Isotope Labelling for Reaction Mechanism Analysis in DBD Plasma Processes
, Jose M. Obrero-Pérez 1, José Cotrino 1,2, Agustín R. González-Elipe 1 and Ana Gómez-Ramírez 1,2,*Round 1
Reviewer 1 Report
The manuscript provided possibly very useful diagnostics techniques to investigate important reaction pathways for plasma catalysis. The submitted manuscript is well written and organized by showing two different examples of plasma gas synthesis, ammonia and wet reforming reaction of methane in parallel.
However, I have one major concern, which the authors should put more effort to address it on the manuscript. I’d like to ask you to explain the intention of using ternary mixture of NH3/N2/D2 rather than N2/D2. It may have meant for proving the inefficient energy consumption by recycling between given gas species rather than actually increasing the number density of desirable product. However, I’d imagine detecting ND3 would have provided more advantage to avoid the intervention from water molecule relatively easily in your mass spectrum analysis. I see there was given description about using leak valve and capillary tube in page 9. Would it be possible to put more detail about your methodology using Mass spectrometry such as schematics of experimental setup and how you’ve obtained the final mass spectra cancelling out the interference from water? It can be included somewhere in the supporting document or briefly in the manuscript. Because we know the adsorbed water in the system also may be desorbed and increase the detected signal at 17 and 18 amu when the plasma power and temperature increase. I believe having more detailed description about this would be beneficial for many researchers working on the plasma gas synthesis applications.
The same concern about Table 1, the consideration on the possible contribution from H2O are not included. In order to count only NHxDy species then it would be necessary to provide a little more detail than what you have described in page 3 line 120-123.
Can authors clarify the paragraph, in page 2, line 82-83? - although slightly changing the proportion of the three gases in order to keep the amount of ammonia constant in the outlet –
It could be misread I’m afraid. Please try to explain clearer why you aimed to control the amount of ammonia to be constant and how much changes you’ve introduced in your input gas.
For the ammonia synthesis, the surface reaction is known to be important and also possibly the 3 body reaction between NH N and H radicals. But you’ve discussed only related to gas phase reaction in page 4. Reaction (1).
Can authors provide more justification for it?
Author Response
The manuscript provided possibly very useful diagnostics techniques to investigate important reaction pathways for plasma catalysis. The submitted manuscript is well written and organized by showing two different examples of plasma gas synthesis, ammonia and wet reforming reaction of methane in parallel.
However, I have one major concern, which the authors should put more effort to address it on the manuscript. I’d like to ask you to explain the intention of using ternary mixture of NH3/N2/D2 rather than N2/D2. It may have meant for proving the inefficient energy consumption by recycling between given gas species rather than actually increasing the number density of desirable product. However, I’d imagine detecting ND3 would have provided more advantage to avoid the intervention from water molecule relatively easily in your mass spectrum analysis. I see there was given description about using leak valve and capillary tube in page 9. Would it be possible to put more detail about your methodology using Mass spectrometry such as schematics of experimental setup and how you’ve obtained the final mass spectra cancelling out the interference from water? It can be included somewhere in the supporting document or briefly in the manuscript. Because we know the adsorbed water in the system also may be desorbed and increase the detected signal at 17 and 18 amu when the plasma power and temperature increase. I believe having more detailed description about this would be beneficial for many researchers working on the plasma gas synthesis applications.
Following this indication of the referee, we include a new figure in the supporting information in order to clarify how we perform the analysis of the different intensities of the mass peak and how, in particular, we remove the concentration of the residual water always present in the MS chamber (see Figure S1 in Supporting Information and lines 129-144 in page 4/12- ain text).
Certainly, another experimental approach would be to use mixtures N2+H2+ND3 instead of N2+D2+NH3. In the two cases, we would observe some similar atom exchange phenomena. However, ND3 is extremely expensive and because of that not practical for conventional laboratory experiments.
We try to clarify why we use the ternary mixture in page 2/12-main text, line 79-98.
The same concern about Table 1, the consideration on the possible contribution from H2O are not included. In order to count only NHxDy species then it would be necessary to provide a little more detail than what you have described in page 3 line 120-123.
