Infrared Thermography as an Operando Tool for the Analysis of Catalytic Processes: How to Use it?
Round 1
Reviewer 1 Report
The authors report their experiences with the design of a chemical reaction chamber equipped by infrared thermography. It enables the detection of spatially resolved temperature distributions in front of a catalyst surface related to chemical reaction intensities. However, the readability suffers somewhat from the previous publication including a series of thermograms and thus demonstrating the principle. The reviewer recommends strongly to include some of these (or other) results to make the concept more clearly. Another critical point is the use of the term “temperature”. It is obviously, and also addressed by the authors, that the IR camera cannot really measure temperatures. The emissivity of the target surface, the temperature of the window and perhaps the hot gases itself influence the temperature measurement. This is not a real problem when only spatial distributions at constant surfaces are of interest. But when real temperatures are needed (to compare with simulations or known enthalpies) or changing surface are observed these issues must be treated carefully. The reviewer think it is worth to publish the manuscript after some revisions. Please regard also the following comments:
Table 1: please note there is no physical reason to limit the application range of bolometer cameras.
Some of these cameras offer temperature ranges upto 1000°C or higher. Using your reference 27 I found the VarioCAM HD research 900suited for temperatures until 2000°C. But the camera must be specified for these high intensities. Since bolometer detector arrays are often cheaper than the semiconductor detectors you should mention this.
The same is also true for the InSb detectors which can be operating in high temperature ranges too but require also some apertures or damping filters.
- 5 line 150: must be Thicker
line 155: please indicate that T stands here for thickness (usually it is temperature)
p.7 table 2: the German Wikipedia describes fused silica (Quarzglas) with transmission upto 5 µm, please check your references, the transmission values must be related to a thickness and a wavelength range
p.8 line 222: The transmission issue has a second aspect which is not regarded so far. A transparency of 60% means you will measure the temperature of the windows with 40% intensity. If the window temperature or temperature transient is changing (it will, because only 60% transmission cause 40% absorption from the inside thermal radiation) you must measure and compensate this. You wrote very imprecisely “challenging the modeling of the heat transfer phenomena”. Please discuss this issue more correctly.
p.9 figure5: it is not really clear, what do you want to show, the different P might be different heating powers?, related to figure 5 right I assume the reaction is going on, thus a constant temperature is expected? And what is this transient hot spot P13? If you measured at different places (P for places?) of your reacting area you should provide a suited thermogram showing these places. The figure caption should describe also what is shown in Figure 5b (you wrote this later in the text, but figures should be readable with their cation or you give an explicit hint to the text)
p.11 l.323: you discussed here figure 6 with simulation results and mentioned again this hot spot with no relation to the actual figure (see comment before)
l.343: decision tree?
p.12 l. 355 ff. the discussion of the hot spot is completely unclear for me at this point, after reading the complete manuscript I suspect you mean the hot zone running over the surface (see my general comment).
p.13 figure 8: figure caption are too short or even wrong
IR spectra are completely different things than temperature fields, thermograms require a temperature scale, how are thermography results integrated in 8.3?
p.14 l. 423 what means measured emissivity? If you compare with the temperature measured inside the window influence is also included in this value, the real emissivity of the catalyst surface might be higher.
p.15 l.440 ff. I don’t know the other studies, but I want to remark that temperature measurements at a changing surface (from oxidized to metallic) are critical because the emissivity is also changing during the surface modification, especially in case of metallic surfaces. Thus, reporting temperatures would require specific calibrations for at least some steps of the entire oxidation process.
The same influence must be considered also in the experiment described later on page 17, where you discussed the occurrence of a new oxidation of the surface. Thus, the presented temperature profiles could also represent emissivity profiles at constant temperature or a mixture of both. However, the comparison with a real temperature measurement at the different positions allows a separation of these effects.
Comments for author File: 
Comments.pdf
Author Response
Reviewer #1
The authors report their experiences with the design of a chemical reaction chamber equipped by infrared thermography. It enables the detection of spatially resolved temperature distributions in front of a catalyst surface related to chemical reaction intensities. However, the readability suffers somewhat from the previous publication including a series of thermograms and thus demonstrating the principle. The reviewer recommends strongly to include some of these (or other) results to make the concept more clearly.
We thank the reviewer for the suggestion. We agree that the readability of the manuscript may significantly improve by adding the thermographs explained in section 4 (case studies). Therefore, we added the thermographs as supplementary information (2 videos) and we better addressed the connection between the thermographs and the explanations in section 4 (page 15, line 455 and page 17 line 523).
