The Effects of Promoter Cs Loading on the Hydrogen Production from Ammonia Decomposition Using Ru/C Catalyst in a Fixed-Bed Reactor

Round 1

Reviewer 1 Report

The unit GHSV is used in presenting the performance of the catalyst. As the employed reactor was a packed-bed reactor, it would follow the performance of a plug-flow reactor. Then the conversions at the same GHSV and temperature are the same within the experimental error. In Figure 9, the conversion with 1.5 mg Ru for 6 ml/min flow should be equal to the conversion with 3 mg Ru for 12 ml/min, as both are the same GHSV. But in the figure the two conversions differ significantly. It is necessary to comment on whether GHSV can be used to estimate the conversion in the present study.

 

The reaction is endothermic and the heat transfer to the 316 stainless steel reactor may not be fast enough to maintain the temperature in the reactor the same as the measured temperature outside, since 316 SS is not a good conductor. This is often indicated by the difference in the conversions for the same GHSV and reaction temperature. The actual catalyst bed temperature might have been considerably lower than the indicated temperature in the manuscript. It is needed to comment on the actual temperatures of the catalyst bed during the experiments.

 

A reaction rate expression for the reaction with the catalyst would be helpful.

 

Author Response

Reviewer 1:

   The unit GHSV is used in presenting the performance of the catalyst. As the employed reactor was a packed-bed reactor, it would follow the performance of a plug-flow reactor. Then the conversions at the same GHSV and temperature are the same within the experimental error. In Figure 9, the conversion with 1.5 mg Ru for 6 ml/min flow should be equal to the conversion with 3 mg Ru for 12 ml/min, as both are the same GHSV. But in the figure the two conversions differ significantly. It is necessary to comment on whether GHSV can be used to estimate the conversion in the present study.

    The reaction is endothermic and the heat transfer to the 316 stainless steel reactor may not be fast enough to maintain the temperature in the reactor the same as the measured temperature outside, since 316 SS is not a good conductor. This is often indicated by the difference in the conversions for the same GHSV and reaction temperature. The actual catalyst bed temperature might have been considerably lower than the indicated temperature in the manuscript. It is needed to comment on the actual temperatures of the catalyst bed during the experiments.

  A reaction rate expression for the reaction with the catalyst would be helpful.

Authors’ reply:

1. The unit in Figure 9 has converted to GHSV in the revised manuscript.
2. I agree that a sharp temperature decrease in the entrance region of catalyst bed may occur due to endothermic reaction of ammonia decomposition (△H = 46.4 kJ/mole) [31]. The “temperature” in this paper refers to the temperature inside of the furnace. A severe cold spot at entrance region of catalyst bed was found in the steam methane reforming (△H = 206.1 kJ/mol) for larger reactor diameter (6 cm) in our paper [15] (Y.C. Chen is one of authors). Comment on the sharp temperature decrease in the entrance region of catalyst bed has been addressed in the revised manuscript.
3. The activation energy is generally used to evaluate the activity for the catalyst. This is given in the revised manuscript. Since the range of temperature and in flow rate is narrow in this experiment, more data are needed for the reaction rate expression for the further experiment.

Reviewer 2 Report

This manuscript studies the effect of Cs/Ru ratio on the performance of Cs-Ru/(Active Carbon) catalysts used in the ammonia decomposition for producing H2. Also, the effect of operation variables (temperature and GHSV) on ammonia conversion for the selected catalyst (that with Cs/Ru=4.5) is studied. Overall, the manuscript has some important deficiencies that advise against it publication, unless it is thoroughly revised as indicated next:    

