Catalytic Reduction of N₂O by CO on Single-Atom Catalysts Au/C₂N and Cu/C₂N: A First-Principles Study

Round 1

Reviewer 1 Report

Su and co-authors have studied by DFT modeling the catalytic reduction of dinitrogen oxide by carbon monoxide on single Cu and Au atoms embedded in C₂N monolayer.

It is found that the highest activation energies for the energetically more favorable mechanism are similar for the two catalysts, 1.21 eV and 1.04 eV for Cu/C₂N and Au/C₂N, respectively. Such barriers are still high and perhaps not so easy to be overcome under the operating conditions of the catalyst. Did the authors calculate the energy barrier for the N₂O + CO reaction in gas phase without participation of catalyst? The result could be compared with the obtained for the studied SAC.

Did the authors performed test calculations with dispersion correction (for example Grimme’s D2 or D3) to check how it will affect the results?  

 

Minor issues:

The color coding of atoms on Figures 1, 2 and 6 is missing. Although for most of the readers it is clear that N is blue, O is red and etc., it will be better to add this information.

 

Line 33: space between “approximately” and “300”

Line 278: remove the space between “Ernzerh of”

Author Response

Dear Reviewer,

 

Thank you very much for your valuable comments concerning our manuscript entitled “Catalytic reduction of N₂O by CO on single atom catalysts Au/C₂N and Cu/C₂N: A first-principles study” (ID: catalysts-2198490). The comments are really helpful for improving our manuscript, as well as the important guiding significance to our research. We have studied all the comments carefully and have made corrections. The main corrections in the paper and the response to the reviewer’s comments are as following by items:

 

1. Did the authors calculate the energy barrier for the N2O + CO reaction in gas phase without participation of catalyst? The result could be compared with the obtained for the studied SAC.

Response: Thanks for Reviewer’s comment. We did not calculate the gas reaction for N₂O + CO reaction, because Dmol3 is not so good at computating small molecule reactions. We gave the experimental activation energy as follows: Indeed, an experimental study done by Loirat et al. found that N₂O reduction by CO has an activation energy as high as 193±8 kJ/mol (about 2.0 eV) between 1076 and 1228 K [12], thus the reaction needs for efficient catalysts to lower the operating temperature. (Please see paragraph 2 in “Introduction” at page 1), and also compare the computational results with experimental one “It is worth noting that the calculated activation enthalpies are much lower than experimental values without any catalyst (about 2.0 eV) [12].” (Last paragraph in “Results and discussion” at page 11)

2. Did the authors performed test calculations with dispersion correction (for example Grimme’s D2 or D3) to check how it will affect the results? 

Response: Thanks for Reviewer’s comment. TS correction was used for dispersion correction in the work, but it was missed in the manuscript. This was added in the revision along with the reference We did not test effect of dispersion correction, because generally DFT is inadequate to describe weak interaction and dispersion correction is necessary for such systems. Previous report also proved that binding energies for the systems with weak interactions were overestimated by pure GGA functionals, and were dramatically improved by applying Grimme or TS dispersion correction to PBE.

“As the generalized gradient approximation (GGA) method is insufficient in describing weak interaction, dispersion correction was added to the system [45]. Transition state (TS) correction was also employed to evaluate dispersion interactions [46].”(Please see the first paragraph in “Computational methods” on page 2).

3. The color coding of atoms on Figures 1, 2 and 6 is missing. Although for most of the readers it is clear that N is blue, O is red and etc., it will be better to add this information.

Response: Thanks for Reviewer’s suggestion. The specification of balls’ color was added in Figure 1, 2, 8, S3 and S4 in revision.

4. Line 33: space between “approximately” and “300” .

Response: Thanks for Reviewer’s suggestion. The space was deleted between “approximately” and “300” (Line 8 in paragraph 1 in “Introduction”)

5. Line 278: remove the space between “Ernzerh of”.

Response: Thanks for Reviewer’s suggestion. It was corrected to “Ernzerhof” in revision (Please see “Computational methods” at Page 2).

 

 

Author Response File: Author Response.pdf

Reviewer 2 Report

See the attached file

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

 

Thank you very much for your valuable comments concerning our manuscript entitled “Catalytic reduction of N₂O by CO on single atom catalysts Au/C₂N and Cu/C₂N: A first-principles study” (ID: catalysts-2198490). The comments are really helpful for improving our manuscript, as well as the important guiding significance to our research. We have studied all the comments carefully and have made corrections. The main corrections in the paper and the response to the reviewer’s comments are as following by items:

 

In the paper ”Catalytic reduction of N2O by CO on single atom catalysts Au/C2N and Cu/C2N: A first-principles study study ”, the authors study, using Density Functional Theory calculations, two mechanisms for the reduction of N2O by CO on single atom catalyst (SAC) Au/C2N and Cu/C2N.

