Effect of Platinum Precursor on the Properties of Pt/N-Graphene Catalysts in Formic Acid Decomposition
Round 1
Reviewer 1 Report
The paper “Effect of precursor on the properties of Pt/N-graphene catalysts 2 in formic acid decomposition” presents the comparative study of Pt/N-graphene catalysts prepared from two different Pt precursors. The study is very well done with comprehensive accurate characterization of catalysts. The work may be published, though some textual changes might improve the perception of the presentation by the readers.
As a general comment: it might help a lot if the paper was restructured to make a direct comparison between one and the other set of the samples, rather than a separate description of the first one, than the second. Also, it would help if the labeling of the samples reflected the preparation method: now one needs to scroll back to the chapter title to find out which sample “0.4% Pt/N-graphene” refers to.
English would need a bit of polishing.
Point by point comments to the text are presented below:
L2 – The title is a bit uncertain as it is not clear a “precursor” for what is meant: for Pt, for the reaction, for N-carbon synthesis. I would recommend modifying it yet keeping short as it is.
L21-44 - The motivation of the process of formic acid decomposition as a solution for the Green Deal sounds unconvincing as the reaction produces 94.5wt% of CO2, which does not sound green for me.
L48-50 – I would avoid mentioning carbon nanotubes in this context as there is a general awareness of their toxicity and cancerogenity in EU. Especially that nanotubes are not studied in this work.
L57 – citation is necessary for size dependent activity of Pt/C.
L108 – “FA” is used without previous definition and is used 2 times further in the text, while “formic acid” is used before and after L108 59 times. Authors should be consistent in using or not the abbreviation.
L124-128 Authors do not mention STEM imaging, which is mostly used in this work, as well as the sample preparation. The later may be crucial as the samples are active in organics (solvents) decomposition and are also sensitive to the atmosphere as follows from XPS data. Authors should at least comment on this.
L174,177 – Fig.2b is a STEM image, not TEM.
L178-179, 184-185 – Authors should provide a proof that there is no atomically dispersed Pt.
L180 Table1 – As soon as authors compare the sizes obtained by CO adsorption and STEM, the later data should be included in the Table 1, accounting for a proper weighted average. % of CO chemisorption should be recalculated in nm for comparison.
L187-190 – EDX map on 3b is not convincing at all and reproduces rather a detector noise than elements distribution. It can be safely avoided as it does not influence discussion and conclusions.
L191 – markers “3a”, “3b” look inconsistent with the style of the rest of the figures.
L194-198 – Fig4 is a STEM image, not TEM
L198 – it is incorrect comparing the size of the atom image to its atomic radius, the microscope is imaging a point object (charge of the nucleus) and the size of the image is solely reflecting the point spread function of the microscope not the “size” of the atom.
L203-210 – the same as for Table1 – once the comparison is done % should be recalculated to nm and corresponding STEM data should be exposed.
L207-208 – authors compare 1% chemisorption data to TEM but do not show corresponding images. IMHO Fig 4 can be extended demonstrating atomic species for all Pt contents.
L248-250 – it is not quite clear how the conclusion of stabilization of Pt on N sites follows from Fig7. Moreover, as follows from Fig5 and the literature, nitrogen substituted atoms in graphene are positively charged (oxidized) – authors should discuss, how oxidized nitrogen can stabilize oxidized/ionic state of Pt atoms.
L257-268 – the discussion around RDF curves is a bit misleading. Authors describe the curves as “Single intense peaks with almost the same intensities located in the range of ~1.1-2.5 Å are observed for all the Pt/N-graphene samples.” However, there are singular peaks at the same ~1.8Å position, the reason for indication of the very broad range is not clear. The plots Y-axis are not calibrated in coordination numbers, while authors do discuss this. In comparison to Pt-N bond length found in the literature (in the range of 2.1-2.2Å) the deviation of observed values is quite significant (0.3-0.4Å), observed value is much closer to Pt-C bond (1.95-1.98Å) – authors should discuss the reasoning for the assignment they make.
L271 – “does have” should probably be “does not have”
L269-272 – in order to make the conclusion about unambiguous proof authors should compare the curves on Fig8 to corresponding curves from the set of samples obtained from Pt(NO3)4, where Pt-Pt bond distances should be observed.
L276-278 – this claim is stretched too much, there is no evidence in the provided data for such conclusion.
