General comments

In this work, the authors characterized ambient OOMs in a densely populated urban site in Shanghai using an ultra-high resolution orbitrap coupled with a nitrate inlet. With its high mass resolving power, CI-Orbitrap gives more accurate identification of molecular composition and better separates N-containing OOMs in complex ambient data. The authors showed that aliphatic 2N-OOMs were the most abundant 2N-OOMs with significant contributions from those derived from long-chain aliphatic compounds, whose fraction was further increased on polluted days. This paper is overall well designed and nicely written. There is one point I am not very convinced about: the authors seemed to directly use the correlation between 2N-OOMs with solar radiation over NO₃ radicals to reference daytime and nighttime formation pathways without considering the effects of meteorological dilution (e.g. the diurnal variation of boundary layer height). It can change or even flip the correlation in some cases. I also have some minor questions as listed below. I recommend publication after these issues are addressed.

Specific comments

Line 40-43: Most field measurements cited here were conducted in the United States and China. There are many available works from Europe, too. For example, the European sites included in Ng et al. (2017) show up to >70% ON in ambient submicron OA. Can the authors also include these data points to the paper, and if possible, more from other parts of the world?

Line 128, 131: Some references are needed for “widely used in previous studies” and “+-50% according to error propagation”.

Line 142: Some references are needed for “one of our companion studies”.

Figure 1a: The font for pie chart percentages is too small (same for other figures).

Figure 1b: How did the authors explain 3N-OOMs fraction decreased when nC>12? 

Line 185: How is the 2N-OOMs enhancement in polluted days compared to the enhancement of total OOMs?

Reference

Ng, N. L., Brown, S. S., Archibald, A. T., Atlas, E., Cohen, R. C., Crowley, J. N., Day, D. A., Donahue, N. M., Fry, J. L., Fuchs, H., Griffin, R. J., Guzman, M. I., Herrmann, H., Hodzic, A., Iinuma, Y., Jimenez, J. L., Kiendler-Scharr, A., Lee, B. H., Luecken, D. J., Mao, J., McLaren, R., Mutzel, A., Osthoff, H. D., Ouyang, B., Picquet-Varrault, B., Platt, U., Pye, H. O. T., Rudich, Y., Schwantes, R. H., Shiraiwa, M., Stutz, J., Thornton, J. A., Tilgner, A., Williams, B. J., and Zaveri, R. A.: Nitrate radicals and biogenic volatile organic compounds: oxidation, mechanisms, and organic aerosol, Atmos. Chem. Phys., 17, 2103–2162, https://doi.org/10.5194/acp-17-2103-2017, 2017.

This manuscript presents measurements of N-containing oxygenated organic molecules (N-OOMs) in a Chinese megacity made using a nitrate-orbitrap chemical ionization mass spectrometer (CIMS), with a focus on N-OOMs containing two N atoms (2N-OOMs). Using an established workflow, the 2N-OOMs are classified based on possible precursors, aliphatics, aromatics and monoterpenes, and the abundances and properties of the 2N-OOM classes are examined as a function of pollution level and time of day. Overall, the results presented in this manuscript are interesting but somewhat incremental given the recent companion paper by Zhang et al. 2022 (DOI: 10.1021/acs.est.1c08346). In several instances, the data/figures also do not appear to support the stated discussions/conclusions. As such, I recommend that publication as a Measurement Report be considered only after the comments detailed below are addressed.

