From the journal Environmental Science: Atmospheres Peer review history

28-Jun-2021

Dear Dr Greaves:

Manuscript ID: EA-ART-06-2021-000043
TITLE: Chemical Functionality at the Liquid Surface of Pure Unsaturated Fatty Acids

Thank you for your submission to Environmental Science: Atmospheres, published by the Royal Society of Chemistry. I sent your manuscript to reviewers and I have now received their reports which are copied below.

I have carefully evaluated your manuscript and the reviewers’ reports, and the reports indicate that major revisions are necessary.

Please submit a revised manuscript which addresses all of the reviewers’ comments. Further peer review of your revised manuscript may be needed. When you submit your revised manuscript please include a point by point response to the reviewers’ comments and highlight the changes you have made. Full details of the files you need to submit are listed at the end of this email.

Please submit your revised manuscript as soon as possible using this link:

*** PLEASE NOTE: This is a two-step process. After clicking on the link, you will be directed to a webpage to confirm. ***

https://mc.manuscriptcentral.com/esatmos?link_removed

(This link goes straight to your account, without the need to log on to the system. For your account security you should not share this link with others.)

Alternatively, you can login to your account (https://mc.manuscriptcentral.com/esatmos) where you will need your case-sensitive USER ID and password.

You should submit your revised manuscript as soon as possible; please note you will receive a series of automatic reminders. If your revisions will take a significant length of time, please contact me. If I do not hear from you, I may withdraw your manuscript from consideration and you will have to resubmit. Any resubmission will receive a new submission date.

The Royal Society of Chemistry requires all submitting authors to provide their ORCID iD when they submit a revised manuscript. This is quick and easy to do as part of the revised manuscript submission process. We will publish this information with the article, and you may choose to have your ORCID record updated automatically with details of the publication.

Please also encourage your co-authors to sign up for their own ORCID account and associate it with their account on our manuscript submission system. For further information see: https://www.rsc.org/journals-books-databases/journal-authors-reviewers/processes-policies/#attribution-id

Environmental Science: Atmospheres strongly encourages authors of research articles to include an ‘Author contributions’ section in their manuscript, for publication in the final article. This should appear immediately above the ‘Conflict of interest’ and ‘Acknowledgement’ sections. I strongly recommend you use CRediT (the Contributor Roles Taxonomy from CASRAI, https://casrai.org/credit/) for standardised contribution descriptions. All authors should have agreed to their individual contributions ahead of submission and these should accurately reflect contributions to the work. Please refer to our general author guidelines http://www.rsc.org/journals-books-databases/journal-authors-reviewers/author-responsibilities/ for more information.

I look forward to receiving your revised manuscript.

Yours sincerely,
Dr Claudia Mohr

Associate Editor, Environmental Science: Atmospheres

************

Reviewer 1

Stewart et al. present an interesting molecular dynamics study of the surface structure of liquid unsaturated fatty acids with particular focus on the comparison of fatty acids with different degrees of saturation.
The article is well-written, is suitable for the journal’s scope, and title and abstract reflect the topic and the content. While the results are in principle very interesting for the community, the following aspects should be addressed before I can recommend the article to be published and therefore recommend a major revision.

Major concerns:

1) The performance of the force fields has not been properly assessed. The authors should provide results, e.g. for the density and the viscosity of the fatty acid liquid systems and compare them to experimental data.
2) Despite an advanced and well-documented equilibration protocol it is not clear whether there is an orientational order present or not. The observation of the positions of functional groups is interesting but cannot replace an analysis of the orientation of the molecules, e.g. using an orientational correlation function. I strongly recommend to strengthen the discussion in this regard.
3) The comparison of C18 fatty acids with different degrees of saturation is in principle very interesting. Could the authors please discuss in detail what makes them comparable and what makes a comparison weaker, such as the molecular structure that is increasingly non-linear with higher degree of unsaturation? Are the molecular dynamics similar or are, e.g., the diffusivities different?

Minor concerns:

1) Abstract: GROMACs -> GROMACS
2) Introduction, second paragraph: “Aersols are in themselves an important source of atmospheric pollution…” Aren’t aerosols themselves pollutants rather than being “a source of pollution”?
3) Introduction, second paragraph: The term “cloud seeding” describes a method in geoengineering of precipitation patterns and seems to be out of context. Do the authors really mean “Aerosols play a crucial role as potential cloud condensation nuclei and for direct and indirect radiative forcing.”?
4) Introduction, end of second paragraph: SOA is formed by either nucleation of low-volatile gaseous compounds or by condensation of such compounds onto primary particles. Atmospheric radicals can react with molecules in the gas phase with the reaction product being less volatile so it can condense and form SOA but they do not generally transform primary aerosol to SOA. The latter process is often referred to as “ageing”.
5) Page 2, second paragraph: Force fields do not “generate the parameters…” but they consist of a set of parameters together with functions that describe the interactions.
6) Page 2, second paragraph: “In general all atom techniques have been found to be of a higher resolution than united-atom approaches.” This is trivial, they are of higher resolution by construction and this is not what the cited references conclude. Higher “resolution”, furthermore, does not necessarily lead to better results. AA and UA concern descriptions of the molecules and are neither “techniques” nor “approaches”. Please delete this paragraph or carefully revise.
7) Page 2, third paragraph: Please define the description in parentheses [#carbon atoms: #C=C double bonds] or describe using Fig. 1
8) Methods, line 6: GROMAS -> GROMACS
9) Methods, second paragraph: Please state the density values obtained from the simulations for all pure fatty acid simulations and compare to experimental values.
10) Methods, end of third paragraph: “… higher computational cost of involved in simulating larger numbers of molecules…” Please correct this sentence.
11) Methods, first pargraph on page 3: Why does an evaporation event end the simulation prematurely? In the setup discussed, any evaporated molecule will re-condense onto the slab eventually.
12) Results and Discussion, first paragraph: “… areas of vacuum,…”, rather “volumes” or “regions” of vacuum?
13) Fig 2: Please provide the figure in appropriate resolution and adapt the size of the four figures in the panel.
14) Fig 3: The percentages of surface coverage of the different functional groups is very interesting. I wonder what the uncertainties are that must be associated with the averages presented in the figure – are the differences statistically significant or not? The analysis method is described in supplemental section S4 but I cannot find any values for \sigma^2(\bar{S}_b). Please provide \bar{S}_b and \sigma^2(\bar{S}_b) in a table and/or include them in the figure.
15) Results and Discussion, page 3, third paragraph: Is the SASA really a good measure in this context? An OH radical is highly polarized and the electronic structure seems likely to have a strong influence on which parts of the system will become involved in a chemical reaction. Can the authors please discuss this aspect?
16) Page 4, third paragraph: Again, is the potential contact of an OH radical a predictor for a reaction? I would expect that the reaction with some parts of the fatty acid molecules are more favorable than with other parts which would modulate the reaction probability. Can the authors please critically discuss the advantages/disadvantages of SASA?
17) Page 4, next-to-last paragraph: Hydrogen bonding between the COOH groups would be enthalpy-favored. Then, at higher temperature, the entropy-favored states with equal surface contacts of COOH and Me would become more dominant, i.e. there should be a temperature dependent shift of the distributions. How can the authors explain that the distributions are nearly temperature-independent?
18) Page 5, second paragraph: These results are really interesting. I’m wondering how the authors can explain the experimental findings of Broekhuizen et al., J. Geopys. Res. 2004 or Dalirian et al., ACP 2017 who find that pure oleic acid particles are not CCN active. With COOH groups exposed to the gas phase there would be hydrophilic sites present, potentially facilitating water condensation. Why do these particles not activate? Can the authors compare to experimental CCN data also for the other fatty acids?
19) Page 5, next-to-last paragraph: It is important to keep in mind that these molecules are not linear but bent due to the sp2-hybridization and that the average shape depends on the number of double bonds. While an oleic acid molecule is very unlikely to expose both ends to the surface, this is much more likely for stearidonic acid. Please add a discussion around this matter and how it can affect the conclusions.
20) Page 5, last paragraph: Ordering of the molecules would also involve their orientation. This is particularly interesting in the current context since elongated molecules can get trapped in amorphous states in MD simulations which prevents equilibrium sampling. Can the authors show, e.g., how the vectors Me-COOH are correlated? (I think gmx rotacf can do something along these lines.)
21) Fig 5: Can the authors please discuss the temperature (in-) dependence, too? It is shown in the figure but not dicussed in the text. I think it is surprising that there are so little changes, particularly between 298 K and 333 K. Can the authors please calculate the viscosity and compare to experimental values of each fatty acid?
22) Page 6: “… evidence of the preferential orientation…” Again, the molecular orientation is not conclusively resolved, only the position of groups. The authors should show the simultaneous position of two groups in the same molecule to draw conclusions about the molecules’ orientations.
23) Conclusions: “This suggests that any preferential orientation of molecules as a result of the presence of the surface is only present in the first layer of molecules at this interface, with no evidence of long-range ordering.” The authors have shown this only along the direction perpendicular to the interface. This means that there is a possibility of extended spatial ordering with the z-axis as a symmetry axis.
24) Conclusions: “… highly surface sensitive experimental techniques…” What about other techniques, such as XPS?
25) Supporting material S1: These tests are interesting. Could the authors please discuss the possibility of bilayer formation, particularly for oleic acid? The rather linear molecules are known to stick together with the COOH groups forming hydrogen bonds and the tails exposed to the gas phase. Is there a periodicity of the density with the bilayer thickness? Could this lead to intrinsic ordering?
26) Supporting material S4: Please state the numerical values for the variances (see also earlier comment).