It aims to analyse both mechanisms, as well as the influence of the use of cheaper metals.

In the paper, both the approach and its development seem to me to be correct, considering the variation of the charges centred on the atoms, and the inter-nuclear distances, to analyse the different intermediates of the reaction. For this purpose, I consider the Mulliken analysis to be sufficient, given that there are large variations in the loads analysed.

1. Now, since they have calculated the vibrational frequencies, it would be interesting to analyse the enthalpies, not just the energy, which is what they seem to present in the paper.

Response: Thanks for Reviewer’s helpful suggestion. The energies, enthalpies along with free energies with zero-point vibrational corrections of all the species were added in the revision. Please see Table 1 at page 5. And energies were also changed to enthalpies in the discussion.

2. They should also correct figure 3, which, while trying to be qualitative, should not show positive values below negative ones.

Response: We are sorry for the careless error. The wrong positions of these bars were corrected in the revision. And according to the third comment, the enthalpies with zero point vibrational corrections were used in Figure 3 in the revision. Please see Figure 3 (page 5) and Figure 9 (page 11).

3. Globally, the theoretical study is correct and the obtained results are of interest for the scientific community, however, the paper could include considerations of Zero point vibrational energies (enthapies) and modify figure 3.

Response: Thanks for Reviewer’s helpful suggestion. The energies and enthalpies with zero point vibrational corrections were added in Table 1, Figure 3 and Figure 9 in the revision. Please see Table 1 at page 5, Figure 3 at page 5 and Figure 9 at page 11.

 

 

Author Response File: Author Response.pdf

Reviewer 3 Report

H. Sun and co-authors, investigated the role of the single Cu and Au atom embedded on the N6-centre of C2N monolayer, for the reduction of N2O in presence of the CO using the first-principles calculations at the ground state DFT level. Calculations are done using DMol³ gaussian type basis set choice. Two types of possible mechanisms are intended to be viable, 1) the formation of the O-M/C2N vs. 2) CO-M/C2N charge centre, and the latter is found to be relatively better in the current gold using the Cu/C2N catalysts than the other one.

 

The work sounds good to the community and would provide a viable fundamental mechanisms to guide exp. synthesis. Unfortunately, the supported data is very weak in quality and arrangement, with several typos. For example, Au-5d orbitals, missing Au(5s) or Cu(4s) projected density of states makes thing more doubtful, despite detail strength of the ground state density functional theory. Hence, I decline it at this stage for publication, but authors are encourage to refresh the arrangement of data sets and also quality figures and tables with following suggestions and comments:

 

1.     Please rearrange and move the computational and mythology sections at the beginning of the article, before results section. Please try to highlight the key words used in the Fig. 2 and Fig. 6. Both figures need a top and side view with better projections. Please read and refer some relevant literature, for example J. Chem. Phys. 154, 134706 (2021).

2.     I do not see any reason of the Fig. 5, since the important case is Cu/C2N model, despite the fact the there is no 5s(4s) electronic density in these DOS plots for Au( or Cu) is provided, which is also true to the Supporting data at the Fig. S5. 

3.     No Bader charge analysis is found in the article, hence it is hard to get convinced the possible charge transfer from the catalyst to the adsorbates or vice versa. The top row of the Table 1 and Table 2, must be correlated with the DOS plots, which I suspect due to the unpaired s-electrons of the metals. Please verify. Please refer to the possible literatures: J. of Power Sources, 556, 232492 (2022), Acc. Mater. Res. 3, 986 (2022), ACS Catal. 12, 11682 (2022).

4.     The key words barrier heights and binding energies are not clearly mentioned in the tables.

5.     Consistent HOMO and LUMO charge density plots are required, for Cu and Au models, at least at the IS1a and TS1a configurations.

6.     Please carefully refresh the typos and grammatical issues. 

Author Response

Dear Reviewer,

 

Thank you very much for your valuable comments concerning our manuscript entitled “Catalytic reduction of N₂O by CO on single atom catalysts Au/C₂N and Cu/C₂N: A first-principles study” (ID: catalysts-2198490). The comments are really helpful for improving our manuscript, as well as the important guiding significance to our research. We have studied all the comments carefully and have made corrections. The main corrections in the paper and the response to the reviewer’s comments are as following by items:

 

1. Please rearrange and move the computational and mythology sections at the beginning of the article, before results section. Please try to highlight the key words used in the Fig. 2 and Fig. 6. Both figures need a top and side view with better projections. Please read and refer some relevant literature, for example J. Chem. Phys. 154, 134706 (2021)

Response: Thanks for Reviewer’s helpful suggestion. The “computational methods” was put before “Results and discussion” (Please see page 2). The keywords were highlighted in Fig. 2 and 6. Please see page 4 and 9. The reference was cited as reference 27 (“Due to the excellent stability, high catalytic activity and remarkable atomic efficiency, SCAs have attracted considerable researchers’ attention in recent years [27-30].” at page 2). Also, the top view of all the structures was added in Figure 2 (page 4), Figure 8 (page 10), Figure S3 (page 3) and Figure S4 (page 4).