Author Response
As a general comment: it might help a lot if the paper was restructured to make a direct comparison between one and the other set of the samples, rather than a separate description of the first one, than the second. Also, it would help if the labeling of the samples reflected the preparation method: now one needs to scroll back to the chapter title to find out which sample “0.4% Pt/N-graphene” refers to.
Answer
According to the reviewer’s comment we modified the name of the samples to include their preparation method, i.e. “0.4% Pt/N-graphene” (H2PtCl6) or “0.4% Pt/N-graphene” (Pt(NO3)4. We hope that this modification will help to understand the text of the manuscript.
English would need a bit of polishing.
Answer
The authors asked additional person proficient in English to proofread the text.
Answers to specific comments are reported below.
L2 – The title is a bit uncertain as it is not clear a “precursor” for what is meant: for Pt, for the reaction, for N-carbon synthesis. I would recommend modifying it yet keeping short as it is.
Answer
Based on this comment the authors modified the title of manuscript adding the word “platinum” to it: “Effect of platinum precursor on the properties of Pt/N-graphene catalysts in formic acid decomposition”
L21-44 - The motivation of the process of formic acid decomposition as a solution for the Green Deal sounds unconvincing as the reaction produces 94.5wt% of CO2, which does not sound green for me.
Answer
The authors made a major revision of Introduction to address this comment.
L48-50 – I would avoid mentioning carbon nanotubes in this context as there is a general awareness of their toxicity and cancerogenity in EU. Especially that nanotubes are not studied in this work.
Answer
The authors took into account this comment. Sentences where the use of carbon nanotubes as the formic acid decomposition catalysts was discussed were excluded from the text. However, one should understand that there are numerous publications devoted to this topic despite the fact that carbon nanotubes are known to be toxic and carcinogenic.
L57 – citation is necessary for size dependent activity of Pt/C.
Answer
A citation for the dependence of the Pt/C activity on the size of platinum nanoparticle was added.
L108 – “FA” is used without previous definition and is used 2 times further in the text, while “formic acid” is used before and after L108 59 times. Authors should be consistent in using or not the abbreviation.
Answer
The term “formic acid” was substituted for the “FA” abbreviation throughout the text.
L124-128 Authors do not mention STEM imaging, which is mostly used in this work, as well as the sample preparation. The later may be crucial as the samples are active in organics (solvents) decomposition and are also sensitive to the atmosphere as follows from XPS data. Authors should at least comment on this.
Answer
The method of ultrasonic spraying in alcohol or hexane is standard for the study of powder samples. We did not notice the effect of this method of powder application on the microstructure of the material. The necessary additions to the text were made.
L174,177 – Fig.2b is a STEM image, not TEM.
Answer
Thank you for the correction, we apologize for the inaccuracy.
L178-179, 184-185 – Authors should provide a proof that there is no atomically dispersed Pt.
Answer
Thank you for your good comment. One of the features of this material is its partial degradation under the beam, especially in STEM mode. We carried out measurements at high magnifications, but did not detect individual atoms. Unfortunately, the images were not saved. However, according to the reviewer’s request, we attach a picture taken in the TEM mode. The series of images do not show clusters or nanoparticles.
High concentration of platinum nanoparticles is observed on the N-graphene surface in the TEM images. In addition, in solution platinum is in the form of hydroxocomplexes. However, a minor fraction of platinum might be present in the atomically dispersed state as well.
L180 Table1 – As soon as authors compare the sizes obtained by CO adsorption and STEM, the later data should be included in the Table 1, accounting for a proper weighted average. % of CO chemisorption should be recalculated in nm for comparison.
Answer
The data on the average sizes of platinum particles obtained from the STEM data were added to Table 1. The data obtained from the CO chemisorption were converted to nm for comparison.
L187-190 – EDX map on 3b is not convincing at all and reproduces rather a detector noise than elements distribution. It can be safely avoided as it does not influence discussion and conclusions.
Answer
We fully agree with the reviewer. In this manuscript we tried to show the uniformity of the distribution of nitrogen over the carbon matrix. However, due to the degradation of the material under the beam, it was not possible to obtain a map of the distribution of chemical elements with sufficiently high quality. We replaced this map with other images illustrating the distribution of the elements.
L191 – markers “3a”, “3b” look inconsistent with the style of the rest of the figures.
Answer
Markers “3a”, “3b” were modified to look consistent with the style of the rest of the Figures.
L194-198 – Fig4 is a STEM image, not TEM
Answer
Thanks for the correction, we apologize for the inaccuracy.