Reviewer 2

In this new contribution, the surface coverage of pure slabs made of by different fatty acids with increasing unsaturation levels is investigated by means of molecular dynamics. Fatty acids are indeed significant molecules when it comes to organic aerosol chemistry (oleic acid has been intensively studied as a proxy for such organic particles). Therefore gaining more insights into the nature of their liquid to gas interface is interesting.
This reviewer cannot comment of the actual level of theory or tools deployed here, and therefore focuses on the atmospheric significance of these simulations, which seems to be low.
My first concern arises from the use of pure fatty acids slabs, without the presence of water for instance. I do realize that adding water may come at some computational costs, but nevertheless some discuss that such pure slabs are not representative of atmospheric interface would be meaningful (especially when submitting a manuscript to Environmental Science – Atmosphere). The presence of water may change the orientation of the fatty acids and therefore the surface coverage of the slab, isn’t it?
Then, the use of the solvent accessible surface area (SASA) analysis is new to this reviewer and probably to many reader of this journal. I would therefore recommend introducing this analysis in such way it can be understood by the non-specialist (including the tuning presented in the SI). In my understanding, this analysis gives the surface coverage from below the interface (i.e., from the condensed phase perspective), how different would it be if the probe hard sphere comes from the vacuum side?
As there is some focus on the surface presence of the carbon double bond, why having selected the size of the probing molecule to OH and not ozone (which reacts efficiently in such unsaturations)? Would this change drastically the outcome of the simulations presented in Figure 3?
Figure 5 does not shown any specific trends for the four fatty acids; is this in contradiction with the SASA analysis?

EA-ART-06-2021-000043 Response to Reviewers Comments

Referee: 1
Comments to the Author
Stewart et al. present an interesting molecular dynamics study of the surface structure of liquid unsaturated fatty acids with particular focus on the comparison of fatty acids with different degrees of saturation.
The article is well-written, is suitable for the journal’s scope, and title and abstract reflect the topic and the content. While the results are in principle very interesting for the community, the following aspects should be addressed before I can recommend the article to be published and therefore recommend a major revision.
We thank reviewer for their kind words, and we agree that the results presented in this paper will be very interesting for the community. We also thank the reviewer for their comprehensive review, detailed comments and questions which we will address point-by-point below, our comments will be indented and italicised for clarity. Many of the points raised by referee 1 are highly technical in nature, so where additional material was needed to address any technical point it has been added to the Supporting Information document. This will allow comprehensive reporting of the technical methodology while allowing the general reader of this highly interdisciplinary journal to focus on the key results and conclusions in the main article.
Major Concerns
1) The performance of the force fields has not been properly assessed. The authors should provide results, e.g. for the density and the viscosity of the fatty acid liquid systems and compare them to experimental data.
Force field performance was assessed at the start of this project. We recognise that force field assessment will be of interest to some readers so have added this information to a new section in the Supporting Information document (section S1).
As we note in S1 many of the commonly used force fields within GROMACS (OPLS, CHARMM, and most AMBER force fields, for example) do not contain the parameters necessary to describe the dihedral interactions in regions of organic molecules where there are multiple C=C groups in close proximity. Three force fields containing all of the necessary parameters (GROMOS43a1, GROMOS54a7 and the General AMBER Force Field) were tested for bulk samples and a comparison of the results of these tests with literature values of densities and viscosities is now given in the Supporting Information section.

2) Despite an advanced and well-documented equilibration protocol it is not clear whether there is an orientational order present or not. The observation of the positions of functional groups is interesting but cannot replace an analysis of the orientation of the molecules, e.g. using an orientational correlation function. I strongly recommend to strengthen the discussion in this regard.

We are grateful that the reviewer has noted the care we put into the equilibration protocol and are happy to include our analysis of molecular orientation to this work. We have added a report of orientation of the molecular vector (Me-COOH, C1-C18) with respect to the surface normal and the rotational autocorrelation function of that molecular vector to the Supporting Information document (section S10). From our analysis it is clear that there is no orientational order present. Our conclusions regarding surface presence/preference of different functional groups are unchanged by this analysis.