Another critical point is the use of the term “temperature”. It is obviously, and also addressed by the authors, that the IR camera cannot really measure temperatures. The emissivity of the target surface, the temperature of the window and perhaps the hot gases itself influence the temperature measurement. This is not a real problem when only spatial distributions at constant surfaces are of interest. But when real temperatures are needed (to compare with simulations or known enthalpies) or changing surface are observed these issues must be treated carefully.
We thank the reviewer also for the attention posed to this detail. As the scope of the paper is to define some guidelines for the use of IR thermography in catalytic applications, we specified this better in the manuscript. This was addressed directly in lines 283-286.
The reviewer think it is worth to publish the manuscript after some revisions. Please regard also the following comments:
Table 1: please note there is no physical reason to limit the application range of bolometer cameras.
We agree on this point and we modified table 1 accordingly
Some of these cameras offer temperature ranges upto 1000°C or higher. Using your reference 27 I found the VarioCAM HD research 900suited for temperatures until 2000°C. But the camera must be specified for these high intensities. Since bolometer detector arrays are often cheaper than the semiconductor detectors you should mention this.
We thank the reviewer for this observation, which we introduced in the manuscript at lines 105-107.
The same is also true for the InSb detectors which can be operating in high temperature ranges too but require also some apertures or damping filters.
This is also a useful observation, which we included in the manuscript at lines 117-119.
- 5 line 150: must be Thicker
This has been now corrected.
line 155: please indicate that T stands here for thickness (usually it is temperature)
We agree that this notation can be confusiong. Therefore we changed the indication of thinkness with t (equation 3) and specified the notation in the text (line 155).
p.7 table 2: the German Wikipedia describes fused silica (Quarzglas) with transmission upto 5 µm, please check your references, the transmission values must be related to a thickness and a wavelength range
We thank the reviewer for noticing this inconsistence. In fact this was a mistake referring to quartz crystal instead of fused silica. The transmission range has been corrected.
p.8 line 222: The transmission issue has a second aspect which is not regarded so far. A transparency of 60% means you will measure the temperature of the windows with 40% intensity. If the window temperature or temperature transient is changing (it will, because only 60% transmission cause 40% absorption from the inside thermal radiation) you must measure and compensate this. You wrote very imprecisely “challenging the modeling of the heat transfer phenomena”. Please discuss this issue more correctly.
We agree that this can be an issue in the measurement of IR spectra. However, it is difficult to draw a general guideline to tackle this challenge, as every specific application is affected by the specific parameters of the system (e.g. heat transfer pattern, reaction enthalpy etc.). Therefore, we modified the sentence to be less generic and to make the reader aware of this possible problem. (lines 229-231)
p.9 figure5: it is not really clear, what do you want to show, the different P might be different heating powers?, related to figure 5 right I assume the reaction is going on, thus a constant temperature is expected? And what is this transient hot spot P13? If you measured at different places (P for places?) of your reacting area you should provide a suited thermogram showing these places. The figure caption should describe also what is shown in Figure 5b (you wrote this later in the text, but figures should be readable with their cation or you give an explicit hint to the text)
We are sorry that figure 5 generates so much incertitude in the interpretation. The intention is to show the effect of inconstant heating (due to the oscillation in the supplied power) to the experimental results. For this reason, we show both the steady state temperature profile and the transient temperature profile of the activation of the Sabatier reaction. P refers to the various measurement points. The points are now made evident in the new panel of figure 5. We added a better explanation in the caption of figure 5. Additionally, we clearly added this explanation in the text (lines 211-216). We thank the reviewer for pointing out the absence of the description of figure 5b, we now added this to the figure caption.
p.11 l.323: you discussed here figure 6 with simulation results and mentioned again this hot spot with no relation to the actual figure (see comment before)
We are sorry for the misunderstanding on the meaning of hotspot. For a better readability of the paper, we added a clear definition of the sense given to the hotspot concept already in the introduction (lines 86-88). This should simplify the understanding of the entire paper. For what concerns figure 6, the reaction hotspot is the predicted higher temperature at the reactor inlet than in the rest of the reactor. This section was made more understandable by a number of new sentences (lines 318-319, 342-343, caption of figure 6, line 355)
l.343: decision tree?
