This paper, Temperature dependent sensitivity of iodide chemical ionization mass spectrometers, describes observations of instrument performance made during laboratory experiments and previous field campaigns to understand unexpected instrumental behavior. Iodide Chemical Ionization Mass Spectrometers (I-CIMS) are used by numerous groups worldwide to measure a variety of species including halogens, NO_Z constituents and oxidized organic compounds. This makes this paper highly relevant to the community. On its face, it seems obvious that higher temperatures would results in less strongly bound clusters falling apart but it is not something you typically see considered in these types of measurements. Since cluster chemistry is important to other ionization schemes beyond iodide, the significance of this work extends to other ion chemistries as well (eg. NH₄⁺, Br^-, Fluoride transfer). One part I feel should be made clearer is which experiments were done with which IMR. I realize the IMR heating experiments were performed with the Aerodyne IMR, but what about the cooling experiments or the room temperature sweep experiments? This is not clear to me and left me confused as to which IMR was used where. I suspect I would not be the only one. I am wondering if the authors could also provide a few more details about the closed loop humidity control. This should be added to the SI. It would give the reader a better understanding of how this system works and the variability of the amount of humidified nitrogen added. Since the I(H₂O)^- cluster is both a product ion and also a reagent ion the temperature dependence of it has important direct and indirect implications on the instrument sensitivity. The manuscript is well written and falls within the scope of AMT. It should be published once the few minor comments are addressed.

Specific Comments

P1 L 23: Sentence should be reworded. If gives the reader the impression that only cooling reduces the sensitivity drift.

P1 L26: Atmospheric trace….

P3 L73: This is a very good point. I am struggling to think any temperature dependent studies beyond some of the initial kinetics papers in 90’s.

P4 L103: The authors should provide a couple of example analytes not merely the reference.

P4 L114: The stabilization of sensitivity assumes the clusters make it the detector without falling apart, also dependent on the operating pressure of SSQ there is potential for chemistry to continue into this region which is never temperature controlled on any of these instruments.

P5 Figure 1: It seems odd to me that the nylon liner on the NOAA IMR does not provide some form of temperature isolation from the outside temperature of the stainless steel fittings. Clearly, it’s just not enough isolation

P5 L131: How much N₂ do you have to flow through the bubbler? Is it the same flow to maintain the same cluster I^-:I(H₂O)^- cluster distribution in the two IMR’s? I am curious because dependent on IMR geometry I have seen this vary greatly in our instrument.

P5 L 138: I’m curious why the authors chose only to normalize by the I(H₂O)^- cluster. Granted it should account for any variation in reagent signal it is merely that often in the literature the reagent is taken as the sum of the two.

P6 L157: From the looks of the temperature profile these experiments, they were not done as ramp and soak type experiments. I’m curious as why this decision was made as opposed to trying to assess the issue at discrete temperatures. Granted this mimics what typically occurs in most measurement trailers or aircraft.

P7 L172: Do you have any idea as to the actual gas temperature in the IMR’s? I am guessing even with heating or cooling they would be very different when the instrument is operated as a PAN CIMS. 

P7 L179: are shown

P8 Table 1:  I am curious why you did not calculate the sensitivities for 2-nitrophenol and phenol if you know what mixing ratio added to the inlet is?

P8 L204: It is not exactly fair to compare the PAN chemistry (particularly the anion chemistry) when the ionization chemistry is a different mechanism than the rest. The authors should specify which channel or drop it and merely discuss the halogens.

P9 L212: Again, I am not sure grouping the PAN anion channel with the others is appropriate.

P9 L215: Why were the sensitivities not determined?

P9 L227:  How much nitrogen needs to be added/removed for the humidity control? Are we talking about 10’s of sscm on a couple of litres or is it more? I am curious if there is any possibility of dilution in the IMR and without that information, it is not clear to the reader.

P10 L230: I(H₂O)^- is also a reagent ion. If it is falling apart in the IMR as a results of increasing temperature that would have a direct effect on sensitivity beyond clusters falling apart downstream. 

P12 L280: This is definitely true.

P12 L283: At what pressure do you operate the NOAA I-CIMS SSQ? 2 mBar or lower?

P13 L294: Are the sensitivities the same using both NOAA IMR and the Aerodyne one? There are definitely geometry differences between the two. Were the cooling experiments only done with the NOAA IMR and heating experiments only done with the Aerodyne IMR?

