One of the reviewers wishes the authors to address the following points, before the paper can be accepted:

Thank you for your detailed answer and comments that have clarified some of my concern. I’d still like to ask for a major revision because I feel that some key characteristics of the experiments are not discussed detailed enough.

1. Surface coverage.
It is of paramount importance to state surface coverages during the experiments early in the manuscript.
* The reader needs to be assured that the experiments probed the SO2-ice interaction and were not done in the multilayer regime of the adsorption isotherms where SO2-SO2 interaction would have been probed.
* High surface coverage can lead to an overload of chromatographic column which results in tailing. By addressing surface coverage in the columns during the experiments early in the manuscript, this artifact needs to be ruled out.
* Surface coverage is also a key parameter to compare different experiments and characterize the experiments. If one wants to compare the findings of this study to other flow tube studies of uptake of acidic gases to ice (such as Crowley’s work on HCl) one need to make sure that similar settings prevailed, most important similar surface coverages.

I hope that the surface coverage can be estimated without using the adsorption isotherms to parameterize the results. It would be most useful to derive the coverage as directly form the chromatographs as possible also giving the opportunity to compare this estimate to the results of the Langmuir-Henry parameterization. As the amount and concentration of injected sample is known, estimates of surface coverage could maybe be gained based on the resolving power of the column to estimate the surface area of ice at which the interaction equilibrium is established as the peak moves through the column OR based on the partition coefficient as given in Eq. 1: k’ = tn/t0 = Conc(stationary phase)/Conc(mobil phase) *vol(stationary phase)/vol(mobile phase).
Your answer to my previous comment makes me confident that neither of these artifacts are an issue, but I’d like you to confirm and describe this in more details based on the raw data (or as close to the raw data as possible). Taken together this should address both the tailing — to rule out column overload — and the shift of the chromatographic peaks with time - to ensure that the weaker interaction (shorter retention time) is not caused by SO2 condensation or multilayer adsorption but by SO2-ice interaction.


2. Temperature trend.
An „anomalous“ temperature trend of SO2 sorption to ice has been described in earlier studies: While most trace gases show increased sorption with decreasing temperature, SO2 uptake increased with increasing temperature. Such behavior seems to be typical for acidic trace gases and has been explained by a combination of the temperature trend of adsorption and that of the acid-base dissociation (S. Zimmermann, M. Kippenberger, G. Schuster and J. N. Crowley, Phys. Chem. Chem. Phys., 2016, 18, 13799–13810). However, we are far from solving this puzzle and adding to this would make this manuscript indeed highly interesting for ACP community. I therefore ask to go into details. Could you show a Figure similar to Figure 2 of chromatograms at constant surface coverage but decreasing temperature? Or, you could base the discussion on Figure 3: At 0.05 Pa SO2 k increases from 210-220 = 220-230 K < 200-210K < 260-270 K < 230-240 = 240-250 = 250-26K. I’m a little concerned about the scatter, that you openly address throughout the manuscript. Restricting this discussion to a subset might be a good idea, or a detailed discussion of surface coverage would probably need to be considered. Any uptake model needs to be able to explain this particular temperature trend. It might help the reader to state and discuss the temperature trend in detail early in the manuscript. (On page 7 line 9, you state that a subset of experiments shows a T dep of k’. This statement is put in the context of discussing the Temkin isotherm. As k’ is an observable, I’d suggest to put this discussion upfront and establish your observed T trend first).



3. Relevance of Henry vs. Temkin/Langmuir
Taken that two uptake-mechanism working hand-in-hand describe your data well, the question arises on the relative importance of either. If I understood correctly, the Temkin isotherm or the dissociative Langmuir explain the change in DHads with surface coverage at low concentrations and the Henry explains the increase in peak area at short residence time (weaker adsorption) at high concentrations. Could you elaborate on the relative importance of either isotherm when going from low to high temperatures and from low to high concentrations. At which Conc and T does the Henry kick in?
In particular, taken that the adsorption sites for dissociation are so very low, the amount of SO2 that adsorbed dissociatively might be negligible compared to the total amount? Could you prove me wrong by discussion the fraction of SO2 the adsorption of which flows Henry and which follows dis-langmuir.

Thank you very much for considering this and I hope you find the comments helpful.

4. Minor comments:
please add a description of Langmuir and Temkin isotherms somewhere early in the text.What are the key basics of these concepts.
p3 line 15: how is the thickness of the ice determined?
p5 line 23: What was the concentration of SO2, methane, hexane, acetone that was dosed to the column?
p 10 ff: Please state concentration and surface coverage of these species as well.
p10 line 10: I’m a little concerned about the tailing and shift that you see with acetone. If the acid-base equilibrium is causing this for SO2, this should be absent in acetone data. Could this rather be an effect of surface coverage?