1. Nitroaromatics. It is assumed that all N atoms in N-OOMs are nitrate groups. For aromatics, however, recent laboratory work by Xu et al. 2020 (DOI: 10.1021/acs.est.0c04780) suggests that nitrate yields are low. Moreover, nitroaromatics are known to be significant products of aromatic oxidation under high-NO₂ conditions. Accordingly, despite the authors stating that nitrophenol peaks were excluded (L269), the contribution of nitroaromatics should be accounted for. Additionally, relevant sections in the manuscript should be rephrased to state that OOMs were classified based on the number of N atoms not nitrate groups (e.g., L45 and L146).
2. Concentration Determination. Given that the calibration factor for H₂SO₄ was applied for all detected OOMs, the detailed derivation (L116–125) is unnecessary.
3. L154: Multistep Oxidation. By this argument, the fraction of 3N-OOM should exhibit a similar dependence with on carbon number (nC), with a maximum at higher nC than the 2N-OOM. A discussion of potential reasons for the absence of this behavior should be included.
4. Percentages. Percentages reported throughout the manuscript should specify the period over which the calculations were performed (e.g., campaign averages, daytime averages, or nighttime averages).
5. Figure 1. It is difficult to distinguish the different shades of red, blue, green, and gray from one another. The authors should consider using hatching/patterns to differentiate carbon numbers within each class.
6. Section 3.2. The analysis in this section does not appear to reflect the key takeaways from Table 1 and Figures 2 and 3. Table 1 indicates that with the exception of aromatic 2N-OOM (2N-OOM_Aro) during the day, the abundances of all 2N-OOM classes during both day and night were higher during the polluted (PL) as compared to clean (CL) period, not just aliphatic 2N-OOM (2N-OOM_Ali) as the authors highlight. The pie charts in Figure 2 show that the relative fractions of the different 2N-OOM classes do not vary considerably as a function of pollution level, especially at night. Figure 3 illustrates that this is also true for the fraction of 2N-OOM_Ali comprised of C_4–10 vs. C_>10 species. As such, the authors should consider restructuring this section to emphasize that pollution levels modulate the absolute concentrations of 2N-OOM but not the relative fractions of 2N-OOM classes or subclasses, whereas daytime vs. nighttime chemistry exerts modest control over the fractional contribution of 2N-OOM classes but does not alter the relative importance.
7. Importance of Nighttime Monoterpene Chemistry. In L200, the authors state that monoterpene 2N-OOM (2N-OOM_MT) “showed overall higher concentrations during nighttime than those in daytime.” In L242, they state that “nighttime chemistry dominated the formation of 2N-OOM_MT…supported by the stronger correlation between 2N-OOM_MT and NO₃ radicals than solar radiation.” However, in Table 1, the daytime and nighttime concentrations of 2N-OOM_MT during the CL period were the same (2.3 x 10⁶molec. cm^-3) and only slightly higher at night than in the day during the PL period (5.3 vs. 4.3 x 10⁶ molec. cm^-3). Furthermore, Figure 4 shows that the R values for 2N-OOM_MT with solar radiation and NO₃ radicals are similarly weak, ~0.15 and ~0.18, respectively, while Figure 5 suggests that the number of 2N-OOM_MT species that exhibit a strong correlation (R > 0.5) with solar radiation is greater than that with NO₃ radicals. As such, these data do not appear to support the conclusion that nighttime chemistry dominated 2N-OOM_MT formation.
8. PMF Results. Given that a combination of “up-to-date understanding of VOC oxidation as well as PMF results” was used to derived the 2N-OOM classes that are the focus of this manuscript, the PMF analysis and factors discussed in L188–191 and L224–241 seem superfluous. The authors should consider moving these discussions to the SI.
9. NO_x vs. NO vs. NO₂. In L282, the authors state that “high NO_x concentrations during the nighttime (Table 1)…could efficiently suppress the RO₂ radicals from autoxidation reactions.” In L298 they state that “nO_eff of 2N-OOM_Aro and 2N-OOM_MT increased with the decrease of NO_x concentrations...likely due to the prevailing of NO_x termination reactions.” In Figure 7, the authors plot nO_eff as a function of NO_x. As shown in Table 1, however, the increase in nighttime vs. daytime NO_x concentrations is driven by NO₂; NO concentrations during both the CL and PL periods are actually lower at night than during the day. Given that reaction with NO not NO₂ modulates the fate of RO₂ (except for acyl RO₂) and that reaction of NO₂ with aromatic alkyl radicals represents an important mechanism of N incorporation, the authors should reanalyze these trends as a function of NO and NO₂ independently.

Minor Comments

L99. A map indicating the location of the field site in the SI would be beneficial.

L142. Reference to companion study is missing.

L143. How were saturation mass concentrations of OOMs calculated?

L164. A detailed description of the approach used to classify the 2N-OOMS should be included.

L158. Reference is missing.

L240. Given that isoprene is not emitted at night, is the hypothesis that this N-OOM is forming from residual isoprene? Does the diurnal profile support this?

