The manuscript entitled, “Desorption Lifetimes and Activation Energies Influencing Gas-Surface Interactions and Multiphase Kinetics,” by Knopf et al., details the importance and impact of accurately measuring or deriving desorption energies to describe trace gas uptake and reaction.  The authors first give an excellent overview of the meaning of desorption and how desorption energies are needed to accurately model multiphase phase kinetics either using simplified resistor models or more sophisticated kinetics simulations.  They show this by changing desorption rates (i.e. by temperature) in K2-Surf simulations of reactive uptake (Figs. 1-5).  The authors then compile an exhaustive list of previously measured desorption energies, from experiment and theory, for various gases onto solid and liquid interfaces.    From this large data set the authors proceed to develop correlations between molecular properties (polarizability, O:C, MW, relative permittivity) and E_des.  The author’s objective is to develop simple ways that E_des can be easily estimated from molecular properties.

Overall, the manuscript is well written and easy to follow.  The amount of data considered and compiled from prior literature is impressive and a great service to the community.

There are a number of comments that the authors should address in their revision.

1. Eq. (3) is the normalized loss rate due to a surface reaction. Thus, shouldn’t the denominator be the sum of desorption and reaction?  Same question for Eq. (4), should the denominator be the sum of desorption and surface-to-bulk transfer?
2. Page 8 line 157. I think an additional sentence is needed to make clear from a physical perspective why accommodation, desorption and surface reaction are intertwined quantities?
3. Page 10 line 196. Given the confusing terminology used in the field, I think a few clarifying sentences are needed to link surface accommodation with thermal accommodation.
4. Page 16 line 337 and Page 25 line 547. The authors include only short paragraphs about liquid substrates.  I agree with the authors that despite some key differences between solid and liquid interfaces the formulations developed in the manuscript nevertheless remain useful.  However, for clarity I do think that the authors need to expand this discussion of liquids a bit to include not only experimental measurements but also theoretical concepts such as interfacial thickness and solvation energies derived from potential of mean force (PMF) calculations in MD simulations.  Recently, for example, desorption and solvation rates/dynamics are directly obtained using these PMF.  For example, see, Cruzeiro, V.W.D., et al. Uptake of N₂O₅ by aqueous aerosol unveiled using chemically accurate many-body potentials. Nat Commun 13, 1266 (2022). https://doi.org/10.1038/s41467-022-28697-8 and Mirza Galib, David T. Limmer, Reactive uptake of N₂O₅ by atmospheric aerosol is dominated by interfacial processes. Science 371, 921-925(2021)
5.  Page 20 line 438. Space between “)A” is needed.
6. Page 31. The example results shown in Fig. 5 for the uptake of a non-reactive species into water is confusing.  The equilibration timescale above E_des > 30 kJ/mol seems entirely dominated by desorption rather than the rate at which the trace gas diffuses below the interface, which should be very fast?  What is assumed about the rate coefficient for surface-to-bulk transfer in this example?  In other words what is assumed about the mass accommodation coefficient in this example?  I believe these details are needed for the reader to assess the actual meaning of the simulations shown in Fig. 5.
7. Page 44 line 983. Fig. 14 cited in the text should be Fig. 15
8. Page 44 line 968. I do not think that the correlation between the glass transition temperature and E_des is robust and physically defensible.  There are many papers (see below*) now showing that the mobility of molecules at glass surfaces can be quite different (i.e. faster) than those molecules in the glass interior. Since desorption is sensitive to the fine details of the interface, which are clearly more complex for a glass, I do not think discussion on page 44 and the associated Fig. 15 is justified.  Unless the authors can make a stronger case, I recommend this entire discussion be removed from the manuscript.  *(Zhang and Z. Fakhraai, Decoupling of surface diffusion and relaxation dynamics of molecular glasses, Proceedings of the National Academy of Sciences, 2017, 114, 4915-4919. Sikorski, C. Gutt, Y. Chushkin, M. Lippmann and H. Franz, Dynamics at the Liquid-Vapor Interface of a Supercooled Organic Glass FormerPhysical Review Letters, 2010, 105, 215701.Tian, Q. Xu, H. Zhang, R. D. Priestley and B. Zuo, Surface dynamics of glasses, Applied Physics Reviews, 2022, 9.)
9. Appendix. I believe that a list of acronym definitions (near the tables) would be helpful for a reader who doesn’t want to search through the text for these.  These could be placed and the beginning of the Appendix or as foot notes to the tables.

