In "Characterisation of the multi-scheme chemical ionisation inlet-2 and the detection of gaseous iodine species", authors Xu-Cheng He and co-workers describe a new version of an inlet (MION) for chemical ionization mass spectrometry (CIMS) that allows for switching between multiple schemes and reaction times. Overall, they are presenting careful work that employed scientifically sound and appropriate methodology. Its findings will be very interesting for the CIMS community, especially of course if using a MION inlet or similar. Unfortunately, the manuscript itself was not prepared as carefully.

General positive points are the abstract, which I believe provides a good summary (except for some ambiguities noted below), as well as the figures, which are of mostly easy to read and of high quality and well-chosen by their relevance to the presented research.

However, for an AMT paper introducing a new CIMS inlet, its description is confusing and substantially lacking important details, as I try to elaborate in my detailed comments. This deficiency is most apparent in Section 2 (Methods), but found also in some parts of Section 3 ("Results") where additional experimental and analytical methodology are described.

In addition, the text will need some proofreading/copy-editing to deal with numerous grammatical errors. Some semantic errors disrupt the reading as well. Nonetheless, the text is in principle understandable.

For these reason, I suggest to reconsider the manuscript only after major revisions.

[Disclaimer: I do not feel qualified to judge the various methods used for the DFT calculations (Section 2.4).]

Specific comments:

Title: Much of this study deals with the Br- ionization scheme, and indeed provides useful insights into that specific scheme in particular, applicable also beyond the MION inlet systems. NO3-, for the most part, is rather used as a reference. Anyway, I would point that out already in the title. E.g., "... using Br- as reagent", or "... using Br- and NO3- as reagents".

Abstract: I suggest disclosing somewhere near the beginning that MION inlets operate (or are at least designed to operate) at atmospheric pressure.

L4 (abstract): "generally more robust operation" ... please be more specific.

L20 (abstract): Should specify if the detection limit for H2SO4 is achieved via Br- or NO3- reagents (or both).

L22 (abstract): again, "generally more robust" is too vague.

L95-102: Somewhat confusing description of the MION inlet. Not the only issue, but also: is "reaction time" the same as "ionisation time"? Most critically maybe: how does MION2 allow for two CI methods with same reaction time, and why was that not possible with the MION1? Please clarify.

After also checking Fig. 1, I think I get it. But now I wonder why only two CI at the same time (and same reaction time) (L101), and not three? (Or even six, if the polarity can be switched quickly as well, which remains unclear. And does "two or more (up to six)" simply mean "up to six" or something more elaborate, possibly including limitations regarding reaction time choice as in MION1?)

Fig. 1: I suggest to indicate the directions of the various flows, to make the drawing easier to comprehend.

L108: How was the sample flow provided?

L110-122: What are typical/required/desired reagent flows, reagent concentrations, and purge flow(s!)?

Also, how are the reagent concentrations facilitated?

Fig. A1 is missing the "purge flow" (and L119-122).

L124-125: How is that "operational stability" manifest or determined? If elaborated on later, please state so. If not, be more specific.

L125-127: How have the ion optics been upgraded? I am not expecting much details, but at least some indication of what type of effort was undertaken and required to increase "reagent ion transmission" (by which the authors might actually mean the amount of reagent ions ending up being detected?)?

Section 2.2.1: What exactly is the "calibration source"? It is referred to multiple times, but never actually specified what it is, except that OH radicals are generated "in it", or how it connects to other parts of the setup. (For example, are the OH radicals actively mixed into that SO2- and I2- containing gas mixture, or is the gas mixture going through a region where OH radicals is generated, and the "source" is the sum of something like that?)

Similarly, Fig. A2 simply refers to a mysterious "calibration box".

Section 2.2.2: 1st paragraph is a general introduction to the problem that may fit better to the Introduction section.

Section 2.2.2: Unclear what was done why, and how the I2 was ultimately supplied during the MION calibration experiments. For example, was the permeation tube output used directly, and the process from dissolution in hexane to quantifying a concentrated solution of the derivative was only to gain knowledge of the I2 permeation tube output rate? (Which is presented as the conclusion of the main paragraph.)

