Evaluation of Luo et al., “Formation and temperature dependence of Highly Oxygenated Organic Molecules (HOM) from ∆3-carene ozonolysis”

The manuscript by Luo et al. characterizes the formation of HOMs from Δ3-carene ozonolysis in two different chamber environments, with and without OH scavenger and at different temperatures and relative humidities. It also includes a box modeling of the formation process of various HOMs from Δ3-carene oxidation.

The paper is well written, nicely structured and provides a good overview over the chemical mechanisms involved and the experimental results. It fits the scope of ACP and I thus recommend publication after addressing the following minor comments.

Minor comments:

line 123: you mention “other untypical reactions”, but you do not give an example or reference. Can you add an example and/or reference to shed more light on these untypical reactions?

line 180: You use the acronym PRAM already two times before defining it. Can you define it at the first occurrence in line 173?

line 212: do you have an explanation, why the there are much more C5 molecules formed for α-pinene ozonolysis, compared to Δ3 carene?

line 243/Figure 4: Can you briefly discuss the much earlier flattening of the RO2 curve compared to the traces for the closed shell products?

line 246/Figure 4: You state that “C10H14O9 began to rise immediately once C10H15O10 was formed” and that C9H12,14O9 and C10H16O9 “started to increase ~3 min later”. Can you indicate either in text and/or in the figure, how you did determine these rise-times. Is it a 50% rise like in the cutoff estimation for CPCs, or 10 %, or another method?

Figure 4: Does normalized signal mean normalized to the reagent ions? You use a variety of normalizations throughout the paper (be it to certain species or time steps), so I would recommend to always clearly stat to what you normalize to. You might also add the unit ncps to indicate normalization to reagent ions.

line 293: You write “Assuming all dimers are ELVOCs at 20 ºC,…”. As it is currently formulated, you assume all dimers are ELVOCs in order to prove that the seven dimers that are not C20H32O11 are ELVOCs, which doesn’t make sense to me. To my understanding you would only need to assume that C20H32O11 is an ELVOC at 20°C and then you can use your method of normalized ratios to show that all seven other considered dimers do behave similarly and can thus also be considered ELVOCs, which would then be the conclusion and not an assumption.

line 293: For this paragraph I would again recommend to more clearly discuss what “normalized” means, since there are several different normalizations involved, which are not properly separated in the text. Especially in the formulation “the normalized ratios of the eight largest dimers to C20H32O11” normalized could be interpreted as normalized to C20H32O11, but in the caption to figure 5 it means normalized to the 10 minute timestamp, which is not mentioned in the main text. Please reformulate this in order to avoid confusion.

Figure 5: Panel a is rather squeezed a bit in the y-axis direction. Can you reduce the y axis range a bit, so that the change in the traces becomes more visible?

Figure 5: In the caption you state that “Larger normalized ratios indicate accumulation in the chamber over time, i.e. a higher volatility”, however in the main text you also mention that “it is crucial to recognize that also differences in the formation pathways can influence the ratios.” So the “i.e” in the caption is a bit too strong of a statement, as it neglects the formation side of things.

Figure 6: These three different shades of red are barely distinguishable. I would recommend to use different colors, like red, orange and blue for the three different temperatures.

Additionally, the colorscale in panel d is confusing, as it is unclear how it relates to fraction and temperature. I would suggest to use red, orange and blue again in panel d, but indicate the fraction with color saturation, so that the three different temperatures are visually distinguished by base color.

line 400: Thank you for also adding data that might not lead to a final conclusion yet, or might be ambiguous. As you wrote, it might stimulate further research into the specific question and is thus important to report.

Figure 9: Add units for Δ3-carene VMR.

line 447: Your results are different to Quéléver et al., 2019, but similar to Simon et al. 2020, who also report an increase in C20/C10 ratio with falling temperatures for the T-range in question. I would add this in the discussion.

SI part:

line ~109: Can you clarify that in batch mode you did not add any flow during the experiment to the teflon bag, but due to sampling flows the volume of the teflon bag gets smaller over the curse of the experiment. Can you give an estimate of shrinkage over the duration of the experiment?

line 133: “species” instead of “specie”

line 137: Can you clarify a bit more how you estimated the 1e10 number for C in this study? (via the reaction time with 1/kt and the reaction rate constants estimated in Ehn et al. 2014?)

line 152: Can you briefly describe what exactly was modeled here in order to get kloss and what input parameters were used for the model?

line 164: How does the d(HOM)/dt term compare to the kloss*HOM term for your yield estimations? Can you indicate the timeperiods that you used to estimate the yield in the experiments and give a range of d(HOM)/dt vs kloss*HOM. It looks like the importance of the d(HOM)/dt term goes down over time.

Figure S5: I would change the y-axis label to “Intensity change due to CO addition” as this describes more directly what is shown in the figure.

Figure S8: Here the word normalized is a bit confusing again:  the data “looks” like being normalized to reagent ions, but as there is no mention of normalizing to the reagent ions, but instead mentioning of a “reference dimer”, the text somehow suggests that the data was normalized to said reference dimer. So I would change the y-axis label from “Normalized signal” to “Signal normalized to reagent ions” or something similar if this is what you are showing. Since this affects multiple figures, maybe you can also just write “Intensity (ncps)” and explain that ncps means normalized to reagent ions in the corresponding figure captions. The main point is that there is no confusion potential as to what you normalize to.

