Gong et al. studied the SOA formation of β-pinene ozonolysis at 298 K, 273 K and 248 K. Coupling the laboratory measurement and chemical modelling, the authors provide insights into the mechanisms of dimers formation in the β-pinene ozonolysis SOA at three studied temperatures. Considering the importance of SOA formation in winter or cold regions or high altitude, the topic is of high interest to the research community in atmospheric science. The study was well designed, but only a subset of data was analysed and presented. The current discussion is mostly limited to the dimers in the particle phase. I would highly encourage the authors to carry out more comprehensive data analysis for the gas- and particle-phase data before they reach the conclusions. I would support the final publication after addressing my comments below.

Major Comments

1. SOA studies at low temperatures are increasingly popular (Huang et al., 2018; Ye et al., 2019; Simon et al., 2020; Gao et al., 2022; Gao et al., 2023), of which the findings inform the SOA formation at high altitudes or colder regions. In particular, several KIT studies (Huang et al., 2018; Gao et al., 2022; Gao et al., 2023) which used the same chamber as this study have reported the effect of temperature on the formation of biogenic SOA and their physicochemical properties. The authors should briefly summarize the findings of the previous studies at least from KIT and then discuss what new values this study is bringing.
2. The O₃ concentration used in the experiments was approximately 1 ppm, which was higher than the typical ambient level. How are the findings of the study applicable to the real atmosphere? Please give a justification. To further strengthen the implications of the work, model simulations should be also carried out for a scenario of ambient level of O₃.
3. If the instrumentation was operated at a temperature which was higher than that in the AIDA chamber, potential losses of particle-phase compounds due to evaporation could occur in the tubing. Please justify how the temperature difference affected the observed gas and particle components in detail.
4. Lines 220 – 231 and Lines 242 - 244: The authors are encouraged to provide more insights into how [HO₂]/[RO₂] affect the SOA formation at different temperatures. Given the gas-phase CIMS measurement was available in the study, the authors should provide detailed analysis on how different [HO₂]/[RO₂] affects the gas phase composition, volatility distributions and consequently SOA formation.
5. Line 235: The monomers contribute to 54 % – 64% of the particle-phase signals and thus were important for the SOA formation in this study. However, there is little discussion about the monomers here. The authors need to provide more details about monomers in the context of SOA composition.
6. Lines 243 – 244: Please provide more discussions about how their gas-phase formation pathways were influenced with the use of the gas-phase data from CIMS.
7. Lines 246 – 249: The authors claim that there is a linear relationship between the particle-phase dimers and monomers in the particle phase, consistent with the findings of Zhao et al. (2018). The analysis of this study was carried out against the particle phase, which is totally different from the one done by Zhao et al. (2018) (i.e., the gas phase). It will be more appropriate if the gas-phase analysis is provided here. In addition, Pospisilova et al. (2020) have observed rapid condensed-phase reactions in the α-pinene ozonolysis SOA under 40 % – 50% RH conditions at 295 K. Different from Pospisilova et al. (2020), the experiments presented in this study were performed at low RH and/or low temperature. Under such conditions, particle viscosity can be substantially high and potentially affect the diffusion of organic molecules (e.g., monomers) in the particle phase. Please provide a discussion about the impact of particle viscosity on the particle phase reactions forming dimers.
8. Lines 256­ – 257: The volatilities of C17-19 dimers are presumably sufficiently low enough and thus they should be primarily in the particle phase even at 298 K. Please quantitatively estimate how much the increase in C17-C19 dimers in particles can be attributed to the decease of volatilities but also the reduced wall losses when temperature decreased from 298 K to 273 K.
9. Lines 270 – 281: Instead of analysing selective compounds, the authors should carry out more thorough analysis with the thermograms. As discussed, the temperature and/or [HO₂]/[RO₂] conditions can affect the production of SVOCs and LVOCs in the manuscript. If so, the particle volatilities changed accordingly. In addition, it is possible that for an identified ion, its desorption temperature varied with the experimental conditions, if isomers of different volatilities were produced. Therefore, it will be useful to add a) the total ion thermograms of particle samples; and b) statistics about the desorption temperatures of all observed ions.