Eq. 1: Q is not defined.

Table 1, Section 3.1: Detection limits are given for MION2/T1, MION2/T2 and Eisele inlets. But for H2SO4, either NO3- or Br- were used on MION2/T1, so, which reagent ion do the reported detection limits correspond to? And the Eisele inlet presumably used NO3-? If so, is there a reason that the NO3- scheme with MION2 was not tested for HIO3, as the Eisele inlet was?

Also, it is unclear at this stage what is meant by "APi1" and "APi2".

~L344: Is fragmentation at atmospheric pressure (as opposed to only in the ion optics) responsible for the HO2 cal factors for T1 vs T2 being only a factor of 2.3 apart?

L360: I am not following the final sentence. As the authors just pointed out, they found (experimentally) that more strongly fragmenting instrument settings reduced sensitivity to HOI, agreeing with somewhat weaker binding between reagent and analyte, compared to the H2SO4 case, expected theoretically. So, why would one anyway blame "iodine chemistry schemes" or "differences in experimental conditions"? As those terms are rather vague, I may just misunderstand what is being pointed at. (Oh, is it differences in chemistries between the cited studies and this study?)

L383: Please provide a reference to that "earlier study".

L420: I disagree with the implication of the first half of this sentence. I agree that, for instance, within a typical day, ambient absolute humidity often does not vary by very much. But within, say, a week, one would expect substantial variations. And more so the longer of a time period is being considered...

Section 3.4: Figure A9 needs some more explanation, maybe via annotations in the figure (photo). Unclear what is what.

Section 3.5: I appreciate that experiments were carried out using two independent detectors and sample sources. But the discussion of the respective differences is awfully short. (And merely from a statistics point of view, a sample size of two is not that much better than a sample size of one.) Hence, is there anything useful to say about differences between APi1 and APi2 (or APi3 for that matter), beyond time since service? E.g., details on tunings, or purity of gas or calibrant supplies, etc.?

~L553: How was DeltaV50 determined? The shapes of the signal-remaining curves (Fig. 6) indicate that for several species the maximum is not obtained at the lowest tested DeltaV. The clearest case is H2O, for which the signal-remaining drops to 50% at ~3-4V, but the curve is steep and a higher reference value (signal-remaining = 1) would likely be obtained at yet "softer" settings (e.g., DeltaV < 2V). Correspondingly, if dV50 is simply the 50% point from Fig. 6, I expect several points in Fig. 7 being "too high".

Fig. 7: It would be very useful if the same color coding was used in Figs. 6 and 7, i.e., same color for same species.

Section 3.7: L568 vs L580 appear to contradict each other, even though using same reference. Is Reaction 4 exo- or endothermic?

Section 4 ("Conclusions"): This Section is really a summary of the results and the major discussion points presented in Section 3. With the exception of one or two sentences, it does not provide any new discussion nor actual conclusions. Consequently, it should be named accordingly. (Whereas Section 3 would be more aptly named "Results and Discussion".)

Technical comments:

L14: More correct, I believe, to write "We calibrated for [...]"

L34: "spectrometer" -> "spectrometry" (for grammar)

L38: missing article ("forming a relatively")

[I will stop commenting on grammatical errors (or semantical errors or typos). Some proofreading/copy-editing service will be more suitable.]

L81: I assume the authors mean the LOD is higher (hence worse), not lower

Fig. A2: stain -> stainless

L150: "calibrator" or "calibration"?

L238-239: I am counting three ways, not two.

The manuscript "Characterisation of the multi-scheme chemical ionisation inlet-2 and the detection of gaseous iodine species" by He and coworkers presents an upgraded version of the multi-scheme chemical ionization inlet including development of a model for gas kinetic studies related to the characterization of the inlet. In addition, it presents a case study for the measurement of iodine compounds using this new inlet.

In general, the topic is well suited for publication in AMT. However, the manuscript in its current form needs some rework prior to this.

There are some parts of the manuscript that are well written, but others are confusing or lack proper descriptions or explanations. The same applies to some figures. The language of the manuscript also needs intensive and proper proof reading in parts. Thus, I have refrained mostly from correcting grammatical errors.