We describe in more detail the experimental conditions utilized and the way of dealing with the experimental MS spectra in order to quantify the concentration of species in the outlet gas. Please see page 4/12-main text, line 129-144.
Can authors clarify the paragraph, in page 2, line 82-83? - although slightly changing the proportion of the three gases in order to keep the amount of ammonia constant in the outlet –
It could be misread I’m afraid. Please try to explain clearer why you aimed to control the amount of ammonia to be constant and how much changes you’ve introduced in your input gas.
We have tried to clarify this point in the new version of the paper. This phrase has been changed to make it more understandable (page 2/12-main text, line 79-98).
For the ammonia synthesis, the surface reaction is known to be important and also possibly the 3 body reaction between NH N and H radicals. But you’ve discussed only related to gas phase reaction in page 4. Reaction (1).
Can authors provide more justification for it?
We agree with the referee in that both surface reactions and plasma processes are contributing to the overall process. Actually, in our previous publication (Ref. [11]) we proposed that surface reactions are involved in the synthesis of ammonia in the DBD reactor. The isotope labelling technique does not only account for plasma processes but also for surface reactions without differentiating between the two. We try to clarify this point in the new version of the paper (page 5/12-main text, new line 183-187).
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments to the manuscript ‚catalysts-405852: Isotope labelling for reaction mechanism analysis in DBD plasma catalysis processes.
General description:
The authors describe the reforming reaction of methane as well as the synthesis of ammonia by labelling reactions with deuterium or heavy water. The core result of this manuscript is that there is no net formation of ammonia and nearly no net transformation of methane to syngas and input energy is mainly consumed by substitution reactions of H with the supplied D.
In context to the literature this results point out that at least non-thermal plasma processes are no effective approach in formation of ammonia or syngas.
Introduction:
· If the main conclusion of the general description is correct (NTP is no effective approach in formation of ammonia or syngas) this main conclusion should be presented more clearly and further discussed with present literature presenting NTP process as the new alternative to the Haber-Bosch – Snythesis or similar.
Results and Discussion:
· A negligible formation of ammonia was indeed proved because MS showed that the final concentration of n2 in the outlet gases remained constant after plasma activation: At least there are no hints whether the process was operated at ambient pressure or at low pressure conditions. However, in case of ambient pressure conditions the amount of N2 is so high that the MS signal goes into signal overflow. Hence, it is unclear how the MS was used to quantify the N2 level in the reaction chamber. Please clarify.
· Figure 1 b) In the range of wave number 1300 – 2000 water shows a significant absorption / transmission characteristics. Considering the peak pattern for the three plasma energy levels a) water occurs in the system and b) te concentration of water declines with higher energy levels. Why is there water at all (not listed in the reactions and the explanations) and b) the presence of water (or remaining OH radicals) for sure is highly relevant in oxidative reactions of NH3, i.e. formation of NOx or nitrate or in combination with catalysts as sulfates or similar. This aspect was for example described in Dobslaw et al. 2017 (https://dx.doi.org/10.1016/j.jece.2017.10.015). Please give a critical statement how to exclude formation of NOx or similar during plasma exposure in presence of water. Please add this critical statements to the manuscript and add relevant literature to the reference list (for example Dobslaw et al., 2017).
Materials and Methods:
· As operational parameters the reactor in case of NH3 synthesis was operated with an asymmetric discharge (30 µs positive side; 120 µs negative side). Please give an explanation for this process parameters.
· Page 8 Line 283: …in Figure 5 is consistent with that…
Author Response
General description:
The authors describe the reforming reaction of methane as well as the synthesis of ammonia by labelling reactions with deuterium or heavy water. The core result of this manuscript is that there is no net formation of ammonia and nearly no net transformation of methane to syngas and input energy is mainly consumed by substitution reactions of H with the supplied D.
This is not really the point. In the case of ammonia we select a reaction mixture where no net new formation of ammonia takes place. This is so because we select a ternary mixture compose of reactants (i.e., N2 and D2) and product (NH3). In the case of the reforming reaction this is not the case since we only use reactants (CH4 and D2O). Actually 50% methane and the corresponding amount of D2O is transformed into H2 and CO at the maximum power. We try to clarify all these concepts in the new version of the paper. Please see carefully changes along sections 2.2, 2.3 and 2.4.
