D. Liu and co-workers have used filter sampling and mass spectrometry to study the composition of secondary organic aerosol (SOA) material formed from the ozonolysis of beta-pinene and limonene at different ozone concentrations. The topic is in principle highly atmospherically relevant, and the manuscript contains many interesting and publishable results. Unfortunately, the manuscript is in several places quite poorly written, and it is in places difficult for the reader to extract the genuinely novel results from the presented data. I’m happy to recommend publication of this manuscript in ACP, but only after a rather thorough revision including substantial copy-editing and scientific proofreading. 

The overall problem in reading the manuscript is that a substantial fraction of its sentences sound sensible, exact and scientific at a quick glance, but on closer inspection are at best very vague, and at worst ill-defined, illogical or meaningless. Sometimes the actual intent and meaning of the writers can be deduced by clever guesswork (e.g. by correcting some rather trivial word order issues), but in other cases even a thorough and repeated reading leaves the reader confused. The following examples are not intended to be an exhaustive list, but simply illustrative examples of this overall problem (note: I mostly gave up on listing problematic sentences after the first 3 pages or so, so the examples are weighted toward the beginning of the text - while the introduction is arguably the most problematic part of the paper this should not be construed to mean that the results - section is problem-free): 

-Line 17, “abundance”, please be more specific - exactly what abundance is meant

-Line 18, “prefer to stabilise”. As written, this would imply that the HOM molecules themselves would preferentially be stabilised (e.g. collisional, thermally, kinetically etc) at some certain [O3], but what is meant is that their yield saturates with respect to [O3] - not the same thing.

-Line 23, “formation of compounds with 10 carbon atoms”. As the precursors themselves contain 10 carbon atoms, no mechanism is needed for their formation. Presumably the authors mean the formation of compounds with 10 C atoms AND SOME NUMBER OF O ATOMS.

-Line 38: “wide range of volatility, which has a strong temperature-dependence”. What is meant by this - that an individual saturation vapor pressure has a strong temperature dependence (generally true), that the variation/range in volatility has a temperature dependece, or perhaps both? Whatever is meant, what is the relevance of the temperature-dependence to the rest of the discussion?

-Line 39: “ambient species are thought to consist mainly of low volatility species”, apart from the grammar issue (species…species) this is not even true as written - the ambient air certainly contains a large range of species (as indeed argued in the previous sentence!) all the way from fully volatile to ULVOC. I assume the authors do have an actual argument here, the presentation is just missing some key assumptions or concepts (e.g. do they mean the species present in ambient SOA?).

-Line 43 “ozonolysis as the most effective”, “as” should presumably be “is”, and also this sentence needs some caveats - for which compounds and in which conditions is ozonolysis the most effective, and compared to what (presumably OH and NO3 oxidation)?

-Line 45, “BVOCs forming ULVOC even in the absence of H2SO4”, it’s not the ULVOC formation in the absence of H2SO4 that is the surprising fact, but the production of aerosol particles in the absence of H2SO4. (Here I again assume the authors do actually mean the latter, they just wrote the sentence incorrectly.)

-Line 52, “high oxygen-containing but low oxidation state determining the oxidation potential”, here I have no idea what the authors mean, this seems to make no sense.

-Line 54, “HOMs refer to the process” - no, HOMs refer to a certain subgroup of the products of this process (and indeed that seems to be the definition the authors are using)

-Line 57, “precursors decomposing into aerosols”, certainly this is not what is going on

-Line 58, “Laboratory studies of HOMs observed by ozonolysis of monoterepenes are closely corresponded to”… this phrasing makes no sense.

-Line 60-65: the science is correct here, the phrasing is just very very poor.