Thank you for noticing this typo, which has been corrected.
p.12 l. 355 ff. the discussion of the hot spot is completely unclear for me at this point, after reading the complete manuscript I suspect you mean the hot zone running over the surface (see my general comment).
We are sorry again for the misunderstanding on this point. We gave a clear definition of the concept of hotspot (see comment above, lines 86-88). The meaning of hotspot is indeed the ‘hot zone running over the surface’ as mentioned by the reviewer. We clarified the meaning of the concept of hotspot also at the beginning of the section ‘Determination of reactor requirements’ (lines 375-376, 380-381, 384-385)
p.13 figure 8: figure caption are too short or even wrong
The caption of figure 8 has been completely revised (lines 421-425).
IR spectra are completely different things than temperature fields, thermograms require a temperature scale, how are thermography results integrated in 8.3?
The thermographs are integrated through the software IRBIS 3.1 PRO provided with the IR camera. In this software, it is possible to insert a complete set of calibration data (i.e. in the entire temperature range considered), which is used as a base for the translation of the IR spectra in temperature profiles. This has been better specified in the manuscript, at lines 447-448.
p.14 l. 423 what means measured emissivity? If you compare with the temperature measured inside the window influence is also included in this value, the real emissivity of the catalyst surface might be higher.
This is a good observation and we thank the reviewer for pointing it out. In this specific case, we always used the same experimental setup, so we used the value of emissivity of the catalyst with the window influence included. This does not create any practical problem in the specific case, but it must be clearly stated in the manuscript. Therefore, we added a couple of sentences to address this point at lines 450-452.
p.15 l.440 ff. I don’t know the other studies, but I want to remark that temperature measurements at a changing surface (from oxidized to metallic) are critical because the emissivity is also changing during the surface modification, especially in case of metallic surfaces. Thus, reporting temperatures would require specific calibrations for at least some steps of the entire oxidation process.
We agree that the emissivity of the catalyst can vary with the oxidation state. However, we used a 0.5% or 2% Ru/Al2O3 catalyst, therefore the amount of active metal whose oxidation state is changed is limited. Hence, we exclude that the total emissivity can vary substantially during the reaction (most of the material is unchanged during the reaction). Additionally, we re-calibrated the system with the completely reduced catalyst and we recorded the same emissivity value. For clarity, this discussion is now present in the manuscript (lines 489-492).
The same influence must be considered also in the experiment described later on page 17, where you discussed the occurrence of a new oxidation of the surface. Thus, the presented temperature profiles could also represent emissivity profiles at constant temperature or a mixture of both. However, the comparison with a real temperature measurement at the different positions allows a separation of these effects.
We thank the reviewer for the comment. Indeed, the emissivity measured before and after the experiments is identical, therefore ruling out the effect of emissivity change (see also comment above).
Reviewer 2 Report
The manuscript written by R. Mutschler and E. Moioli describes an interesting study on infrared thermography experiment design principles. The provided information is logical and illustrative. The significance of presented issues is quite high taking into account prospective correct results of such experiments and can be treated as semi-model procedure of the IR thermography experiment design. However, the manuscript itself has some typos as well as a question must be answered before publication. Therefore I recommend a minor revision before acceptance.
- Why the SF (safety factor) was equal to 4? Some instructions/explanation of the SF selection should be provided (line 156).
Minor corrections:
Line 114: antimonide instead of antimonite
Line 130: “and” in the Figure caption should be removed
Table 1: Rows in the Table 1 are not clear enough, the assignment of specific values to particular Detector type is not straightforward.
Line 150: Thicker instead of Ticker
Line 156: Symbol “T” for thickness in eq. (3) should be also explained here.
Line 343: tree instead of three
Line 421: The sentence “Please…” should begin with a capital letter.
Figure 12 caption: subscripts should be checked.
Author Response
Reviewer #2:
The manuscript written by R. Mutschler and E. Moioli describes an interesting study on infrared thermography experiment design principles. The provided information is logical and illustrative. The significance of presented issues is quite high taking into account prospective correct results of such experiments and can be treated as semi-model procedure of the IR thermography experiment design. However, the manuscript itself has some typos as well as a question must be answered before publication. Therefore I recommend a minor revision before acceptance.
We thank the reviewer for the appreciation of our manuscript.
Why the SF (safety factor) was equal to 4? Some instructions/explanation of the SF selection should be provided (line 156).