The authors build up on (their) previous work (Knopf and Ammann, 2021), explaining desorption lifetime and desorption energy and its importance for a molecular based approach for describing trace gas uptake and surface reaction on aerosol particles. Here, they provide an overview over the different experimental and modeling approaches to determine desorption energies. In addition, they survey and compile an impressive list of desorption energies and use these data and calculations to come up with a parametrization based on compound mass, polarizability and O:C ratio. 

The manuscript is well written and it will certainly serve the community by providing not only the compilation of desorption energies but also illustrating its relevance for a molecular understanding of reactive uptake.

I would like the authors to consider a few comments below for the revision of the manuscript.

(1)  As a non-specialist in this particular field, after reading the introduction, I am still struggling to understand the role of the surface accommodation coefficient, α_S. As the (the authors) Knopf and Amman (2021) write: “In the context of atmospheric sciences, adsorption is commonly described by the surface accommodation coefficient, which is the probability that a molecule undergoing a gas kinetic collision is adsorbed at the surface”. Hence, in the case of a “simple” physisorption, this probability need to somehow relate to the desorption lifetime as the authors explain starting in line 87: “In case of physisorption, E_des is equal to the negative value of the enthalpy of adsorption with a correction for the change in degree of freedom between gas and adsorbed phase”. I certainly will benefit from an expanded explanation on why we cannot then set the accommodation coefficient equal to one very generally, once we base the desorption process on the Frenkel equation (eq. 1). For example in equations 3, and 4 there remains this molecular interpretation of the terms Γ_s and Γ_sb with the first-order desorption rate based on the desorption energy, but there is also this unexplained (in terms of molecular properties) surface accommodation term. Furthermore, what follows for the surface accommodation coefficient when we assume reversible adsorption (line 252)?

(2) At the end of section 3.2. “Gas adsorption by solid surfaces” the authors correctly discuss that most often atmospheric particles may have a condensed aqueous solution on its surface. They also state that in these cases one should consider the uptake process as proceeding on liquid substrates. However, the high vapor pressure of the relevant liquids does not allow easily to measure desorption kinetics (line 542). The authors suggest to use nevertheless the same concept although they admit (line 563) here the hydrogen bonding network is of particular importance and this may depend on the solutes being present. May be the authors could come back to this problem in their conclusion section?

(3) Same section “Gas adsorption by ice”: I recommend to cite the review by Huthwelker et al. (2006) here for those who are interested in experimental techniques and available data and a discussion with a different focus.

(4) line 793 ff: I suggest to have similar figures for the different substrates (ice, water, aqueous, solids) like Fig.8 in the SI using the parametrization of eq. 16 to show that there are no significant physical state of substrate specific differences.

(5) line 800 ff, Correlation desorption energy enthalpy of vaporization: Could it be that for the atmospherically very relevant liquid substrates, a parametrization based on this correlation including O:C (Fig. 11(c)) is as good as eq. (16) in particular for substances with O:C > 1? In this context: the statement in line 821 that E_des and ΔH_solv are better correlated than E_des and ΔH_vap may be not true if using O:C for a parametrization as well.

(6) line 968 ff: “glass transition”: I do not feel that the correlation between galss transition temperature and E_des goes much beyond that both correlate with molecular mass. I think this section deviates very much from the more solid molecular picture of the other sections and should be omitted.

References:

Huthwelker, T, Ammann, M. and Peter, T.: The Uptake of Acidic Gases on Ice, Chem. Rev., 106, 1375-1444, 2006

Knopf, D. and Ammann, M.: Technical note: Adsorption and desorption equilibria from statistical thermodynamics and rates from transition state theory, Atmos. Chem. Phys., 21, 15725–15753, 2021