Comparing Proton Transfer Reaction (PTR) and Adduct Ionization Mechanism (AIM) for the Study of Volatile Organic Compounds
, Bianca Bonato 1,*, Valentina Simonetti 1, Silvia Guerra 1, Laura Ravazzolo 2, Gabriela Gjinaj 1, Marco Dadda 1 and Umberto Castiello 1
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis manuscript presents a well-designed and timely comparative study of two direct-injection mass spectrometry (DI-MS) techniques, PTR-MS and the novel AIM-MS, for the fingerprinting of the volatilome of pea plants. The research question is relevant, the experimental setup in a custom phytotron is robust, and the data analysis workflow is generally sound. The study provides valuable insights, particularly highlighting the broader detection capability of the AIM reactor for untargeted analysis. The findings have significant potential for applications in plant phenotyping, chemical ecology, and precision agriculture. However, the manuscript requires revisions to enhance its clarity, methodological rigor, and overall impact before it can be considered for publication. The major points concern the depth of the literature review, the completeness of methodological details, and the optimization of data presentation.
- Revise for concision. For example, in the abstract, break down dense sentences. The sentence "Analyzing changes... crop protection." could be split into two or three shorter sentences for better flow.
- While the introduction adequately explains the importance of plant VOCs and the principles of PTR-MS, the coverage of the AIM-MS technique is somewhat superficial. As a newer technology, a more thorough review of its recent applications, especially in plant science or environmental bio-monitoring, would strengthen the rationale for the study.
- Figure 4 (Van Krevelen Diagrams): The distinction between putative compound classes is a key result. Ensure that the legends, axis labels, and fonts are consistent and of a high resolution across all figures to improve readability. The current figures are functional but could be more polished.
- Perform a thorough check to ensure all figure captions and references follow a consistent style as required by the journal (e.g., use "0.5 L/min" consistently, not "0,5 L/min").
- Explicitly state the versions of the key Python libraries (e.g., pandas, scipy, numpy) used for statistical testing and data normalization. This is crucial for computational reproducibility.
Author Response
1. Revise for concision. For example, in the abstract, break down dense sentences. The sentence "Analyzing changes... crop protection." could be split into two or three shorter sentences for better flow.
R1. We thank the reviewer for the suggestion, the sentence has been splitted.
2. While the introduction adequately explains the importance of plant VOCs and the principles of PTR-MS, the coverage of the AIM-MS technique is somewhat superficial. As a newer technology, a more thorough review of its recent applications, especially in plant science or environmental bio-monitoring, would strengthen the rationale for the study.
R2. Thank you for your valuable comment. We have expanded the section describing the AIM technique by incorporating numerous additions. The explanation of the AIM methodology is now much more detailed, providing a clearer overview of its characteristics. In addition, we have included specific information on the classes of compounds to which AIM is sensitive, based on findings from a previous study that describes the specificity of the AIM-MS techniques. Please refer to p.3 L118-132, p.11 L446-450.
3. Figure 4 (Van Krevelen Diagrams): The distinction between putative compound classes is a key result. Ensure that the legends, axis labels, and fonts are consistent and of a high resolution across all figures to improve readability. The current figures are functional but could be more polished.
R3. Figure 4 has been updated. Please refer to p.5.
4. Perform a thorough check to ensure all figure captions and references follow a consistent style as required by the journal (e.g., use "0.5 L/min" consistently, not "0,5 L/min").
R4. Done, thank you.
5. Explicitly state the versions of the key Python libraries (e.g., pandas, scipy, numpy) used for statistical testing and data normalization. This is crucial for computational reproducibility.
R5. The versions of the key phyton libraries used for processing and statistical analysis has been added to the text. Please refer to p.13 L506 and p.14 L530.
Reviewer 2 Report
Comments and Suggestions for AuthorsI think that the manuscript entitled “Comparing Proton Transfer Reaction (PTR) and Adduct Ionization Mechanism (AIM) for the study of volatile organic compounds” by Avesani et al. (molecules-3949150) is flawed with too many too serious drawbacks to be published.