1. The authors have included well up-to-date references on SRM, but have not made a well-up-to-date review of the state of the art on the subject of the manuscript, which is the decomposition of ammonia for H2 production. Except for Ref. Nº 30, there are no references from the last 10 years related to the subject of the manuscript. A search in Scopus with the words "ammonia decomposition hydrogen" returns at least 50 references since 2018, and there are 2 reviews on the subject (from 2013 and 2019). Accordingly, authors should update the introductory section.
2. The novelty of the manuscript is not clear, are at least it is not well stressed. Authors claim that “the optimal loading of promoter of Cs on Ru/AC (activated carbon) catalyst in the fixed-bed reactor is not investigated before”. But the effect of Cs-Ru ratio of a catalyst supported on carbon used for the ammonia decomposition reaction has been addressed in a previous manuscript (Int. J. Hydrogen Energy, 38 (2013) 3233-3240) from authors themselves. Nothing is commented in this manuscript of that previous work, which showed an optimal Cs/Ru ratio lower (3) than that obtained in this manuscript (4.5). Do the authors have any reason for this difference?
3. What is the reason for the improvement in the activity of the catalysts upon Cs introduction? Does Cs improves the dispersion of Ru? Does Cs cause any change in the oxidation state of Ru? The use of some other characterization techniques, such as TEM images for measuring the crystal size of metal/metal oxides would give valuable information, or perhaps XPS technique for analysing possible changes in the oxidation state. Some deeper discussion on the beneficial effect of Cs upon Ru is in fact expected.
4. Likewise, a comparison of the activity obtained with the optimum catalyst in this work should be compared to that obtained with other catalysts in literature under similar operating conditions (specially, comparison with the catalysts used in the 2013 authors’ manuscript mentioned above).
5. Authors mention the presence of some impurities (Na, Cu, O). Do the authors have some explanation for the origin of these impurities?
6. Figure 9 shows results of different Ru loading on ammonia conversion at 350 ºC for 2 different values of ammonia flowrate, and Figure 10 shows results for a given Ru loading and different ammonia flowrates. In fact, the 3 curves correspond to the effect of the same variable, GHSV. All the results of those curves corresponding to 350 ºC should be better plotted on a unique Figure showing the effect of GHSV on ammonia conversion.
7. Presentation: The manuscript is messy and unkempt:

7a) Figure 1 is mentioned on page 3, but it appears on page 8, after Figures 2-5.

7b) Equations 1 (ammonia decomposition reaction) and 2 (ammonia conversion) appear on page 8, but the calculation of ammonia conversion was previously explained on page 5 (lines 148-151), and, moreover, the symbols denoting ammonia consumption rate and ammonia inlet flowrate are missed on line 208.

7c) There are repetitive sentences: Lines 139-141 should be removed (the picture with the equipment has been previously explaines and, moreover, is correspond to Fig 2, not Fig 3 as states in those lines); the information in lines 202-204 of page 7 has been previously indicated in the text; line 241 on page 11 is improperly located, and it is repetitive; n previously

7d) The order of the sections in the manuscript do not follows the Template corresponding to the journal. Specifically, sections “Materials and methods” should be places after Results and discussion. Moreover, sections “Authors Contributions” and “Funding” and missed :.

7e) References: The abbreviated manes of the journal should be in italics, and the year in bold letters.

According to all the previous comments, the manuscript requires major revision

Author Response

Responses of Reviewers’ comments

Reviewer 2:

This manuscript studies the effect of Cs/Ru ratio on the performance of Cs-Ru/(Active Carbon) catalysts used in the ammonia decomposition for producing H2. Also, the effect of operation variables (temperature and GHSV) on ammonia conversion for the selected catalyst (that with Cs/Ru=4.5) is studied. Overall, the manuscript has some important deficiencies that advise against it publication, unless it is thoroughly revised as indicated next:    

1. The authors have included well up-to-date references on SRM, but have not made a well-up-to-date review of the state of the art on the subject of the manuscript, which is the decomposition of ammonia for H2 production. Except for Ref. Nº 30, there are no references from the last 10 years related to the subject of the manuscript. A search in Scopus with the words "ammonia decomposition hydrogen" returns at least 50 references since 2018, and there are 2 reviews on the subject (from 2013 and 2019). Accordingly, authors should update the introductory section.

Authors’ reply: We have included 15 up-to-date references (2015~2021) on ammonia decomposition, including 2 reviews and one from Nature Commun. One earlier review paper is also included (Yin wt al. 2004).

2. The novelty of the manuscript is not clear, are at least it is not well stressed. Authors claim that “the optimal loading of promoter of Cs on Ru/AC (activated carbon) catalyst in the fixed-bed reactor is not investigated before”. But the effect of Cs-Ru ratio of a catalyst supported on carbon used for the ammonia decomposition reaction has been addressed in a previous manuscript (Int. J. Hydrogen Energy, 38 (2013) 3233-3240) from authors themselves. Nothing is commented in this manuscript of that previous work, which showed an optimal Cs/Ru ratio lower (3) than that obtained in this manuscript (4.5). Do the authors have any reason for this difference?

