The article describes quantitative reaction kinetics calculations performed for the reaction between Criegee Intermediates and hydroacetonitrile. The calculations are well done, and the reaction in question is atmospherically relevant, as evidenced by the GEOS-Chem modelling, and therefore I gladly recommend the publication of this manuscript in Atmospheric Chemistry and Physics.

However, I have a few minor issues I hope that the authors will address first, mostly but not exclusively relating to presentation and clarity of the manuscript.

Suggestions related to calculations & results:

Lines 160-162: “We mainly consider the most feasible mechanism in detail in this work because the enthalpy of activation via TS1 at 0 K is at least 5 kcal/mol lower than those of other reaction pathways by M11-L/MG3S (See Fig. 1). Moreover, the intermediate product M1 formed has a larger enthalpy of activation of –58.33 kcal/mol in Fig.1”

While I agree that the five-ring closure is probably the most competitive reaction, I would also like to point out that you have a barrierless reaction with negative T-dependence 1-2 orders of magnitudes slower than the collision limit. This is evidently a case where the entropy also matters, and this raises the possibility of the H-shift reaction (TS1a + TS1b) being possible as a minor channel, due to it likely being less entropically restricted than the ring formation. I suggest you add a table or a plot to the supplementary of the Gibbs free energies of TS1, TS1a, and TS1b over the considered T range to check if if might be relevant.

The Supplement is missing the xyz geometries for all species in Figures 3 & S1. Please add them.

And here’s the big one:

There’s a combined experimental/computational article on Criegee + C≡N reactions published quite soon after yours (https://doi.org/10.1021/acs.jpca.2c07073. Full disclosure: I am one of the co-authors on this article), that found quite different decomposition channels for the 5-membered ring product. Our decomposition channels have higher 0 K energies than yours, but they are presumably more entropically favourable, which especially matters for the product distribution when the intermediate product forms with 51-58 kcal/mol worth of excess energy. What’s more, now that I got aware of your calculations on the topic, I tried to do a saddle point search on your M2-TSb transition state using wB97X-D3 based on the figure, without success. This makes me question whether this N-C-O ring closure saddle point may actually be an artefact produced by M11-L. I would be very interested in seeing calculations of this barrier using some other levels of theory as well, and preferably also a comparison to our CH2O ejection saddle point. Please also report entropies in addition to 0 K energies, since all of these saddle point are well below the energy of the free reactants, and since we have shown that only a small fraction of the five-membered ring stabilized for CH2OO + Acetonitrile.

Discussion-related comments:

In general: You use abbreviations of “Higher level” and “lower level”, but you write out “enthalpy of activation at 0 K” a total of 11 times throughout the manuscript. How about abbreviating it as H_0?

In the abstract and on line 53, you say that are “designing” a computational strategy, but it seems to me that you are simply using a previously establised computational strategy that you have already utilized in several published papers. Please rephrase.

Lines 21-22: “hydroxyacetonitrile (HOCH2CN) has been recently identified as a C2H3NO isomer by using I− chemical ionization mass spectrometry (I-CIMS) instrument detection.”

You do not need an instrument to identify HOCH2CN as an isomer of C2H3NO. It’s in the chemical formula.

Lines 21-28: “However, previously several field studies had misattributed the C2H3NO signal to methyl isocyanate (CH3NCO) by using CIMS. CIMS is insensitive to the detection of isomers, and thus cannot differentiate between the isomers of CH3NCO and HOCH2CN. CH3NCO had been detected in chemicals released from biomass burning, such as wildfires and agricultural fires, as well as in bleach cleaning environments. Therefore, the atmospheric sources of CH3NCO from previous investigations are actually the sources of HOCH2CN.”

I am unable to follow the argumentation here. If I may simplyfy, it seems that it says “HOCH2CN is an isomer of CH3NCO. Therefore previous observations of CH3NCO must have been HOCH2CN in reality.” Slightly more evidence is needed to claim this.

