Support Effect of Ga-Based Catalysts in the CO₂-Assisted Oxidative Dehydrogenation of Propane

Round 1

Reviewer 1 Report

The paper attempts to present a desirable area in dehydrogenation at present, that is, in the presence of CO2. However, I think the authors have missed teh mark somewhat. I have made comments and queries in the annotated version that is attached but the authors should also pay attention to teh following.  The catalysts were only tested with CO₂ and not without so it is difficult to access the catalytic performance of the catalysts with and without CO₂. This will also aid in establishing the role of CO₂ in the reaction. In lines 231 to 239 in the manuscript, the authors talk about the formation of formic acid and the removal of H2 in this way. However, based on the comparison of the CO₂ conversion and the propane conversion there is only a small fraction of the H₂ that would be removed. What about the rest of it?

The role of CO₂ is not clearly defined. In one case, it is meant to remove coke and in another, it aids in removing hydrogen. Some evidence is provided for teh latter but teh argument appears weak. No TGA data is presented to substantiate the former claim.

No ICP data is provided to establish Ga loading. Understandably, it is low, but how do the authors know how much was incorporated in each of the catalysts?

 I don’t understand how to interpret the data in Figure 6. Is this a time dependent monitoring? It is not clear. The numbers given in the figure are not clearly explained in the text.

Comments for author File: Comments.pdf

Author Response

1. The catalysts were only tested with CO₂and not without so it is difficult to access the catalytic performance of the catalysts with and without CO₂. This will also aid in establishing the role of CO₂in the reaction.

RESPONSE:

Thank you for this remark. The advantages of CO₂ as a “soft oxidant” for the dehydrogenation of alkane have been widely reported in the literature,^1-4 such as (i) inhibited deep oxidation of the alkane and improved olefin selectivity, (ii) mitigated anthropogenic greenhouse gas emissions, (iii) enhanced equilibrium conversion by removing hydrogen through the reverse water-gas shift (RWGS) reaction, and (iv) reduced coke formation and decelerated deactivation.

To verify the role of CO₂ in the reaction, we tested Ga/ZSM-5 (28) in the absence and presence of CO₂, and the results are shown in Figure R1. It can be seen that in both cases there is almost no significant difference in the conversion of propane on Ga/ZSM-5 (28). We speculate that the competitive adsorption of CO₂ and propane in ZSM-5 zeolite.^{5, 6} In order to further investigate the role of CO₂ in the ODHP reaction, we performed TGA characterization on spent Ga/ZSM-5(28) after reaction in the presence and absence of CO₂. According to the TGA data (See Figure R2 below), we can see that the addition of CO₂ in ODHP reaction can remove the coke. We have added the above results to the supplementary information and have refined the relevant statements in the manuscript.

COMMENT：

2. In lines 231 to 239 in the manuscript, the authors talk about the formation of formic acid and the removal of H₂in this way. However, based on the comparison of the CO₂conversion and the propane conversion there is only a small fraction of the H₂ that would be removed. What about the rest of it?

RESPONSE:

As previously reported by our group,⁷ on the zinc-modified zeolite, CO₂ can react with part of H₂ generated by dehydrogenation of alkanes to formic acid. However, in this study, i.e., on the Ga-modified catalysts, the removal of hydrogen by CO₂ was not significant, and we speculate that this is because less hydrogen was produced on the Ga-modified zeolites compared to the zinc-modified zeolites, resulting in less removal of hydrogen by CO₂. According to our DFT calculations (which is under the preparation to be submitted), although the activation energy barrier required for the C-C bond breaking on gallium is slight higher than that for C-H bond breaking, the process of C-C and C-H bond breaking is very competitive. While, it is very different in Zn-modified catalyst, in which the activation energy barrier required for the C-H bond breaking is much lower, resulting in the dehydrogenation mainly occurred. Therefore, the cleavage of alkanes occurs mainly in the Ga-modified zeolites, which is also confirmed by the selectivity of gallium- and zinc-modified SSZ-13 zeolites (Figure R2), so the hydrogen generation itself is much less on the Ga-modified zeolites, and therefore CO₂ plays a less significant role in the removal of hydrogen in the current study. In this paper, the main role of CO₂ is to remove coke, as shown in the TGA results (Figure R2). We have revised the relevant statement in the manuscript about the role of CO₂ over Ga-modified catalysts in the ODHP reaction (CO₂-assisted oxidative dehydrogenation of propane).

