efficient photocatalysts.
(1) It is difficult to control the coating of carbon layer and introduce the heteroatoms. In general, enough and intimate contact interface is the key factor ensuring the efficient transfer of photo-generated carriers. While metal oxide nanoparticles are generally loaded on carbon materials, only a small fraction of the nanoparticle surface is in direct contact with the carbon materials. The core-shell structures could provide a three-dimensional (3D) intimate contact and maximize the interface between carbon layers and metal oxide nanoparticles. But, the uniformly coating carbon layer on metal oxide nanoparticles is still a challenging task. Recent studies reported that heteroatom-doped carbon could further improve the conductivity of carbon materials (Adv. Mater. 2015, 27, 6021-6028 Mater. 2016, 26, 5708-5717; Carbon 2017, 118, 511-516). Realizing the co-doping of double heteroatoms is even harder. Therefore, it is desirable but challenging to fabricate the ultrafine metal oxide nanoparticles coated by double heteroatoms along with co-doped carbon layers. (2) It is also difficult to recycle the ultrafine photocatalyst nanoparticles from solution. Though the ultrafine nanoparticles display excellent photocatalytic performance, the difficulty in recycling ultrafine nanoparticles may increase the cost of photocatalytic reactions and result in secondary pollution of water. While magnetic photocatalysts may facilitate the recycling of photocatalysts to some extent, the reaction conditions of magnetic photocatalysts might be limited, because the dispersion of magnetic photocatalysts in solution can only use mechanical stirring and cannot use magnetic stirring. Therefore, a rational design of ultrafine nanoparticles with high reactivity and easy recyclability is highly desirable for extending their photocatalytic application.
To date, only a few studies could overcome abovementioned two bottlenecks using one photocatalytic material system. In our study, we developed a simple and efficient method for fabricating the double-shelled hollow rods assembled by N,S-codoped carbon coated In 2 O 3 ultrafine nanoparticles (N,S-C/In 2 O 3 DHR) using MIL-68-In MOF as the template and BIT as the modulator in one step. The obtained N,S-C/In 2 O 3 DHR possesses both uniform N,S-codoped carbon layers and easy recyclability from solution through simple centrifugation. In addition, the double-shelled hollow structure of N,S-C/In 2 O 3 DHR not only improves light harvesting by simultaneously increasing the reflection on the outer shell and the diffraction on the hollow cavity to generate more electrons and holes, but also provides more active sites for photocatalytic reactions. Therefore, we have achieved very high photocatalytic performance by using N,S-C/In 2 O 3 DHR. The present research puts forward a useful protocol for fabricating efficient photocatalysts, exhibiting its novelty and significance to warrant the publication in Nature Communications. [Redacted] 1. The authors might want to emphasize the novelty and significance of the work. Particularly, how good is the catalytic performance by comparing with reported system? Response: Thanks for the suggestion. We have addressed the novelty and significance of our work in the response to the general comments above. Here, we would like to further discuss about it. Light absorption efficacy is still the primary factor limiting the practical applications of photocatalysis. Therefore, improving the sensitivity of photocatalysts to light is the key to the practical applications of photocatalysis. In literature, high-power Xe lamp and LED lamp were used as the photocatalytic light sources, such as 300W Xe lamp (Angew. Chem. Int. Ed.   2016, 55, 4697-4700; Appl. Catal. B 2019, 243, 10-18 Soc. 2016, 138, 12719-12722). In our work, we used 3W blue LED lamp as the photocatalytic light source, indicating that the as-prepared N,S-C/In 2 O 3 DHR possesses highly improved sensitivity to light. As shown in Figure R1, the as-prepared N,S-C/In 2 O 3 DHR also exhibited a certain degree of photocatalytic activities under green light (3W LED lamp, 525 nm) and yellow light (3W LED lamp, 590 nm). These results mean that the N,S-C/In 2 O 3 DHR broadens the response range of light to wide wavelength. Figure R1 had been added in the revised supporting file as Figure S15, and corresponding descriptions have been added in the revised manuscript (page 13).