Authors’ reply:

1. The research on ammonia decomposition in National United University (NUU) was initiated by me (Y.C. Chen), who invited Prof. Hsueh and Prof. Liu to join the research. This research stoped when the research assistant and the student left. I didn’t know that they continued to do it in Dept. Chemical Engineering, and they didn’t notice me the paper publication (Int. J. Hydrogen Energy, 38 (2013) 3233-3240) (the authorship of Y.C. Chen at that paper is low). I admit that I un-intentionally found that paper several years ago, but I didn't read that paper before submitting this manuscript. It is my fault, since I am too late to determine for submitting the manuscript to the special issue in “Catalyst” and didn’t read the related papers and the earlier inappropriate description in the research object wasn’t removed. The research object is modified in the revised manuscript.
2. Some discrepancies between the previous paper and the current manuscript are the following:

(a) The previous paper emphasized the preparation of Ru-Cs catalyst, and the Ru/C is prepared from RuCl₃ or Ru(NO)(NO₃)₃ and carbon powder, while the commercial 5 wt.% Ru/C catalyst is used here to guarantee the quality of Ru/C.

(b) The optimum Cs/Ru loading is mild different. At molar ratio Cs/Ru = 6, the ammonia conversion in the previous paper is still higher than that of the pure Ru/C case, while it is smaller than that of the pure Ru/C case in this manuscript.

(c) Ammonia conversion is an important characteristic for ammonia decomposition. The temperature effect on the ammonia conversion is provided in this manuscript, while no such information in the previous paper.

(d) The mechanisms for the performance of ammonia conversion at the low and high Cs/Ru loadings are interpreted as following in the revised manuscript, while the discussion about the mechanism is very limited in the previous paper.

(e) The percentage of atom numbers for each element of C, Ru, and Cs in SEM pictures is given in this work to make sure that even at the micro-scale, the molar ratio of Cs/Ru is still maintained at nearly the same as the overall molar ratio during the blending processes. This information wasn’t provided in the previous paper.

(f) The ammonia conversion is near 100% at 400°C in this manuscript, while it is 90% in the previous paper.

(g) The activation energy is at range of 28-46 kJ/mole in this work, which is lower than that (54-72 kJ/mole) in the previous paper.

3. What is the reason for the improvement in the activity of the catalysts upon Cs introduction? Does Cs improves the dispersion of Ru? Does Cs cause any change in the oxidation state of Ru? The use of some other characterization techniques, such as TEM images for measuring the crystal size of metal/metal oxides would give valuable information, or perhaps XPS technique for analysing possible changes in the oxidation state. Some deeper discussion on the beneficial effect of Cs upon Ru is in fact expected.

Authors’ reply: The mechanisms for the performance of ammonia conversion at the low and high Cs/Ru loadings are interpreted as following in the revised manuscript.

There are some factors that affect the performance of ammonia decomposition for the Ru/C catalyst. One beneficial factor for the performance is that the N₂ dissociating barrier is reduced by the promoter of the Cs atoms, surrounding around the Ru crystallite and adsorbed on the carbon substrate [52]. Some other factors are the transport resistance for NH₃ gas diffuse to the Ru surface, the resistance for the electronic conducting from the N₂ dissociating site to the carbon surface, and new compound formation. For low Ru loading of the 5 wt% Ru/C catalyst, the Ru crystallites are well dispersed on the carbon surface.

When the Cs/Ru loading is lower, the mechanism for the improvement of NH₃ conversion is basically the same with that for ammonia synthesis. This is so-called “hot ring promotion” (electronic) [53], where the Ru crystallite is surrounded by the Cs atoms that form a single layer of ring structure. Because the N₂ dissociating barrier is the rate limiting step [54], the promotion occurs at contact points between the Ru crystallites and Cs atoms [52], that is, the electronics produced from the N₂ dissociating sites are easier to reach the carbon substrate.

When the Cs/Ru loading is higher, the Cs atoms surrounding the Ru crystallite are increasing. The transport resistance becomes pronounced due to the reduction of contact area between Ru and NH₃ gas and the reaction sites are reduced. Since we didn’t find the new compound formation from Figures 3 and 5 for molar ratio of Cs/Ru = 6, the decreasing of ammonia conversion for molar ratio Cs/Ru > 4.5 may be attributable to remarkable increasing of the transport resistance and the electronic conducting resistance, and the reduction of reaction sites. The picture that that the Cs atoms form a single layer of ring structure at the carbon substrate for the low Cs/Ru loading may be needed to modify for the high Cs/Ru loading. This suggests that the information about the detail structure between the Cs atoms on Ru crystallite and carbon substrate is needed to well realize the mechanism at higher molar ratio of Cs/Ru.