COMMENT：

3. The role of CO₂is not clearly defined. In one case, it is meant to remove coke and in another, it aids in removing hydrogen. Some evidence is provided for the latter but the argument appears weak. No TGA data is presented to substantiate the former claim.

RESPONSE:

As answered in the previous questions, CO₂ has various advantages such as enhancing equilibrium conversion and reducing the coke in ODHP reactions. However, in this study, the role of CO₂ in removing hydrogen, and enhancing equilibrium conversion is not obvious due to the low amount of hydrogen production compared to zinc-modified catalysts, and the main role of CO₂ is to remove coke and improve the stability of the catalysts.

In addition, The TGA profiles of Ga/ZSM-5(28) after reaction in the absence and presence of CO₂ are shown in Figure R2. As can be seen, performing the CO₂-ODHP reaction without CO₂ led to slightly more coking compared to when CO₂ was absent. The results demonstrate that the addition of CO₂ in ODHP reaction can remove the coke. We have added the TGA results to the supporting information and analyzed the data in the manuscript. 

COMMENT： 

4. No ICP data is provided to establish Ga loading. Understandably, it is low, but how do the authors know how much was incorporated in each of the catalysts?

RESPONSE:

Thanks for the comment. ICP-OES measurements of the catalysts Ga/Na-SSZ-13(27), Ga/Na-SSZ-39(9) and Ga/Na-ZSM-5(28) before and after the reaction was carried out and the results are shown in Table R1. From the results, we can see that the gallium content of these catalysts is basically around 0.22-0.23%, which is higher than the theoretical loading of 0.1%. Because we prepared the Ga-modified samples by incipient wet impregnation at a liquid-to-solid weight ratio of 4, so the results are as expected. Interestingly, it can be seen that after the reaction, the gallium content of the catalysts all have a significant decrease, indicating that the gallium species in zeolites may not be so stable and its loss occurred during the aggressive reaction condition. We have added the results of ICP-OES characterization to the supporting information and have analyzed the results in the manuscript.

COMMENT：

5. I don’t understand how to interpret the data in Figure 6. Is this a time dependent monitoring? It is not clear. The numbers given in the figureare not clearly explained in the text.

RESPONSE:

Thank you for this remark. Figure 6 shows the results of in situ FT-IR characterization of the catalysts in real time during the reaction. The numbers given in the figure may be disturbing to the reader. we have refined Figure 6, as shown below.

• Detailed Comments:

COMMENT：

6. (Line 112) How do you know this?Did you do TGA. Please clarify.

RESPONSE:

Thank you, we have modified this paragraph (see below or manuscript) to include an analysis of the TGA data, please see comment 3 for the TGA results.

“In addition, the catalyst gradually deactivated with the reaction proceeding, which, combined with the TGA and ICP results, may be due to the loss of Ga species and im-plying a partial deactivation of Ga species deposited by coke.”

COMMENT：

7. (Line 124) What does this mean? Please clarify. what about the Ga? is it playing no role?

RESPONSE:

The incorporation of gallium into zeolites introduces Lewis acid sites,⁸ which are active for the breaking of C-H bond. Ga/SiO₂, however, does not exhibit the same high activity as other materials probably due to its topological structure and differences in acid properties.

COMMENT：

8. (Line 128) How do you know this? Did you do the reaction without CO₂to compare? If not, it should be done to see the relative response.

RESPONSE:

Thank you, we have tested Ga/ZSM-5 (28) in the absence and presence of CO₂, and the results were shown in Figure R1 (See comment 1 for details).

COMMENT：

9. (Line 124 – Line 132) Is the role of CO₂just to burn off coke? A comparative test needs to be done without CO₂to check. Did monitor the H₂ produced?. This should be done to ascertain the effect of CO₂ in the reaction.

RESPONSE:

Thank you, due to the limitations of the sensitivity of the detector of our GC, we could not accurately detect the amount of hydrogen produced. However, according to our explanations in the above comments, the role of CO₂ is mainly to remove coke and improve the stability of the catalyst, and the effect of CO₂ in removing hydrogen in this study was not significant.

COMMENT：

10. (Figure 1) Reasons for low conversionfor these catalysts? SA, acidityGa dispersion??

RESPONSE:

In this study we have focused on the performance of gallium-modified zeolites with different topologies in PDHP reactions in relation to their topologies, acid amounts and other factors. For the reasons of low activity of Ga/SiO₂ and Ga/Al₂O₃ we believe that the amount of acid and the structure have certain effect.