Figure S10: Can you add the RH values in the axis labels for panels d and e so that the reader directly knows which RH values are compared. It might be clear for the authors which experiment name corresponds to which conditions, but much less clear to the reader.

Figure S13: You do not discuss this figure in the main or SI text. Can you either add some discussion or remove the figure? You mention SOA yields in your introduction, but do not come back to them later in the text, although you have a figure of modelled SOA yields in the SI. You write in line 97f in the SI that your study “exclusively” focusses on HOM formation in the gas phase, while other studies focus on particle phase properties of the Δ3 carene system, but Figure S13 (and also Figure S12) somewhat contradicts that. You could exchange the “exclusively” for “mostly” and also briefly discuss Figure S13 in a similar fashion as you discuss Figure S12. Can you compare the modelled SOA yields to measured ones, or is this the focus of one of the other studies? (It would be perfectly fine if the main SOA yield discussion is part of another study, however, with Figure S13 you make the reader curious also about SOA yields and the corresponding measurements, so you should at least mention where the reader can find this discussion)

Figure S14: You do not mention what the color of the bars mean. (C number, but could you add that to the caption?)

General comments:

This paper presents new experimental results elucidating the product distribution of highly oxygenated molecules from ozonolysis of d-carene, combined with a box model to help determine oxidation mechanisms and interpret changing yields with temperature. Particularly because of its differences to the more widely studied a-pinene ozonolysis, this study can add valuable new experimental constraints to our understanding of mechanistic reasons for varying biogenic SOA formation. There are some opportunities to improve the figures and language, which I elaborate below. I recommend publications after minor corrections.

Specific comments:

p.3 line 86-87: suggest to reword the phrase “meaning the sources and sinks ... each experiment”, current phrasing is a bit confusing. Here you are explaining the different between batch mode and continuous chamber operation. How about something like, “meaning all reagents are injected in a single batch at the start of the experiment and products accumulate progressively” or similar? It’s the idea of sources and sinks “accumulating” that doesn’t feel clear to me.

p. 4 line 91: do you mean that the same instruments were used as in the COALA chamber? Specify, or briefly discuss key differences, maybe refer to SI.

Line 94: Briefly describe the Bianchi 2019 criteria. Does this mean >6 O atoms?

Line 120: “two of the RO can recombine” – is this really a recombination process, or does the ROOR just persist rather than cleaving the O-O bond?

p. 7 discussion around lines 160-165: I think it would be helpful if you show & label SOZs and 3-caronic acid in your Scheme 1. Also maybe include branching ratios / yields used in the model on the figure

p.8 around line 194: Maybe it’s a good idea to mention the key relevant conditions from Li 2019, e.g. O3 concentration, temperature, RH?

p. 9 around line 198: does this mean more auto-oxidation steps are possible in d-carene than in a-pinene? Or just different product distribution?

Line 199: “exclusively abundant” -> “only found” : if this is indeed what you mean here, this phrase would be clearer to me

Line 200-201: explain m and n notation.

Around line 210: Suggest to add some discussion of Figs 2b – 2e here before moving on?

p.10 around line 217: why would ring-opening not be dominant in OH oxidation also?

p.12 Fig. 4: Mention what the red dashed line is, I guess the only signal attributed to a radical?

p.13 line 305: what does “transition from semi-volatile to low volatility” mean? a certain gas to particle ratio?

p.14, Figure 5: blue and yellow lines are not visible on panels b, c, and d. In caption, list panel label before each description, not at end.

p.15 around line 342: Does any in this list of papers have any information about the stability of different ROORs for different MT precursors? How much do they range? I could imagine not much, because the R chain lengths are ~ the same, but does a different degree of oxidation matter to how stable the ROOR product channel is? Or are these differences more likely to be driven by the exact arrangement of the closest few atoms to the O-O bridge?

Around line 343: List the key differences between your and Quelever's experimental and instrument settings

p.18 Figure 8: Recommend to make model lines thicker in panels d, e, f. I didn't see at first until reading further than the C10H16O7 model line goes down after CO injection, because it's buried behind the pink and blue measurements points. Also suggest to add a vertical line indicating CO injection, or mention the time in the caption

p.19 around line 414: Can you speculate another reason that C10H16O7 does not decrease? Is there another source? It does seems to “react” to the injection , dipping in concentration before climbing back up.

Figure 9: Observed molar yields seem to be missing on this plot. Should definitely include

p.20 line 434: absence -> absence and presence, right?

Line 443: yes, warrants further investigation, but your observed decrease at lower temps matches your model well, right? Say something about this

Technical corrections:

p.2 line 33: hydroxyl radical (OH), or nitrate radical (NO3)

line 38: suggest to remove “a vital class of BVOCs”

p.3 line 75: suggest to change “set-ups” to “chambers”

p.5 line 124: suggest “unique to particular RO structures”

line 130: “both” -> “either”; line 131: “and” -> “or” & suggest to omit “as well”

p.7 line 153: researches -> research

line 167-169: “ The first-generation ... that is assumed to be ... and three other isomers, all C10 RO2, that ...”

line 172: reacting -> reacts

Line 182: Zenodo link missing?

p.9 line 208: (0A) should probably be (20A)?

p.11 line 235: “they have” -> “the product has”

line 239: “undetermined” -> “unclear”

p.13 line 29: identity condense behavior -> identical condensation behavior?

p.15 line 334: spell out CS

line 353: omit “could”

p.20 line 452: oxidation -> ozonolysis