In context to the literature this results point out that at least non-thermal plasma processes are no effective approach in formation of ammonia or syngas.
Again, this is not really the point. We demonstrate, and quantify, the existence of reversal processes transforming products present in the reactor into reactants and therefore consuming energy inefficiently. We believe that our results, showing the existence of these inefficient processes, open the way to alter the reaction processes in ways that could increase the overall energy efficiency. We introduce a brief comment on this point at the end of the paper.
Introduction:
· If the main conclusion of the general description is correct (NTP is no effective approach in formation of ammonia or syngas) this main conclusion should be presented more clearly and further discussed with present literature presenting NTP process as the new alternative to the Haber-Bosch – Snythesis or similar.
According to our comments above, this is not the main conclusion. We rephrase these conclusions to clarify the implications of our work.
Results and Discussion:
· A negligible formation of ammonia was indeed proved because MS showed that the final concentration of n2 in the outlet gases remained constant after plasma activation: At least there are no hints whether the process was operated at ambient pressure or at low pressure conditions. However, in case of ambient pressure conditions the amount of N2 is so high that the MS signal goes into signal overflow. Hence, it is unclear how the MS was used to quantify the N2 level in the reaction chamber. Please clarify.
We describe in more detail the experimental conditions. Plasma reaction is carry out at ambient pressure, but analysis of product by MS is made at low pressure. Sample (i.e. outlet flow of gases coming from the reactor) is dosed into the MS chamber using a capillary tube and a leak valve under conditions that are known not to produce any change in the proportion of gas molecules. Although this is a conventional measurement technique, we describe them in detail in the Supporting Information where we present a detailed scheme of the experimental set up (Figure S4).
· Figure 1 b) In the range of wave number 1300 – 2000 water shows a significant absorption / transmission characteristics. Considering the peak pattern for the three plasma energy levels a) water occurs in the system and b) te concentration of water declines with higher energy levels. Why is there water at all (not listed in the reactions and the explanations) and b) the presence of water (or remaining OH radicals) for sure is highly relevant in oxidative reactions of NH3, i.e. formation of NOx or nitrate or in combination with catalysts as sulfates or similar. This aspect was for example described in Dobslaw et al. 2017 (https://dx.doi.org/10.1016/j.jece.2017.10.015). Please give a critical statement how to exclude formation of NOx or similar during plasma exposure in presence of water. Please add this critical statements to the manuscript and add relevant literature to the reference list (for example Dobslaw et al., 2017).
The series of peaks in the range 1300-2000 cm-1 in the IR spectra are due to NH3. In the supporting information (Figure S2) we present now IR spectra of the mixtures N2+D2 and N2+D2+NH3 and clearly demonstrate that the three sets of peaks appearing in the spectra are due to NH3 (see page 4/12-main text, line 129-153). Then, when isotopically exchanged molecules start to be formed the three characteristics zones where NH3 present bands experience a modification due to the contribution of NH2D, NHD2 or ND3 as explained in the text. We must note that there was no water in the reaction mixture and that, therefore, there is no source of oxygen that might produce some NOx. It is also confirmed by the absence of peak m/z=30 (NO) in the MS (see Figure 1a). Nevertheless, the reference you provided has been included as an example of VOCs removal applications in the Introduction (new Reference [6]).
Materials and Methods:
· As operational parameters the reactor in case of NH3 synthesis was operated with an asymmetric discharge (30 µs positive side; 120 µs negative side). Please give an explanation for this process parameters.
Page 8 Line 283: …in Figure 5 is consistent with that…
There was a mistake in the text, actually we wanted to mean 80 µs positive side and 120 µs negative side. The use of squared profiles instead of sinusoidal provides conditions more favorable for the efficiency of the process, because of the higher Vrms in the former case. We try to clarify this point in the new version of the Supporting Information.
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors tried to introduce a new method for the evaluation of plasma chemical reactions. However, the current format is not satisfactory to guarantee the merits of the new method.
1． Title is confusing. Recently, plasma-catalysis is one of important issues that combined plasma and catalyst in a same volume. However, In the current, no catalyst was used. It seems like that the authors intended to describe plasma-induced chemical reactions as catalysis. To avoid confusion, it is highly recommended to change the title.
