This is the review of the manuscript entitled “Deposition freezing, pore condensation freezing and adsorption: three processes one description?” by Mária Lbadaoui-Darvas, Ari Laaksonen, and Athanasios Nenes. This study aims to gain a deeper understanding of deposition ice nucleation by means of molecular simulations. Deposition ice nucleation is mechanistically compared to the recently suggested pore condensation freezing (PCF) and how it is related to adsorption. The results of this study support the application of the so-called adsorption nucleation theory (ANT) to describe deposition ice nucleation in place of classical nucleation theory (CNT), PCF, and other parameterizations. A black carbon substrate, with and without pores, serves as a surrogate of an ice nucleating substrate. The simulations show that a quasi-liquid layer remains on the water cluster on the surface, i.e., adsorption and ice nucleation occur in a disordered phase. Also, the input parameters for ANT that describe the interaction between water or ice and black carbon are very similar suggesting that an ANT description might hold for the water nucleation (adsorption) as well as for deposition ice nucleation.

Further fundamental understanding of ice nucleation is crucial, and this study nicely fills this gap. Hence, in terms of topic and theme it fits in the scope of the journal Atmospheric Chemistry and Physics. In general, I am in support of publishing this study. However, I have a couple of revision requests the authors should address.

I think it should be stressed that the terminology/definition of deposition ice nucleation is historically macroscopically defined (before the application of in situ microscopy and MD simulations). For many current experimental techniques, this still has validity. Though on a molecular level this may not be true. Following the conventional definition, it is “deposition ice nucleation” and not “deposition freezing”. Liquid (macroscopic) water freezes, but deposition ice nucleation does not involve (following convention) bulk liquid water. If the authors by purpose mix these two definitions and generate a novel terminology, since they observe deposition ice nucleation to originate from a liquid-like water cluster, then this has to be discussed. However, this seems not to be the case since this term is readily used. Also, I would not challenge the conventional definition based on one simulation study only. Hence, “deposition freezing” should be exchanged for deposition ice nucleation throughout the manuscript.

The other issue regarding terminology is to call the water clusters “droplets”, “dropletwise”, etc. I see that the authors struggle with this issue as well, trying also “nanodroplets” or “nanophase”. In this community droplets are usually defined to be 10s of micron in size. The “nanodroplets” forming inside the pore are about 2 nm or smaller. Typically, we call those entities clusters. It may not even be clear if this cluster size possesses bulk-liquid water properties (surface tension, etc.)? I am also not entirely sure how to name those condensed nanometer-sized islands of water but naming those “droplets” is unfortunate and ambiguous. Liquid-like or ice-like water clusters may be an idea. Maybe “nanodroplets” works to convey the idea but I feel this is not ideal either.

From the abstract and introduction, one would expect some analysis using ANT, i.e., deriving ice nucleation rates, etc. However, this study makes the case that ANT can be applied to deposition ice nucleation based on the simulation results. As written, this fact may not be so clear, and the overall confidence is only supported by this study looking at one idealized substrate. Maybe in some instances the text could convey a more exploratory study. I do not disagree with the authors; I suggest being a bit more conservative. Especially when reading the model methods. Many caveats are discussed (which I appreciate, and this does not minimize the impact of the study) but it feels a bit counter (i.e., weaker) to the introduction. This is maybe something the authors could consider.

Specific comments:

Line 22: The authors could cite here the recent review by (Knopf and Alpert, 2023).

Line 30: I doubt that (DeMott et al., 2010) discuss in detail nucleation theory and rates relevant for this paper and they do not discuss specifically deposition ice nucleation. Other literature would be needed in this place.

Line 37: “…adsorbed water can exist….”

Line 40: At this point it is not clear what you mean by “whereas other locations that collect pre-critical clusters might have an opposite effect.” Why do they have an opposing effect?

Line 47: Here it suddenly switches to immersion freezing. I recommend keeping the focus on deposition ice nucleation throughout the introduction. Also, I am not sure if I agree with this statement. When CNT is expressed in terms of water activity, intrinsic parameters like contact angle, interfaces, etc. are considered. See, e.g., the work by Knopf and Barahona groups. In fact (Knopf and Alpert, 2023) show that deposition ice nucleation may be well described using water activity as for the case of homogeneous ice nucleation and immersion freezing.

Line 55-57: Missing words, empty brackets?

Line 73-75: As mentioned above, considering water activity in CNT descriptions might account for these issues (Knopf and Alpert, 2023; Koop et al., 2000; Barahona, 2015, 2014; Knopf and Alpert, 2013).

Line 81: Period missing?

Line 81-86: A long sentence, maybe too long. Also, this statement is too general. Careful literature review will show that there are several studies (some employ nanoscale resolution) which do not corroborate PCF occurring in observed deposition ice nucleation experiments, e.g. (Wang et al., 2016). I would avoid “in reality” and write “…’freezing’ could be pore condensation…”.