The SF is set equal to 4 to avoid rupture in all the foreseen practical situations during the experiments. However, this choice is inevitably arbitrary and can be changed anytime according to different considerations. Therefore, we better specified this point at line 163-165.
Minor corrections:
We thank the reviewer for the attention to these minor issues to be addressed.
Line 114: antimonide instead of antimonite
Corrected
Line 130: “and” in the Figure caption should be removed
And removed
Table 1: Rows in the Table 1 are not clear enough, the assignment of specific values to particular Detector type is not straightforward.
The intervals are reduced to one single value per detector type
Line 150: Thicker instead of Ticker
Ticker corrected
Line 156: Symbol “T” for thickness in eq. (3) should be also explained here.
We added the explanation of the symbol, modified in t
Line 343: tree instead of three
Tree corrected
Line 421: The sentence “Please…” should begin with a capital letter.
The sentence at line 421 is present by mistake. It has been removed.
Figure 12 caption: subscripts should be checked.
Subscripts are now correct
Reviewer 3 Report
The authors describe the advantages of IR thermography for the in situ investigation of heat-generating catalytic reactions. This work could be of interest for a specialized audience.
I would suggest to add, as a decisional point to the flowchart, "is enough heat produced by the reaction(s) of interest?"
Line 139: “As the target of thermography experiments is the investigation of the IR emission of a catalyst UNDER OPERATION”.
Line 146: not «absorbant» but «absorbing».
Line 150: correct «Ticker» as «Thicker».
Table 2: correct «GaAS» as «GaAs».
Figure 5a: what is the origin of the oscillations? Thermal noise or actual chemical reactions? If it’s noise, how to take that into account?
Line 332: the phrase is incomplete.
Figure 10b: does the green line really refer to He?
Supplementary materials: a “Video S1” is mentioned, which is however not available.
Author Response
Reviewer #3:
The authors describe the advantages of IR thermography for the in situ investigation of heat-generating catalytic reactions. This work could be of interest for a specialized audience.
We thank the reviewer for the appreciation of our manuscript.
I would suggest to add, as a decisional point to the flowchart, "is enough heat produced by the reaction(s) of interest?"
We thank the reviewer for the suggestion. We added the decisional point accordingly.
Line 139: “As the target of thermography experiments is the investigation of the IR emission of a catalyst UNDER OPERATION”.
Added at line 145.
Line 146: not «absorbant» but «absorbing».
Corrected
Line 150: correct «Ticker» as «Thicker».
Corrected
Table 2: correct «GaAS» as «GaAs».
Corrected
Figure 5a: what is the origin of the oscillations? Thermal noise or actual chemical reactions? If it’s noise, how to take that into account?
The oscillations are unfortunately due to noise originated by the controller of the heater. In order to take that into account, we have to be sure that the phenomenon of interest is well visible despite the noise (sufficiently high signal to noise ratio). This is now explained at line 225.
Line 332: the phrase is incomplete.
This is a problem of non-optimal conversion of the manuscript, the end of the sentence is present below figure 6.
Figure 10b: does the green line really refer to He?
Yes, the green line refers to He that is progressively depleted in the reactor. It is an indicator of the residence time distribution in the reaction.
Supplementary materials: a “Video S1” is mentioned, which is however not available.
This is also a problem in the conversion of the manuscript, as the sentence on supplementary materials was automatically added, despite no videos were provided. However, in the revised version two videos were provided (S1 and S2).
Round 2
Reviewer 1 Report
Things are now much clearer, but some details could be improved, mainly phrases:
l.215 ff: I am not a native speaker, but I associate a “hot spot” with a clearly defined point shaped area and an even temperature. Larger hot regions are regions or areas. I can only say what is usual in the field of thermography, where hot spot are typically a result of heating with a laser spot. Later, you explained the term (in l.376), you could shift this above. However, using this term for a larger region with temperature gradients is completely uncommon.
Figure 5 contains in fact 3 parts. The upper image is a thermogram (where is temperature scale is missed so far), which was recorded at a certain time, thus you should provide this time (probably it is early or the stationary regime). The curves in Figure 5 a) are temperature transients, profiles is usually used for spatial distribution (as 6a or 11a). And the numbers are very small in the thermogram and hard to recognize, please think about a proper design for this image (because you don’t want to cover the thermogram with large descriptors)
- 343: … first third of the reaction chamber
- 344: background is usually the background behind and around an object, I suspect you mean the other part of the reaction plate?