2． The definition of RE is not clear. I have read the text several times but I cannot understand why RE can be useful parameter for evaluating the process performance. For example, what is the rationale for the first sentence in page 1 of 11?. What about the following cases?
Newly formed NH3 à experience exchange reaction to form NDH2
What do you say for this case based on RE?
The elementary reactions of plasma are quite complicated, and I think that the method suggested by the authors too much simplified. If the merit of the method is not guaranteed, then the rest of discussions (Figs 3 and 4) may loose its importance.
3. The Lissajous figure in Figure 5 is quite different from the normal parallelogram. The authors need to explain this in more detailed way.
Author Response
The authors tried to introduce a new method for the evaluation of plasma chemical reactions. However, the current format is not satisfactory to guarantee the merits of the new method.
In the new version of the manuscript, we have tried to clarify the fundamentals of the method and hope that it may result more understandable.
1． Title is confusing. Recently, plasma-catalysis is one of important issues that combined plasma and catalyst in a same volume. However, In the current, no catalyst was used. It seems like that the authors intended to describe plasma-induced chemical reactions as catalysis. To avoid confusion, it is highly recommended to change the title.
We partially agree with the referee. Surface reactions certainly taken place in pellet moderating DBD reactors and this surface contribution can be taken as “catalysis” (we briefly discuss this point in our previous paper, Ref. [11].). However, this catalysis is not equivalent to that occurring onto a conventional catalyst surface induced thermally. Nevertheless, since the “catalysis” (i.e., surface related) aspects of the process are only incidentally dealt with in the paper we agree in removing the “catalysis” term in the title.
2． The definition of RE is not clear. I have read the text several times but I cannot understand why RE can be useful parameter for evaluating the process performance. For example, what is the rationale for the first sentence in page 1 of 11?. What about the following cases?
Newly formed NH3 à experience exchange reaction to form NDH2
What do you say for this case based on RE?
We try to explain better the RE concept by illustrating it with the example of NH3, NDH2, ND2H and ND3. We hope that the concept is clearer in the present version of the paper. Please read again sections 2.2, 2.3 and 2.4, where several changes have been done.
The elementary reactions of plasma are quite complicated, and I think that the method suggested by the authors too much simplified. If the merit of the method is not guaranteed, then the rest of discussions (Figs 3 and 4) may loose its importance.
The method is a simplification that, as indicated in the text, underestimates the effect of intermediate processes in the overall reaction. Nevertheless, it has the virtue of providing semiquantitative information about the occurrence of intermediate processes and, most importantly, its simplicity and empirical basis. Accurate descriptions of plasma processes are usually achieved through the set-up of plasma models where all intermediate reactions and their relative importance and kinetics are incorporated. Obviously, the isotope labeling technique cannot compete in accuracy with such model descriptions, but can provide some preliminary empirically-based data accounting for intermediate process reactions. See page 6/12-main text, line 231-236.
3. The Lissajous figure in Figure 5 is quite different from the normal parallelogram. The authors need to explain this in more detailed way.
This is true. The particular shape of these Lissajous curves is due to the particular squared shape of the voltage curve and due to the packed bed reactor configuration. We clarify this point in the new version of the paper.
Round 2
Reviewer 3 Report
The reviewer still cannot understand the importance of RE as an evaluation parameter for energy efficiency in plasma chemistry. The reply from the author is still not enough to grasp the merits of RE. It may stand for the “degree of exchange reaction” rather than the degree of unnecessary reactions. Furthermore, the isotope molecules are known to have different reaction rate (see Journal of chemical phys, V124, 104303, 2006 for ozone synthesis reaction). In other words, the reaction data cannot be the same between isotope molecules.
Another points is that, isotope-exchange reaction has less to do with NH3 synthesis or CH4 reaction.
Considering these points, the proposed method seem to be less important as evaluation method. Perhaps it will be better to change the focus like “study of elementary reactions using isotope molecules” for example.
The reply to the Lissajous figure is also not clear. For example, temperature change in the plasma reactor is known to change the shape of Lissaous figure (for example, J phys Chem C, V111, 5090-5095, 2007). Did you see any significant temperature change with the plasma turned on?
Titles in the reference are mixed with upper and lower cases.
Ref 18 has no information of Journal, volume, pages, and year.