3) The comparison of C18 fatty acids with different degrees of saturation is in principle very interesting. Could the authors please discuss in detail what makes them comparable and what makes a comparison weaker, such as the molecular structure that is increasingly non-linear with higher degree of unsaturation? Are the molecular dynamics similar or are, e.g., the diffusivities different?
The fatty acids investigated in this study were chosen so as to be as closely related as possible, except for the differing numbers of double bonds. Thus each of the acids has the same number of carbon atoms and is unbranched. The authors, however, accept that the choice to look at acids with cis- double bonds inevitably means that the more unsaturated species would be more bent and that the properties of the samples could affected by this. Cis- acids were chosen, as cis- double bonds are generally more commonly found in naturally occurring fatty acids. The authors have noted the bent nature of the acids in the main text (page 3) and have referred the reader to an additional discussion around this point in the Supporting Information document (section S5). Our simulation protocols ensure that the systems have time to reach equilibrium despite differing diffusivities. As such we do not believe any differences between the molecular dynamics of the acids to be sufficient as to affect the conclusions of the paper.

Minor Concerns
1) Abstract: GROMACs -> GROMACS
1. The authors apologise for this typographical error and have corrected it.

2) Introduction, second paragraph: “Aersols are in themselves an important source of atmospheric pollution…” Aren’t aerosols themselves pollutants rather than being “a source of pollution”
2. The authors have updated the wording of this text so as to avoid confusion.

3) Introduction, second paragraph: The term “cloud seeding” describes a method in geoengineering of precipitation patterns and seems to be out of context. Do the authors really mean “Aerosols play a crucial role as potential cloud condensation nuclei and for direct and indirect radiative forcing.”?
3. The authors have updated the text so as to clarify this point.

4) Introduction, end of second paragraph: SOA is formed by either nucleation of low-volatile gaseous compounds or by condensation of such compounds onto primary particles. Atmospheric radicals can react with molecules in the gas phase with the reaction product being less volatile so it can condense and form SOA but they do not generally transform primary aerosol to SOA. The latter process is often referred to as “ageing”.
4. The authors have updated the wording of this text so as to avoid confusion about the generation of SOA. The text in the third paragraph already notes that the process of oxidation chemically ages atmospheric aerosols.

5) Page 2, second paragraph: Force fields do not “generate the parameters…” but they consist of a set of parameters together with functions that describe the interactions.
5. The authors have deleted this paragraph as suggested in 6.

6) Page 2, second paragraph: “In general all atom techniques have been found to be of a higher resolution than united-atom approaches.” This is trivial, they are of higher resolution by construction and this is not what the cited references conclude. Higher “resolution”, furthermore, does not necessarily lead to better results. AA and UA concern descriptions of the molecules and are neither “techniques” nor “approaches”. Please delete this paragraph or carefully revise.
6. The authors have deleted this part of the text as suggested.

7) Page 2, third paragraph: Please define the description in parentheses [#carbon atoms: #C=C double bonds] or describe using Fig. 1
7. The authors have removed the parentheses to avoid confusion and believe that the differences between the structures of the acids are clear from Figure 1.

8) Methods, line 6: GROMAS -> GROMACS
8. The authors apologise for this typographical error and have corrected it.

9) Methods, second paragraph: Please state the density values obtained from the simulations for all pure fatty acid simulations and compare to experimental values.
9. A comparison of the fatty acid densities has been included as part of the force field comparison that has been included in response to major concern 1, see section S1 in Supporting Information. The text has been updated to direct the interested reader to the Supporting Information.

10) Methods, end of third paragraph: “… higher computational cost of involved in simulating larger numbers of molecules…” Please correct this sentence.
10. The authors apologise for this typographical error and have corrected it.

11) Methods, first paragraph on page 3: Why does an evaporation event end the simulation prematurely? In the setup discussed, any evaporated molecule will re-condense onto the slab eventually.
11. It was found that if the temperature was too high the program output an error message and the simulation stopped. When analysing the trajectory files produced before this point it was found that these crashed runs were the ones for which the slab had divided into two (characterised by the partial density values for all atoms having another minimum that did not correspond to the edge of the slab). This line has, however, been removed from the main text to avoid confusion. Evaporation events are unwanted because they would not represent the very low vapour pressures of the liquids, and would make the SASA analysis more complex.

12) Results and Discussion, first paragraph: “… areas of vacuum,…”, rather “volumes” or “regions” of vacuum?
12. “Areas of vacuum” has been updated to regions of vacuum”, as suggested.

13) Fig 2: Please provide the figure in appropriate resolution and adapt the size of the four figures in the panel.
13. An updated, higher resolution, figure has been provided.

14) Fig 3: The percentages of surface coverage of the different functional groups is very interesting. I wonder what the uncertainties are that must be associated with the averages presented in the figure – are the differences statistically significant or not? The analysis method is described in supplemental section S4 but I cannot find any values for \sigma^2(\bar{S}_b). Please provide \bar{S}_b and \sigma^2(\bar{S}_b) in a table and/or include them in the figure.
14. The differences in surface coverage for different functional groups is statistically significant, the stability of the surface configuration means that the uncertainties are small. A table of absolute and error values of the SASA analysis has been provided in the Supporting Information (S9) for the SASA analysis that was used to make Figure 3.