-Line 66, “HOMs can provide nucleation conditions for early growth”: no, the HOMs can nucleate (or not), or participate in early growth (or not), this phrasing about providing conditions is meaningless

-Line 70 “an exocyclic bond as the second-most-abundant VOC” - no it’s beta-pinene (not its exocylic double bond) that’s the VOC

-Line 80 “nucleation rate of monoterpene SOA” - SOA is by definition material that has already formed particles (e.g. nucleated), what is presumably meant (though not measured here directly) is the nucleation rate of monoterpene-derived SOA precursors, such as the HOM discussed above. These are not the same thing. 

-Line 94, “obtain reaction mechanism”, this is incorrect - the mass spectra just gives the molecular compositions, these can then often be used to indirectly infer something about the reaction mechanism but claiming that the mechanism is “obtained” is wrong. 

-Line 207: “high ozone concentration tends to increase oxygen reaction”, this seems to be almost trivially true - maybe the authors need to specify what is meant by “oxygen reaction” here. 

-Line 208, “may indicate the importance of ozone… by ozonation”. Ozone is by definition important in ozonation, so this sentence as written is tautologically true (and thus meaningless). 

-Line 215, “peak intensity of MW” - presumably what is meant is the MW (molecular weight) corresponding to the peak of highest intensity (not the same thing). 

-Line 215, “maximum proportion”, what is meant by proportion here?

-Line 235, “limonene is preferred to proceed oxygenate and accretion reaction”, what is meant by this?

-Line 279, “front” should presumably be “former”

-Line 323-326, The first sentence here lists possible gas-phase formation pathways for (HOM) monomers and dimers. The second sentence then mentions rapid particle-phase reactions of some of the monomer and dimer types. The third sentence says “Due to the high activity of these pathways, the dimers…are expected to be distributed directly onto particles after gas-phase production”. While I agree that the dimers will be “distributed directly onto particles”, surely this is due to their low volatility, and not the rapidity of the subsequent reactions? How could a dimer still in the gas phase be affected by the reactions going on in the particle phase?

-Line 339, “monomers with CH2”, what is meant by this?

-Line 347, “The presence of…” this whole sentence does not seem to make sense - please rephrase, rewrite and give more background as this is potentially a quite important point! The next sentence is also very long and hard to parse - please split it up into two or more sentences. 

-Line 375, “almost hardly”, just “hardly” is enough

-Line 378, “related to the way of broken bonds” please rephrase

***

Other technical or notation-related questions and suggestions:

-Please note in the abstract that measurements were made from filter samples (as many readers may initially assume the gas phase is being probed).

-Line 17 in the abstract, “5-13% higher than” - should there be a comma here? Without the comma this literally means that the abundance (see above for a note on this term) of HOMs in limonene SOA was 5 to 13 percentage points higher than the abundance of HOMs in beta-pinene SOA (a claim not actually validated by the numbers in the text). With a comma this just more vaguely implies that the former is in general higher than the latter (which would seem to be true). 

-Line 90, what is meant by “gradient concentrations”? Just that there were three different concentrations? OR something else?

-Do the authors expect ROOR or ROOH - type compounds to fragment in their ionisation setup, as recently predicted even for milder chemical ionisation (https://doi.org/10.5194/amt-15-1811-2022)? If not, why not?

-Line 145, please redefine BDE here (it’s defined in the intro but that is easily missed)

-What are the Int_i weights used in equation 4? I couldn’t find the actual numbers anywhere. Also on line 205 the weighed average is mentioned with reference to Table 1, but I don’t see the weighted average numbers in that table. 

-What are the likely error margins of the (useful but crude) volatility estimate of equation 5? At least a couple of orders of magnitude I assume, since functional group identities are completely ignored?

-Line 204, “plausible different partition and agglomeration kinetics”, can you be a little more specific, and e.g. suggest which mechanisms could lead to the observed difference?

-Line 212, “number of organic molecules” - first I thought this was a mistake, but it seems the authors actually do mean the number of distinct molecules, i.e. the number of different elemental compositions measured (above some noise threshold). This might be good to specify explicitly here, to avoid misunderstandings (e.g. that “number” would mean “number concentration” or similar). 

