General Comments:

The manuscript by Sun et al. presents the mass spectral characteristic of cloud water samples throughout a long-lasting cloud event by FT-ICR-MS, and attempts to shed light on the potential influences of in-cloud aqueous phase reactions, which are currently uncertain for the formation of SOA. They show that CHON with aromatic structures are the most abundant type in cloud water, suggesting their enhanced formation in cloud. Their results also indicate distinctly differences between day and night, which is most probably attributed to diurnal differences in aqueous chemistry.

Such observation could provide valuable cloud chemistry data for the community, and has the potential to be published after considering my comments. The major weakness is the limited dataset, thus the authors have to clearly indicate in the discussion of the diurnal difference of cloud chemistry between day and night, since there could be other factors contributing to such difference.

Specific Comments:

-Introduction: Overall it is OK, but it would be better to include the aqueous formation mechanisms related to CHON and CHOS.

- Lines 172, “the current understanding that aqueous-phase reactions generally increase the degree of oxidation (Ervens et al., 2011).” Please also include the reasons to this understanding. Does such aqueous reactions refer to in cloud processing?

- Lines 182, “The O/C ratios and OSC of CHO collected during the daytime is slightly lower than the nighttime…”. What about the influence of primary emission? Since the samples collected during the daytime and nighttime may originally presents different characteristics without oxidation.

- Lines 195, is there any result of aromaticity related to traffic emission or other sources, in addition to coal combustion and biomass burning? Since the present OA molecular does not correspond to these sources, i.e., coal combustion and biomass burning as discussed.

- Lines 251, It is an interesting result that coal combustion contributes to S-containing formulas in cloud water more significantly compared with CHO and CHON. Is there any other evidence to support the demonstration, such as the correlation between CHOS with the concentration of so2 or sulfate?

- Lines 278, “For CHO, the most abundant C17H26O4 in cloud water is not detected in the PM2.5 samples, suggesting a formation by the in cloud aqueous-phase reactions, although the contribution from BVOCs cannot be ruled out.” Reasons should be discussed for such a contradiction.

This work provides a comprehensive analysis of the compounds in cloud water as well as the interstitial PM2.5, and characterize the distribution of different groups of species by using FT-ICR-MS; daytime and nighttime comparison was also made. Based on such analysis, aqueous-phase processing and the reactions involved were inferred to enhance our understanding of the aerosol chemistry. The paper is fairly well written and provide useful information and knowledge regarding the cloud water organics, this reviewer however has a series of comments to be addressed first before its acceptance

• Line 51: as you stated here, “chromatographic and spectroscopic techniques only determined ~20% of all kinds of organics”, then what does the FT-ICR-MS perform? Even though it has a super high mass resolution, is it being able to determine all existing species? If not, organics with what functionalities are preferred to be detected? What is the fraction of determined species to the total? How does this bias affect your interpretation? I think this issue should be clarified in your manuscript.
• Section 2.2: Even though instrumental details are included in the supplement, I think some key information, for example, the mass resolution, and how to remove background organics, etc can be briefly described here, as well as the IC and TOC/TN analysis.
• As described in (2), background organics or impurities during sample storage and treatment, might be detected as FT-ICR-MS is highly sensitive and has ultrahigh resolution. The number of molecules in cloud samples seem to be much higher than those in PM2.5 samples, I am wondering how do these excess compounds come from except from possible aqueous-phase processing?
• Line 170-179: This reviewer thinks that cloud cycling might need to be considered, as the interstitial PM2.5 sampled here may contain aqueous oxidation products inside cloud droplets as cloud droplets in reality cycle per few minutes. Therefore a high O/C value might be observed in PM2.5 samples rather than cloud water. Whether or not aqueous processing could enhance the oxidation degree of organics depends on the ageing time. In a short time scale, the organic oxidation degree could increase and the more oxidized species may fragment into low oxygenated ones given enough time.
• Line 286-292: Similar for OS, typically OS can be produced more efficiently in aerosol water rather than liquid water, yet no statistical difference are observed here, as there are repeated cycling between cloud water and interstitial PM2.5.

Specific comments

• Line 25-26, do you mean CHON and CHO-containing species? It is not clear. The last sentence in Line 26 is not a full sentence.
• Line 28: A recent paper by Wang et al (2021, 118:e2022179118) demonstrates that aqueous-phase oxidation of aromatic species could be a source of SOA, this might be a supporting evidence that “CHON with aromatic structures are abundant in cloud water”
• Line 69: Consider to add citation， Ye et al., Atmos Environ 2020;223:117240, which determines the organic acids produced from aqueous-phase oxidation of a certain precursor.
• Line 144: RA means relative abundance, this reviewer somehow think the authors can directly use "relative abundance", it is easier to understand than RA。