First of all, in my opinion the classes of compounds claimed to have been detected (amino acids, sugars, lipids, phenols) cannot be regarded as volatile (line 435: “…enabled the detection of a diverse range of putative compounds, including lipids, amino acids, carbohydrates, and phenols, though the specific signals and compound classes varied between the two techniques”). Since these are non-volatile compounds, the question is how could they have been detected by the used techniques.
Secondly, the claimed chemical formulae of the detected m/z signals (supplemental tables 1-4), claimed to correspond to the [M+H]+ or [M]+ ions, often do not make any sense. Since the whole work is based on the detected m/z values it calls into question the whole manuscript. I will describe several examples.
- Supplemental table 1:
No. 159, formula C8H9O3 does not correspond to mass 105 (it is 153).
No. 160, formula C8H10NO3 does not correspond to mass 106 (it is 168).
No 190, formula CH12P3 – there is no compound which would correspond to it.
No 202, formula C7H7S – there is no compound which would correspond to it.
No 415, formula C3H9O9 – there is no compound which would correspond to it.
- Supplemental table 2:
No 608, formula C13H14NO4 – the detected accurate m/z 248,07959 does not match this formula.
No 614, C11H15N5O2 - the detected accurate m/z 249,132721 does not match this formula.
- Supplemental table 3:
No 4, formula C8N3 – there is no compound which would correspond to it.
- Supplemental table 4:
No 21, formula C6H11N4 – there is no compound which would correspond to it.
No 91, formula C17H19 – there is no compound which would correspond to it.
No 98, formula C11H17O5 – there is no compound which would correspond to it.
There are a number of strange claims in the manuscript e.g. that PTR-MS (which is the ionization method) provides adequate resolution (line 85) and enables the separation of the compounds (line 90).
Some of the cited references do not support the claims, e.g., lines 78: “…given that requires rather long analysis time for the chromatographic separation [18].” – reference 18 does not support this claim; line 381: “…wide variety of phenolic compounds, including flavonoids, isoflavonoids, and phenolic acids, which can be converted into VOCs and released into the atmosphere, contributing to the plant's aroma and its interactions with the environment [2,8,49].” – references 2, 8, 49 do not support this claim.
Author Response
1. First of all, in my opinion the classes of compounds claimed to have been detected (amino acids, sugars, lipids, phenols) cannot be regarded as volatile (line 435: “…enabled the detection of a diverse range of putative compounds, including lipids, amino acids, carbohydrates, and phenols, though the specific signals and compound classes varied between the two techniques”). Since these are non-volatile compounds, the question is how could they have been detected by the used techniques.
R1. We thank the reviewer for this valuable comment. We agree that amino acids, sugars, lipids, and phenols are generally considered non-volatile or semi-volatile compounds and are therefore not expected to be directly detected as intact molecules by PTR- or AIM-MS. In our case, these compound classes were putatively assigned based on the detected m/z features and their tentative annotation using available databases (e.g. ChemCalc, Composition finder tool of Tofware Software). We are aware that the gas-phase ions measured here may also represent volatile fragments, derivatives, or thermal/ionization products and that their interpretations are therefore limited to trends in elemental composition rather than definitive structural identifications. We have clarified this in the new version of the manuscript to avoid any misunderstanding making clear that the assignments are tentative and based on putative chemical classes inferred from fragmentation patterns or database matches, not on direct detection of the compounds. See p.8 L299-301; p.13 L511-513.
2. Secondly, the claimed chemical formulae of the detected m/z signals (supplemental tables 1-4), claimed to correspond to the [M+H]+or [M]+ ions, often do not make any sense. Since the whole work is based on the detected m/z values it calls into question the whole manuscript. I will describe several examples.
Supplemental table 1:
No. 159, formula C8H9O3 does not correspond to mass 105 (it is 153).
No. 160, formula C8H10NO3 does not correspond to mass 106 (it is 168).
No 190, formula CH12P3 – there is no compound which would correspond to it.
No 202, formula C7H7S – there is no compound which would correspond to it.
No 415, formula C3H9O9 – there is no compound which would correspond to it.
Supplemental table 2:
No 608, formula C13H14NO4 – the detected accurate m/z 248,07959 does not match this formula.
No 614, C11H15N5O2 - the detected accurate m/z 249,132721 does not match this formula.
