Review on  article https://doi.org/10.5194/acp-2021-20 by Zhang et al., Formation and Evolution of Secondary Organic Aerosol Derived from Urban Lifestyle Sources: Vehicle Exhaust and Cooking Emission

This article characterizes secondary aerosol formation potential from vehicles and cooking that are two very important aerosol sources in urban areas. The secondary aerosol formation potential from these sources remains poorly characterized, so the topic is timely and important. Experiments done in laboratory and field seem to be comprehensive and novel in many aspects. The article is nicely written and relatively easy to follow, however it would be important to revise the English language. Also, the experiments done and instruments used need to be described better. Also, the PMF results (e.g. how the number of factors was decided, how they were identified, how many factor solutions were tested and why the were not chosen etc) needs to be described better in order to reader to understand the results. 

Detailed comments

L20-21: is the source of SOA the traffic and cooking or city dwellers? please clarify sentence

L24: define POA

L24: what instrument was used to measure SOA/POA in this case? AMS or SMPS?

L26: define “vehicle group” and “cooking group”

L35: replace could with can

L36: I propose reformulating: POA is directly emitted into ambient air through several sources such as coal combustion, biomass burning, vehicle exhaust, cooking procedure.

L39: change to: but models typically fail to simulate..

L47-49: Please reformulate by changing "would -> can be".

L52: I am not sure if Manchester meets the definition of a Megacity.. please check..

L52: replace lifestyle source with SOA source. 

L65: replace Lab with laboratory all through the manuscript

L83, 102: define China V

L102: Is this vehicle or engine that you are using to produce the exhaust? describe the engine/vehicle (manufacturer, model, engine size, aftertreatment, mileage, year, fuel, lubricant oil) in detail. All these have a major impact to SOA formation. Also, if engine, maybe add information that this is not exactly same as vehicle, assuming there is no catalysts (like nowadays almost all have). Please add discussion about the differences between engine produced SOA and vehicle produced SOA.

L102. Accronym GDI should be Gasoline direct injection

L104: please, describe the used sampling and dilution setup in detail. How the dilution air was cleaned? dirty dilution air can be major source of SOA. Was the vehicle exhaust taken directly from tailpipe? and cooking fumes from the room air? if so, what it the influence of this extra dilution/aging in the room air in the case of cooking fumes? For the cooking, maybe explain how the boiling water acts as a blank. Was a blank/zero measurements done for the vehicle? Also, I find it hard to find how these measurement points were taken. According to table there was 6 points with different OH concentrations, and 3-5 repetitions for each. Please clarify how long each point was, were the results of repetitions averaged? if so, maybe give standard deviations to values in figures/tables to describe the variability between repetitions.

L117: This chapter is quite unclear and missing quite many details. Please, explain the AMS and GoPAM in more detail, e.g. in separate chapters and the rest shortly like it is now.

L120: I think the reference describing Aerodyne instrument should be: 

L120: I think the reference describing Aerodyne instrument should be: DeCarlo, Peter F., Joel R. Kimmel, Achim Trimborn, Megan J. Northway, John T. Jayne, Allison C. Aiken, Marc Gonin, et al. “Field-Deployable, High-Resolution, Time-of-Flight Aerosol Mass Spectrometer.” Analytical Chemistry 78, no. 24 (December 2006): 8281–89. https://doi.org/10.1021/ac061249n.

L128: I think this should state that the measured CO2 concentrations (Model 410i, Thermo Electron Corp.) were used to conduct CO2 correction for AMS data in order to reduce the CO2 interference to organic fragments in mass spectra of HR-ToF-AMS.

L125-126: You are saying that “The SMPS-1 determined the mass concentration of POA, while the SMPS-2 determined the mass concentration of aged OA, and their mass difference could be regarded as the SOA.”. are these results shown in somewhere? 

L143: The ionization efficiency (IE), relative ionization efficiency (RIE) and collection efficiency (CE) were determined individually before data processing. How were these determined? please give the results.

L146-147: please give the range of CE values and average CE value

L147: SOA was defined earlier as SMPS2-SMPS1. How does this results gained from PMF compare to SMPS results?

L177: I don’t really understand what this means. Please clarify the sentence: “The mixing and wall loss conditions have already met our experiment needs.”

L183: should this be secondary aerosol formation potential?

L184: How do you know the functionalization increased as the photochemical age increased?

L184-185: How the SOA is defined here? is this all mass (POA+SOA) after the go:PAM or is it the SOA separated by PMF or calculated from the SMPS? please, define the terms you use and use them systematically. Include the same information to the figure captions.

L186: define term “mass growth potential”

L191-194: are these aromatics, cycloalcanes, fatty acids etc compounds you measured or found from literature? if literature, please formulate the sentences so that it is clear that this is information found from literature.

L219: please define f43 and f44

L222: remove apparently

L247-249: are the fx fractions average fractions for all measurement points with different OH-exposure and different speeds? maybe give range or standard deviation to describe the variation within the dataset.

Figure/table captions: please, add all necessary information to figure/table caption. E.g. is the shown data average values, are the shown bars standard deviations, which instrument was used to measure data, etc.

Figure 2. Define what data was used for SOA/POA in the figure.