Lu et al applied a CI-Orbitrap to a field campaign in urban Shanghai. The motivation for the work is to offer higher mass resolving power for oxygenated (and nitrogenated) organic matter (OOM) in aerosol compared to previous works using CI-API-TOF. The CI reagent is the nitrate anion. They used positive matrix factorization to estimate the source contribution of the OOM that has 2 nitrogen groups identified on the analyte molecular composition. The paper offers interesting insight into nitrogen-containing aerosols and urban chemistry. I would suggest publication after the following comments are addressed.

Major comments:

Line 127. The sensitivity of H2SO4 was used for all OOM analytes; the only justification given was that this has been “widely used” in other studies. Can the authors offer more compelling justification? The authors go on to say that H2SO4 has the highest sensitivity of all low-volatility vapors, which seems to contradict that H2SO4 is a good surrogate for sensitivity. They didn’t disclose the range of sensitivity discrepancies between H2SO4 and OOM proxy compounds. The error margin of 50% is estimated based on error propagation from equation 2, which I find confusing considering that the error margin for the sensitivity estimate is not known… I suggest the authors to do one of two things:

(1) Please cite at least one study that quantitatively demonstrates how different the sensitivities of authentic standards of OOM-like compounds are compared to H2SO4 and give that sensitivity error range separately. Clearly state the dependence of the sensitivity on relative humidity.

(2) put some authentic OOM surrogate standards in your mass spectrometer at different relative humidity and different m/z to quantitatively determine the range in errors from the sensitivity estimation alone. It is appropriate for CIMS instruments do this calibration prior to field measurements.

Line 249-255: The nighttime chemistry in Shanghai is fascinating due to the high NO levels at night, compared to a rural location where NO levels dropped to near zero after sunset (https://acp.copernicus.org/articles/16/7623/2016/acp-16-7623-2016.pdf). I suggest to state this clearly so the reader can have better context for the chemistry, e.g., in line 249 “Under the nighttime conditions observed in Shanghai (Table 1), it is estimated that monoterpenes primarily react with NO3, and the fate of nighttime RO2s is dominated by NO. Therefore, formation of…”

The authors should offer quantitative support for their suggested mechanism by doing a quick kinetic calculation based on IUPAC rates and the values in Table 1. In any case, with those concentrations of NO, I can estimate that NO reacts with something like >99% of the RO2s, consistent with the mechanism that the authors have discussed.

In an interesting contrast, Bates et al (ACP 2022, https://doi.org/10.5194/acp-22-1467-2022) did a model of the more rural location and found that nitrated RO2s were reacting primarily with HO2 and RO2. The result is that for the site studied in Bates et al, they might see more 2N-OOM with carbon number 20 (due to RO2+RO2 --> ROOR reaction, where each RO2 has one nitrate group) when in urban locations, the 2N-OOM may be more prevalent for carbon number 10 for the monoterpene + NO3 chemistry (where the 2N product is from the RO2+NO --> RONO2 reaction). This is worth adding to the discussion, as it highlights that mononitrates and dinitrates from this chemistry may be prevalent at multiple locations for different reasons.

Lines 256-259: This discussion is a bit confusing. Can the authors pick one monoterpene and give the result of the quantitative assessment of reactivity? E.g., for 22.8 ppb of ozone and 1.3 ppt of NO3, X% of monoterpene reacts with NO3 and Y% of reacts with ozone? Also Table 1 shows that the polluted nighttime case is 6.2 ppt, it would be worth calculating for both clean and polluted nighttime cases.

Line 262: I think the Figure S8 offers clearer evidence to support the NO3-related claims of the paper, whereas some of the other figures (e.g., Fig. 4) do not tell a clear story, as evidenced by Reviewer 2’s comments. It might be worth to bring S8 to the main text and to offer a correlation factor R2 analysis for these two data products?

General comment is that the nitrogen compounds from aromatics should also consider the nitroaromatics.

Minor comments

Line 44. What does “under different atmosphere” mean?

Page 4, when citing mass resolving power please also note the mass that this number is calculated at. Each m/z in the spectrum has a different mass resolving power associated with it. For example, 12000 m/dm at m/z 400.

The authors use both “mass resolving power” and “mass resolution” throughout to refer to the same entity, it would be more correct and consistent to just use “mass resolving power.”

Line 290: Please change to “…known that NO is also critical in determining the fate of RO2…” NOx (NO+NO2) is ambiguous, and will not matter much if all of the NOx is NO2. Same comment on line 294, 299, 303 etc. The authors should re-do this analysis with NO not NOx.