Supplemental table 3:
No 4, formula C8N3 – there is no compound which would correspond to it.
Supplemental table 4:
No 21, formula C6H11N4 – there is no compound which would correspond to it.
No 91, formula C17H19 – there is no compound which would correspond to it.
No 98, formula C11H17O5 – there is no compound which would correspond to it.
R2. We thank the reviewer for the careful examination of the chemical formulae reported in the supplemental tables.
We acknowledge that, in a few cases (e.g., C8H9O3 and C8H10NO3), incorrect annotations were the result of human error, and these have now been corrected in the revised version. We apologize for this oversight.
Regarding other m/z signals for which the reviewer notes that no corresponding known compounds seem to exist, we would like to clarify that the large amount of data acquired required the use of automated putative formula assignment software (ChemCalc and Composition finder tool of Tofware software). This algorithm proposes possible elemental compositions solely based on the exact m/z values and ionization mode, and therefore some improbable formulae may occasionally appear. The main objective of the study was to compare the mass spectral richness and distribution of detected ions across techniques, not to identify each compound. The Van Krevelen analysis and interpretation are based on the overall elemental composition trends inferred from the software output and used solely for high-level visualization of chemical space. To avoid misunderstanding, we have now further clarified in the manuscript that formula assignments are putative and software-generated and that the study does not claim structural identification at the compound level. We believe it is appropriate to retain these putative assignments in the supplementary tables, as they reflect the complete dataset obtained during our analyses. Importantly, these few uncertain annotations do not affect the validity of the manuscript’s conclusions. In the main text and discussion, we only considered and interpreted compounds that were statistically significant and could be reliably associated with specific chemical classes, as supported by existing literature.
3. There are a number of strange claims in the manuscript e.g. that PTR-MS (which is the ionization method) provides adequate resolution (line 85) and enables the separation of the compounds (line 90).
R3. We thank the reviewer for this observation and agree that the wording in the original manuscript could be misleading. PTR-MS itself does not provide chromatographic separation or mass resolution. Our intended meaning was that PTR-MS, when used with a high-resolution mass spectrometer, could provide real-time quantification of VOCs in complex mixtures. The text has now been revised to better explain the aspect related to mass resolution and avoid implying that PTR-MS performs chromatographic separation. Please refer to p.2 L84-96.
4. Some of the cited references do not support the claims, e.g., lines 78: “…given that requires rather long analysis time for the chromatographic separation [18].” – reference 18 does not support this claim; line 381: “…wide variety of phenolic compounds, including flavonoids, isoflavonoids, and phenolic acids, which can be converted into VOCs and released into the atmosphere, contributing to the plant's aroma and its interactions with the environment [2,8,49].” – references 2, 8, 49 do not support this claim.
R4. We do not agree with the reviewer.
- About the sentence “given that requires rather long analysis time for the chromatographic separation [18]”: quoting from reference [18] when talking about the coupling of MS with GC/LC/CE: “These approaches require rather long analysis time for the chromatographic separation (typically 10–60 min)”
- About the sentence “…wide variety of phenolic compounds, including flavonoids, isoflavonoids, and phenolic acids, which can be converted into VOCs and released into the atmosphere, contributing to the plant's aroma and its interactions with the environment [2,8,49].”: references [2] and [8] have an extensive chapter describing, in general, the synthesis of plant VOCs and the way they are release and interacts with the environment (e.g. “successful pollination/fertilization triggers a decrease in overall scent emission mediated by the phytohormone ethylene”) ; reference[49] has an entire section on Volatile Compounds Responsible for Aromatic Features in medicinal and aromatic plants including phenolic compounds.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe paper titled "Comparing Proton Transfer Reaction (PTR) and Adduct Ionisation Mechanism (AIM) for the Study of Volatile Organic Compounds" gives a comparative analysis of two DI-MS methods—Proton Transfer Reaction (PTR-MS) and Adduct Ionisation Mechanism (AIM-MS)—for the study of volatile organic compounds (VOCs) released by pea plants. The study is practically comprehensive, well-arranged, and aptly timed, given the recent advent of AIM-MS as a technique for real-time monitoring of VOCs. Nevertheless, certain technical, statistical, and interpretational shortcomings restrict the solidity of its findings. The work presents new information on VOC fingerprinting and volatilomics in plants but needs further validation, replication, and quantitative comparison prior to acceptance for publication.