Zhang et al. measured SOA formation from emissions from a GDI engine and four types of food cooking emissions that were exposed to OH radicals in a Go:PAM. Equivalent atmospheric aging timescales of up to 5 days were studied. The SOA/POA ratio was approximately 100 for the oxidized vehicle exhaust and 2 for the oxidized cooking exhaust. Higher SOA oxidation states were observed for the oxidized vehicle exhaust. AMS spectra of the oxidized emissions were examined and compared to ambient LO-OOA, MO-OOA, and COA factors resolved using PMF. The studies are well motivated. In its current form, there are too many important experimental details that are missing for me to support publication in ACP. 

Comments

1. Critical details about the sampling inlets between the source emissions and the OFR (e.g. tubing length, diameter, material, residence time) that could influence the penetration efficiency and/or delays in transmission to the OFR (e.g. Pagonis et al, 2017) are missing. If these conditions vary between vehicular exhaust and cooking exhaust measurements, that is one of several potentially important variables that could complicate direct comparison of results between the two studies.
2. It is not clear to me why different OFR conditions were used in the vehicle exhaust and food cooking experiments. For example, the residence time was 110 s and the RH was 44-49% in the GDI exhaust studies, compared to 55 seconds’ residence time and RH = 18-23% in the food cooking studies. This makes it more challenging to directly compare results from the different studies – for example, although there may be overlap in OH exposure between the different experiments, timescales for gas-to-particle condensation and any humidity-dependent heterogenous chemistry are different.
3. More detailed information about the gas-phase measurements are necessary to interpret the results. For example, it is not clear to me whether more SOA is formed from aging of vehicle emissions because the VOC concentrations are higher, the SOA yields of those VOCs are larger, or both. Please add table(s) showing the list of VOCs that were measured from each source, their emission factors relative to CO₂, and their OH rate coefficients that were used to calculate the total external OH reactivity.
4. L96: A field study at IAP is mentioned here, but it is not clear until much later (L254) that results from this study are (I think) already published in the Li et al. (2020a) paper that is referenced much later in the manuscript.
5. L112: Given the presumably large emission factors of unsaturated fatty acids in the cooking emissions, and their corresponding fast reaction rate coefficients with ozone, it would have been useful to conduct control measurements to measure the ozonolysis products of the cooking emissions. Why were those experiments not performed here?
6. L136: The information in Tables 1-2 indicates that the sample line temperature was 20-25^oC, presumably at or close to room temperature. How did the authors determine that this temperature was sufficient to “prevent freshly warmed gas from condensing on the pipe wall”?
7. L137: What are the particle backgrounds when the lamps are turned on with ozone and humidified air flowing through the Go:PAM? Simply flowing dry purified air and ozone through a dark OFR is likely insufficient to clean out the OFR between experiments that employ OH as the oxidant. In this case, the background concentrations are probably significantly underestimated because as soon as the lamps are turned on, there is the potential to generate SOA from the OH oxidation of background contaminants that are not reactive towards O3
8. L159: In addition to the OFR conditions that were summarized in Tables 3-4, the actinic flux at 254 nm (or, alternatively, the ratio of O₃ measured before and after photolysis at 254 nm) is also a required input to the OFR254 OH exposure estimator. Please add this information to Tables 3-4 or describe in the text.
9. L161-L165: This discussion is confusing and in places incorrect. I don’t understand why acetylacetone is mentioned at all if it is not present in the emissions, whereas other specific VOCs that were measured are not discussed here or anywhere else in the paper. Also, I suspect ozone is likely an important oxidant for some species, especially unsaturated fatty acids that are presumably important components of the cooking emissions as discussed in the paper. For example, for oleic acid, at OH and O₃ exposures of 2.7E11 and 5E15 molecules cm^-3 s (OFR conditions in Line 5 of Table 4), assuming effective OH and O₃ rate coefficients of 3.5E-11 and 2.1E-15 cm³ molecule^-1 s^-1 (Renbaum et al, 2012), the estimated fractional oxidative loss of oleic acid to O₃ is ~0.55. Thus, O₃ may actually be the major oxidant for several important compounds emitted from the cooking sources. 
10. L203: The O/C ratio is insufficient by itself to associate the mass spectra of the SOA with ambient PMF factors.
11. L207 and L227-L228: These statements are too speculative, and references are made to other source characterization studies that are not directly relevant to the sources that were characterized here. Molecular speciated measurements of VOCs were performed with GC-MS as described in L168-L172 that are not discussed in the text. Those measurements should either support (or not) this interpretation of the results. And the fatty acids that are mentioned have high effective OH rate coefficients (e.g. Renbaum et al., 2012), so it is not obvious to me how the statement that “cooking produces more hardly oxidized acids” is justified.
12. L225: Is “souring” a typo? If not, I am not certain what this statement means.
13. Figures 2-3, 7. I find it confusing/distracting to have two different photochemical age and/or two SOA:POA axis scales on the same figure.
14. Figure 6: I think that the numbered symbols represent the integer m/z values of the average vehicular and cooking exhaust AMS spectra, but this should be made clearer in the legend or the caption.

References

L. Renbaum-Wolff and G. D. Smith,“Virtual Injector” Flow Tube Method for Measuring Relative Rates Kinetics of Gas-Phase and Aerosol Species, J. Phys. Chem. A, 116, 6664−6674, dx.doi.org/10.1021/jp303221w, 2012.

Pagonis, D., Krechmer, J. E., de Gouw, J., Jimenez, J. L., and Ziemann, P. J.: Effects of gas–wall partitioning in Teflon tubing and instrumentation on time-resolved measurements of gas-phase organic compounds, Atmos. Meas. Tech., 10, 4687–4696, https://doi.org/10.5194/amt-10-4687-2017, 2017.