Refs 19-20, pages are missing.
Author Response
The reviewer still cannot understand the importance of RE as an evaluation parameter for energy efficiency in plasma chemistry. The reply from the author is still not enough to grasp the merits of RE. It may stand for the “degree of exchange reaction” rather than the degree of unnecessary reactions. Furthermore, the isotope molecules are known to have different reaction rate (see Journal of chemical phys, V124, 104303, 2006 for ozone synthesis reaction). In other words, the reaction data cannot be the same between isotope molecules.
We partially agree with these referee remarks and have included the expression “degree of exchange reaction” instead of the previous ones utilized in the paper. However, although we are aware of differences in reaction rate depending on the isotope present in a given molecule (we cite reference suggested by the referee) we think that these differences are second order in the context of the processes taken place in a DBD atmospheric pressure plasma. It is well known that high AC voltages at relatively low frequency (i.e. conditions utilized in DBD discharges) produce a non-equilibrium plasma with very high electron temperatures and high vibrational temperatures, while rotational and translational temperatures are relatively low (A. Fridman, Plasma Chemistry. Cambridge: Cambridge University Press, 2008). In other words, most energy is directly applied to directly break chemical bonds. Under these conditions, the little difference in bond energy existing because of substituting H by D in a chemical bond must be negligible regarding the effect of plasma on the excitation processes of molecules. We comment briefly about this point in the paper citing the reference mentioned by the referee. See page 6/13-main text, line 234-240.
Another points is that, isotope-exchange reaction has less to do with NH3 synthesis or CH4reaction.
In both cases isotope exchange reactions are taking place. Also in both cases, these reaction processes contribute to decrease the energy yield of the overall reaction and to decrease the conversion yield. The difference between both type of reactions is that isotope exchange increases rather linear with the applied power in the case of the ternary mixture N2/D2/NH3, while it does not progress so efficiently in the case of the reforming reaction. We try to make it clearer in the new version of the paper introducing this point in the conclusions of the paper. See page 11/13-main text, line 361-366.
Considering these points, the proposed method seem to be less important as evaluation method. Perhaps it will be better to change the focus like “study of elementary reactions using isotope molecules” for example.
We do not agree with this point. We think that isotope exchange, practically never used in the context of plasma, has actually much potential to analyze reaction mechanisms, particularly to assess the energy efficiency of plasma reactions. We are not studying elementary reactions like in the paper suggested by this referee, but a complete plasma process which, involving a large series of elementary reactions, induces many different isotope exchange reactions. We try to make this point clear and stress it in the conclusions. See page 10/13-main text, line 349-351.
The reply to the Lissajous figure is also not clear. For example, temperature change in the plasma reactor is known to change the shape of Lissajous figure (for example, J phys Chem C, V111, 5090-5095, 2007). Did you see any significant temperature change with the plasma turned on?
We are aware of the importance of temperature on the efficiency of DBD reactions (i.e., detected through possible changes of Lissajous curves) and therefore all reported measurements have been taken at steady state conditions (i.e. constant temperature, which after stabilization reached approximately 60º C in our case) in the case of the ternary mixture. When we studied the reforming we even heat the reactor at 130ºC to avoid water condensation and, indirectly, provide constant temperature conditions. The reference indicated by the referee compares Lissajous curves taken at 100 ºC and 500ºC. In our case, within a given experiment with the ternary mixture, changes in temperature were smaller than 10ºC and therefore with a negligible effect on efficiency. Nevertheless, we comment about this point in the new version of the paper and cite this reference. See page 9/13-main text, line 313-315.
Titles in the reference are mixed with upper and lower cases.
Ref 18 has no information of Journal, volume, pages, and year.
Refs 19-20, pages are missing.
We thank the referee for drawing our attention to these misspelling that have been now corrected.
Round 3
Reviewer 3 Report
I would like to comment two more points (basically the same things that I have raised previously).
The meaning of RE in eqs 3 and5 is closer to “relative concentration” than the “number of events”.
I strongly suggest the authors to reconsider the “wording” for the definition or explanation.
Basically the reaction with N2/D2/NH3 system is about balance between forward and reverse reactions. It is very natural to see the increase of various products from N2/D2/NH3 as increasing input power.