Specific comments:

L98-L100: The description of the geometry and the inlet is not represented in Figure 1. For readers unfamiliar with the inlet system of Aerodyne/ToFWerk CIMS instruments it might be difficult to picture what the authors describe (e.g., distance from injection port to pinhole).

L101ff: How is this achieved? And how do you define reaction time? Please clarify!

L103ff: Is 25cm the standard configuration for the connecting pipe between the two sources, or why was this length used in this work?

L139: Bubblers tend to produce not only gaseous water vapor but also micro droplets, which could act as sink for trace gases. Was there a filter/trap to prevent possible droplets from entering the sample gas stream?

L212: What was the inner diameter and residence time inside the quartz tube?

L213: What is the wavelength of the used LED(s)?

L263: What is the "It-product”? Please shortly mention the definition for the general reader. In your phrasing the time dependence is completely omitted.

L269: Rephrase “… 1) simulating connected two flow reactors …”.

Table 1: What is the difference between APi1 and APi2 for Tower 1?

Table 1: How do reaction times of the towers compare to reaction times typically for an Eisele inlet?

Figure 3: Why is it that the measured values for HOI deviate from the line fit at the lower end, and in every case in the same way (in this figure as well as well as in Figure A7)? Is it because those measurements are close to the LOD?

L344: “… will be shown …”: Please change wording to point directly to the appropriate section.

L355: What is the “signature of HOI”? I guess, what you want to say, is that a fraction of HOI*Br^- is de-clustered or loses its charge. Please rephrase.

L357: “… a relatively fragmenting …” Relatively fragmenting compared to what? Please rephrase.

L390: What does “relatively easier” mean?

L403: “Despite …” please rephrase this sentence, it is difficult to understand.

L406: What do you mean with “humidity tolerance”? Please rephrase.

L414: If the H₂SO₄ is lost to the walls it is effectively removed from the gas phase and can as such no longer be detected by any method. I would not call that a systematic error.

L425ff: Isn’t the consequence then to not use Br^- for the more distant MION source?

L458ff: How do your findings regarding dilution and the detection of SO₂ fit to your previous statement in L440? Or was that statement exclusively for a low-pressure system?

L491: What does this sentence mean? Can one lower the detection limit by a more thorough analysis of the obtained data, or is it possible to “guess” concentrations of specific compounds even below the LOD? What would be the significance of such an educated “guess”?  

L495: What is the “softness” of the tuning of the MS system?

L496: What significance does a LOD determination at optimum conditions have if the actual measurement conditions deviate strongly from those conditions?

L499: The explanation for APi1 and APi2 needs to be given in Table 1 too!

L511: I do not understand why getting similar LODs from both the MION1 and the MION2 system suggests that the newer one is an improvement over the older one (regarding the LODs). This could also be simply an issue of instrument performance.

L520: Are you making a point that your MS systems were not well performing?

L521ff: “Nevertheless, the achieved…” I believe this sentence is quite questionable.

L520-L523: I would suggest removing or rewriting this paragraph.

Figure 6: It is somewhat confusing that your diagrams are indicated as a), b), and c), but the labelling is never explained. What adds to the confusion is that you mention three different groups in the text. However, those three groups do not actually correspond to the labelling in Figure 6.

L593: “… for reactions 5 and 6 and they are …” should read “… for reactions 5 and 6 are …”

Section 3.7: I find that this section seems to be quite distant from the remaining context of this manuscript. Also, it does not really add to the characterization of the MION2 inlet. For example, would it not possible to do these measurements with the MION1 inlet? The topic is prominently featured in the title but compared to the rest of the manuscript quite weakly presented. Most of the section explains the scientific background and previous measurements, with only the last, short paragraph presenting a laboratory study with some results by the authors.

Figure A7: Again, both fit lines for HOI do not really seem to represent the data points, hinting at either a linear fit not being the best representation, or at an additional unaccounted factor/bias. Maybe a short discussion would be helpful (See also comment to Figure 1)