-Line 231, “crack” has a very definite meaning in hydrocarbon chemistry, “fragment” is probably the word needed here.

-Line 249, “It seems” - could the authors speculate on the reason/mechanism of this inhibition? Could for example the OH produced by ozonolysis begin to cause more fragmentative oxidation at high O3 levels?

-Line 259, “overoxidation” - not a well-defined concept - I think I understand what is meant but the authors should still spell it out

-Lines 270-280, “dimer” often misspelled as “dimmer”, please correct

-Line 273, “This may be due”: this is almost certainly the explanation (already included in standard chemical mechanisms such as MCM) so the sentence could be a bit stronger here. 

-Section 3.4, accretion reactions are probably going on BOTH in the gas phase and the particle phase (and potentially some gas-phase accretion products are, in parallel, broken up in the particle phase). So the measured particle-phase “dimers” may be quite different from the original gas-phase ones. (This is  certainly implied already by the present discussion, but could be explicitly stated.)

-RA_HOM in the conclusions is not defined, please spell out what is meant. 

-The last paragraph of the conclusions seems to be very general, and unrelated to the specific study performed here. While the presented arguments are correct, would it fit better in the discussion or even the introduction? 

-Figure 8, WHAT are the markers colour-coded to? Please specify. 

-Figure 10, all the proposed channels are oxygen-increasing, and ALL of them proceed through the initial (well-established) Criegee intermediate => vinyl hydroperoxide => vinoxy radical + OH => peroxy radical sequence. I thus fail to understand why some of the channels are denoted “Criegee channels” and “OIR” while others are not. (I would call the right-hand-side for example alkoxy channels - “hydroperoxy channel” is an acceptable name for the left-hand-side.) Also, wouldn’t the alkoxy radical form equally well in reactions with HO2 or NO - or are these concentrations known to be low by the authors? Finally, the postulated H-shift from a C4 carbon in the (incorrectly labelled, see above) “Criegee channel” of b-pinene does not seem very plausible due to the associated ring strain - would not the H-shift from the alcohol carbon be the most likely one here also (as in the “hydroperoxide channel”)?

The manuscript entitled “Large differences of highly oxygenated organic molecules (HOMs) and low volatile species in SOA formed from ozonolysis of β-pinene and limonene” reports chemical composition of SOA, particularly, HOM in particle-phase formed from ozonolysis of b-pinene and limonene. The SOA was formed in a flow tube with ~5 min reaction time. SOA composition was determined via filter collection followed by water extraction and analysis using Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR-MS). This study investigated the effect of ozone concentration and compared the difference of the chemical composition between SOA formed in b-pinene and limonene ozonolysis. It was found that for b-pinene ozonolysis, as O3 concentration increased, particle size, OS, relative abundance (intensity-based) of HOM, HOM dimer abundance, and the fraction of LVOC in HOM increased; the relative abundance (intensity-based) of HOM monomer kept almost invariant. For SOA formed in limonene ozonolysis, as O₃ increase, particle size increase, relative abundance of HOM stabilize, of HOM dimer, of HOM monomer remained generally stable; the fraction of LVOC OS, O/C, n(O), DBE, and MW first increased and then stabilized or slightly decreased with O3 increase. At the same O3 level, SOA formed in b-pinene ozonolysis had lower OS, MW, O/C, DBE, lower relative abundance of HOM, of HOM monomers, of HOM dimers, and more numbers of unique compounds (especially those with low OS_c) than SOA form limonene ozonolysis.

This study addresses the chemical composition SOA determined on molecular level, which is an important and challenging topic of atmospheric chemistry. The manuscript fits the scope of ACP. I have the following comments for the authors to consider before publication.

