In this manuscript, the authors present an attempt to quantify toluene oxidation intermediates by establishing quantitative and semi-quantitative calibration methods for I-CIMS. Specifically, the authors established semi-quantitative equations for four distinct categories: monophenols, monoacids, polyphenol or diacid species, and species with multiple functional groups. This classification method enhances the accuracy of the semi-quantitative approach (R2 from ~0.50 to >0.88). Overall, the research goal of this manuscript is novel and has practical atmospheric significance. The description of the calibration methods and experimental results is logical and comprehensive. After the authors address the minor comments below, the manuscript can be published in AMT.

1. Line 42, “H3O+ ions are used for the detection of VOCs”. This description here is not accurate. Conventional PTR has the ability to detect some I/SVOCs, although not many species. The newly developed Vocus or Fusion PTR can detect more I/SVOCs, some of which are oxygenated compounds (e.g., Atmos. Meas. Tech. 2019, 2403-2421).

2. Lines 53-76, some key references are missing when introducing the calibration method of CIMS and its research progress. For example, Li et al. (Environ. Sci. Technol. 2021, 12841-12851) used 22 organic standards to calibrate I-CIMS and reduced the uncertainty in total organic carbon concentrations to ~20%-35% when combining the voltage scanning approach.

3. Line 172, “humidity” should be “relative humidity”.

4. Section 3.1, it would be more straightforward to list the sensitivities in a table, probably in Table S1.

5. Line 230, the sections and figures in the SI should be presented in order in the manuscript.

6. Figure 2, it is difficult to match the data points to compound names. Adding some arrows may help.

7. Lines 363-370, could the authors show the time series of C7H8O4, C7H10O4, C7H10O5, C4H4O3, and C5H6O3 separately somewhere in the SI? It would be helpful to see the ratio of C5H6O3 to C4H4O3 as well.

8. Lines 401-403, there is no need to repeat these numbers in the conclusion.

9. Check the capitalization of the first letter. Some examples: Line 60, “Discovered”; Line 149, “Computational”; Line 213, “2,6-Xylenol and Texanol”; Compound names in Figure 2; Line 288, “Salicylic acid, Citric acid”.

In their manuscript "Optimizing Iodide-Adduct CIMS Quantitative Method for Toluene Oxidation Intermediates: Experimental Insights into Functional Group Differences," the authors present an in-depth look at understanding the sensitivity of chemical ionization mass spec using iodide as the reagent ion, and propose a method to improve calibration using calculation of binding enthalpies and classification by functional groups. Overall, I think it is a well-written paper that presents its findings well and is well grounded in the current literature. The topic is of interest to the readership of this journal and I generally support its publication here. I have one major comment/concern that needs to be addressed, but otherwise I believe it is publishable with fairly minor revisions.

Major comment:

The authors present a method for calibration based on binding enthalpies, an idea that has been demonstrated before but is elaborated upon nicely here. The conclusion, at times implied and at times fairly explicit, is that this approach will provide lower uncertainty that other methods, particularly the voltage scanning method currently sometimes employed. I do not dispute significant concerns and limitations of the voltage scanning method and I am not trying to make a strong pitch for it per se, but I believe the authors are somewhat too rosy about their method and have not properly discussed the limitations or provided a fair assessment of its uncertainty. Examples of my concerns in how the results are being viewed optimistically are below.

Line 301. In Figure S6, there are a number of compounds with low binding energies and high sensitivities. These do not seem to be any of the classified compounds (in Figures 6b-6e, none go down to enthalpies below 10). These points seem to be driving a lot of the poor correlation, so is the observed improvement in R2 because of a true improvement, or more becomes some outliers seem to be excluded. It is also a little hard to tell, but from Figure S6 it looks like most things are on the same line, except the multi-functional compounds, is this correct? In other words, if you just exclude the multi-functional compounds from Figure 6a and do the fit, do you get a similarly high R2? I'm not sure that would negate some of the conclusions of the authors, but it seems an interesting fact if I am interpreting correctly.