Strengths of the Paper
To me, this appears to be the initial comparative study utilizing AIM-MS in plant volatilomics, responding to a genuine analytical void.
The authors have carried out thorough parameter analysis and exercised strict analytical control, such as baseline correction.
They have utilized an open workflow with Tofware and Python normalisation.
The results have been correlated with findings and ion chemistry in contemporary literature.
This research may form a basis for transportable AIM-MS applications in precision agriculture.
Major Comments
Please indicate the number of independent replicates since it is not clear from the text.
PTR and AIM are carried out under varying physical and chemical conditions; the outcomes are qualitative, therefore it cannot be stated if the outcomes depict equivalency or not.
The detected compounds should have been confirmed by a well-established method such as GC–MS or MS/MS for full confirmation.
Mann–Whitney U tests are not sufficient for time-series data; employ PCA or PLS-DA with replicate clarification.
Assertions of AIM sensitivity and range should be quantitatively verified or reformulated as hypotheses.
The authors mentioned that these techniques are portable and could be taken out for off-site analysis, but still, it needs power, and the size from the figure seems big for being a portable instrument.
Minor Comments
Please correct decimal separators (0.5 L/min) and eliminate placeholder text.
Enhance the clarity of figures; add a comparative summary figure (PTR vs AIM).
Simplify long sentences and correct typographical errors.
When referring, provide uniform formatting and add other recent studies on VOC validation.
Author Response
Major Comments
Please indicate the number of independent replicates since it is not clear from the text.
PTR and AIM are carried out under varying physical and chemical conditions; the outcomes are qualitative, therefore it cannot be stated if the outcomes depict equivalency or not.
The detected compounds should have been confirmed by a well-established method such as GC–MS or MS/MS for full confirmation.
R1. The number of independent replicates are now reported at p.9 L362-364. Regarding the qualitative nature of PTR-MS and AIM-MS: the goal of this study was not to perform compound identification or quantification, but to compare their ability to explore chemical space (m/z features) in an untargeted screening context. In the section “Data processing, statistical analysis, and annotation” we now further state that our outputs reflect differences in compounds populations, not absolute concentrations or identities. Please refer to p.8 L324-326. Concerning compound confirmation via GC–MS or MS/MS: while we agree that GC–MS or MS/MS would be required for structural identification, the scope of this work was to evaluate method performance and chemical coverage rather than to identify compounds. This is the reason why we used the term “putative” to describe the detected compounds. Nonetheless, we recognize this as a limitation and indeed we have mention this on the conclusions. Please refer to p.14.
Mann–Whitney U tests are not sufficient for time-series data; employ PCA or PLS-DA with replicate clarification.
R2. We thank the reviewer for the suggestion; however, we think that our approach is robust. We used Mann-Whitney U test to compare data of each growth chamber from different days. In this way we performed a comparison of ion feature distributions between time points. The main conclusions of the study are based on overall chemical space trends, Van Krevelen analysis, and feature comparisons across techniques.
- Assertions of AIM sensitivity and range should be quantitatively verified or reformulated as hypotheses.
R3. The performance of AIM-MS in this study was evaluated based on the observed detection of VOC features. While this provides an indication of the types of compounds detectable under the applied conditions, it does not constitute a quantitative assessment of sensitivity.
- The authors mentioned that these techniques are portable and could be taken out for off-site analysis, but still, it needs power, and the size from the figure seems big for being a portable instrument.
R4. We thank the reviewer for this comment. The sentence in question was intended to highlight the potential for future applications using portable AIM-MS systems. We did not claim that the current system is portable. Current instruments require power and are relatively large; however, technological advances are moving toward miniaturization and in-field deployment (e.g. https://www.tofwerk.com/introducing-the-vocus-eiger/). We further clarified this issue within the new version of the manuscript. Please refer to p.8 317-319.
Minor Comments
Please correct decimal separators (0.5 L/min) and eliminate placeholder text.
Enhance the clarity of figures; add a comparative summary figure (PTR vs AIM).
