|The authors have made substantial changes to the manuscript in response to comments by myself and another reviewer, and the manuscript is much improved. I still have some, relatively minor, issues with the manuscript, as detailed below. Once these are addressed I can recommend publication in ACP.|
While the claims of “stabilisation”, “promotion”, “enhancement” etc have been toned down, and I especially appreciate the rewriting of the title, there are still a number of sentences that I feel are not really supported by the computed data.
-First, in the last sentence of the abstract, please explicitly spell out the conditions where the claim holds, i.e. replace the last part of the sentence by something like “promotes new particle formation in the atmosphere, at least in conditions where the SUA concentration is significantly larger than the SA concentration”.
-On line 259, the authors state that the negative hydration free energies suggest that “hydration stabilises the clusters”. This does not really follow, at least not if “stabilisation” is meant to mean stability toward the evaporation of acid or base molecules, not just toward the evaporation of water. Having a strongly negative free energy for the AB + W => ABW reaction may OFTEN correlate with having lower evaporation rates for the hydrated cluster (i.e. ABW => AW + B or ABW => BW may be slower than AB => A + B) but this is not necessarily the case: if the A + W or B + W reactions have even more negative free energies, then hydration will actually INCREASE the evaporation rate of AB, instead of decreasing it. (Examples of this can likely be found even in the authors’ own data. Also note that this applies for any number of A, B & W molecules in the cluster, not only the three-molecule example I discuss above). So, to recap: negative hydration free energies just mean that the cluster likes to be hydrated. It doesn’t in and of itself imply anything about the more general “stability’” of the cluster, and claiming such an implication is incorrect.
-Line 272: again this “energetically favourable” term: technically true by some definition, but very very misleading when applied to cases where the free energy of the addition reaction is only marginally negative (i.e. not even close to the approximately -10 kcal/mol threshold needed for the evaporation rate not to vastly exceed the formation rate at trace-gas concentrations).
-Line 287: sometimes the described interactions are indeed “synergetic” (should that be “synergistic”?) , but in many cases it seems they seem to rather be “antagonistic”, for example when the presence of SUA actually hinders SA-DMA clustering. This fact, that the multi-component interactions can go “both ways” in terms of their effects of clustering, could be noted here.
-Line 293-297: here the authors engage in some sort of “bait and switch” argumentation which is frankly quite intellectually dishonest. First they acknowledge that the free energy of adding SA to SA*DMA*SUA is higher than that of SA addition to SA*DMA. And then they state - as if this somehow invalidated the former conclusion - that “the free energies for adding SA to SA*DMA*SUA*(W)x clusters are negative”. Well yes, even the addition of SA to SA*DMA*SUA without water molecules was NEGATIVE - it was just less negative than the competing addition to SA*DMA. The crucial point here is that the SA + SA*DMA*SUA*(W)x reaction free energies (with x = 1…5) are all ONLY VERY SLIGHLY negative - actually even less negative than the x = 0 case. So in other words the SA molecule will very very rapidly evaporate from the (SA)2*DMA*SUA*(W)x clusters. Again, the “energetically favourable” argument made here may be technically true (by some, fairly irrelevant, definition of favourable), but completely and absolutely misleading: the (SA)2*DMA*SUA*(W)x clusters are actually very very unstable with respect to SA loss, and this should be honestly admitted here!
-There seems to be something wrong with equation 9 or how it is defined: if this the [cluster] variable is supposed to be the absolute concentration a cluster with x SA molecules, y W molecules, z AM molecules, n DMA molecules and m SUA molecules, then the expression should read
[cluster] = [SA]^x[W]^y[AM]^z[DMA]^n[SUA]^m exp(-dG/RT)
If the [cluster] variable is instead some relative concentration measure (as implied in some parts of the text), then this needs to be explained/defined better. Since I don’t understand exactly what the authors are calculating, I can also not evaluate the arguments on lines 349-351, where the computed values are used to support an argument about SUA “likely contributing to further growth” - however given the issue with the discussion on lines 293-297 described above I’m a bit suspicious about this claim too.
-The Ho et al reference seems to be about the urban atmosphere in China - this is unlikely to be representative of global dicarboxylic acid concentrations (which are very likely lower). It may well be that such global estimates do not exist - but then the authors should openly admit this, and also acknowledge that the SUA concentrations that they use likely correspond to the upper end of the global range. (This will require some amendments to the discussion on page 17 - also the use of the phrase “typical abundances” on page 19 is not really justified).
-The authors suggest on line 379 that “the presence of organic acids typically increases the dipole moments of clusters”. This may be generally true, however for comparing the competing mechanisms SA*DMA and SA*DMA*SUA, it’s not clear that the SUA-containing clusters will have necessarily have higher dipole moments than the corresponding SUA-free clusters. Since the authors actually have a huge amount of cluster data (including dipole moments) perhaps they could present the dipole moments for all the clusters in Table 2, to see if the argument actually holds for the specific systems studied here?
-On line 380, the authors again use the (often only very slightly) negative addition energies to imply that “SUA likely stabilises the SA*base clusters”. This is problematic in two ways: first, as discussed above, the authors own data indicates that for example the (SA)2*DMA*SUA*(W)x will actually lose SA very rapidly, and second, as discussed above in the context of hydration, even a strongly negative free energy for e.g. AB + SUA => (A)(B)(SUA) does not necessarily mean that the loss rate of A or B is decreased - it can even increase, if the binding of SUA to A or B individually is even stronger. I suggest just removing this whole sentence.
-Analogous to the introduction, please specify also in the conclusion that the “promoting” effect the authors claim to have discovered (which I am overall very sceptical about overall given the very high SA loss rates from the only SUA-containing 2-SA clusters studied in the paper) is only important when the SA concentration is relatively low, AND the SUA concentration high (i.e. it’s not enough for the conditions to just be “polluted”, they have to be “polluted” in a very particular way).
-The Kurtén et al 2008 paper (which by the way seems to be missing from the reference list even though it’s cited) is not a good primary reference for atmospheric DMA concentrations (it’s a purely computational study).