Lines 41-42: “Criegee intermediate is considered to be a key intermediate due to its effect on the atmosphere (Chhantyal-Pun et al., 2020b).”

This sentence adds nothing that was not already said in the previous sentence. You can remove it easily.

Lines 49-51: “While reaction kinetics of Criegee intermediates are prerequisite for elucidating their chemical processes and finding new sink pathways in the atmosphere, their kinetics are very limited and even unknown.”

It’s unclear what is meant by this sentence, but it seems that it is trying to say that our knowledge of the reaction kinetics of Criegee intermediates is very limited. This is nonsense, as there are plenty of articles on the topic, including all those that were cited in this paragraph. Please rephase to make it clearer what you mean.

Line 55: “the quantitative enthalpy of activation at 0 K acted as high level is used to calculate the rate constant by using conventional transition state theory without tunnelling”

This sentence is nonsensical. It looks like you have changed your mind about what the subject should be mid-sentence.

On Line 61, a citation in needed for GEOS-Chem.

Line 68-69: “The basic requirement in kinetics calculations is the quantitative enthalpy of activation that is determined…”

You are making it sound like reaction kinetics is all about determining the entrahlpy of activation, when the statistical mechanics description of gas-phase chemical reactions also crucially depends on entropies in thermalized conditions at non-zero temperatures, as well as excess energy, collisional stabilisation, and fall-off effects in non-thermalized conditions. And then there are of course tunneling effects. Having a quantitatively accurate enthalpy of activation does not matter if all of these physical effects are modelled poorly, and I presume you know this, considering how much effort you have put into getting them right.

Line 109: “ktight was calculated by using a dual-level strategy discussed below”

The higher and lower levels of theory were already described previously in the paper. Therefore, you are causing a lot of confusion for the reader by referring to “a dual-level strategy discussed below“ here. It made me think “Wait, wasn’t the dual-level strategy already described? Is this a different dual-level strategy? Was I supposed to read this before the previous page?” etc. Please rephrase.

Line 112-113: “the reaction coordinate s is obtained by defining the distance between one pivot point on one reactant and the other pivot point on the other”

How were these pivot points determined? Was it the centre of mass of each molecul or something else?

Line 129: “More details were provided in Tables S4 and S5.” (Also on Line 226)

1. I think you mean “Tables S5 and S6”.

2. What you have in these tables can not exactly be described as “more details”. I would describe it as a factorization of the rate coefficient into each component present in equation (2).

Table S2 requires an explanation of what the two lambdas represent, and how the accurate anharmonic ZPEs were calculated to determine the reaction-specific scaling factors.

Line 205: “The calculated enthalpy of activation at 0 K of P1 is – 205 86.55 kcal/mol, indicating the unimolecular isomerization is –5.61 thermodynamically driven”

Please rephrase this. It is unclear what number you are referring to with “-5.61 themodynamically driven.”, and what you mean by it.

Line 222-223: “Therefore, the enthalpy of activation at 0 K for every reaction can only be quantitively obtained by specific calculations.”

Here it is again unclear what you mean. Do you mean that the barriers must be calculated separately for each reaction instead of estimated based on literature data? I’m sure everyone in the field agrees that it is better to do so in principle, but this does not invalidate the staregy of making rate estimates whenever possible. Please either rephrase, or possibly remove the sentence completely.

Lines 230-231: “The temperature-dependent Arrhenius activation energies also have been fitted by using eqn (4) as listed in Table 3.”

Typically Ea is a constant in the Arrhenius equation. While this is already impled, you should add that you are fitting to the Arrhenius-like equation k = A exp(-Ea(T)/RT) instead of k = A(T) exp(-Ea/RT), for example.

Line 235: “which provides the evidence for the negative temperature dependence of the rate constants of R1.”

The activation energy does not provide evidence for anything. Your quantitative rate calculations provide the evidence, and the T-dependence of Ea is just those results fit to a simple function.

Lines 236-244: There are several parts here where you forget to say “at 298 K” when citing a specific value of k.