Simplify long sentences and correct typographical errors.
When referring, provide uniform formatting and add other recent studies on VOC validation.
R5. We have corrected all decimal separators (e.g., 0.5/L min) and removed placeholder text. Figure 4 has been clarified and improved in readability. A graphical abstract has now been added, which already illustrates the comparison between PTR and AIM; therefore, a separate comparative summary figure might no be necessary. Long sentences have been simplified, and typographical errors corrected. Studies on VOC validation are the most recent on the literature that we have found.
Reviewer 4 Report
Comments and Suggestions for AuthorsGeneral Assessment
The manuscript presents a comparative evaluation of two direct-injection MS techniques (PTR-MS and AIM-MS) for real-time monitoring of VOC molecules emitted by pea plants. This manuscript is clearly written, methodologically sound, and provides nice datasets obtained under well-controlled experimental conditions. The topic may be of interest to the Molecules readership, as it bridges analytical chemistry and plant metabolomics in the molecule level.
However, the manuscript in its current form overstates the analytical advantages of AIM-MS over PTR-MS without sufficiently considering the reagent-ion-dependent selectivity intrinsic to adduct ionization. The rationale for the comparison, the interpretation of the observed compound-class differences, and the framing of the conclusions all require further clarification and refinement. I recommend major revision before the manuscript can be considered for publication.
Major Comments:
- Authors need to clarify the scientific rationale for the comparison between PTR-MS and AIM-MS. The Introduction provides an extensive background on VOC analysis but does not clearly articulate the analytical limitations of PTR-MS that the AIM approach is designed to overcome. The motivation currently appears primarily technical (such as AIM is new, therefore we are testing.). Please strengthen the manuscript by specifying which types of VOCs or ionization mechanisms are expected to differ between the two methods, based on known chemistry or previous studies.
- Reagent-ion dependence of AIM selectivity is not sufficiently explained and discussed. A key omission is the discussion of how the choice of reagent ion governs AIM-MS performance. Although the AIM reactor can use a range of ions (such as, dor example, C6H6+, NH4+, NO3-, and many more), this study employed only benzene cation (C6H6+). The manuscript should explicitly acknowledges that AIM results therefore reflect the behavior of C6H6+ towards certain class of VOCs, not the intrinsic capability of the AIM system itself. Furthermore, because C6H6+ preferentially forms stable adducts with nonpolar or unsaturated VOCs via pi-pi or dispersion interactions but exhibits poor adduct stability with polar oxygenated compounds due to charge transfer, its use naturally biases detection toward lipid-like and hydrocarbon species. Thus, the observed broader compound diversity under AIM may result from reagent-specific selectivity rather than an inherent instrumental superiority.
- Mechanistic explanation for reduced phenol detection is not properly provided. The authors attribute the reduced phenol signal in AIM-MS to “pi-pi interactions forming overly stable aromatic dimers.” This explanation is chemically unconvincing. The pi-pi stacking typically enhances, not suppresses, adduct formation. A more plausible mechanism may involve charge-transfer ionization, where C6H6+ transfers its charge to phenols (whose ionization energy is lower than benzene), leading to rapid fragmentation or neutralization (please note that this is also my speculation). Revising this section with a mechanistically sound interpretation or experimentally testing alternative reagent ions would significantly strengthen the discussion.
- Scope and framing of conclusions need to be revised. The concluding statement that “AIM-MS outperforms PTR-MS for VOC detection” is too general. Conclusion should accurately reflect the data and position AIM and PTR as complementary rather than competing techniques.
Minor points:
- Please indicate (i) the number of biological replicates per condition, (ii) whether the observed differences were consistent across replicates, and (iii) how background subtraction (control chamber) affected low-intensity signals. These details are necessary to evaluate the statistical robustness of the reported differences.
- The Introduction could benefit from a short paragraph explicitly stating that this work compares the two reactors under their respective optimal conditions, rather than under equivalent ionization environments.
- The Discussion would be clearer if Figure 4 (Van Krevelen plots) were referenced in parallel with compound-class summaries.
- The explanation of “photo-absorbent benzene” in Section 2.2 could be expanded to include a brief rationale for its selection (photoionization efficiency, reagent stability).