15) Results and Discussion, page 3, third paragraph: Is the SASA really a good measure in this context? An OH radical is highly polarized and the electronic structure seems likely to have a strong influence on which parts of the system will become involved in a chemical reaction. Can the authors please discuss this aspect?
While the presence of a highly polarized radical could indeed alter the equilibrium structure of a system, due to the short interaction time of a gas phase collider with the surface there is not sufficient time for the whole surface to rearrange, i.e. it would not change the configuration of the surface (revealed by the SASA analysis) to allow e.g. more COOH or C=C to be present. Any changes would be of a very small scale such as a small torsional motion in a Me to allow more favourable alignment of a C-H bond to the incoming radical. Any such rearrangement is also likely to be the prelude to reaction, aligning bonds to approach the required transition state structure. Due to the highly interacting nature of OH it would be unlikely to penetrate the surface any further than the SASA analysis, which uses a hard sphere. As such we have presented a description of the native structure of the interface.
Text has been added to the Results and Discussion section on page 4 to highlight the difference between the inert SASA probe and a polarising incoming species. Additional discussion of these affects is included in section S6 of the supporting information.
See also the response to point 16 below.

16) Page 4, third paragraph: Again, is the potential contact of an OH radical a predictor for a reaction? I would expect that the reaction with some parts of the fatty acid molecules are more favorable than with other parts which would modulate the reaction probability. Can the authors please critically discuss the advantages/disadvantages of SASA?
15&16. The SASA analysis carried out as described in the main text probes the presence of different groups at the vacuum-liquid interfaces over the course of the production runs. It gives information about which groups are present at these surfaces when the sample is isolated in a vacuum and cannot give any information about how the structure of the interface is altered by the presence of incoming species, especially polarizing ones, such as OH radicals. Such information is not accessible via the MD techniques used in this study, and an accurate treatment would require a higher (and computationally prohibitive) level of theory, such as a full quantum mechanical or quasi-classical treatment of the reaction site (QM/MM).
The relative surface coverages of different groups does, however, give an indication of the numbers of reactive groups present at the interface and therefore the likelihood of the slab to be reactive towards atmospheric species, as compared to a sample with a purely statistical coverage of different groups at the interface, and this is one of the main aims of this work. It should be noted that the SASA analysis involves a probe of a similar size to an OH, the authors do not claim that this probe mimics OH exactly, as the probe is a hard sphere and does not have any electronic structure. The size of the probe was, however, chosen so as to give an indication of the accessibility of the surface to molecules that are the same size as this common atmospheric radical. The presence of particular groups at the surface does not, of course, necessarily mean that they will react with an OH radical, or indeed any other species, encountering the surface. There are other factors, such as the trajectory of the OH radical and the intrinsic reactivity of the surface groups towards an incoming species, however, the much higher than statistical surface preference of largely unreactive methyl groups is an indication that such interfaces would be less reactive towards OH radicals than if the surface groups coverage were as in the bulk sample.
The authors have also (in response to a question from referee 2) included a SASA analysis for the case of a probe the same size as ozone, which shows the same results as the smaller probe.

17) Page 4, next-to-last paragraph: Hydrogen bonding between the COOH groups would be enthalpy-favored. Then, at higher temperature, the entropy-favored states with equal surface contacts of COOH and Me would become more dominant, i.e. there should be a temperature dependent shift of the distributions. How can the authors explain that the distributions are nearly temperature-independent?
17. The authors agree that increasing the temperature should favour a more random distribution of groups both throughout the slab and at the surface, and the annealing process showed that this is certainly the case at very high temperatures. Over the temperature range of the production runs, however, this effect is very subtle. The authors have noted this as an interesting point in the main paper on page 5.
The effect mentioned by the referee here would be more pronounced if a greater extent of ordering or large-scale structure was present (possibly expected by the referee given their other comments e.g., 20, 22, 23, & 25). As we have noted elsewhere there is a lack of such large-scale structure, and this may explain why there are only subtle temperature effects.