Minor Comments

1. The abstract is too long and descriptive for the readers. It needs to be more concise and compact for better readability.
2. Line 11: Specify what are the SCIs scavengers used in this study.
3. Lines 14 – 15 and Lines 256­ – 257: Whether the changes in SOA yield were partially attributed to partitioning as well as wall losses or not? Please give an estimation about how the changes in SOA yield was affected by partitioning and wall losses.
4. Line 20: It will be good to mention that the [RO2]² was derived from the model simulations.
5. Line 37: Please briefly summarize the current knowledge about how temperature impacts the chemical reaction mechanisms regarding SOA constituents.
6. Lines 33 – 35: References are too old. There are a series of temperature studies (Huang et al., 2018; Ye et al., 2019; Simon et al., 2020; Gao et al., 2022; Gao et al., 2023) in the past few years. Please include up-to-date citations here.
7. Line 44: Please specify “several trace species”.
8. Line 57: It is unclear that whether “RO₂+RO₂ reactions” mean self or cross reactions here but also many places in the manuscript.
9. Line 105: Why only semi-volatile products? How about compounds of low volatility?
10. Section 2.2: How was the FIGAERO-CIMS calibrated? Please provide additional details.
11. Lines 129 – 130: What were the mass loadings of the collected filter samples? Delaying desorption on the FIGAERO which arises from too high mass loading can affect the thermogram analysis and interpretation (Huang et al., 2018).
12. Lines 162 – 167: Model sensitivity tests should be carried out to investigate the impacts of negative or positive temperature dependence of RO₂+RO₂ reactions.
13. Section 3.2: How were the volatilities and gas-particle partitioning were calculated in the model? What were the particle wall loss rates?
14. Line 185: “indicating that the wall loss rates of organic acids were higher at higher temperature” is redundant.
15. Line 201: Please comment why new particle formation still occurred.
16. Lines 202 – 203: Considering the appearance of new particle formation and potential size-dependent variations in the organic-to-inorganic ratios, it can be technically challenging to calculate the SOA density. I would suggest the authors provide detail descriptions to explain how they derived the SOA density.
17. Line 206: Was the uncorrected SOA mass concentration used to calculate the SOA yield?
18. Line 208: Were the previously reported SOA yield for β-pinene ozonolysis obtained at 298 K and low [HO₂]/[RO₂]? What is the range of the reported SOA yield?
19. Line 210: Please estimate the extent to which the partitioning process was promoted due to the decrease in temperature. This can be simply done by comparing the model simulations with and without accounting for the temperate dependence of volatilities.
20. Lines 216 – 219: Whether the temperature dependence of SOA formation was observed in SOA particles other than β-pinene ozonolysis ones in previous studies?
21. Line 219: “which contribute to the formation of SVOCs and LVOCs” is unclear. How does it link to the observed temperature dependence of SOA formation / yield?
22. Line 227: Please specify the unwanted interferences if high concentrations of FA were used.
23. Line 229: How was the scavenging efficiency of more than 70% estimated?
24. Line 234: What are the other compounds in Fig S6? Depending on the experimental condition, they account for considerable fractions of the particle samples.
25. Line 235: What is the use of the mass defect plots here?
26. Lines 236 – 237: I disagree that the formation mechanisms of C20 dimers were not discussed due to their lower abundance. The authors should briefly discuss why the production of C20 dimers was low across the experimental temperatures.
27. Line 270: What is the fraction of monomers showing two peaks at different temperatures?
28. Line 284: It is unclear that how the authors came to the point that the contribution of β-pinene-derived SCIs to dimers formation through bimolecular reactions was significant.
29. Lines 293 – 294: Provide a table in SI to show the molar yields of HCHO and nopinone and experimental temperature for this study and literatures.
30. Figure 6: Please indicate which experiments were included for the analysis here.
31. Line 329 ­– 334: Figure 6 shows that non-SCIs-controlled dimers are important at 298 K. It is confusing that how higher RH inhibited 40% of the dimer formation at 298 K, if the non-SCIs-controlled pathway is more important than the SCIs-controlled one at that temperature.
32. Line 354: Please justify why “20%” and “40%” were chosen for the classification.
33. Figure S13: There is one additional marker beyond those listed in legend. What does the solid marker in purple represent?
34. Lines 360 – 361: The variations in the temperature dependences are not obvious in Figure S13.
35. Lines 362 – 363: Which figure shows the influence of the variation of [HO₂]/[RO₂] on C18 dimers?
36. Line 366: The fractions from the gas and particle phase should be showed in Figure 9A.
37. Lines 369 – 371: Please provide the time series of nopinone in the supporting information.
38. Line 374: It is unclear what are these two possible reasons here.
39. Lines 387 – 388: Why 75 s^-1 and 15 s^-1 were chosen in the box model?
40. Line 407: There is little discussion about how temperature affect the compounds’ volatilities in the manuscript.
41. Lines 409: Please specify what the positive- and negative-temperature-dependent processes are.
42. Line 410: Provide citations for “Such influence with varying temperatures could also exist in other VOCs oxidation systems…”.
43. Line 413: Please clarify ”… a constant yield is not sufficient to represent the SOA formation potential of β-pinene in models”.