- 381: I suspect you mean the temperature difference or temperature increase, the term “extension of the hotspot” describes more a lateral size
Figure 5: the term “must be identified” suggests that the points must be found within a data set, must be defined is more suited, 8.2 shows no IR spectra but thermograms which require also a temperature scale. I note that the temperature scale in the hot zone is probably saturated. You could select your temperature scale in such a way, if you want to highlight the reaction zone, but you should mention it. Otherwise one assumes that the temperature within the reaction zone is very even (perhaps it is? I don’t know). Numbers and letters are again very small, please make it readable or skip it.
- 443 ff.: the camera calibration includes a specific setup and will be provided by the camera supplier usually. But this is valid only for emissivity=1 and no transmission losses between camera and observed surface. What you did is to estimate an effective or apparent emissivity, which allows you to determine the real surface temperature. Please rephrase this part to reach a correct description.
Fig. 11 b) caption: see my comments before about terms transients and profiles, perhaps you could use flow profile, time profile, timeline, time course …
Author Response
Reviewer #1:
Things are now much clearer, but some details could be improved, mainly phrases:
l.215 ff: I am not a native speaker, but I associate a “hot spot” with a clearly defined point shaped area and an even temperature. Larger hot regions are regions or areas. I can only say what is usual in the field of thermography, where hot spot are typically a result of heating with a laser spot. Later, you explained the term (in l.376), you could shift this above. However, using this term for a larger region with temperature gradients is completely uncommon.
We thank the reviewer for the observation and sharing the experience developed in the field of thermography. This is then a mismatch between the language used in thermography and in chemical reaction engineering. In this latter field, the term hotspot defines the region of the reactor where the temperature is higher than the target value, due to the heat released from the reaction. We now understand the source of the misunderstanding and we added a sentence explaining this concept at lines 220-221. We hope in this way to avoid misunderstanding with the readership of the paper. Additionally, the concept of hotspot is clearly defined in the introduction, lines 86-88
Figure 5 contains in fact 3 parts. The upper image is a thermogram (where is temperature scale is missed so far), which was recorded at a certain time, thus you should provide this time (probably it is early or the stationary regime). The curves in Figure 5 a) are temperature transients, profiles is usually used for spatial distribution (as 6a or 11a). And the numbers are very small in the thermogram and hard to recognize, please think about a proper design for this image (because you don’t want to cover the thermogram with large descriptors)
We thank the reviewer for the suggestions to improve figure 5. We added the temperature scale and the indication of the time to the figure 5a. We significantly increased the numbers on the thermograph and on the corresponding figure 5b.
- 343: … first third of the reaction chamber
This was corrected
- 344: background is usually the background behind and around an object, I suspect you mean the other part of the reaction plate?
We agree that this may be confusing and we changed to “ higher than the baseline of the external heater”
- 381: I suspect you mean the temperature difference or temperature increase, the term “extension of the hotspot” describes more a lateral size
Modified to “difference between the set and the measured temperature” (line 385-386)
Figure 8: the term “must be identified” suggests that the points must be found within a data set, must be defined is more suited, 8.2 shows no IR spectra but thermograms which require also a temperature scale. I note that the temperature scale in the hot zone is probably saturated. You could select your temperature scale in such a way, if you want to highlight the reaction zone, but you should mention it. Otherwise one assumes that the temperature within the reaction zone is very even (perhaps it is? I don’t know). Numbers and letters are again very small, please make it readable or skip it.
“must be identified” is in effect not an appropriate in this case, we modified it as ‘must be selected’. Figure 8 was completely restructured to improve the comprehension and avoid confusion generated by low-quality pictures.
- 443 ff.: the camera calibration includes a specific setup and will be provided by the camera supplier usually. But this is valid only for emissivity=1 and no transmission losses between camera and observed surface. What you did is to estimate an effective or apparent emissivity, which allows you to determine the real surface temperature. Please rephrase this part to reach a correct description.
We thank the reviewer for the suggestion. We rephrased the section at lines 444 ff. to make this concept clearer.
Fig. 11 b) caption: see my comments before about terms transients and profiles, perhaps you could use flow profile, time profile, timeline, time course …
This was modified into “ion current change over time”
Author Response File: Author Response.pdf
Reviewer 3 Report
The manuscript may now be suitable for publication in Catalysts.
Author Response
We thank the reviewer for suggesting our manuscript for publication