COMMENT：

11. (Line 153) What does this imply? Can the authors please explain how the topology affects the catalytic behavior, but surface area and pore volume does not.

RESPONSE:

Our results show that in CO₂-ODHP reaction Ga-modified zeolites with different topologies exhibit large differences in activity. We believe this is mainly due to the differences in the topologies of zeolites. For instance, taking Ga/ZSM-5 (28) and Ga/SSZ-13 (27) as examples, their specific surface areas and pore volumes are around 700 m²_·g^-1 and 0.3cm³_·g^-1 respectively, but the propane conversion rate of ZSM-5 is three times higher than that of SSZ-13, suggesting that surface area and pore volume are not significant for the activity of Ga-modified zeolites, we have revised the statement in the manuscript to make it more precise.

“The above topological properties indicated that between the individual samples, the topological structure of zeolites was an essential factor, but the surface area and pore volumes unaffected have a minor effect on the performance of CO2-assisted oxidative dehydrogenation of propane.”

COMMENT：

12. (Line 171 - Line 173) The NH₃-TPD (Figure S2) should be deconvoluted if possible to better define the type of acid sites. Deconvolution should help better define this, it is too vague

RESPONSE:

Thank you for your valuable comment. We have made peak differentiation for the results of NH₃-TPD, as shown below. The relative acid amounts of weak, moderate strong acid and strong acids for the two zeolites are shown in Table S1.

COMMENT：

13. (Line 178) Why NH₃? this is meant to be pyridine IR!

RESPONSE:

Thank you. We have made revisions to the manuscript. Please refer to the details below.

“The species adsorbed at 1440 cm^-1 over the Ga-containing samples were assigned to pyridine molecules on Lewis acid sites.”

COMMENT：

14. (Line 180) Where is the data for the pristine materials?

RESPONSE:

Thank you for this comment. The Py-IR spectra of the pristine materials have been added to the supplementary materials, please see below. Also, we have amended the statement in the manuscript.

“The species adsorbed at 1440 cm^-1 over the Ga-containing samples were assigned to pyridine molecules on Lewis acid sites. The incorporation of Ga species into the zeolite structure makes the peak at 1440 cm^-1 stronger compared to the pristine materials (see Figure S5).”

COMMENT：

15. (Line 259) Was the supernatant liquid analyzed for Ga?

RESPONSE:

Thank you for this comment. According to the inference of the ICP characterization results, gallium is present in the supernatant liquid. Because we prepared the Ga- modified samples by incipient wet impregnation at a liquid-to-solid weight ratio of 4.  According to the results of ICP, the gallium loading of the samples are all about 0.22-0.23%, which is far below the value of 0.4%, so we can judge the presence of gallium in the supernatant liquid.

Reference

1. Lawson, S.;  Newport, K. A.;  Axtell, A.;  Boucher, C.;  Grant, B.;  Haas, M.;  Lee, M.;  Rezaei, F.; Rownaghi, A. A., Structured Bifunctional Catalysts for CO2 Activation and Oxidative Dehydrogenation of Propane. ACS Sustainable Chemistry & Engineering 2021,9(16), 5716-5727.
2. Farsad, A.;  Lawson, S.;  Rezaei, F.; Rownaghi, A. A., Oxidative dehydrogenation of propane over 3D printed mixed metal oxides/H-ZSM-5 monolithic catalysts using CO2 as an oxidant. Catal. Today 2021,374, 173-184.
3. Liu, J.;  Zhang, Z.;  Jiang, Y.;  Jiang, X.;  He, N.;  Yan, S.;  Guo, P.;  Xiong, G.;  Su, J.; Vilé, G., Influence of the zeolite surface properties and potassium modification on the Zn-catalyzed CO2-assisted oxidative dehydrogenation of ethane. Appl. Catal., B 2022,304.
4. Zhang, Y.; Aly, M., Effect of CO2 on activity and coke formation over gallium-based catalysts for propane dehydrogenation. Appl. Catal., A 2022,643.
5. de Oliveira, J. F. S.;  Volanti, D. P.;  Bueno, J. M. C.; Ferreira, A. P., Effect of CO2 in the oxidative dehydrogenation reaction of propane over Cr/ZrO2 catalysts. Appl. Catal., A 2018,558, 55-66.
6. Xu, B.;  Zheng, B.;  Hua, W.;  Yue, Y.; Gao, Z., Support effect in dehydrogenation of propane in the presence of CO2 over supported gallium oxide catalysts. J. Catal. 2006,239(2), 470-477.
7. Liu, J.;  He, N.;  Zhang, Z.;  Yang, J.;  Jiang, X.;  Zhang, Z.;  Su, J.;  Shu, M.;  Si, R.;  Xiong, G.;  Xie, H.-b.; Vile, G., Highly-Dispersed Zinc Species on Zeolites for the Continuous and Selective Dehydrogenation of Ethane with CO2 as a Soft Oxidant. ACS Catal 2021,11(5), 2819-2830.
8. Ji, Y.;  Shi, B.;  Yang, H.; Yan, W., Synthesis of isomorphous MFI nanosheet zeolites for supercritical catalytic cracking of n-dodecane. Appl. Catal., A 2017,533, 90-98.

