R. Dörich and co-workers have presented convincing laboratory evidence that iodide-CIMS instruments efficiently detect HNO3 as NO3- - but ONLY in the presence of ozone (which converts I- into IOx- ions, which in turn react with HNO3). This finding has substantial implications for field studies using such instruments, and suggests that some previous measurements have been incorrectly interpreted. This is one more example of how the powerful tool of chemical ionisation should be used very carefully, with due consideration of possible side reactions, including indirect pathways such as that discovered here. The study is definitely worth publishing in AMT. I have only very minor corrections and questions as described below.

-On line 135, reaction (R6) should presumably be reaction (R7), i.e. I- + HNO3 not I- + H2O. 

-On line 153, the product should presumably be IO3- not IO2-. 

-Line 159: Maybe mention already here the the O2 concentration in the quoted studies was MUCH less than that in the atmosphere (or in these measurements) - I had a hard time reconciling the dominance of IO3- with the stated rate coefficients, since I kept assuming 0.2 atm O2… Also, is it the lack of an IO3- + O2 reaction that drives the equilibrium toward IO3-?

-Could the authors use e.g. gas-phase acidity / proton affinity data to estimate thermodynamic parameters (at least endo/exothermicity) for reactions R13-R15 (and also R19-R21)? 

-Could the authors speculate about the reasons for the differences in rate coefficients for reactions R13…R15? The ion size seems to play a role, but is that enough to explain a difference of a factor of 3 between IO- and IO3-?

-Line 221: “shut of” should be “shut off”

Review of “Iodide-CIMS and m/z 62: The detection of HNO3 as NO3− in the presence of PAN, peracetic acid and O3”, amt-2021-57

This is a very nice manuscript that explores a large dynamic signal that occurs at mass to charge ratio 62 (NO3-) in the iodide CIMS instruments. This signal has long puzzled users of the iodide ion for detection of trace gases, and this manuscript adds extremely valuable information to that ongoing debate. The manuscript is clearly written and organized and presents a well thought out convincing argument assigning signal at 62 to HNO3 detection via the IOx- anion. This will be a valuable addition to the existing literature on iodide CIMS and I fully support its publication pending minor changes.

 It has been my own experience, having run a similar set of experiments, that addition of significant O3 to the iodide CIMS instrument while holding HNO3 addition constant will result in observable depletion of IHNO3- observed at m190, even when normalizing to iodide. The authors convey nicely that the impact of ozone on signal detection will be highly variable depending on the instrumental conditions used, and that alone could be the difference here. However, considering the authors present HNO3 data from the field using observations at m/z190, the data should exist from the laboratory experiments to validate that this does not occur in their system. It would be beneficial to this manuscript if the authors could include a figure similar to 7a but looking at the signal at m/z 190. This information is important to validate some of the assumption throughout the manuscript such as on page 8, line 232 where it is indicated that “negligible depletion of HNO3” occurs, a fully testable assumption considering the ability to detect HNO3 at m/z 190. It would also be useful in understanding the comparisons of the m62 and m190 data at the end of the manuscript from the ambient observations.

I am having a difficult time understanding how the authors are able to retrieve an ambient HNO3 signal from the NO3- signal. The signal that occurs at that mass is dependent on a changing HNO3 ambient concentration as well as a changing O3 concentrations. It would seem rather difficult to retrieve the real concentration unless you make an assumption that at any given point HNO3 is not changing and only O3 is changing, or the other way around. The only way this may work is if the authors used a measurement of true O3 collected during the ambient flights, however it is unclear from the description that this was the case, perhaps those details are just missing. If an ozone measurement was used what instrument provided that data. I would like to see more details of how the concentration profiles in Figures 9 and 10 are determined, and what assumptions were made. If the author agree that one can not retrieve true HNO3 unless a true ozone measurement is used it would benefit the readers to explicitly state this. It is possible that a reader could interpret the data presented here to mean that by using the Iodide CIMS measure of IOx- and NO3- one can quantitatively measure HNO3.

I am in general surprised by the lack of instrument response to ambient features in the data in figures 9, 10 and S1. HNO3 should have a significant inlet response time relative to O3 in the CIMS regardless of the inlet conditions. In figure S1 in particular there appears to not be any lag in the observations. This quality in addition to the exact correlation to O3 makes me believe that you are really just detecting HNO3 that is made inside of the instrument. Otherwise, I would expect some deviation between the ambient O3 and HNO3 from 12:00 to 15:00 in figure S1. Can the authors comment on the time response of their measurements? What does a laboratory time response on m190 look like relative to m62?