Technical Comments

1. It will be better to use “L”, “M”, and “H” to indicate “low”, “middle” and “high” in Figures. E.g., 298L, 298M, 298H.
2. Lines 18 – 19: “Increasing [HO₂]/[RO₂] inhibited dimers and SOA formation with a higher sensitivity at lower temperatures.” What is the meaning of higher sensitivity in this sentence?
3. Line 37: It should be “volatilities”.
4. Line 40: Please use another word instead of “performances”.
5. Line 43: “fate” should be plural here.
6. Line 66: It should be “low volatilities”.
7. Lines 68 – 70: The sentence is too long to follow.
8. Lines 79 – 81: Provide references.
9. Line 110: Please replace “low-oxidized products” with “lightly oxidized products”.
10. Lines 175 – 177: Provide the wall loss rate of β-pinene although negligible.
11. Line 188: It should be “the impact of vapor wall losses”.
12. Line 197: Please indicate that Figure 2 only shows the timeseries data for 298 K in the text. In addition, same types of plots should be made for other experimental conditions but shown in the supporting information.
13. Line 206: The abbreviation Y_SOA is redundant and only used for one time.
14. Line 210: It should be “gas-to-particle partitioning”.
15. Figures 5 and 7: Please increase the gaps between two neighbouring categories.
16. Line 283: It should be “reduction” instead of “reduce”.
17. Line 306: Specify “which was critical for the results”.
18. Lines 291 – 293: Please add plots showing the formation of HCHO and nopinone as a function of the consumed β-pinene for the 248 and 273 K conditions in Figure S11.
19. Lines 319 – 320: Please indicate that the dimers are the “chosen” or “selective” dimers.
20. Line 333: What is the meaning of “less water effects”?
21. Figure 9: The colour codes are confusing. The colours used for a (L), b(M), and c(H) in subplots (B-D) should be different from those used for different temperatures in the subplot (A).
22. Lines 371 and 372: Th sentence “The [HO₂]/[RO₂] impact… from C9-SCIs reactions.” is unclear. Please rephrase the sentence.
23. Lines 377 – 379: The sentence “Since the interference of water vapor… and concentration of SCIs” is too long to be understood.
24. Can you use “CIs” instead of Criegee intermediates throughout the manuscript after the abbreviation was introduced? Same applies for RO₂.
25. The full name of HO₂ need to be introduced when it is mentioned at the first time in the abstract as well as the main text.