4. Likewise, a comparison of the activity obtained with the optimum catalyst in this work should be compared to that obtained with other catalysts in literature under similar operating conditions (specially, comparison with the catalysts used in the 2013 authors’ manuscript mentioned above).

Authors’ reply: The comparison of the activation energy with the previous paper and other papers are given in Table 2.

5. Authors mention the presence of some impurities (Na, Cu, O). Do the authors have some explanation for the origin of these impurities?

Authors’ reply: The impurities most likely come from the commercial products of 5 wt.% Ru/C catalyst and CsNO₃ powder during their preparation, since our blending processes are simple, as shown in Section 3.1.

6. Figure 9 shows results of different Ru loading on ammonia conversion at 350 ºC for 2 different values of ammonia flowrate, and Figure 10 shows results for a given Ru loading and different ammonia flowrates. In fact, the 3 curves correspond to the effect of the same variable, GHSV. All the results of those curves corresponding to 350 ºC should be better plotted on a unique Figure showing the effect of GHSV on ammonia conversion.

Authors’ reply: We agree that the effects from both the inlet flow rate and catalyst loading can be well to represented by the single parameter of GHSV. It is revised in the revised manuscript.

7. Presentation: The manuscript is messy and unkempt:

Authors’ reply: We was too hurry to submit the manuscript in at midnight Jan. 31 before the deadline. The manuscript is revised to remove typing and other errors.

7a) Figure 1 is mentioned on page 3, but it appears on page 8, after Figures 2-5.

Authors’ reply: This is the first time we submits a manuscript to “Catalyst”. We used the format that was used for the submission to Int. J. Hydrogen Energy, where all the pictures are put together orderly and all of them are placed in the bottom of the text in the original draft submitted to “Catalyst”. The Editorial office help us to convert our earlier manuscript to current format, but Figure 1 was placed in the wrong place, and the symbols denoting ammonia consumption rate and ammonia inlet flow rate are missed on line 208 during that conversion. Now Figure 1 has been moved to the right place in the revised manuscript.

7b) Equations 1 (ammonia decomposition reaction) and 2 (ammonia conversion) appear on page 8, but the calculation of ammonia conversion was previously explained on page 5 (lines 148-151), and, moreover, the symbols denoting ammonia consumption rate and ammonia inlet flowrate are missed on line 208.

Authors’ reply: The redundant description on page 5 (lines 148-151) is removed, and the symbols are added in the revised manuscript.

7c) There are repetitive sentences: Lines 139-141 should be removed (the picture with the equipment has been previously explaines and, moreover, is correspond to Fig 2, not Fig 3 as states in those lines); the information in lines 202-204 of page 7 has been previously indicated in the text; line 241 on page 11 is improperly located, and it is repetitive previously

Authors’ reply: (a) We forgot to delete the statement “A picture of the ammonia decomposition performance measurement system used in the experiment is shown in Figure 3.”, when we determined to remove the photograph picture of the measurement system. Now it is removed. (b) The repeat text in in lines 202-204 of page 7 is deleted. (c) “The EDS intensity spectrum pattern” in line 241 has been changed to “The XRD pattern”.

7d) The order of the sections in the manuscript do not follows the Template corresponding to the journal. Specifically, sections “Materials and methods” should be places after Results and discussion. Moreover, sections “Authors Contributions” and “Funding” and missed :.

Authors’ reply: The “Experiment” in the previous manuscript has been changed to “Materials and Methods”, and it places after “Results and Discussion” in the revised manuscript. The sections “Authors Contributions” and “Funding” are added in the revised manuscript.

7e) References: The abbreviated manes of the journal should be in italics, and the year in bold letters.

According to all the previous comments, the manuscript requires major revision

Authors’ reply: They have been corrected in the revised manuscript.

Round 2

Reviewer 2 Report

The authors have made a thorough revision of the work.They have answered quite satisfactorily to the questions and improvement aspects raised by this reviewer. The paper can now be published