- Minor English polishing is recommended (e.g., avoid repetitive “might/could” phrasing in the Discussion).
Author Response
Major Comments:
1. Authors need to clarify the scientific rationale for the comparison between PTR-MS and AIM-MS. The Introduction provides an extensive background on VOC analysis but does not clearly articulate the analytical limitations of PTR-MS that the AIM approach is designed to overcome. The motivation currently appears primarily technical (such as AIM is new, therefore we are testing.). Please strengthen the manuscript by specifying which types of VOCs or ionization mechanisms are expected to differ between the two methods, based on known chemistry or previous studies.
R1. The comparison aims to explore how differences in ionization chemistry affect the detectable VOC space, rather than simply testing a new instrument. The study evaluates which VOC classes or molecular formulae are preferentially detected by each technique, providing insight into method-specific chemical coverage in complex plant VOC mixtures. The scientific rationale for comparing PTR-MS and AIM-MS has now been further clarified in the Introduction. Please refer to p.2-3.
2. Reagent-ion dependence of AIM selectivity is not sufficiently explained and discussed. A key omission is the discussion of how the choice of reagent ion governs AIM-MS performance. Although the AIM reactor can use a range of ions (such as, dor example, C6H6+, NH4+, NO3-, and many more), this study employed only benzene cation (C6H6+). The manuscript should explicitly acknowledges that AIM results therefore reflect the behavior of C6H6+ towards certain class of VOCs, not the intrinsic capability of the AIM system itself. Furthermore, because C6H6+ preferentially forms stable adducts with nonpolar or unsaturated VOCs via pi-pi or dispersion interactions but exhibits poor adduct stability with polar oxygenated compounds due to charge transfer, its use naturally biases detection toward lipid-like and hydrocarbon species. Thus, the observed broader compound diversity under AIM may result from reagent-specific selectivity rather than an inherent instrumental superiority.
R2. We thank the reviewer for this comment. We agree that the AIM-MS results in this study reflect the behavior of the benzene cation (C₆H₆⁺) as the reagent ion, and do not represent the full intrinsic capability of the AIM system. Our comparison with PTR-MS is limited to this configuration, and the observed VOC coverage should be interpreted in that context. We added a specification in the “introduction” and the “discussion” (p.3 L132; p.11 L446-450) sections.
3. Mechanistic explanation for reduced phenol detection is not properly provided. The authors attribute the reduced phenol signal in AIM-MS to “pi-pi interactions forming overly stable aromatic dimers.” This explanation is chemically unconvincing. The pi-pi stacking typically enhances, not suppresses, adduct formation. A more plausible mechanism may involve charge-transfer ionization, where C6H6+ transfers its charge to phenols (whose ionization energy is lower than benzene), leading to rapid fragmentation or neutralization (please note that this is also my speculation). Revising this section with a mechanistically sound interpretation or experimentally testing alternative reagent ions would significantly strengthen the discussion.
R3. We thank the reviewer for the comment. We agree with his/her view. We have revised the manuscript to reflect this mechanism. We have removed the speculative explanation regarding aromatic dimer formation.
4. Scope and framing of conclusions need to be revised. The concluding statement that “AIM-MS outperforms PTR-MS for VOC detection” is too general. Conclusion should accurately reflect the data and position AIM and PTR as complementary rather than competing techniques.
R4. We appreciate the reviewer’s comment and fully agree that AIM-MS and PTR-MS are complementary techniques in the current state of the art. However, we respectfully disagree with the suggestion that our conclusions are too general. Our statement that “AIM-MS outperforms PTR-MS for VOC detection” refers specifically to the experimental conditions described in this study, namely, our setup, the selected VOCs, and the investigated plant species. We did not intend to imply a general superiority of AIM over PTR, but rather to highlight the observed differences under our specific conditions. Emphasizing these differences does not suggest competition between the two techniques; on the contrary, it underlines their complementarity and the value of understanding their respective strengths and limitations.
Minor points:
1. Please indicate (i) the number of biological replicates per condition, (ii) whether the observed differences were consistent across replicates, and (iii) how background subtraction (control chamber) affected low-intensity signals. These details are necessary to evaluate the statistical robustness of the reported differences.