In addition, the N,S-C/In 2 O 3 DHR could be easily recycled from the solution, due to the formation nature of assembling N,S-codoped carbon coated In 2 O 3 ultrafine nanoparticles (~10 nm) into double-shelled hollow rods (~7 μm). The excellent photocatalytic performance of N,S-C/In 2 O 3 DHR is benefited from its novel structure. (1) The hollow structure can improve light harvesting by simultaneously increasing the reflection on the outer shell and the diffraction on the hollow cavity. (2) The N,S-codoped carbon layers, uniformly distributed on the surface of In 2 O 3 nanoparticles, can guarantee highly efficient electron transfer and improve the separation of photo-generated electron-hole pairs. (3) The double shells, composed by N,S-codoped carbon coated In 2 O 3 ultrafine nanoparticles, can provide more reactive sites for the photocatalytic reactions. We then compared the photocatalytic activity of N,S-C/In 2 O 3 DHR with N,S-C/In 2 O 3 NP, N-C/In 2 O 3 NP, In 2 O 3 DHR, In 2 O 3 HD and commercial In 2 O 3 to reveal the influence of composition and structure on photocatalytic activity, showing that N,S-C/In 2 O 3 DHR is the best photocatalyst among them in terms of catalytic performance. We also expanded the substrate scope ( Figure R2) using N,S-C/In 2 O 3 DHR as the photocatalyst, once again presenting its high catalytic performance. Corresponding experimental results had been added as Figure 4d in the revised manuscript.  2. The authors mentioned that their catalysts could easily be recovered. However, this means the catalysts have very poor colloidal dispersity. Undoubtably, this should not be called as an advantage! If the catalysts cannot be well dispersed, how good can their catalytic activity be? Response: Thanks for the comments. The dispersity is indeed an important property of photocatalysts, which cannot be neglected. The N,S-C/In 2 O 3 DHR could be easily recycled from the solution, but this does not mean its poor dispersity. As shown in Figure R4, only a small fraction of N,S-C/In 2 O 3 DHR settled during the first 4 hours without stirring, indicating the deposition process of N,S-C/In 2 O 3 DHR was slow. On the other hand, during the photoreactions, the solutions with N,S-C/In 2 O 3 DHR were continuously stirred with dynamoelectric stirrers. As shown in Figure R10, the N,S-C/In 2 O 3 DHR well dispersed in solutions under continuous stirring. Moreover, easy recycling of N,S-C/In 2 O 2 DHR is due to its own gravity, which is different from the magnetic photocatalysts. Compared to the magnetic photocatalysts, the N,S-C/In 2 O 3 DHR has greater application prospects, since it is not restricted by stirring conditions (mechanical stirring or magnetic stirring). In order to avoid potential misunderstanding, Figures R4 and R10 have been added in the revised supporting file as Figures S17 and S18, respectively. The corresponding descriptions have also been added in the revised manuscript (page 15).  uniformly coated by a carbon layer as efficient photocatalysts is an effective solution in achieving the goal. The uniformly coated carbon layer can act as photo-generated-electron acceptor to improve the separation of photo-generated electrons and holes. Recent studies reported that heteroatom-doped carbon could further improve the conductivity of carbon materials, which is beneficial for improving the transfer of photo-generated electrons (Adv. Mater. 2015, 27, 6021-6028; Nano Energy 2016, 19, 373-381; J. Mater. Chem. A 2017, 5,  22964-22969). Nanoparticles with small sizes may provide more active sites for light absorption and reactions. As abovementioned, there are two difficulties in synthesizing such kinds of photocatalysts. (1) It is difficult to uniformly introduce heteroatoms into carbon layers, especially introducing double heteroatoms. (2) Recycling ultrafine photocatalyst nanoparticles from solution is also difficult. Since traditional techniques have unavoidable disadvantages limiting their practical applications, developing a simple method to fabricate efficient photocatalysts with excellent catalytic activity, high stability and easy recovery at the same time has been highly sought after. In this work, we successfully synthesized double-shelled hollow rods assembled by N,S-codoped carbon coated ultrafine In 2 O 3 nanoparticles (N,S-C/In 2 O 3 DHR) by using MIL-68-In as the template and BIT as the modulator in one step. The N,S-C/In 2 O 3 DHR possesses excellent photocatalytic activity, improved sensitivity to light (3W blue LED as light source), high stability (catalytic efficiency has not been reduced and the morphology and composition of catalysts have not changed after multiple cycles), and easy recovery (achieved by simple centrifugation) at the same time. In order to more clearly express the importance of this work, corresponding descriptions have been added into the abstract of the revised manuscript.