References:

Gao, L., Song, J., Mohr, C., Huang, W., Vallon, M., Jiang, F., Leisner, T., and Saathoff, H.: Kinetics, SOA yields, and chemical composition of secondary organic aerosol from β-caryophyllene ozonolysis with and without nitrogen oxides between 213 and 313 K, Atmospheric Chemistry and Physics, 22, 6001-6020, 2022.

Gao, L., Buchholz, A., Li, Z., Song, J., Vallon, M., Jiang, F., Möhler, O., Leisner, T., and Saathoff, H.: Volatility of secondary organic aerosol from β-caryophyllene ozonolysis over a wide tropospheric temperature range, Environmental Science & Technology, 2023.

Huang, W., Saathoff, H., Pajunoja, A., Shen, X., Naumann, K.-H., Wagner, R., Virtanen, A., Leisner, T., and Mohr, C.: α-Pinene secondary organic aerosol at low temperature: Chemical composition and implications for particle viscosity, Atmospheric Chemistry and Physics, 18, 2883-2898, 10.5194/acp-18-2883-2018, 2018.

Pospisilova, V., Lopez-Hilfiker, F. D., Bell, D. M., El Haddad, I., Mohr, C., Huang, W., Heikkinen, L., Xiao, M., Dommen, J., and Prevot, A.: On the fate of oxygenated organic molecules in atmospheric aerosol particles, Science Advances, 6, eaax8922, 2020.

Simon, M., Dada, L., Heinritzi, M., Scholz, W., Stolzenburg, D., Fischer, L., Wagner, A. C., Kürten, A., Rörup, B., and He, X.-C.: Molecular understanding of new-particle formation from α-pinene between− 50 and+ 25° c, Atmospheric Chemistry and Physics, 20, 9183-9207, 2020.

Ye, Q., Wang, M., Hofbauer, V., Stolzenburg, D., Chen, D., Schervish, M., Vogel, A., Mauldin, R. L., Baalbaki, R., and Brilke, S.: Molecular composition and volatility of nucleated particles from α-pinene oxidation between− 50 c and+ 25 c, Environmental science & technology, 53, 12357-12365, 2019.

Zhao, Y., Thornton, J. A., and Pye, H. O.: Quantitative constraints on autoxidation and dimer formation from direct probing of monoterpene-derived peroxy radical chemistry, Proceedings of the National Academy of Sciences, 115, 12142-12147, 2018.

This manuscript presents the results from a series of chamber experiments designed to investigate the formation of dimers from stabilized criegee intermediates (SCIs) as well as peroxy radicals (RO₂). The presented experimental protocol for the chamber experiments appears well-considered, and the presented experimental data is of great quality. This work also involves box-modelling of the chamber experiments to investigate aspects of the system not able to be directly measured. Again, the modelling approach employed here seems sound, with good model-measurement agreement where presented.

The topic of dimers’ contribution to SOA is of increasing interest and the results presented in this manuscript will be useful to a range of researchers when considering the implications of such processes in ozone-impacted environments. The work is of high quality and many of the comments below simply address the presentation of the results. However, there are some potential weaknesses in the analysis, particularly when attempting to provide information on the formation pathways of dimers.

I support the publication of this manuscript after the following comments are addressed.

CIMS Calibrations:

Throughout the manuscript, the authors discuss the concentrations of various compounds measured with I^--CIMS. The authors note in the experimental section that they calibrated the CIMS data to pinic acid. This is a reasonable approach given the lack of available standards for the vast majority of compounds discussed, but I believe that the authors should better highlight the large impact that this could have on the reported concentrations. Previous work has shown that compound responses to I^--CIMS can vary by orders of magnitude, even for similarly structured compounds (Lee et al., 2014). The authors should note this at the end of Section 2.2. This calibration issue can also be problematic if the species distribution of a specified group changes as the result of a change in experimental conditions. For example, in Figure 4, each C16 compound will have a different I^--CIMS response, so if the distribution of these compounds changes when the temperature changes (as the authors argue), then the average response of the C16 group will change. This could increase or decrease the observed drop in dimer signal depending on the change in average response of this group.