R1. Thank you for pointing this out. We now clarify these aspects in the “Methods” section. For each tested method we analyzed three biological replicates, each corresponding to a separate plant chamber. Please refer to p. 9 L362-364. Only features that remained significant across all biological replicates after FDR correction were retained. Background subtraction and replicate-consistency filter removed low or unstable signals. Significant features were merged across replicates using an inner join, meaning that only VOC features significant in all biological replicates were retained. Thus, our results reflect reproducible VOC patterns across replicates and are not driven by one sample. Please refer to p.14 L533-534
2. The Introduction could benefit from a short paragraph explicitly stating that this work compares the two reactors under their respective optimal conditions, rather than under equivalent ionization environments.
R2. We have now added this information at p.3 L132.
3. The Discussion would be clearer if Figure 4 (Van Krevelen plots) were referenced in parallel with compound-class summaries.
R3. We are unclear regarding the the reviewer’s request. In general, the graph has been updated for clarity and the Van Krevelen together with the frequency histogram give a complete representation of the compound composition. We would be happy to apply further changes if requested.
4. The explanation of “photo-absorbent benzene” in Section 2.2 could be expanded to include a brief rationale for its selection (photoionization efficiency, reagent stability).
R4. We thank the reviewer for this suggestion. We have expanded the description to include a brief rationale for the selection of photo-absorbent benzene, highlighting its high photoionization efficiency under our UV source and its stability as a reagent, which make it suitable for consistent and reproducible AIM-MS measurements. Please refer to p.11 L446-450.
5. Minor English polishing is recommended (e.g., avoid repetitive “might/could” phrasing in the Discussion).
R5. Done, thank you.
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript is accepted without any further changes.
Author Response
We thank the Reviewer for the valuable comments
Reviewer 2 Report
Comments and Suggestions for AuthorsI maintain my negative opinion concerning the manuscript entitled “Comparing Proton Transfer Reaction (PTR) and Adduct Ionization Mechanism (AIM) for the study of volatile organic compounds” (molecules-3949150).
- The title suggests that some compounds were studied, but in the text of the manuscript no compounds are mentioned.
- Detection of putative classes of compounds (e.g. line 24 – “putative lipids”, line 159 – “putative phenols chemical class”, line 495 – “putative compounds, including lipids, amino acids, carbohydrates, and phenols”, and in many other places in the manuscript), in my opinion, does not meet the scientific criteria applicable in the Molecules.
- The only two compounds mentioned (aldehyde 3,5-octadien-2-one and ester trans-3-hexenyl butyrate, page 8) have been claimed to be detected only on the basis of detected ions at given accurate m/z value (line 255: “Among putatively annotated signals, the m/z values of 124.098114 and 170.112244 detected with AIM were putatively annotated as C8H12O and C10H18O2, respectively.”). First of all, the accurate m/z may confirm the elemental composition but these values are not enough to identify a given compound. In the supplementary material, table 2 and 4, the values m/z 124.098114 and 170.112244 have been assigned as “unknown”. This is an unacceptable discrepancy between the manuscript text and the obtained experimental data reported in the supplementary material. Furthermore, even the claimed elemental composition has been wrongly determined, since the values 124.098114 and 170.112244 in no case confirm the formulas C8H12O and C10H18O2 (these formulas correspond to the values 124.08882 and 170.13069).
- In the authors’ response to the first revision, they agreed that amino acids, sugars and lipids cannot be regarded as volatile compounds, but in the revised manuscript they are still regarded as such.
The manuscript contains more mistakes/discrepancies but it is unnecessary to mention all of them, since these described above clearly indicate that the manuscript molecules-3949150 has been badly prepared.
Author Response
1.The title suggests that some compounds were studied, but in the text of the manuscript no compounds are mentioned.
R1. We respectfully disagree with the reviewer’s statement that “no compounds are mentioned in the manuscript.” This is a strong assertion that does not accurately reflect the content of our work. As clearly indicated in the title, our study focuses on volatile organic compounds (VOCs), and this is precisely what we investigated by analysing the air from different growth chambers equipped with two complementary analytical reactors (PTR and AIM).