Has the data from Figure 10b been corrected for scrubber zeros? Is this actually signal over zero. The way the data correction is explained for 9b where an approximate exponential of the Po HNO3 source is subtracted out would suggest that there are not real zeros removed from the ambient data, which would include source HNO3. If real zeros have not been used how do you know any of the features shown in figure 9b or 10b are real ambient signals. If real zeros were used why didn’t they account for the Po source background?

This ion chemistry is ultimately a three-body reaction as indicated in page 7, line 216. Therefore, the result should be quite dependent on IMR pressure. The iodide CIMS literature is filled with instruments running across a very large range of pressures, sometimes exceeding 100 mbar, do the authors have any data or comments on the impact of pressure on this ionization mechanism.

The authors comment on the potential issues with past data sets using iodide CIMS to measure PAN, due to the secondary acetate ion chemistry that is occurring. It is my experience that many if not all of the airborne systems employ a constant standard addition of 13C PAN to the instrument. I believe that this addition would account for any secondary ion loss due to IOx- chemistry. I think the authors should add a comment in their discussion on this point on pages 9/10. While there are likely many dataset that have not used such an internal standard to track sensitivity changes it is not appropriate to question the data set that have used that method. A related question is do the authors believe such a system would indeed correct for these ion chemistry issues? That could be a good couple sentence discussion to add here.

On page 12, line 369 the authors state that 190 does not show any correlation with O3. However, in the same figure, #10, it is clear that 190 correlates with m62. It was previously argued in describing figure 9, using panel C that it is expected that HNO3 should correlate with O3 even after correction for this unique ion chemistry. Additionally, there are clearly features in 10b that correlate with the ozone signal shown in 10a. Please elaborate on the disagreement between these two statements.

On page 6 there is a discussion on the residence time in the flow tube and comparison to similar work. I encourage the authors to reconsider their calculation of the residence time in their flow tube as presented. With a critical orifice on a cylinder there will be a jetting effect of the air through the region such that your reaction timescale will not be equivalent to the laminar sweeping of the volume. Rather it will be dependent on the velocity and flow dynamics of that jetting effect. This likely results in a significantly shorter reaction time than the volumetric calculation for most of the analyte molecules being sampled. It is not necessarily a discussion that is needed here but should be considered when trying to understand the difference being discussed, and leveraged as a potential answer for some of the disagreement observed.

Figure 1. I am interested in the flow dynamics in the figure with the initial sampling line. It appears that at some point your inlet could be connected to a pump exhaust line in the far left of the diagram. Even if overflowing the inlet with N2, depending on the flow characteristics and pressures I could see a scenario where pump exhaust would flow out the inlet past your sampling point. Perhaps something is missing or I am interpreting this diagram incorrectly.

Why do the authors believe the HNO3 to 62 appears nonlinear in the inset figures on Fig 2 and 8?

Figure 9 and 10. These figures are difficult to interpret because there is not legend given in the figure and the colors are reused for different species. In 9, I believe the label attitude should be black and without reading the caption the reader has no idea what the pink line is for example. Blue is used in 3 places in Fig 9 for three different things.

Page 8, line 238 needs a space between the and IOx

Page ,8, line 241, suggest adding the word “of” in front of 100 greater.

Page 9, line 265, there is a close parenthesis missing.

Page 12, line 371, calibrations should be singular

The paper reports an investigation of the influence of the presence of HNO3 on the detection of NO3- and N2O5 when using chemical ionization mass spectrometry with I- as the reactant ion. Investigations are reported for especially the effect of ozone and humidity (that may affect the reactions through direct association of water molecules to ions). Additionally, the effect of PAN or PAA, that can release acetate anions causing extra reactions with HNO3, are reported.  Finally, examples of field data, where the significance of the HNO3 trace gas is important for data interpretations of I-CIMS data, is given. The reported investigations wer motivated by observations from air borne measurements using an I-CIMS where the magnitude and variation of the signal at mass 62 (NO3-) seemed to contradict previous beliefs on the measurement sensitivity to trace amounts of HNO3.

The theme of systematic sources of inaccuracies in CIMS measurements is of high relevance in general to atmospheric measurements and the presented results give important information on this aspect for the particular case of HNO3 trace gases in I-CIMS.  As such the paper is of high relevance.