Author Response File: Author Response.pdf

Reviewer 2 Report

Recommendation: Publish after major revisions noted.

Comments:
In this work, the authors discussed the support effect of Ga- based catalysts in the CO2-assisted oxidative dehydrogenation of propane. They claimed that the appropriate acidity in the catalyst could positively promote the dehydrogenation activity of this reaction. While the overall conclusions are reasonable, the manuscript has several deficiencies that need to be addressed before publication.

Detailed comments are listed below:

1)    Page 2 (line 55-57): It is too brief and misleading to talk about the direct dehydrogenation of light alkanes. There are many published works showing environmentally friendly and non-noble metal based catalysts for the direct dehydrogenation of alkanes to olefins (e.g., 10.1021/acscatal.0c03536, 10.1021/jacs.8b05060, 10.1021/jacs.0c07792, etc.).

2)    What is the particle size and distribution of Ga?

3)    I would suggest the authors putting reaction conditions in the captions of Figure 1 and 2.

4)    The authors need to put error bars in Figure 1.

5)    Did the authors calculate the equilibrium conversion of propane under their reaction condition? The equilibrium conversion of propane at 600C is slightly lower than 60% (10.1038/s41467-018-03793-w). The propane conversion over Ga/Na-SSZ-39(9) is very closed to the equilibrium conversion.

6)    When the authors compared selectivity, all catalysts must be kept at similar propane conversions.

7)    TOF needs to be calculated to compare the activity between each catalyst.

8)    Ga atoms have the possibility to be lost under such aggressive reaction condition. ICP-OES measurements of the catalyst before and after reaction may be helpful to understand the elemental stability of Ga.

Author Response

1. Page 2 (line 55-57): It is too brief and misleading to talk about the direct dehydrogenation of light alkanes. There are many published works showing environmentally friendly and non-noble metal based catalysts for the direct dehydrogenation of alkanes to olefins (e.g., 10.1021/acscatal.0c03536, 10.1021/jacs.8b05060, 10.1021/jacs.0c07792, etc.).

RESPONSE:

Thank you for pointing that out, we have added a comment of the application of environmentally friendly and non-noble metal based catalysts in the direct dehydrogenation of alkanes to olefins, as shown below.  

“Recently, some environmentally friendly and non-noble metal based catalysts, such as Al₂O₃ loaded with Ni₃Ga nanoparticles, Ga-modified γ-Al₂O₃ and Fe-containing MFI siliceous zeolite have also been used in direct dehydrogenation of alkanes and have demonstrated excellent performance, however, these studies are also still at the stage of laboratory research and are still a long way from practical applications.” 

COMMENT：

2. What is the particle size and distribution of Ga?

RESPONSE:

We thank the reviewer for the valuable suggestion. In response, we have done TEM characterization of Ga-based catalysts to investigate the particle size and distribution of Ga. Please refer to Figure 7 in manuscript or below. Also, we have inserted a description of particle size and distribution of Ga in the manuscript, please see below or in the manuscript.

“The energy dispersive system (EDS) maps of aluminum, silicon and gallium of five samples are shown in Figure 7(a-e), the distribution of these species is uniform, suggesting that the gallium is uniformly distributed in these carriers. Taking the case of the Ga/SSZ-39(9) catalyst, the Ga₂O₃ particles with an average size of ca. 13.8 nm are distributed on the zeolite surface, as shown in Figure 7(f).”

COMMENT：

3. I would suggest the authors putting reaction conditions in the captions of Figure 1 and 2.

RESPONSE:

Thank you for this suggestion. We have added a description of the reaction conditions to the legend of Figure 1 and Figure 2, and details are shown below.