4. When different amounts of competitive or modulator ligand were involved, the morphology can be controlled (J. Am. Chem. Soc. 2011, 133, 15506-15513; Angew. Chem. 2009, 121, 4833 -4837). Would the shape change with different amounts of BIT? If not, the effect of doping amount should be performed. Response: Thanks for the useful suggestions. We agree with the reviewer's comments that the morphology of MOFs can be controlled when different amounts of competitive or modulator ligand were involved. According to the reviewer's suggestion, we changed the dosage of regulator (BIT) to control the morphology of N,S-codoped MIL-68-In. As shown in Figure  R11, after changing the amount of BIT, the obtained N,S-codoped MIL-68-In could maintain the rod morphology. When further increasing the amount of BIT, the length-width ratio of N,S-codoped MIL-68-In rods gradually decreased. Figure R10 has been added in the revised supporting file as Figure S1. The corresponding descriptions (page 5) and references (J. Am. Chem. Soc. 2011, 133, 15506-15513   5. Some information is lack of consistency. e.g. "Hence, annealing N,S-codoped MIL-68-In at 500 o C in a vacuum atmosphere with a ramping rate of 2 °C min −1 was sufficient to ensure complete conversion from MIL-68-In to In 2 O 3 ." (line 109-111 in the main text) and "N,S-C/In 2 O 3 DHR was synthesized via the calcination of the obtained N,S-codoped MIL-68-In at 550 °C in vacuum atmosphere with heating rate of 5 °C min-1 for 45 min." (line 13-14 in the supporting information) What are the exact calcination temperature and heating rate? Response: Thanks for the kind reminder. The exact calcination temperature, time and heating rate were 550 °C for 45 min with heating rate of 5 °C min -1 . We have corrected these information and unified corresponding descriptions in the revised manuscript (page 5 and supporting file page S1).
Response: Thanks a lot for the kind suggestion. In order to provide more convincing evidence, dynamic light scattering (DLS) has been used to characterize the dispersity of the catalysts. Figure R1 shows a plot with mean count rate as a function of different sedimentation time without (a) and with stirring (b). The count rate measured by DLS declines with the decrease of particle number in the solution [Mol. Pharmaceutics 2013, 10, 3392]. As shown in Figure  R1a, when the samples were not under stirring, the count rate data of commercial In 2 O 3 NP remained basically unchanged for 12 h. However, the data of N,S-C/In 2 O 3 DHR sharply decreased over 12 h. The results indicate that N,S-C/In 2 O 3 DHR can be easily recycled from the solution by its own gravity. When the samples were under continuous stirring for 12 h, the count rate data were almost unchanged ( Figure R1b), indicating that the N,S-C/In 2 O 3 DHR can be well dispersed in solutions under stirring. These data correspond well with the photos of Figures S17 and S18.
In order to more clearly show the dispersity of the catalyst, Figure R1a and R1b have been added in Figures S17 and S18, respectively. Corresponding experimental details have been added to the supporting information file (Page S3). Figure R1. Plots of mean count rate for different sedimentation time (detected by using dynamic light scattering). (a) Samples without agitation, and (b) samples under stirring. 6. The response to Q3 should also be reflected in the introduction. Response: Thanks a lot for the kind suggestion. The response of Q3 has been reflected in the abstract and introduction of the revised manuscript (Pages 1, 3, and 4). Specific ideas are as follows: The challenge of this work is to develop a simple and effective approach to fabricate efficient photocatalysts assembled from codoped carbon uniformly coated metal oxide nanoparticles with excellent catalytic activity and easy recovery at the same time. However, there are two difficulties in synthesizing such kinds of photocatalysts. (1) It is difficult to uniformly introduce heteroatoms into carbon layers, especially introducing double heteroatoms.
(2) It is difficult to balance the ease of recovery with the surface area. In this work, we have successfully synthesized double-shelled hollow rods assembled by N,S-codoped carbon coated ultrafine In 2 O 3 nanoparticles (N,S-C/In 2 O 3 DHR) by using MIL-68-In as the template and BIT as the modulator in one step. The N,S-C/In 2 O 3 DHR possesses excellent photocatalytic activity and easy recovery at the same time.