L198 – it is incorrect comparing the size of the atom image to its atomic radius, the microscope is imaging a point object (charge of the nucleus) and the size of the image is solely reflecting the point spread function of the microscope not the “size” of the atom.
Answer
Thank you for the correction, we apologize for the inaccuracy. Appropriate correction have been made in the text.
L203-210 – the same as for Table1 – once the comparison is done % should be recalculated to nm and corresponding STEM data should be exposed.
Answer
The data on the average sizes of platinum particles obtained from the STEM data were added to Table 2 as well. The data obtained from the CO chemisorption were converted to nm for comparison.
L207-208 – authors compare 1% chemisorption data to TEM but do not show corresponding images. IMHO Fig 4 can be extended demonstrating atomic species for all Pt contents.
Answer
Thank you for your suggestion. An image with a 1% Pt sample was also added to the manuscript.
L248-250 – it is not quite clear how the conclusion of stabilization of Pt on N sites follows from Fig7. Moreover, as follows from Fig5 and the literature, nitrogen substituted atoms in graphene are positively charged (oxidized) – authors should discuss, how oxidized nitrogen can stabilize oxidized/ionic state of Pt atoms.
Answer
The formation of donor-acceptor bonds is typical for platinum and palladium. These bonds are essential for the formation of their complexes, e.g. [Pt(NH3)4]Cl2 and [Pd(NH3)4]Cl2. A nitrogen atom has an electron pair whereas a platinum atom has an empty d orbital. When a donor-acceptor bond is formed, a nitrogen atom donates an electron pair acting as a donor. Meanwhile, a platinum (or palladium) atom provides an empty orbital acting as an acceptor.
The nature of the chemical bond formed between palladium atoms and N species in NCNTs is discussed in more detail in the following paper:
Arrigo, R.; Schuster, M.E.; Xie, Z.; Yi, Y.; Wowsnick, G.; Sun, L.L.; Hermann, K.E.; Friedrich, M.; Kast, P.; Hävecker, M.; Knop-Gericke, A.; Schlögl, R. Nature of the N–Pd Interaction in Nitrogen-Doped Carbon Nanotube Catalysts, ACS Catal. 2015, 5, 2740. https://doi.org/10.1021/acscatal.5b00094
“The nature of the interaction of those N species with a Pd metal center immobilized onto NCNTs is of σ-type donation from the filled π-orbital of the N atom to the empty d-orbital of the Pd atom and a π back-donation from the filled Pd atomic d-orbital to the π* antibonding orbital of the N atom. We have found that the interaction of pyridine N with Pd is characterized by a charge transfer typical of a covalent chemical bond with partial ionic character, consistent with the chemical shift observed in the Pd 3d core level of divalent Pd”.
L257-268 – the discussion around RDF curves is a bit misleading. Authors describe the curves as “Single intense peaks with almost the same intensities located in the range of ~1.1-2.5 Å are observed for all the Pt/N-graphene samples.” However, there are singular peaks at the same ~1.8Å position, the reason for indication of the very broad range is not clear. The plots Y-axis are not calibrated in coordination numbers, while authors do discuss this. In comparison to Pt-N bond length found in the literature (in the range of 2.1-2.2Å) the deviation of observed values is quite significant (0.3-0.4Å), observed value is much closer to Pt-C bond (1.95-1.98Å) – authors should discuss the reasoning for the assignment they make.
Answer
The authors totally agree with the reviewer’s comment. These data were poorly presented. So, the discussion of the RDF curves was subjected to a major revision.
L271 – “does have” should probably be “does not have”
Answer
Yes, the mistake was corrected.
L269-272 – in order to make the conclusion about unambiguous proof authors should compare the curves on Fig8 to corresponding curves from the set of samples obtained from Pt(NO3)4, where Pt-Pt bond distances should be observed.
Answer
Data on platinum metal as a reference sample were added to the manuscript. An EXAFS spectrum obtained for Pt foil was added to Fig. 8. The distance attributed to the Pt-Pt bond can be clearly seen in this spectrum.
Unfortunately, the authors do not have such data for the series of samples prepared from Pt(NO3)4. Such data can be obtained only when the authors get beam time on synchronous irradiation in the future.
L276-278 – this claim is stretched too much, there is no evidence in the provided data for such conclusion.