Line 112: It is crucial to conduct atomistic and coarse-grained molecular simulations as discussed in (Knopf and Alpert, 2023) and shown in (Roudsari et al., 2022). They can yield different results while atomistic simulations are likely the preferred method, if feasible.

Line 127: What do you mean by energetic background? This is not a thermodynamic expression. Maybe just state the parameters you assess?

Line 146-147: Could you please elaborate here. The target vapor pressure corresponds to the adsorption layer structure”? At such high vapor pressure, one would have multiple layers of water? I assume, this is what you want?

Line 159: What do you mean by “deterministic dynamics”?

Line 168: Figure 1 instead 4?

Line 170-175: Long sentence with a lot of information. Again, a rather strong caveat. Maybe split up this information.

Line 194: sigma parameter has no units?

Line 242: What is “dropletwise phase”? Does this expression exist? See also general comment.

Line 247: You discuss Fig. 5 before Fig. 4?

Figure 3: Panel (b) is missing?

Line 277: Does this statement depend on pore size? Is it correct to generalize this?

Line 281: Maybe semicolon after “point I.” to make this easier to follow?

Line 287: “In neither case is complete pore filling required for freezing.” I think this is a very interesting and important finding of this study. It challenges the PCF mechanism. I am wondering why this is not mentioned in the abstract.

Figure caption 4: I am not sure what is the difference between freezing onset and initial stage. Onset is an “initial’ condition in a way. Not sure how to better (unambiguously) define this. Panel (b) is not described? And panel indicator “(b)” in caption text should be “(c)”?

Line 319-320: Several MD studies suggest that the critical nucleus forms in second water layer, e.g., (Cox et al., 2013). I assume the nucleus attaches after it forms?

Line 324-327: This sentence could be split in two to make understanding easier. From where comes the insight that layer-by-layer ice growth (is it growth or nucleation?) represents barrierless freezing? Please add references. Though if it is growth, then we talk about other energies than that needed for nucleation?

Line 327: What do you mean with “it shifted to the left”?

Line 360: What is meant by (kal)?

Line 361 -362: The A-values have no units?

Line 458: Clearly these 2 nm clusters are not supercooled droplets in the conventional application of terminology.

Line 479: empty citation?

Line 489: Was ITIM defined previously?

Line 493: (kal)?

Technical corrections:

Line 18: omit first “of”

Line 39: superfluous space

Line 55: “…implemented in regional…”

Line 102: solid-vapor phase? I suggest to use hyphens for expressing interfaces.

Line 117: Omit “even”.

Line 195: superfluous space

Line 205: space missing

Figure caption 2: Shaded areas not visible in my pdf file.

Line 231: double citation.

Line 282: superfluous space

Line 318: “dropletwise”?

Line 325: reach

Line 393: space needed before “for”

Line 442: adsorption

Line 444: …superior compared to that on…

Line 470: proven

Line 480: superfluous space

Line 480: molecules, respectively,

References

Barahona, D.: Analysis of the effect of water activity on ice formation using a new thermodynamic framework, Atmos. Chem. Phys., 14, 7665-7680, 10.5194/acp-14-7665-2014, 2014.

Barahona, D.: Thermodynamic derivation of the activation energy for ice nucleation, Atmos. Chem. Phys., 15, 13819-13831, 10.5194/acp-15-13819-2015, 2015.

Cox, S. J., Raza, Z., Kathmann, S. M., Slater, B., and Michaelides, A.: The microscopic features of heterogeneous ice nucleation may affect the macroscopic morphology of atmospheric ice crystals, Faraday Discuss., 167, 389-403, 10.1039/c3fd00059a, 2013.

DeMott, P. J., Prenni, A. J., Liu, X., Kreidenweis, S. M., Petters, M. D., Twohy, C. H., Richardson, M. S., Eidhammer, T., and Rogers, D. C.: Predicting global atmospheric ice nuclei distributions and their impacts on climate, Proc. Natl. Acad. Sci. U. S. A., 107, 11217-11222, 10.1073/pnas.0910818107, 2010.

Knopf, D. A. and Alpert, P. A.: A water activity based model of heterogeneous ice nucleation kinetics for freezing of water and aqueous solution droplets, Faraday Discuss., 165, 513-534, 10.1039/C3FD00035D, 2013.

Knopf, D. A. and Alpert, P. A.: Atmospheric ice nucleation, Nat. Rev. Phys., 10.1038/s42254-023-00570-7, 2023.

Koop, T., Luo, B. P., Tsias, A., and Peter, T.: Water activity as the determinant for homogeneous ice nucleation in aqueous solutions, Nature, 406, 611-614, 10.1038/35020537 2000.

Roudsari, G., Pakarinen, O. H., Reischl, B., and Vehkamaki, H.: Atomistic and coarse-grained simulations reveal increased ice nucleation activity on silver iodide surfaces in slit and wedge geometries, Atmos. Chem. Phys., 22, 10099-10114, 10.5194/acp-22-10099-2022, 2022.