The supplementary information is not available, and as I learned from the editors, this information was actively removed by the authors prior to the review process. The supplementary information is in fact heavily needed to critically address the content of the paper, in particular

• Page 2 - Fig. S1 – should illustrate the data that motivated the re-investigation of the sensitivity of the I-CIMS detection of N2O5 to presence of the HNO3 trace gas
• Page 4 - Fig. S2 – should show details of the experimental calibration
• Page 9 - Fig. S3 – should give more information on the data corresponding in fig. 7
• Page 12 - Fig. S4 + Fig. S5 – should show data from The CAFE-Europa flights

I consider the removal of the supplementary information, where essential information is given and to which reference in made throughout the manuscript, as rather unserious and problematic. On this account, I would strongly hesitate to consider the paper for acceptance.

General comments to the paper.

While the introduction is well written and easily accessible, the sections “experimental details” and “Laboratory characterization” could well be work over again for logic argumentation and clarity in the presentation.

I have the following specific comments, if the editors at all find the paper relevant for publication after the removal of the supplementary information.

Comment 1: Title and Abstract( e.g. line 11,14)

The meaning of PAN is not given. Should be mentioned the same way as PAA. In line 14 either use PAN and PAA or give the chemical formulas for both – a least be consistent.

Comment 2: Page 2, Line 34-45 – several references to literature is missing . The statement “In very well known series of reactions  …” must be backed up with a reference where one can find the reaction rate constants for these reactions.

The statement on the “non-gas loss process” should also be backed by references

The statement on rapid photolysis of NO3 should also be quantified with a reference and an actual number for the photolysis rate.

Comment 3: Page 3, line 81.

I believe a rate of 380 s-1 would correspond to ~3 ms rather than ~2 ms as stated, or simply state 2.63 ms to keep the number of significant digits ?. I am also missing the reference that tells where the stated reaction rates comes from, i.e, where do the numbers  380 s^-1 and  1940 s^-1 come from. To appreciate the importance of these timescales, the authors should also specify the transport time of the various ions though the instruments sectors. I realize that some of this information can (partly) be reconstructed from the description of pages 6-7, but it should be clearly stated in this place, which would also ease the reading of the pages 6-7 a lot.

Comment 4: page 3, line 86.  

Give reference to the origin of the stated reaction rate constants.

What was the actual concentration of added NO ?

Comment 5: page 5, line 133.

To appreciate that it is correct to normalize all signals to the primary ions signal, the reader need to be assured that the intensity of this peak (mass 127) is not affected (i.e. only marginally) by the reactions taking place. The authors needs to quantify this more precisely. I am in particular puzzled, since the [I(H2O)]/[I] ratio changes dramatically (approximately a factor of 7 (stated as 6, line 116)), so a least humidity most be important here ?

Comment 6: page 5, line 133-135.

Could you expand on the explanation why reaction R6 (association of water to I-) is affected by the concentration of O3. This may be true in some indirect way, but to understand that, it really requires a more explicit explanation at this place in the paper.    

Comment 7: page 5, line 146.

The statement “This could be confirmed … m/z 62” seems to reflect an action that the authors speculate could prove their point that the first mentioned explanation for the sensitivity to O3 is not likely.  Is this speculation or did you really do the suggested measurement ? It would be good to see the suggested evidence. (A figure in the missing supplementary information would be fine)

Comment 8: page 5, R10.

On the right hand side, IO2- should be IO3-

Comment 9: page 6-7 - discussion of the observed intensities of IOx-.

Given the rate coefficients of the various reactions, it seems straightforward to calculate the steady state distributions of I-, IO-, IO2-, and IO3- under the various conditions shown in figure 3. Following all the argument that ends on line 200, I believe it would worthwhile to do such a (simple) calculation and compare the result to the data in figure 3.

Moreover, following the discussion one page 7-8, in line 242, the statement is made that “ confirming that the detection of IO3- in our experiment is inefficient” (see also line 223-24). I am puzzled if the efficiencies of the various IOx- components may affect the actual relative intensity ratios between them. The authors needs to clarify this issue.

Comment 10: figure 7a and equation 2.  

Please explain the idea of suggesting the form in equation 2 to represent the data. I suppose it represents a short of saturation – but please clarify this more explicitly. Also the description in line 251 that “is clearly non-linear” is not really true: except for the lowest curve (18.5ppbv) all displayed curves are in fact linear to a good approximation as also suggested by eqn. 2, which is indeed almost linear at low [O3].