Furthermore, several compounds are explicitly discussed in the manuscript. For example, compounds detected with AIM at m/z 124.098114 and PTR at m/z 171.107834 were putatively annotated as C₈H₁₂O and C₁₀H₁₈O₂, respectively. These annotations are consistent with previous literature identifying C₈H₁₂O as the volatile aldehyde 3,5-octadien-2-one from pea cultivars Crécerelle and Firenza, and C₁₀H₁₈O₂ as the ester trans-3-hexenyl butyrate from pea cultivar Aragorn under non-stressed conditions.
Therefore, the manuscript does include compound-specific information, both in terms of detected VOCs and their putative biochemical identities.
2. Detection of putative classes of compounds (e.g. line 24 “putative lipids”, line 159 “putative phenols chemical class”, line 495 “putative compounds, including lipids, amino acids, carbohydrates, and phenols”, and in many other places in the manuscript), in my opinion, does not meet the scientific criteria applicable in the Molecules.
R2. We respectfully but firmly disagree with the reviewer’s assertion that the detection of putative classes of compounds does not meet the scientific criteria applicable to Molecules. The use of the term putative is standard and widely accepted in metabolomics, volatilomics, and mass-spectrometry-based studies when referring to molecular features that cannot be confirmed with analytical standards but can be annotated based on accurate mass, fragmentation patterns, and established databases.
We clearly distinguish between putative annotations and confirmed identifications, following internationally recognised guidelines. In this context, reporting putative compound classes, including lipids, amino acids, carbohydrates, and phenols, is not only scientifically valid but also common practice in Molecules, as reflected in numerous previously published studies employing non-targeted MS approaches.
Importantly, our phrasing never claims definitive structural identification where such confirmation is not possible; instead, it transparently communicates the confidence level of the annotations. This approach ensures scientific rigor rather than diminishing it.
3. The only two compounds mentioned (aldehyde 3,5-octadien-2-one and ester trans-3-hexenyl butyrate, page 8) have been claimed to be detected only on the basis of detected ions at given accurate m/z value (line 255: “Among putatively annotated signals, the m/z values of 124.098114 and 170.112244 detected with AIM were putatively annotated as C8H12O and C10H18O2, respectively.”). First of all, the accurate m/z may confirm the elemental composition but these values are not enough to identify a given compound. In the supplementary material, table 2 and 4, the values m/z 124.098114 and 170.112244 have been assigned as “unknown”. This is an unacceptable discrepancy between the manuscript text and the obtained experimental data reported in the supplementary material. Furthermore, even the claimed elemental composition has been wrongly determined, since the values 124.098114 and 170.112244 in no case confirm the formulas C8H12O and C10H18O2 (these formulas correspond to the values 124.08882 and 170.13069).
R3. We thank the reviewer for this helpful observation. Regarding the feature at m/z 124.098114, we confirm that its assignment as unknown is correctly reported in the Supplementary Material. The incorrect value in the main text (124.088127) is the result of a typographical oversight caused by the two mass values being listed on adjacent lines in our supplementary table. We apologize for this mistake and are grateful to the reviewer for pointing it out.
Concerning the other feature, we acknowledge that the manuscript did not clearly explain its origin. This signal was detected using the PTR reactor, and we have now clarified in the revised manuscript that this sentence was intended to follow from the previous one, highlighting that the combination of the two reactors provides the most comprehensive and complementary coverage of detectable masses. The text has been revised accordingly.
Please refer to lines 255–264 on page 8 in the revised manuscript.
4. In the authors’ response to the first revision, they agreed that amino acids, sugars and lipids cannot be regarded as volatile compounds, but in the revised manuscript they are still regarded as such.
R4. Where amino acids, sugars, and lipids are mentioned, it is in the context of putative compound classes detected in the headspace or as ion signals, and they are clearly reported as such without implying volatility. We have maintained precise terminology throughout the manuscript to ensure scientific accuracy, and we believe the revised version correctly reflects this distinction.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors have addressed all the comments, and the paper can be accepted in its present form for publication.
Author Response
We thank the Reviewer for the valuable comments
Reviewer 4 Report
Comments and Suggestions for AuthorsI think the revised manuscript has been improved and publishable in Molecules.
Author Response
We thank the Reviewer for the valuable comments