This publication details a set of laboratory experiments examining the effect of temperature on a chemical ionization reactor employing the iodide anion as a reagent. The experimental results are then used to correct a field data set for ambient temperature variations.

In some ways this is an unusual manuscript. It contains information that typically would comprise the SI of a paper detailing the results of the ambient measurements. But I think that hiding many of these experimental instrumental details in the SI has previously been a disservice to the community. SI sections are rarely read nor widely disseminated, and such important details should be out in the open as these instruments proliferate across atmospheric chemistry.

However, this manuscript should be more clearly written and organized than it currently is.

I support publication eventually after the authors address the items listed below. More experiments should not be necessary, but I suggest major rewrites for clarity. Additionally, there are several places that the authors speculate about instrumental details and results without supporting evidence. These should be clarified with much more precise language.

Specific Comments

L87: “This discrepancy…” The authors should add proof of this speculation via a citation or data.

L95: The connection between this work and Lopez-Hilfiker’s voltage scanning method is not clear to me, nor is the reason for this paragraph.

L108: Typically the last paragraph of the introduction focuses on what the forthcoming paper has done. Instead, this manuscript introduces another paragraph referencing previous temperature control strategies. This paragraph is disorienting. It should be moved to earlier in the introduction or later in the results. Also, the authors reference the FIGAERO as an IMR temperature control strategy, which is incorrect. The FIGAERO is an inlet that goes on the front of the IMR and is independent of the IMR itself, separate from the IMR’s temperature regulation or lack thereof.

L117: My understanding of the Tofwerk instrument is that it is an OEM instrument, onto which others can install ionization sources. So this would seem to not be a “modified commercially available TOF”, which the authors directly contradict anyway in L123 where they say the “TOF has not been modified.”

L120 seems to refer to the IMR, but these running conditions in Lee et al. are specific to Iodide and Bertram et al. used pressures of 20-100 mbar. This entire paragraph should be edited for clarity.

L121 “It is worth noting that higher pressure IMR systems will likely be more susceptible to thermal effects due to the increased residence time” certainly seems possible but is still speculative and should be supported in some way.

L140 I was surprised here to read that this work also uses the ARI IMR since the entire introduction focuses on the NOAA-built ionization source. There is no prior introduction to the ARI IMR and no discussion on its specifics, nor a citation where it’s referenced here. I suggest adding more detail on the ARI IMR to the main methods section (more than just the tiny table in the SI). Further, regarding the L117 comment, is the NOAA instrument any different than the commercially modified one if it sometimes employs the ARI IMR?

Section 2.1: The authors start this paragraph off making a claim that a constant IMR but a changing TOF temperature would cause temperature-dependent ion chemistry in focusing ion optics. But it doesn’t appear that the authors conducted this experiment. Could they please support this claim?

Supplemental figures 1 and 2 should be cleaned up and moved to the main paper as a multi-panel figure referenced in Sections 2.2 and 2.3. This is a short paper and its entire value is that it is not hidden in the SI (see general comment above).

L239: The text references slopes on the graph, which I think is a good idea, but there are no slopes. The authors should add slopes to Figure 3, or if too busy, add a plot with an example regression. I find section 3.2 to be the most interesting, but it is short on details and specifics.

L258: I’m skeptical that HCl “fall[s] close to the 1:1 line”. It has essentially the same measured reaction enthalpy as Phenol and HCN within the margin of error, but a different literature enthalpies. Is this a small influence of ligand switching? Why does nitrophenol fall so far off the line and is not discussed other than having a low overall sensitivity? This section could benefit from more analysis and textual interpretation.

L277: These field temperature swing would be a useful SI figure and the authors should have plenty of good examples.

L308: “Temperature control of the IMR region can help reduce the impact, but de-clustering can occur further in the instrument ion optics.” The link between collisional fragmentation and temperature dependence is not clear to me as this is written.

All over: sensitivities in units of Hertz or ions/s should be defined with an extraction frequency for the TOF.

Technical Comments

L99: I’m not sure I follow the sentence at L99. Please clarify or add more detail: “Additionally, product ion formation dependence on temperature may make these voltage scanning determinations difficult to interpret”

L51: “refined restriction policy” – This could be clearer

L241: The nist webbook should be a citation in common citation format

L299: “dependence of sensitivity on temperature”