18) Page 5, second paragraph: These results are really interesting. I’m wondering how the authors can explain the experimental findings of Broekhuizen et al., J. Geopys. Res. 2004 or Dalirian et al., ACP 2017 who find that pure oleic acid particles are not CCN active. With COOH groups exposed to the gas phase there would be hydrophilic sites present, potentially facilitating water condensation. Why do these particles not activate? Can the authors compare to experimental CCN data also for the other fatty acids?
18. The papers mentioned by the reviewer also include data on CCN formation tests on linoleic acid, with similar results to the oleic acid work. The authors of this paper have been unable to find any data for such experiments using linolenic or stearidonic acid, however, the incredibly low surface area coverage of COOH groups in our work on these acids suggests that they would not be CCN active. The authors are unable to comment on why the COOH found to be present at the surface of the oleic and linoleic acid in these studies does not activate. This is beyond the scope of the present work as the MD simulations here are only able to predict the surface structure in the absence of incoming molecules and are not able to give any mechanisms for the interactions between these surface sites and incoming species.

19) Page 5, next-to-last paragraph: It is important to keep in mind that these molecules are not linear but bent due to the sp2-hybridization and that the average shape depends on the number of double bonds. While an oleic acid molecule is very unlikely to expose both ends to the surface, this is much more likely for stearidonic acid. Please add a discussion around this matter and how it can affect the conclusions.
19. It is true that the more unsaturated fatty acids are more bent and thus it might in theory be more possible that they would have both their COOH and methyl groups present at the surface at the same time. However, the surface area presence of COOH groups in stearidonic and linolenic acid is so low (c. 1-2 molecules per frame have an atom from this group at the surface) that this would only be the case in for a very small minority of molecules. In any case it is the total surface coverage of groups that is the focus of the work here, as this is where the atmospheric relevance of this study lies, in finding which groups could be potentially exposed to an incoming species, regardless of which molecule they are part of.
We have added information to discuss this, along with major concern 3, in section 5 of the supporting information.

20) Page 5, last paragraph: Ordering of the molecules would also involve their orientation. This is particularly interesting in the current context since elongated molecules can get trapped in amorphous states in MD simulations which prevents equilibrium sampling. Can the authors show, e.g., how the vectors Me-COOH are correlated? (I think gmx rotacf can do something along these lines.)
20. A discussion of the correlation of the molecular (Me-COOH) vectors, as well as the angular distribution of these vectors with the simulation axis perpendicular to the interface, has been added to the supporting information section S10. And noted in the main text on page 6. The short answer here is that the molecules show no evidence of being trapped in an amorphous state.

21) Fig 5: Can the authors please discuss the temperature (in-) dependence, too? It is shown in the figure but not dicussed in the text. I think it is surprising that there are so little changes, particularly between 298 K and 333 K. Can the authors please calculate the viscosity and compare to experimental values of each fatty acid?
21. The temperature independence of the data has been discussed as part of the response to minor concern 17, and the viscosities have been added to the Supporting Information section S1 as part of the comparison of force fields carried out in response to major concern 1.

22) Page 6: “… evidence of the preferential orientation…” Again, the molecular orientation is not conclusively resolved, only the position of groups. The authors should show the simultaneous position of two groups in the same molecule to draw conclusions about the molecules’.
22. The authors apologize for any confusion caused by the use of the word orientation. The main focus of the paper was which groups would be preferentially found at the interface, and an increased number of methyl groups present at the surface necessarily means that there is likely to be an excess of COOH groups in the sub-interfacial region, as these are the tail ends of those molecules whose methyl head-groups are at or protruding from the interface. The authors wished to highlight that from the z-density analysis discussed non-bulk distributions of functional groups were only seen at the region immediately next to the surface, and that the presence of the surface and/or the enhanced COOH concentrations in the subinterfacial regions did not appear to cause a non-bulk distribution of functional groups at areas further inwards.
The text has been updated to avoid confusion and to direct the interested reader to supporting information sections S10 & S11 which contain a detailed discussion of the (lack of) orientation.

23) Conclusions: “This suggests that any preferential orientation of molecules as a result of the presence of the surface is only present in the first layer of molecules at this interface, with no evidence of long-range ordering.” The authors have shown this only along the direction perpendicular to the interface. This means that there is a possibility of extended spatial ordering with the z-axis as a symmetry axis.
23. A partial density analysis along the x and y directions of the slabs has been carried out and has found no ordering in either direction. A line has been added to the end of the discussion on page 7 to clarify this and the corresponding graphs are presented in Section S11 of the Supporting Information.

24) Conclusions: “… highly surface sensitive experimental techniques…” What about other techniques, such as XPS?
24. The authors have chosen to suggest reactive atom scattering as a method of experimentally probing these results as this method most closely mimics the situation that they are simulating here- that is the focus is on the atoms that are present at the surface of the liquid that would be exposed to an incoming radical. Reactive atom scattering is commonly carried out using OH radicals, which due to their high reactivity will only probe the outermost atoms of the surface. The SASA probe size was chosen so as to mimic the size of an OH radical and only provides surface areas for the outermost layer of atoms, therefore reactive atom scattering would be a particularly good test of the SASA results described here.
A sentence has been added to the conclusions on page 7 to clarify this.