2. I do not see any reason of the Fig. 5, since the important case is Cu/C2N model, despite the fact the there is no 5s(4s) electronic density in these DOS plots for Au( or Cu) is provided, which is also true to the Supporting data at the Fig. S5.

Response: Thanks for Reviewer’s comment. Actually, two catalysts are both important. FMO and PDOS plots were put in the main text, and s electronic density was considered in the revision. Please see Figure 5 (page 7) and Figure 7 (page 9).

 

3. No Bader charge analysis is found in the article, hence it is hard to get convinced the possible charge transfer from the catalyst to the adsorbates or vice versa. The top row of the Table 1 and Table 2, must be correlated with the DOS plots, which I suspect due to the unpaired s-electrons of the metals. Please verify. Please refer to the possible literatures: of Power Sources, 556, 232492 (2022), Acc. Mater. Res. 3, 986 (2022), ACS Catal. 12, 11682 (2022).

Response: The authors are very thankful to Reviewer’s comments. Since Bader charge analysis is unavailable in Dmol3 package, the charge values we used in the manuscript were mulliken charge. Reviewer’s comments reminded us to check DOS results and we found that we set wrong parameters about spin multiplicity. We re-calculated DOS and took s orbitals into account in the revision. Please see Figure 5 (page 7) and 7 (page 9). We are also grateful for the references Reviewer suggested, especially J. Chem. Phys. 154, 134706 (2021), which is excellent example in excavating the nature of systems from DOS results. Some discussions between DOS, charge and interaction were added in “Results and discussion”. Please see the discussions marked in red at page 5-8.

 

4. The key words barrier heights and binding energies are not clearly mentioned in the tables.

Response: The authors are grateful for Reviewer’s comment. We gave the definition of adsorption energy, but we used binding energy during describing MS3a(Cu) in the main text. We changed “binding energy” into “adsorption enthalpy” to keep consistent with the definition. For energy barrier, according to Reviewer2’s suggestion, activation enthalpy was used to describe energy barrier, and the definition was added in the end of “Computational methods” (page 2-3) as follows:

“The adsorption enthalpy (H_ads) was defined as follows:

H_ads = H_total-(H_M/C2N + H_adsorbate)

where H_total, H_M/C2N, and H_adsorbate, represent the enthalpies of the gas molecules adsorbed on the catalyst, M/C₂N, and gas molecules, respectively. The more negative the E_ads value, the higher the stability of the gas molecules on the catalyst.

The activation enthalpy (H_a), i.e. enthalpy barrier, which is one of the most important factors to evaluate the catalytic activity, was defined as fowllows:

H_a = H_TS-H_IM1

where H_TS and H_IM1 represent the enthalpies of the transition state and intermediate IM1 in the reaction step of IM1→TS→IM2.”

 

5. Consistent HOMO and LUMO charge density plots are required, for Cu and Au models, at least at the IS1a and TS1a configurations.

Response: The authors are thankful for Reviewer’s comments. We have two aspects for understanding this issue, but we are not sure which one is correct.

1. Reviewer suggested that we should add a FMO charge density plots for the rate-determinant step of two reactions. The plots were shown in Figure S5 and S6. Although there is only slight difference between HOMO charge density and LUMO one for each species, changes between different species along the reaction are significant. It can exhibit clearly the dissociation and formation of chemical bonds.
2. “Charge density” in the comment means the wave function density, ψ², which corresponds to the isovalue. In the original manuscript, the figures were plotted with same isovalue, but the color of the orbits in the plot is same with that of the N atoms, which is not easy to distinguish them, making readers feel that their wave function density settings are different. So, we changed orbital color to pink and yellow and altered ratio of ball radius to stick ones to eliminate possible misunderstanding, and isovalue was given in the figures title (The isovalue is setting as default 0.03 a.u.) in the revision. FMO plot for Cu/C₂N system was put in the main text for the convenience of comparison. Please see Figure 4 (page 7) Figure 6 (page 9).

We speculated number 2, but we were not very sure. So we gave the two solutions in the revision.

 

6. Please carefully refresh the typos and grammatical issues.

Response: Thanks for Reviewer’s comments. We have improved English writing in the revision.

Author Response File: Author Response.pdf