These calibration issues are difficult to overcome, and the approach the authors have taken is reasonable. However, the authors should discuss the potential effect of calibration issues where possible, or at least acknowledge the potential issue.

Box Models and Mechanisms:

Section 3 outlines the procedure used for modelling in this work, however there are several points in the manuscript that make it unclear how many models were run and what mechanisms were used. The authors should use Section 3 to consolidate a description of all of the model runs performed and any alterations to the MCM.

• Line 295 – From here, the authors note a number of modifications to the MCM. It is not clear where these changes are used. Are all of these changes used in the models of each experiment (e.g. the data presented in Figure 1)? If so, then these changes should be noted in Section 3 when the mechanism is introduced. Otherwise, the changes should be described as “proposed changes”, or something similar, to make it clear that this chemistry isn’t implemented in the current models. Additionally, if these changes are not implemented in the current models, then the authors may consider implementing them to test the effect on model predictions (e.g. by reproducing Figure 11 with the added chemistry).
• Line 376 – Here, there is mention of “a box model” but it isn’t clear whether this is a separate set of models or the same models as before.
• Lines 385 and 391 – New reactions and rates are listed here but it isn’t clear which models they’re used for. Again, if these are implemented in all of the model runs presented then they should be described alongside the original mechanism description in Section 3.

Use of Zhao et al. 2018:

In Section 4.2, the authors attempt to use the methodology of Zhao et al. 2018 to show that the measured particle-phase dimers do not result from the reaction of particle-phase monomers. There are three issues with the authors’ use of Zhao et al. here:

• Zhao et al. are discussing the formation of dimers in the gas phase from RO2 radicals. This means that the application of this methodology to the particle phase with closed-shell products must be suitably justified.
• A lack of a quadratic relation does not preclude the formation of the dimers from the monomers. I believe that such a quadratic relationship in Zhao et al. suggests the formation of dimers via RO2 due to a second-order formation process represented by RO2+RO2-->ROOR and a first-order loss process such as rapid uptake to the particle phase. So, if the dimers were formed from the monomers via multiple reaction steps, or there were additional losses of the particle-phase dimers (as is likely to be true in the particle phase), then this quadratic relationship would not hold.
• The presence of a linear relationship does not preclude the formation of the dimers from the monomers. Zhao et al. state that their observation of a linear relationship shows that “the dimers identified here are not due to a simple association (i.e., clustering) of closed-shell monomers.” This cannot be generalized to a general statement on the formation of dimers in the particle phase without significant justification.

Correlation with [RO₂]²:

In Section 4.4, the authors attempt to show that the gas-phase dimers measured are not formed predominantly via RO2 by correlating the dimer signal with [RO₂]². A lack of linear correlation is taken as evidence that this is not the case. The authors should explain why they expect a linear correlation with [RO₂]², or reference this methodology in another paper, as it is not currently clear.

In fact, this analysis seems to be opposed to the analysis presented in Zhao et al. 2018 which has been previously misinterpreted. Zhao et al. plot RO2 concentrations against dimer concentrations (in the same way as in Section 4.4) and use the quadratic relationship as evidence of ROOR formation from RO2. Figure 8 does not appear to show a quadratic relationship as in Zhao et al., so the authors should attempt to explain this discrepancy.