25) Supporting material S1: These tests are interesting. Could the authors please discuss the possibility of bilayer formation, particularly for oleic acid? The rather linear molecules are known to stick together with the COOH groups forming hydrogen bonds and the tails exposed to the gas phase. Is there a periodicity of the density with the bilayer thickness? Could this lead to intrinsic ordering?
25. The authors have found no evidence of bilayer formation in the z density plots shown in Figure 5, or the rotational autocorrelation function and angle-to-z axis results in Figure S7. These plots show no evidence of the degree of orientational alignment that would be required for bilayer formation.
It should be noted that the z-density plots in Figure S1 of the Supporting Information section have been created using the symmetrise option in GMX density (whereby the positive and negative z components of the slab are averaged), and this symmetrical appearance may look at a first glance like evidence of bilayer formation.

26) Supporting material S4: Please state the numerical values for the variances (see also earlier comment).
Absolute and error values have been provided in the Supporting Information S9.


Referee: 2
Comments to the Author
In this new contribution, the surface coverage of pure slabs made of by different fatty acids with increasing unsaturation levels is investigated by means of molecular dynamics. Fatty acids are indeed significant molecules when it comes to organic aerosol chemistry (oleic acid has been intensively studied as a proxy for such organic particles). Therefore gaining more insights into the nature of their liquid to gas interface is interesting.
1) This reviewer cannot comment of the actual level of theory or tools deployed here, and therefore focuses on the atmospheric significance of these simulations, which seems to be low.
1. We thank the referee for their candour, and believe that referee 1 has been very thorough on the technical aspects here. We believe that these simulations are an important part of the puzzle that will help to fully understand the chemistry of the atmosphere, indeed Referee 2 has stated that ‘Fatty acids are indeed significant molecules when it comes to organic aerosol chemistry (oleic acid has been intensively studied as a proxy for such organic particles). Therefore gaining more insights into the nature of their liquid to gas interface is interesting.’. There are also numerous occasions in the text where we discuss the works relevance to the atmosphere and previous work carried out to study all aspects of it. It is also worth noting that referee 1 has stated that the article ‘is suitable for the journal’s scope’ and ‘the results are in principle very interesting for the community’. Further discussion can be found in our responses to this referees questions below.

2) My first concern arises from the use of pure fatty acids slabs, without the presence of water for instance. I do realize that adding water may come at some computational costs, but nevertheless some discuss that such pure slabs are not representative of atmospheric interface would be meaningful (especially when submitting a manuscript to Environmental Science – Atmosphere). The presence of water may change the orientation of the fatty acids and therefore the surface coverage of the slab, isn’t it?
2. Pure oleic acid samples have commonly been used as a proxy for unsaturated components within aerosols, with aerosol mimics in laboratory studies often involving only oleic acid,1–4 without the presence of water or other components. The choice of using pure samples also allows benchmarking with known physical properties, as well as comparisons with the laboratory studies. While the authors acknowledge that the presence additional components other than just pure fatty acids within aerosols would possibly affect the positioning and orientation of these fatty acid molecules, the modelling of multiple species within a sample would make the simulations both more complex and difficult to benchmark. The MD studies here are therefore relevant atmospherically, both in terms of being a computational counterpart to laboratory studies of oleic acid as an important component of aerosols, and in terms of investigating the general trends in surface properties of chemical groups. Using pure acids allows us to see the effect of change only one particular aspect of a species, here the level of unsaturation, which could be useful when looking at other members of the large group of organic molecules that have been reported in aerosols.

3) Then, the use of the solvent accessible surface area (SASA) analysis is new to this reviewer and probably to many reader of this journal. I would therefore recommend introducing this analysis in such way it can be understood by the non-specialist (including the tuning presented in the SI). In my understanding, this analysis gives the surface coverage from below the interface (i.e., from the condensed phase perspective), how different would it be if the probe hard sphere comes from the vacuum side?
3. As we stated in response to referee 1 the more technical aspects of this work are in the supporting information document. We have added further information to both the main text (pages 3 & 4), and supporting information to further explain the SASA procedure, and references have been given for the methods used. For clarity the probe used in the SASA procedure is on the outside (vacuum or non-condensed) side of the molecules/system it is analysing.



4) As there is some focus on the surface presence of the carbon double bond, why having selected the size of the probing molecule to OH and not ozone (which reacts efficiently in such unsaturations)? Would this change drastically the outcome of the simulations presented in Figure 3?
4. The authors have chosen to use a probe size that is similar to an OH radical, as this is the most common atmospheric radical5 and is also commonly used in reactive scattering experiments, thus allowing for the potential that our results to be verified by future experiments. The authors have, however, carried out additional tests using a larger probe size that represents the size of an incoming O3 molecule, and these are now included in the Supporting Information (S6). From these results it can be seen that the main conclusions of the paper are not altered by this change in probe size.