Figure 1: “(a) Propane conversion and (b) CO₂ conversion over Ga-based catalysts. Reaction conditions: m_cat = 0.5g, T = 600 ^◦C, P = 0.1 MPa, GHSV = 7200 mL gcat^-1 h^-1, CO₂:C₃H₈:N₂ = 1:1:18.”

Figure 2: “Products selectivity over Ga-based catalysts within 60 min. Reaction conditions: m_cat = 0.5g, T = 600 ^◦C, P = 0.1 MPa, GHSV = 7200 mL gcat^-1 h^-1, CO₂:C₃H₈:N₂ = 1:1:18.”

COMMENT：

4. The authors need to put error bars in Figure 1.

RESPONSE:

We have made corresponding changes to Figure 1 according to your requirements. The revised Figure 1 can be seen in the manuscript or below.

COMMENT：

5. Did the authors calculate the equilibrium conversion of propane under their reaction condition? The equilibrium conversion of propane at 600C is slightly lower than 60% (10.1038/s41467-018-03793-w). The propane conversion over Ga/Na-SSZ-39(9) is very closed to the equilibrium conversion.

RESPONSE:

Thank you for this remark. The equilibrium conversion of propane under our reaction condition could not be calculated since our reaction was carried out on a fixed bed, which is a non-closed system. The equilibrium conversion of propane in the literature (10.1038/s41467-018-03793-w) were obtained through HSC Chemistry 8 software, utilizing a Gibbs free energy minimization algorithm, which are different from our reaction conditions and therefore are not applicable to our reaction system.

COMMENT：

6. When the authors compared selectivity, all catalysts must be kept at similar propane conversions.

RESPONSE:

As it is difficult for us to control the reaction conversion of both catalysts at the same level in this reaction, it is difficult to compare the selectivity of the catalysts at the same conversion and we have modified the manuscript accordingly.

COMMENT：

7. TOF needs to be calculated to compare the activity between each catalyst.

 RESPONSE:

Thank you for this remark. The turnover frequency (TOF) for the three catalysts was calculated from TOF=r_Ac/n_OF. Therewith, r_Ac is the total amount of propane converted per weight of Ga per second, and n_OF the amount of Ga at the surface. The latter is calculated from n_OF = n_total ∗ D, where n_total is the total amount of Ga and D the degree of dispersion (i.e., the surface-to-volume ratio), estimated by TEM.¹ (at standard conditions: T = 673 K, P_total = 0.1MPa).

COMMENT：

8. Ga atoms have the possibility to be lost under such aggressive reaction condition. ICP-OES measurements of the catalyst before and after reaction may be helpful to understand the elemental stability of Ga.

RESPONSE:

We thank the reviewer for the valuable suggestion. ICP-OES measurements of the catalysts Ga/Na-SSZ-13(27), Ga/Na-SSZ-39(9) and Ga/Na-ZSM-5(28) before and after the reaction was carried out and the results are shown in Table R1. From the results, we can see that the gallium content of these catalysts is basically around 0.22-0.23%, which is higher than the theoretical loading of 0.1%. Because we prepared the Ga-modified samples by incipient wet impregnation at a liquid-to-solid weight ratio of 4, so the results are as expected. Interestingly, it can be seen that after the reaction, the gallium content of the catalysts all have a significant decrease, indicating that the gallium species in zeolites may not be so stable and its loss occurred during the aggressive reaction condition. We have added the results of ICP-OES characterization to the supporting information and have analyzed the results in the manuscript.

Reference

1. C. Mohr, H. H., P. Claus a, The influence of real structure of gold catalysts in the partial hydrogenation of acrolein. J. Catal. 2003,213, 86-94.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The revised manuscript has addressed many of my concerns. thank you. In your response, you mention Figure R2 but I think you mean Figure S5 or Figure S6 where appropriate.

Author Response

Reviewer 1

RESPONSE:

Thank you for your comments and approval. Figure R2 in the response should be Figure S7, and we have double-checked the legend of each Figures in the manuscript and supplementary information.

Reviewer 2 Report

Recommendation: Accepted.

Comments:
The revised manuscript has been significantly improved! I am glad to see the authors carefully addressed most of my comments noted before. I would recommend accepting the manuscript in present form.

Author Response

Reviewer 2

RESPONSE:

Thank you for your approval.