Table 2: The footnotes a & b are unnecessary if you use the subscripts kR1 and kR2 for the reaction rates instead of k1 and k2.

Lines 248-250: “Additionally, we have found that recrossing, tunneling transmission, and torsional anharmonicity effects for the reaction of R1 and R2 can be negligible because they are close to unit, as listed in Table S5 and S6.”

There is nothing wrong with this sentence, but I want to point out that all of this is to be expected for a barrierless reaction for two small molecules where the torsional modes of freedom (the O-H, C-C(OH) and for syn-Criegee the -CH3 bond rotations) undergo no change during the reaction.

Line 262-264: “In the atmosphere, Vereecken et al. have evaluated that the concentrations for stabilized Criegee intermediates are in the range between 104 and 105 molecule cm−3, especially in the Amazon rainforest region, where sCls could reach a maximum concentration of 105 molecule cm−3 (Novelli et al., 2017).”

1. Vereecken et al. is missing a citation.

2. The cited Novelli et al. article discusses measurement of [Criegee] in boreal forest environments, not rainforest environments.

3. The wording leaves it unclear if which Criegee concentration you trust more, Vereecken’s estimate or GEOS-Chem’s model. Please also motivate your judgement.

Typos and grammatical errors:

In the Abstract:

“species containing C≡N group” → “containing a C≡N group.”

“that the rate constants of” → “that the rate constant of”

“at below 260 K.” → “below 260 K.”

Line 55: “dual delve” → dual-level

Line 60: “the corresponding OH radical” → “the corresponding OH radical reactions”

Line 87: “The reliable density functional method was chosen” → “A reliable density functional method was chosen”.

I hope you were not trying to imply that M11-L is the only reliable density functional method. :)

Line 116: “performed by M06-CR/MG3S ” → “performed using M06-CR/MG3S”

Line 139: “there are three different…” → “there are three different functional groups in HOCH2CN:”

Line 141: “Three of the is” → “Three of them are”

Lines 147-148: The words “addition of” are entirely superfluous in the sentence describing the two addition reactions, due to the use of “is added to” later.

Line 249: “can be negligible” → “are negligible”

This manuscript describes new quantum chemical and computational kinetics calculations that show that reactions of carbonyl oxides are likely to play a significant role in the atmospheric transformations of hydroxyacetonitrile. The calculations are of high reliability and the conclusions are of substantial interest for understanding atmospheric reactions of wildfire products. I recommend publication of the manuscript and have some suggestions for clarification. First, the basis for assessing that hydroxyacetonitrile is in fact a significant atmospheric product in wildfire burning is half described — either the authors should simply refer to the publications that treat the analytical chemistry behind the assignment and its revision (which are lucidly described in the referenced reports) or they should go into more detail here. I would lean towards the first option, but as it is in the current manuscript the story is very confusing. Second, the manuscript treats simply the control of atmospheric removal of hydroxyacetonitrile, but the implications of this reaction for atmospheric chemistry depend also on the fate of the products and on the (still highly uncertain) tropospheric concentration of carbonyl oxides. The present calculations show a sequence of energetically accessible unimolecular transformations (not really “decomposition”) of the initial product. I am wondering if the authors could describe for completeness the lowest bimolecular channels of the reaction, and whether the fate of any of the proposed isomeric products would be expected to have different consequences for the atmosphere or environment than the others. Also, given that the reaction with OH would be competitive in many environments, would the products of that reaction have different implications than the products of this reaction? I understand that this stretches the boundaries of the work, and perhaps the authors will consider it out of scope, but in an atmospheric chemistry journal I think a bit of additional context in the conclusion section would be fitting.

An editing error I noticed: the citation to Haring 1942 is to the *review* of the book that the authors meant to cite, not the book itself. I believe the citation should be

Samuel Glasstone, Keith James Laidler, Henry Eyring, The Theory of Rate Processes: The Kinetics of Chemical Reactions, Viscosity, Diffusion and Electrochemical Phenomena; 611 p. McGraw-Hill Book Company, Incorporated, 1941