Answer
Indeed, the presentation of the EXAFS data obtained on the Pt/N-graphene catalysts was poor. The text was substantially revised. The authors hope that after revision this evidence for this conclusion will be convincing.
Reviewer 2 Report
In this work, the authors intend to illustrate the effect of the precursor on the catalytic performance of Pt/N-graphene in formic acid decomposition. Two different precursor solutions Pt(NO3)4 and H2PtCl6 were listed, the different precursor concentrations were adjusted to generate different loadings of formic acid decomposition catalysts. However, one must point out that some experimental results were not believable, eg: the rough quality of the electron microscopy figure and the absence of C-N bond energy C1s XPS in Figure 5. Meanwhile, the mechanism for the precursor effect was not reliable, and the manuscript was not well organized with confusing logical relationships. In my opinion, this paper could NOT meet the journal requirements of Catalysts.
Author Response
Comments and Suggestions for Authors
In this work, the authors intend to illustrate the effect of the precursor on the catalytic performance of Pt/N-graphene in formic acid decomposition. Two different precursor solutions Pt(NO3)4 and H2PtCl6 were listed, the different precursor concentrations were adjusted to generate different loadings of formic acid decomposition catalysts. However, one must point out that some experimental results were not believable, eg: the rough quality of the electron microscopy figure and the absence of C-N bond energy C1s XPS in Figure 5. Meanwhile, the mechanism for the precursor effect was not reliable, and the manuscript was not well organized with confusing logical relationships. In my opinion, this paper could NOT meet the journal requirements of Catalysts.
Answer
Thank you for your comments on our manuscript. The authors improved the quality of the microscopy images. The electron microscopy images were replaced by images of better quality.
According to the reviewer’s comment the authors performed deconvolution of the C1s XPS spectrum in Figure 5. The deconvolution demonstrates the presence of the component attributed to the C-N bond in the spectrum.
The text of the manuscript was substantially revised.
Reviewer 3 Report
The authors must show studies of conversion vs time in order to study deactivation. Conversion vs temperature it is not enough to study this.
Authors mentioned that they have studied deactivation with flow system. How did they perform that? They have to show in detail the experimental set-up and the reactor system and show explicitly how they performed deactivation studies.
I do not understand why the authors show the conversion vs temperature for each degree (celsius) ?
The authors must be clear with the novelty of their work. Also they must explain their findings in more detail. Why the selectivity decreases with temperature.
Author Response
Comments and Suggestions for Authors
The authors must show studies of conversion vs time in order to study deactivation. Conversion vs temperature it is not enough to study this.
Answer
To address this comment, the authors performed additional studies of the catalyst stability for Pt/N-graphene catalysts. Experiments where the formic acid conversion was measured as a function of time were performed at 150 °C for 5 h. The obtained results are reported in Fig. 14. One can see that the catalytic activity did not change for 5 h. The data on the catalytic activity observed in these experiments are in a good agreement with those obtained in the temperature-programmed heating mode.
An additional axis showing reaction time was added in Figures 10-13 where the dependence of conversion on temperature is reported.
Authors mentioned that they have studied deactivation with flow system. How did they perform that? They have to show in detail the experimental set-up and the reactor system and show explicitly how they performed deactivation studies.
Answer
The authors did not have a goal to study the catalyst deactivation and, therefore, did not report it. The catalysts were stable for at least 5 hours. The temperature dependence of the selectivity is not related to the catalyst deactivation.
I do not understand why the authors show the conversion vs temperature for each degree (celsius)?
Answer
The authors do not show the conversion vs temperature for each degree (celsius). A conventional technique was used in the study where conversion was measured as function of temperature using temperature-programmed heating. In our installation conversion was automatically measured approximately every 5 degrees. There are numerous publications where a similar technique was used.
The authors must be clear with the novelty of their work. Also they must explain their findings in more detail. Why the selectivity decreases with temperature.
Answer
It is shown in the manuscript that the platinum precursor has a significant effect on the properties of obtained catalysts. The novelty of this study is related to development of a method for synthesis of catalysts on a nitrogen-containing carbon support using H2PtCl6 as the precursor. This method made it possible to synthesize atomically dispersed platinum catalysts with high selectivity in a wide range of platinum concentrations. The investigation of their catalytic properties demonstrated that atomically dispersed platinum has high activity, selectivity and stability in the formic acid decomposition to hydrogen and CO2.