General comments

1. Some formulations of the manuscript are not easy to follow (e.g. lines 18-19, 82-83, 214-216, 247-248, 295-297, 305-306). And there are a number of grammar mistakes. I suggest the authors to polish the language throughout the manuscript.

Specific comments

1. Can the intensity-based abundance directly translate to concentration? (e.g. L220, Fig.3, and Fig.4).

In another word, do all compounds have same sensitivity in MS so that peak intensity is directly proportional to the concentration?

1. How much are the organic aerosol concentrations for the various O3 levels? OA concentration can affect the partitioning of gas-phase species and thus interpretation of the dependence of chemical composition on O3 concentration.
2. 7 vs. Fig. 9, why is there no ULVOC in HOMs?
3. The foci of the abstract, conclusion, introduction is not exactly the same. . (extracted using water). only WSOC?
4. L17, “5-13%” is not shown in the main text. How is this number obtained?
5. L120, I suggest noting that the organic compounds are water soluble ones as only water is used for extraction.
6. L192, why the standard of O/C<0.7 is used rather than n_O<7 for the classification of HOMs(Bianchi et al., 2019)?
7. L200, “which may be due to the formation of high molecular weight and low-volatile dimers” such a statement is not supported by an evidence. I suggest either omitting this or citing the figures on dimers in this study.
8. L207-208, “…tends to increase oxygen reaction” is not clear.
9. L209, how the “abundance of organic peroxides” are obtained?
10. L228-229, is it possible that dependence of composition on O₃ is related to the OH produced in ozonolysis as SOA formed via b-pinene ozonolysis likely does not contain C=C double bonds and thus can less likely react with O₃?
11. L237, “as well as low H/C ratio organic molecules”, I suggest citing Fig. 5 here. Otherwise, it is hard to follow.
12. L259, what does the overoxidation mean? Also by dissociation, fragmentation might be a better word.
13. L341, how the conclusion “β-pinene increases the possibility of carbonyl formation at high ozone concentrations” is not clear.
14. L358, why it is attributed to “the particle-phase chemistry” rather gas-phase reactions?

Technical comments

1. L24, “evolution mechanism of monoterpenes”, do you mean evolution mechanism of monoterpene-derived SOA?
2. L67, “The more abundant atmospheric β-pinene and limonene” is not clear?
3. L205, this statement is for b-pinene.
4. L261, the last “or” should be “and”.
5. L274, a “of” is missed before “dimmer”.
6. L316, many or more?
7. L370-371, I think that these sentences are not directly relevant to the main findings of this study.
8. L374, “with” is not correctly used here.
9. 8, color bar is missed

Reference

1. Bianchi, F., Kurten, T., Riva, M., Mohr, C., Rissanen, M. P., Roldin, P., Berndt, T., Crounse, J. D., Wennberg, P. O., Mentel, T. F., Wildt, J., Junninen, H., Jokinen, T., Kulmala, M., Worsnop, D. R., Thornton, J. A., Donahue, N., Kjaergaard, H. G., and Ehn, M.: Highly Oxygenated Organic Molecules (HOM) from Gas-Phase Autoxidation Involving Peroxy Radicals: A Key Contributor to Atmospheric Aerosol, Chem. Rev., 119, 3472-3509, 10.1021/acs.chemrev.8b00395, 2019.

The authors describe a set of experiments in which volatile organic compounds (VOCs), either limonene or beta-pinene, were oxidized by ozone in a flow tube. The subsequently formed aerosol particles were then analysed using different techniques. The main analytical tools used in the work are the scanning differential mobility particle sizer (SMPS) for measuring the size distribution of the formed aerosol particles, and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) to characterize their chemical composition from filter samples. 

In its current form, I find that the manuscript lacks sufficient detail of the experimental procedure to support its conclusions. I feel that more data would be required to validate the results. Also, I would strongly suggest making the data used for the manuscript freely available. In addition, some of the introduction and conclusions sections are rather generic. For these, I would suggest for the authors to have a thorough look at the reference material, and form a coherent story based on that. Below, I have listed a set of concerns that I think need to be addressed before publication can be considered.