5) Figure 5 does not shown any specific trends for the four fatty acids; is this in contradiction with the SASA analysis?
5. There is no contradiction between the two analyses. SASA analysis only examines the outermost layer of atoms within the sample, as the interface is not atomically smooth this would correspond to positions z>5.5 nm in the partial density analysis in Figure 5. The most visually obvious result observed in the SASA results (Figure 4) is comparing the COOH present at the surface for oleic and linoleic acids with that for linolenic or stearidonic acids: the linolenic & stearidonic acids have almost no COOH, whereas there is significantly more COOH at the surface for the former acids. Comparing this with the partial density analysis in Figure 5, it can be seen that for stearidonic and linolenic acids the blue COOH line has already reached ≈0 for z > 5.5 nm, whereas for oleic and linoleic acid it is visibly above zero, and thus available at the surface. The SASA analysis shows similar CH3 coverage for all acids, and this is observed in Figure 5. With regards to differences in HC=CH and CH2 coverage on moving to more unsaturated acids (as per Figure 3), any trends would not be visible in Figure 5, as each of the curves has been area normalised to one, and as such trends stemming from the different numbers of particular groups within molecules would not be visible.




1 Hearn, J. D.; Lovett, A. J.; Smith, G. D. Ozonolysis of Oleic Acid Particles: Evidence for a Surface Reaction and Secondary Reactions Involving Criegee Intermediates. Phys. Chem. Chem. Phys. 2005, 7 (3), 501–511. https://doi.org/10.1039/b414472d.
2 Smith, G. D.; Woods, E.; DeForest, C. L.; Baer, T.; Miller, R. E. Reactive Uptake of Ozone by Oleic Acid Aerosol Particles: Application of Single-Particle Mass Spectrometry to Heterogeneous Reaction Kinetics. J. Phys. Chem. A 2002, 106 (35), 8085–8095. https://doi.org/10.1021/jp020527t.
3 Morris, J. W.; Davidovits, P.; Jayne, J. T.; Jimenez, J. L.; Shi, Q.; Kolb, C. E.; Worsnop, D. R.; Barney, W. S.; Cass, G. Kinetics of Submicron Oleic Acid Aerosols with Ozone: A Novel Aerosol Mass Spectrometric Technique. Geophys. Res. Lett. 2002, 29 (9), 71-1-71–74. https://doi.org/10.1029/2002gl014692.
4 Asad, A.; Mmereki, B. T.; Donaldson, D. J. Enhanced Uptake of Water by Oxidatively Processed Oleic Acid. Atmos. Chem. Phys. 2004, 4 (8), 2083–2089. https://doi.org/10.5194/acp-4-2083-2004.
5 Li, M.; Karu, E.; Brenninkmeijer, C.; Fischer, H.; Lelieveld, J.; Williams, J., Tropospheric OH and stratospheric OH and Cl concentrations determined from CH4, CH3Cl, and SF6 measurement, npj Clim. Atmos. Sci. 2018, 29, 1–7.

23-Aug-2021

Dear Dr Greaves:

Manuscript ID: EA-ART-06-2021-000043.R1
TITLE: Chemical Functionality at the Liquid Surface of Pure Unsaturated Fatty Acids

Thank you for submitting your revised manuscript to Environmental Science: Atmospheres. After considering the changes you have made, I am pleased to accept your manuscript for publication in its current form. I have copied any final comments from the reviewer(s) below.

You will shortly receive a separate email from us requesting you to submit a licence to publish for your article, so that we can proceed with publication of your manuscript.

You can highlight your article and the work of your group on the back cover of Environmental Science: Atmospheres, if you are interested in this opportunity please contact me for more information.

We will publicise your paper on our Twitter account @EnvSciRSC – to aid our publicity of your work please fill out this form: https://form.jotform.com/211263048265047

For tips on how to publicise your research, please visit: https://www.rsc.org/journals-books-databases/about-journals/maximise-your-impact/

Discover more Royal Society of Chemistry author services and benefits here: https://www.rsc.org/journals-books-databases/about-journals/benefits-of-publishing-with-us/

Thank you for publishing with Environmental Science: Atmospheres, a journal published by the Royal Society of Chemistry – the world’s leading chemistry community, advancing excellence in the chemical sciences.

With best wishes,

Dr Claudia Mohr

Associate Editor, Environmental Science: Atmospheres

Reviewer 2

In my opinion, the authors have addressed properly all comments by the reviewers.