The observed change of the selectivity with the temperature increase is related to changes in the contributions of the metal and the carbon support in the formic acid decomposition. At low temperatures (100-200 °C) formic acid dehydrogenation takes place on atomically dispersed platinum with high selectivity whereas no reaction occurs on N-graphene. However at temperatures above 200 °C N-graphene starts to contribute to the formic acid decomposition. The data presented in Fig. 13 demonstrate that N-graphene has lower selectivity in the formic acid dehydrogenation than platinum. As a result, the selectivity to hydrogen and CO2 goes down, although this decrease is entirely reversible.
Reviewer 4 Report
The Manuscript Numbered “Catalysts-1881393 " and titled “Effect of precursor on the properties of Pt/N-graphene catalysts in formic acid decomposition" focuses on the effect of platinum precursor on the synthesis of a novel catalytic system of Pt/N-graphene for the gas- phase formic acid decomposition to pure hydrogen. However, the manuscript suffers from some critical issues and can be considered for publication after correcting the typological errors and major corrections.
Comments to the authors:
- The authors should provide the details for the N-graphene synthesis.
- Replace palladium in line 121 to be platinum.
- The title of section 3.1. (Line 164) should be TEM study of N-graphene, the title of section 3.2. (Line 171) will be TEM study of Pt/N-graphene catalysts prepared using Pt(NO3)4 and the title of section 3.3. (Line 158) will be TEM study of Pt/N-graphene catalysts prepared using H2PtCl6.
- Why the authors did not use the same wt% of Pt in the two precursor samples?
- Mention the definition of the abbreviation (HAADF).
- Improve the discussion of section 3.3.
- Merge Fig.3 and Fig.4 to be in one Fig.
- The accurate method to calculate the Pt particle size is TEM. How you calculate the average particle size?
- Give a reference for the paragraph (Line 234-236).
- The name of samples in Fig.7 should be 0.4 % Pt/N-graphene, not 0.4% Pt/C.
- Correct the name of catalysts in Table 3 to be 0.4% instead of 0,4%.
- The authors should provide a comparison table of using different catalysts in formic acid decomposition to Hydrogen from the literature.
- Correct the typological and Grammarly errors.
Author Response
Comments to the authors:
- The authors should provide the details for the N-graphene synthesis.
Answer
A more detailed description of the N-graphene synthesis procedure was added to the text.
- Replace palladium in line 121 to be platinum.
Answer
The mistake was corrected.
- The title of section 3.1. (Line 164) should be TEM study of N-graphene, the title of section 3.2. (Line 171) will be TEM study of Pt/N-graphene catalysts prepared using Pt(NO3)4 and the title of section 3.3. (Line 158) will be TEM study of Pt/N-graphene catalysts prepared using H2PtCl6.
Answer
Thank you for the comments. The corresponding changes were made.
- Why the authors did not use the same wt% of Pt in the two precursor samples?
Answer
The authors used almost the same Pt concentrations in the two precursor samples
- Mention the definition of the abbreviation (HAADF).
Answer
Thank you for your comments on our manuscript. The corresponding changes were made to the text of the manuscript.
HAADF - High-angle Annular Dark-field
- Improve the discussion of section 3.3.
Answer
The authors tried to improve discussion both in section 3.3 and throughout the rest of the manuscript.
- Merge Fig.3 and Fig.4 to be in one Fig.
Answer
Thank you for your helpful comment. We expanded and changed the set of images. We hope that the current position of the images will be convenient for reviewers and readers.
- The accurate method to calculate the Pt particle size is TEM. How you calculate the average particle size?
Answer
Thank you for your comments on our manuscript. The particle size distribution was calculated using microscopy images using ImageJ software.
- Give a reference for the paragraph (Line 234-236)
Answer
The requested references [40,41] were added.
- The name of samples in Fig.7 should be 0.4 % Pt/N-graphene, not 0.4% Pt/C.
Answer
The name of the samples in Fig. 7 was changed according to this comment.
- Correct the name of catalysts in Table 3 to be 0.4% instead of 0,4%.
Answer
The mistake was corrected.
- The authors should provide a comparison table of using different catalysts in formic acid decomposition to Hydrogen from the literature.
Answer
The authors made a comparative analysis of different catalysts reported in the literature in the formic acid decomposition to hydrogen. For our 1 % Pt/N-graphene TOF is equal to 0.42 s-1. This value exceeds those reported for all known catalysts on carbon supports.
Round 2
Reviewer 2 Report
I suggest accepting the manuscript.
Reviewer 4 Report
No comments