Line 308-314. This comparison in uncertainty is unfair, in part because it is fairly circular and in part because it is optimistic. Essentially, in this study, a calibration curve is made between binding enthalpy and sensitivty for subsets of compounds, then that calibration is reassigned to the same compounds and found to not have a lot of error. It is true this would not work well if the correlation between binding enthalpy and sensitivity were poor, but that is all this shows, which the R2 has already shown (consider for example, mono-phenols, for which a line is drawn for only 4 points, then this line is reapplied to these four points to demonstrate low uncertainty). This also depends on selecting the correct classification for the mass, as it is demonstrated in Figure 3 and S6 that without classification there is a wide range. It's not clear to me how to assign a classification to compounds that were not introduced by standards (nor is it really discussed in detail). A third paper by Bi et al. in 2021 that I think is not cited here (10.5194/amt-14-3895-2021) suggests that photolysis of most precursors, including aromatics, produces many isomers for each formula. How would one go about assigning the proper classification to an ion in the absence of standards, when it is possible that the ion contains multiple distinct structures which may have different functional groups? Of course all methods have their limitations, but the 40% estimate presented here is a best case that really only applies for compounds that were used to generate the calibration curve. I don't understand how the 40% could be considered an upper limit (it seems to me an upper limit would be caused by mis-asigning classification, in which case Figure 3 implies the error could be as high as 2 orders of magnitude).

Relatedly, in Figures 2 and 3 and discussion thereof - do the authors have an idea of why multi-functional compounds are so much more sensitive for a given binding enthalpy, but this is not true for diacids? Why, for instance is pinonic acid so much more sensitive than fumaric acid, though the second acid groups is likely more binding than the keto group; this could be a sterics issue due to the double bond, but the same question applies to adipic or glutaric acid, which is on a much lower slope than pinonic acid - why would the diversity of functionality matter? Or put another way, how come 2 acid groups doesn't "count" as multi-functional?

Line 349. What do the authors mean less differences? Are they again referring to the comparison to calibrants? As an example of my concern, see Figure S10. How is relative sensitivity assigned here? For example, there are 3 points that fall noticeably off the line at dV50= 4, 5.5, and 6.5. One explanation is that dV50 does a bad job of capturing sensitivity. The other explanation that does not seem to be considered here, is that dV50 is correctly estimating sensitivity, and the relative sensitivity estimated by the binding enthalpy approach is incorrect (e.g., assigned to the wrong classification). To be very clear, I do not mean to imply either case is correct, but rather that there are limitations to the proposed method developed here that are not really being considered, and could be providing spurious understanding.

Technical comments:

Line 38. "identify" isn't really the right word here. It has been shown (e.g., Riva et al. 2019, Isaacman-VanWertz et al. 2018, as cited) that current CIMS can see essentially all the reactive organic compounds, but they are generally only classified by molecular formula, not really identified, which to me implies some knowledge of molecular structure. This applies at line 100 as well, where I think "identified" should be changed to "classified by exact molecular weight."

Line 59. Typo on semi-quantitative

Line 60. "Discovered" should not bed capitalized

Line 103: "and making the absolute" is incorrect grammar and should be corrected

Section 2.2. The order of this section is a little confusing. In the beginning of the first paragraph is a dscussion of what is "often" done for calibration, then a mention of humidity, and only then is it made clear what is being done here. Reorganize so that a description of the actual calibration system comes right after the description of what is often done.

Line 126. What is the humidity correction equation? Similarly, in line 145, what is the mass transmission correction equation? Are these listed in the SI somewhere that I missed? I gather it is like the data in Figure S4, but some mention should be made here.

Line 158. Typos in this phrase

Line 176. How did the wall of the chamber form HONO in the NO-free experiments?