Additional Comments:

• Line 148 – “…this reaction was calculated in the model and was regarded as too slow to make a difference to SCIs reactions”. It is unclear how the authors determined this to be true. Were these rates implemented into the mechanism and shown to be insignificant? If so, then some data should be presented from such a model run to illustrate the insignificance. Alternatively, if this conclusion is simply reached by the magnitude of a rate calculated with typical CO and SCI concentrations being small then this should be stated along with the calculated rate.
• Line 176 – “…and found that the wall loss of β-pinene was negligible at all temperatures”. The authors should provide some indication of the magnitude of this loss (even if it is very small) either via a calculated loss rate or a plot of the concentrations not changing over a reasonable time period. The same is true for nopinone and HCHO discussed in the following sentence.
• Line 179 – Calculating the O3 loss rate after β-pinene was added to the chamber introduces a potential artifact resulting from the reaction of ozone with secondary oxidation products as opposed to just the chamber walls. Do the authors have data showing the wall loss of O3 in the absence of any organic compounds (either before β-pinene was added to the chamber or in a separate experiment where no β-pinene was added to the chamber)? If not, then the authors should justify why it is reasonable to calculate the O3 loss rate in the presence of organic compounds, such as the oxidation products of β-pinene.
• Line 182 – How was the loss of FA and C9H14O4 calculated? From Figure S5 it seems like the authors introduced the species into an empty chamber and then observed the decay. If so, this should be explained.
• Line 182 – Should “FA and C9H14O4 (pinic acid and homoterpenylic acid)” actually read “FA and C9H14O4 (formic acid and homoterpenylic acid)”?
• Line 187 – Figure S5 shows data for C9H14O4 but not for FA. Why is this? Would the authors be able to provide similar data for FA?
• Line 253 – The authors should clarify that Figure 4B shows the results from the ‘L’ experiments at each temperature.
• Line 271 – The uncertainty in the assignment of terpenylic acid, pinic acid and homoterpenylic acid to signals corresponding to C8H12O4 and C9H14O4 should be more explicit here. The authors cannot make a definitive assignment (unless these thermograms were obtained from chemical standards, which I don’t believe is true). As such, the chemical names should not just be listed in brackets next to the formulae, but rather should include some phrase indicating a tentative assignment, e.g. “C8H12O4 (corresponding to the formula of terpenylic acid) and C9H14O4 (corresponding to the formulae of pinic acid and homoterpenylic acid).”
• Line 283 – “…leads to a reduce of more than 40% in total dimers from 248 K to 298 K…” The authors should provide a reference to a figure that illustrates this (e.g. Figure 5)
• Line 305 – How do the results from Berndt et al. compare to the results from the models presented in this manuscript?
• Line 345 – The authors should make it clear that the discussion of ROOR focusses on the gas-phase ROOR, as opposed to the particle-phase dimers that have been discussed up until this point.
• Line 362 – “It is intriguing to find that at low temperatures, the variation of [HO2]/[RO2] has such a big influence on C18 dimers, which are significantly contributed by SCIs reactions.” The authors should provide a reference to Figure 7 which illustrates this result.
• Line 367 – The authors should state how they obtained the 90% value. Was this obtained from the models or by comparison of particle and gas-phase dimer measurements?
• Line 369 – The authors should outline how they “evaluated the influence of C9-SCIs reactions with CO at lower temperatures.”. Was this by looking at nopinone concentrations in the H, M, and L experiments in the same manner as with C9H14O4 in figure S14?

Minor Comments:

• Line 51 – “Would this lead to a more important role of SCIs in SOA formation in winter and colder regions of the atmosphere?” It is unusual to phrase this as a question here. If this is an aim of the current study then it should be stated explicitly (e.g. “This study will provide insight into whether this will lead to a more important role of SCIs in SOA formation in winter and colder regions of the atmosphere”).
• Line 52 – “vital in the atmospheric radical circle, and reactions” should read “atmospheric radical cycle”.
• Line 145 – The latest MCM version is v3.3.1, not 3.3.2 as stated here. Also, it may be of interest that the MCM is now located at http://mcm.york.ac.uk/ , the Leeds website does still work, but the York site is faster.
• Line 190 – I don’t think that the introduction paragraph here is necessary. At a minimum it should be converted to the future tense for clarity (This discussion will begin with the ... then this will be explored... etc).
• Line 283 – “…SCIs leads to a reduce of more than…” should read “…SCIs leads to a reduction of more than…”
• Figure 6 – I find the presentation of this data confusing. Particularly the hashed bars, such as C17H22O6, where it is difficult to see the 298K bar. Some potential actions to improve the readability of this figure may be: change the direction of the hashes on one temperature set which would allow the two to be better distinguished, adding a statement to the figure caption indicating that the bars for each temperature are overlaid for each compound, changing the figure colours to provide better contrast, splitting the bars out to not overlap (in a similar fashion to figure 5 or figure 7).
• Line 327 – There should be a new paragraph between these two sentences to indicate that the discussion has shifted from temperature effects to RH effects.
• Figure 4B – The y-axis label is cut off