Wang, B., Knopf, D. A., China, S., Arey, B. W., Harder, T. H., Gilles, M. K., and Laskin, A.: Direct observation of ice nucleation events on individual atmospheric particles, Phys. Chem. Chem. Phys., 18, 29721-29731, 10.1039/C6CP05253C, 2016.

Review of “Deposition freezing, pore condensation freezing and adsorption: three processes one description?” by M. Lbadaoui-Darvas, A. Laaksonen, and A. Nenes, submitted to Atmos. Chem. Phys.

General comments and assessment:

The authors present atomistic and coarse-grained molecular dynamics and grand canonical monte carlo simulations of deposition freezing on graphitic surfaces that are either flat or contain a pore. The main result is that water molecules from the gas phase first form supercooled liquid-like “droplets” when adsorbing to the surface, before a critical ice nucleus forms heterogeneously. Based on these observations, the authors conclude that deposition freezing could be described by the adsorption nucleation theory (ANT).

The study is scientifically relevant and of high quality; it represents an important contribution to understanding atmospheric ice nucleation on the atomistic level. The manuscript is overall well written and the results are mostly presented in a clear way. However, there are quite many technical issues in the text and the figures, listed below, that the authors need to address. I therefore recommend the Editors accept the manuscript after the necessary revisions.

Specific comments:

The atomistic simulations of the graphite/water interface, used to determine the FHH parameters employ the TIP5P water potential. While TIP5P overall predicts structural and thermodynamic properties of water and ice quite well, it does underestimate the density difference between water and ice Ih [see e.g. Vega et al.,  J. Phys.: Cond. Matter 7, S3283–S3288 (2005)]. Do you believe that using a water model yielding more realistic differences in density between liquid and ice phases, in conjunction with the fixed definition of the layer thickness, could lead to systematic differences in the average interaction energies obtained, in addition to possible differences in interaction energies resulting from the different water model?  

I would encourage the authors to review their variable names and formulae. Throughout the manuscript, sub- or superscripts in variables that do not denote indices are set in italics (e.g. A_{FHH}, k_B, \sigma_{OC}, V_{filled}, N_{ice}, etc.). These sub- and superscripts should be changed to roman font. In addition, several variable names appear unnecessarily long or convoluted to me (e.g., in eq. 2, the number of carbon atoms “NC” could be “N_\mathrm{C}” and the number of water molecules in i-th adsorbed layer, “NW_{Li}” could simply be “N_i“, so that the mean interaction energy per unit area in the i-th adsorbed layer could simply be denoted “E_i”.

Technical corrections:

l.14: not only the extension … is possible -> not only is the extension … possible

l. 39: clusters , -> clusters,

l.55: implemented regional -> implemented into regional

l.57: two broken references

l.69: …with direct links to the molecular-scale interactions…

l.139: …a graphite slab with a hemispherical or cylindrical pore consisting…

l.164: numercially->numerically

l.171: remove comma after carbon

l.176: remove spurious “because”

l.178: any estimating -> estimating any

l.194: units missing for \sigma_OC^LJ (I assume nm)

l.206: space missing before reference.

Figure 2 caption: describe arrows and label water and graphite in panel (a). Shaded areas mentioned in the caption not visible in panel (b).

l.230: increase of adsorbed water molecules

l.231: re-format reference

Figure 3:  “b)” is missing in the figure. A period is missing at the end of the first sentence in the caption. Carbon atoms are not mentioned in the caption.

l.299: steepness -> slope?

Figure 4: “teal” and “blue” curves are very hard to distinguish; please change one of the two colors! “gray” curve appears black to me, especially as the graph has a gray background. Period at the end of the first sentence of the caption is missing. “adsobing” -> adsorbing. “out-of-pore nucleation” -> ice growing out of the pore?

Figure 5: panels (a) and (b) in the actual figure are not references in the caption. Instead, the caption references a “top” and “bottom” panel, which are both part of panel (a) in the actual figure. The content of panel (b) is not explained at all in the figure caption, but it is referenced in the main text…

l.327: shifted towards the left -> shifted to smaller time values?

l.345 remove “makes”

l.355 “and a A_{FHH}” -> and A_{FHH}

l.360 broken reference

l.381: missing math environment 3k_{B}T/2

l.385/Fig 6: in the text, E_i has the unit of energy (kJ/mol), in the figure the unit of energy/area (kJ/mol/nm^2). Please fix this.

l.386: missing math environment k_{B}T

Figure 6: caption: add explanation that “LnS” denotes the interation between the surface and the n-th adsorbed layer in the graphs. add information that the means and standard deviations(?) (in parantheses) of the distributions are also shown as insets on the graphs of the distributions. Check whether the unit of energy/surface area is the correct one for E_i (see l.385 in the main text)

l.393: missing space after 30º

l.399: ice made up TIP5P water molecules -> the ice phases of the TIP5P water model?

l.401: box plots obtained from the time series… -> box plots of the distribution…

Figure 7: caption: box plots of the time series… -> box plots of the distribution…

l.425:  “the potential that deposition freezing can be described” -> “that deposition freezing can/could be described”