Figure 1. Bar charts (and stacked charts) cannot be used with logarithmic axes because there is no real zero so the bottom of the axis is arbitarily selected. The visual size of each bar does not actually represent relative difference. For example, one could set the axis to start at 10^-10 and then all the bars look basically the same. Similarly, bars 2 and 3 differ by 100 units, while pars 5 and 6 differ by 10,000 units, yet the difference in their relative areas is the same. This figure should be remade as a scatter plot.

Line 191-192. Which of these bars or points are these three named compounds? The reference to the figure implies I should be able to tell, but they are not labeled, and two of the three of them are not mentioned in the caption.

Line 197. I'm a bit confused. Are the compounds shown in Figure 1 those that are lacking standards? If so, how are these categorized by structure? This sentence, and indeed this paragraph, somewhat confuses me about which points are those lacking standards and which are for standards, and how the former is being classified. Edit this paragraph for clarity around known and unknown compounds.

Line 206-207. The described increase in sensitivity is not really clear to me from Figure 1, as reference. Which compounds contain only a single active keto group? There is no indication of that in the figure. I guess maybe they mean the furanones? Although, really one of those is an ester group, not a ketone, and the structure of furanones is never described so many readers may not realize these are ketones.

Line 208-216. The number of significant digits on sensitivities seems optimistic, is it really known to 6 digits for e.g.,allylacetic acid? I think limiting it to maybe 2 would be more realistic, given the amount of uncertainty in these instruments.

Line 218. What do they authors mean by "detection of species at the molecular level"?

Line 255. "taking the phenolic pathway for toluene" is incorrect grammar

Line 260. For compounds with no known structure (i.e., all compounds not introduced as standards), how does one classify the molecular formula? Especially given that one formula could contain many compounds with different structures

Line 289. "tend" should be "tends"

Figure 3. See note for Figure 1 about bar charts with log axes.

Line 342. How are structures assigned for calculating binding enthalpies of unknown products?

This manuscript describes the development, characterization, and application of an optimized semi-quantitative method for toluene oxidation intermediates using iodide-adduct CIMS. The method was established based on the linear correlation between instrument sensitivity to the iodide addition and theoretically calculated binding enthalpies of the formed iodide-analyte adduct for different species categories. Compared to the previous semi-quantitative method based on binding enthalpies of iodide adducts, the categorized semi-quantitative method proposed in this study appears to have a higher quantification accuracy. Although this method was developed particularly for the toluene oxidation products, the concept can be also applied to the development of the quantification method for other oxidation system. This work is scientifically sound and the manuscript is well written. I recommend its publication in AMT after several minor comments are addressed.

L100: What was the TIC value of the iodide CIMS under typical operating conditions?

L109-110: It would be more informative and readable if the method for the generation of the calibration gas (i.e., using gas cylinders, permeation tubes, or atomizing standard solutions) can be added for each standard in Table S1.

L114: Suggest also providing the corresponding RH values at ambient pressure.

L117-119: Did all the standards dissolved into the solvent turn into the gaseous form after atomization? Were any particles detected in the atomized gas flow?

L185: As the sensitivity of iodide ionization is strongly dependent on the RH, the authors should clearly indicate the RH conditions (dry or wet) under which the sensitivity values shown in Figures 1-3 were estimated.

L307-308: Figure S8 provides a nice validation of the proposed semi-quantitative method, I would suggest moving the figure to the main text.

L309: “with absolute deviations below 40%”. This value is actually the relative deviation.

L312-313: How were these uncertainties quantified?

L338-339: There may be isomers for the oxidation products identified here. Did the sensitivity values estimated here take into account the sensitivity differences between potential isomers?

L349: Replace “less” by “small”. Also, “absolute deviations” should be “relative deviations”.

L357: How was the "satisfactory" defined or determined here?

L360: Were the concentrations of toluene oxidation products corrected for the RH effect?    

L371-372: References should be provided for this statement.