- The reaction conditions. The authors list in Section 2.1 that they used a 7 l flow tube. This was operated with flows of 1 lpm N2 to evaporate VOC (MFC 2 in Fig. 1) and around 2 lpm synthetic air through the ozone generator (MFC 3). Was MFC 1 used at all? The listed flows add up to 3 lpm: at this flow rate, the residence time in the tube would be 2 min 20 s, but the authors state a residence time of 5 min. Where does this discrepancy come from?

- In addition, a large fraction of the total flow consisted of nitrogen. However, HOM are formed in autoxidation, i.e., in the reaction with atmospheric O2. Did you consider this effect?

- I suppose the listed ozone concentrations are without VOC in the flow tube. You have relatively imprecise estimates for the VOC concentrations. You could get better values by monitoring the ozone concentration change upon adding the VOC and calculating back from there. Did you do this?

- The reaction rate of limonene with ozone (2e-16 cm3 s-1) is over ten times that of b-pinene with ozone (1.5e-17 cm3 s-1). So, for similar starting concentrations, you would, as a first approximation, expect over ten times more limonene to react. And so, assuming similar SOA yields, you would expect ten times more SOA from the limonene reaction. Limonene would be expected to form even more aerosol than this, as it typically has a higher SOA yield, and SOA yields typically increase with increasing reaction rate. Yet, you observe similar aerosol production, both in size distribution (Fig. 2) and in mass (Table S1). In mass, you even get clearly less mass for limonene for most conditions. How do you explain this?

- I'm very crudely estimating that in Fig. 2, you have around 15 000 particles per cubic centimetre for both limonene and b-pinene in the high ozone case. Assuming a mass mean diameter of 90 nm and a density of 1.5 g/cm3, this would correspond to a mass concentration of some 70 ug/m3. In Table S1, you collect around 200 ug of SOA in an hour. This would mean collecting around 3 m3 per hour, or 50 litres per minute. However, your total flow is much smaller. Am I missing something? Or how do you get those numbers? Could you please provide the data for Fig. 2, or the whole SMPS data?

Additional point by point comments on the text:

Intro, lines 38-65: the text flows poorly here, and is rather generic. I would reformulate your message and write it down again. 

Line 60: the term ULVOC was not used by Bianchi et al., and has only been introduced later

Line 77: "opposite trends in SOA formation potential": what does this mean?

Line 80: "Nucleation rate of monoterpene SOA" does not really make sense here: nucleation rate and SOA yield are two separate (though connected) concepts. Here you seem to refer to SOA yield

Lines 82-83: I don't understand this sentence

Line 84-85: The same applies to low-volatility vapours. And the high ozone here means a high oxidation rate, and a high initial particle production rate. This makes condensation to particles a competitive sink for vapours, as opposed to wall loss. 

Line 90: what does "gradient" mean here? 

Line 102: what about the total flow, and MFC1? Already with the numbers here, 1/3 of the flow is N2, which will already drop the O2 concentration in the flow tube considerably. This may influence HOM formation, as they are formed in autoxidation, i.e., with atmospheric O2

Line 103: What was the flow through MFC1? This is not listed, but with 1 lpm through the VOC and 2 lpm through the O3 generator, the total flow is already 3 lpm. In a 7 l flow tube, this should give a residence time of 2 min 20 s. 

Line 105: how did you vary the ozone concentration?

Lines 107-108: at 1 ppm, the ozone lifetime towards limonene oxidation is around 200 s, and 2500 s for b-pinene. This means that in the 5 min almost all of the ozone should react in the limonene case, while in the beta-pinene case, only around 10 % reacts. Combined with the higher SOA yield of limonene, we should see much more SOA in the limonene case. This is, at least to some extent, seen in the SMPS data, but the opposite is seen in the collected mass. As the majority of the article is based on the filter analysis, this should be addressed

Line 115: what flow rate is this? To the SMPS?