Instead, the CH4/D2O system is about forward reaction. It is well-known that the reaction efficiency generally become saturated as input power is increased. In other word, the patterns of Fig 4 is quite common just like the first-order reactions in any kind of reactors (bioreactor, photocatalyst, plasma reactor etc.). Again, the discussion of RE on lines 282-284 is to simplified.
For example, the following reactions may take similar patterns to Fig 4 (at lease for wide range of input power).
: NOx removal efficiency vs input power (or specific energy),
VOC removal vs input power
CO2 conversion vs input power,
Ozone synthesis vs input power
Author Response
(x) I would not like to sign my review report
( ) I would like to sign my review report
English language and style
( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style
Yes | Can be improved | Must be improved | Not applicable | |
Does the introduction provide sufficient background and include all relevant references? | ( ) | ( ) | ( ) | (x) |
Is the research design appropriate? | ( ) | ( ) | (x) | ( ) |
Are the methods adequately described? | ( ) | ( ) | (x) | ( ) |
Are the results clearly presented? | ( ) | ( ) | (x) | ( ) |
Are the conclusions supported by the results? | ( ) | ( ) | (x) | ( ) |
Comments and Suggestions for Authors
I would like to comment two more points (basically the same things that I have raised previously).
The meaning of RE in eqs 3 and 5 is closer to “relative concentration” than the “number of events”.
We used the term “number of events” instead of “number of intermediate reactions” because we are aware that using the isotope labeling technique we do not “count” all these intermediate reactions. We specify that the estimations made based on the RE concept are semiquantitative, Yet, no other reported simple method provides a simple way to get a direct assessment of the intermediate processes that are inefficient to get the products of the plasma reaction.
It is true that the RE value is calculated from the “relative concentration” of the molecules that have undergone isotope exchange. However, it goes a step forward as it incorporates the number of times that exchanges have taken place and the variation in flow between inlet and outlet in the case that the reaction entails a change in the number of molecules.
In the new version we try to clearly differentiate between “intermediate reactions” and REs and, more specifically, stress the limitations, assumptions and possibilities by the use of REs. Please see carefully changes along section 2.2, specially in lines 232-242.
I strongly suggest the authors to reconsider the “wording” for the definition or explanation.
We really do not know what alternative name use. Since the definition according to eq. (3) and (5) is quite clear and offer no doubts, we think that is better to keep the same wording. Nevertheless, in this new version and even risking redundancy, we have tried to clarify these concepts one more time. Please see carefully changes along section 2.2.
Basically the reaction with N2/D2/NH3 system is about balance between forward and reverse reactions. It is very natural to see the increase of various products from N2/D2/NH3 as increasing input power.
According to the definition used in the paper (similar to the current definition in reaction synthesis), products are the stable molecules obtained from the stable reactants molecules (i.e. N2, H2). In our case, we do not directly explore the process “reactant” to “product” transformation but how and into which extent “product” molecules (i.e., NH3) experience isotope exchange processes. From the estimation of the exchange events (e.g., RE and TRE) we try to figure out the possibilities of using plasma DBD technique for the chemical synthesis of products. We try to clarify all these concepts with a new writing of the text. Please see carefully changes along section 2.2.
Instead, the CH4/D2O system is about forward reaction. It is well-known that the reaction efficiency generally become saturated as input power is increased. In other word, the patterns of Fig 4 is quite common just like the first-order reactions in any kind of reactors (bioreactor, photocatalyst, plasma reactor etc.). Again, the discussion of RE on lines 282-284 is to simplified. For example, the following reactions may take similar patterns to Fig 4 (at lease for wide range of input power):
NOx removal efficiency vs input power (or specific energy),
VOC removal vs input power
CO2 conversion vs input power,
Ozone synthesis vs input power
We have tried to clarify the meaning of the sentences in these lines. We must also point out that the observed saturation in Figure 4 refers to the isotopic exchange estimated from the evolution of the TRE values as a function of power and not to the overall reaction yield (i.e., transformation of reactants into products). We point out that precisely this saturation in TRE opens the way to increase the reaction yield by increasing power. A new writing of the text tries to clarify these ideas. Please see carefully changes along section 2.4. (lines 288-293).
Author Response File: Author Response.pdf