References

Lee, B. H., Lopez-Hilfiker, F. D., Mohr, C., Kurtén, T., Worsnop, D. R., and Thornton, J. A.: An iodide-adduct high-resolution time-of-flight chemical-ionization mass spectrometer: Application to atmospheric inorganic and organic compounds, Environ. Sci. Technol., 48, 6309–6317, https://doi.org/10.1021/es500362a, 2014.

Gong et al. investigated the formation of sCI- and RO₂-related dimers from the ozonolysis of β-pinene at different temperature. The authors found that temperature could influence the mechanism and production distribution of sCI or RO₂ reaction based on the analysis of experimental and modelled results. Overall, the topic of this manuscript is interesting, and the designed experiments are reasonable. However, the discussion is not in-deep. I would recommend the author to provide more explanations on experimental results. This manuscript would be published if the following comments could be addressed.

Specific comments

1. Line 19. How to define low and high [HO₂]/[RO₂] conditions?
2. Abstract should be written in a more concise manner.
3. Line 41. Which trace gas species?
4. Line 80. This manuscript focused on dimer formation from RO₂ or sCI reaction. The authors mentioned that different isomers of monoterpenes have different reaction mechanisms due to their different molecule structures. It is necessary to briefly summarize the current knowledge about RO₂ and sCI from β-pinene ozonolysis.
5. Lines 75-84. It would be better to describe the research topic, focus and content of this manuscript in the end of the Introduction.
6. Line 100. Again. How to define L, M, and H conditions based on [HO₂]/[RO₂] ratio? Please specify that.
7. Line 108. In order to ensure that SOA precursors were consumed to a negligible level, the initial amount of ozone was generally 3–4 times higher than that of SOA precursors. In this manuscript, the mixing ratios of β-pinene and O₃ were ~17 ppb and ~1100 ppb, respectively. Ozone of excess amount could initiate heterogenous reaction and lead to the changes in SOA composition. Why did the authors use such high ozone concentration? Reasonable explanations should be shown in the manuscript.
8. Line 196. It is illogical to describe the result first and then introduce Figure 2 in the second sentence.
9. Line 207. SOA yield is related to its mass concentration. Both experimental condition and SOA mass concentration in previous studies should be noted.
10. Line 225. The authors showed experiment results about the dependence of SOA formation on [HO₂]/[RO₂] at different temperature. However, discussions about the influencing mechanism of [HO₂]/[RO₂] on SOA yield were not provided in detail. The authors should give more explanations based on MS-measured data and modeled data.
11. Line 229. Could the authors measure the amount of sCI? How did the authors calculate the proportion of the scavenged sCI?
12. Line 283. Is the proportion of the scavenged sCI determined to be 70% in all experiments?
13. Line 296. How did the authors measure HCHO and nopinone? Based on the yield of HCHO and nopinone, the authors carried out some updates in the MCM mechanism. Please provide some justifications about MCM updates. Did the author compare the modelled results before and after the update?
14. Lines 346. RO₂ could participate in dimer formation via its self or cross reaction. How did the author consider the RO₂ self reaction?
15. Line 692. There is no experimental information in the caption.
16. Many sentences in this manuscript are too long to obtain useful information. Please double check.