Lines 116-117: what experimental tests? And what are these results on aqueous phase radicals that are mentioned a few times, but not really presented?

Line 117: what wall loss was negligible? For sure there are different types of wall losses, both of vapours and of particles

Line 149: About the isotopic peaks: will this not lead to overestimation of some compounds, as isotopic signals from neighbouring masses overlap with them?

Line 161: I don't really understand where this comes from. For example, C10H14O11 would fall under highly oxidized organic compounds, and I can't really think of any realistic examples of more oxidized HOM. Also, the method does not distinguish between -OOH group and two -OH groups, while the latter is twice as oxidized. Do you still think this classification makes sense?

Lines 189-190: The reference talks about ambient measurements, where volatilities were estimated with the SIMPOL parametrization. So, I would not use it for such a general statement. 

Line 191: Do you mean that these compounds were assigned as HOM monomers and dimers? 

Lines 200-201: As the ozone increases, the whole reaction rate increases. And if SOA yield stays the same, or at least doesn't go dramatically down, the size and concentration of the particles should increase. So, this does not point to any specific compounds or mechanism yet

Lines 201-202: Do you mean that organics promote the formation and growth, and thus increase the survival probability of the particles? Now it sounds like you mean that the specific organics themselves survive, which, as far as I know, is unknown

Lines 207-208: "increase oxygen reaction": what does this mean? 

Lines 209-210: Is this your result, or someone else’s?

Line 214: I only see a slight decrease. Also, please avoid using "significantly" unless talking about statistical significance

Line 227: Abundance or relative abundance? I would expect the abundance of pretty much anything to go up with ozone in these conditions

Line 237: "condensation reaction channel": what does this mean?

Line 251: You are using quite a narrow definition of HOM, if I understand correctly (the predefined formulae on line 191). Looking at Fig. 5, I think many more would also qualify as HOM.

Line 251: We can't see from the plot that they are dimers. Also, it might be useful to label some of the largest peaks in the plot

Line 252: What is RA? Relative abundance? I don't think it's defined 

Lines 323-324: add references

Lines 329-330: The reference talks about slightly different element number ranges. Please be careful with references: now it sounds like Pospisilova et al would be talking about these exact numbers

Lines 330-332: Hard to follow sentence

Line 336: High probability based on what?

Line 337: where do these compounds come from here?

Figure 10: is this all based on work of others? Or do you have some more support for these channels in particular? If not, I would leave this figure out

Line 342: You are measuring monoterpene oxidation here, so presumably all of these are monoterpene oxidation products

Line 343: Do you mean that C10H16O9 was only detected in b-pinene SOA?

Lines 347-348: this is very speculative

Lines 349-351: Hard to follow sentence. Also, C10H16O8 does not necessarily come from C10H15O8, it may also have other sources. And it only forms from C10H15O8 upon reaction with HO2. Finally, I don't think detecting these products "verifies" this pathway. At maximum, it does not contradict the pathway, but does not rule others out either. For instance, C20H30O12 could also form from C10H15O10 + C10H15O4. 

Lines 355-358: First you talk about formation of oligomers from closed shell monomers. But the Berndt et al reference is for gas-phase formation of accretion products from RO2-RO2 reactions

Lines 358-359: What observation? The reference is not about particle phase chemistry, but gas-phase

Line 363: I would also expect highly oxidized dimers in this range

SI:

Table S1: These numbers don't make sense to me. At low O3, why do you get 70 ug of SOA in 0.4 h for beta-pinene, but only 30 ug in a full hour for limonene? And in nearly two hours, still only 30 ug? Limonene should have a higher SOA yield. Also, the SMPS shows more mass for limonene. 

Table S1: median-->medium

Fig. S1: for each ozone concentration, are the results the average of the two filters?