In their manuscript, ‘Partitioning of ionic surfactants in aerosol droplets containing glutaric acid, sodium chloride, or sea salts,’ Bain et al. describe surface tension measurements of microdroplets with different compositions containing ionic surfactants. Like previous measurements from the same researchers, the critical micelle concentrations (CMCs) of ionic surfactants in microdroplets are significantly larger than those macroscale solution due to bulk depletion effects. Two separate partitioning models are used to explain the surface partitioning behavior of surfactants in microdroplets: a Simple Kinetic Model and a Monolayer Model. While both models predict similar ‘effective CMCs’ in microdroplets, the simple kinetic model is found to predict the surface tension in droplets significantly better than the monolayer model. Overall, the manuscript is well-written and the experimental results are interesting (albeit unsurprising). After addressing the comments and suggestions below, the manuscript is suitable for publication in Atmospheric Chemistry and Physics.

Comments:

Figure S2. While most of the macroscale parametrizations of surface tension vs SDS concentration yield similar curves that match experimental data, the monolayer model parametrization for CTAB in NaCl and sea salt solution is not good and does accurately fit the experimental data. It is surprising this is not mentioned in the main text and that the ‘best fit’ parameters are still used to predict droplet surface tension. The authors need to address why the monolayer model does not accurately the macroscale experimental data and discuss how this influences microdroplet surface tension predictions.

Line 228. Jacobs et al. (DOI: 10.1021/acs.jpca.4c06210) recently presented a simple expression to predict the effective CMC of surfactants in microdroplets using only the droplet size and macroscale Langmuir parameters. It could be interesting to include a comparison of the experimental effective CMCs to those predicted using the simple expression.

Figure 2 should include error bars that represent experimental uncertainty (x) and uncertainty in size for the model results (y).

I appreciate that the experimental data were not ‘normalized’ to the expected minimum surface tension for a surfactant system. However, an obvious experimental artifact/limitation that precludes direct comparison to the two models is the formation of a compressed when droplets collide. Fig. S5 demonstrates that the Langmuir parametrization may extend past the surface excess concentrations that are possible in macroscale solution. Thus, I wonder if it is possible to use this calculate the surface excess concentrations (and by extension surface tensions) in droplets pre-merge geometrically by assuming equal sized droplets and that all molecules at the surface remain at the surface during the measurement.

Line 262. Should read ‘…with decreasing CMC.’

Consider adding a horizontal line to Figure 4b showing what the fractional coverage is at the CMC.

General Comments:

In this work, the authors present new results on the surface tension of picoliter droplets (radii of about 8 µm) containing an ionic surfactant and cosolutes, and the comparison between these experimental data and two partitioning models. The authors provide a detailed presentation of the method, as well as extensive use of models used and experimental results obtained. The results support previous findings from similar experiments. This manuscript contains also an in-depth discussion of the drawbacks of the experimental set-up and the Monolayer model used to predict surface tension. The conclusion that NaCl can be used as a substitute for salt in sea spray aerosols for surface tension studies will be very useful to the community. I have minor comments and questions to the authors, detailed thereafter.

Specific Comments:

l.20: The work of (Jacobs, Johnston, et Mahmud 2024) on ionic surfactants is also consistent with your results. You can modify this sentence to take account of this work.

l.69: “atmospheric proxy anionic or cationic surfactants” and l.181: “CTAB and SDS were chosen as commercially available cationic and anionic representatives of atmospheric surfactants.” To what extend are they good proxies? For example, in (Gérard et al. 2016; 2019), the critical micelle concentrations (CMC) of the aerosol extracts from filters collected in coastal regions are in the range of a few 100s µM, whereas the CMCs of SDS and CTAB are about 5-10 mM and 1 mM, respectively. Could you comment on that?

It is not clear to me why you write this much about divalent cations. I believe that monovalent cations, and surely anions, also alter the surface properties of surfactants by salting-out effects. According to the Hofmeister series, sodium has a larger salting-out effect than calcium (Hyde et al. 2017). In addition, sulfate, which is also included in the sea salt mimic, has an even stronger salting-out effect than chloride, therefore I wonder why you only mention calcium and not sulfate. Could you clarify this point?

l.89: Where does the a_w = 0.99 comes from?

l.106: “The concentration of surfactant in the droplet was then determined using the molar ratio of surfactant to cosolute in the nebulized solution” In (Bzdek et al. 2020), you justify this assumption on three grounds:

1. “collection of nebulized aerosol containing glutaric acid and Triton X-100 had the same RI and surface tension as the initial solution and the residual solution in the nebulizer, which is only expected if the glutaric acid:Triton X-100 ratio is conserved upon nebulization”
2. “if the ratio were not conserved, one might expect substantial variability in retrieved surface tensions among measurements produced from solutions with the same primary solute:surfactant ratio. In fact, the surface tension values of droplets produced by nebulization of the same solution were uniformly consistent (<±3 mN·m−1) over several weeks even using different nebulizers”
3. “changed. Third, in previous work with mixtures of glutaric acid and NaCl, the retrieved droplet RI and corresponding surface tension gave excellent agreement with model predictions and macroscopic solution measurements, indicating the relative ratio of glutaric acid and NaCl in solution was conserved” upon nebulization

I must say that I am very surprised that the solute:surfactant ratio remains constant upon nebulization. You have very strong arguments to argue that this ratio is not changed in droplets, yet I would expect aerosol produced by nebulization to be enriched in surfactants, as observed by (Faust et Abbatt 2019). The reason probably lies behind the large volume of droplets you analyze, or comes from your nebulizer. However, I would strongly advise to any future work to check this assumption, especially for works measuring surface tension of smaller droplets. You checked thoroughly the surface tension of nebulized solution with Triton X-100 for your work in (Bzdek et al. 2020), did you take the same precautions with SDS and CTAB in this work? Could you comment more on your nebulization system?

L.117: You set n=2 for 0.9 M glutaric acid solute, but glutaric acid dissociates into its conjugated base and a proton. At such a high concentration, wouldn’t it be preferable to consider this solute as an electrolyte, and set n=1? In addition, setting n=2 leads to lower values for Γ_max for both SDS and CTAB with glutaric acid compared to other cosolutes (see Table 1). However, if you set n=1, Γ_max=4.32 10^-6 and 2.02 10^-6 mol m^-2 for SDS and CTAB, respectively, which makes more sense to me compared to the values of Γ_max obtained with other cosolutes.

On the method used to measure surface tension of droplets. Your results show that the minimum surface tension measured beyond CMC in picolitre droplets is lower than the minimum surface tension reached in macroscopic solutions. In this study, you sum up this observation in the sentence “When the concentration is greater than the effective droplet CMC, nonequilibrium surface concentrations impact the measured surface tension”, and call it a bias in the result section (l.239-241). In (Bzdek et al. 2020), you explain that this “may be due to the presence of the surfactant film on the surface of the droplet, which […] can decrease the droplet oscillation frequency from its true value, resulting in a retrieved surface tension below the true value. Another possible explanation is that the composite droplet surface is slightly enriched in surfactant compared to a fully equilibrated surface. Coalescence produces a composite droplet with a smaller total surface area than the two initial droplets. If surfactant diffusion away from the composite droplet surface is slower than the timescale of the shape oscillation (10 to 100 μs), the composite droplet surface will not reestablish equilibrium, resulting in a tighter packing of surfactant molecules and a reduction in surface tension.” Is there any particular reason why the surface of composite droplets would be enriched with surfactants only at concentrations above the CMC, but not below the CMC? At all concentrations, the droplet resulting of the fusion of two smaller droplets has a lower total surface area (by a factor 2^2/3), and a higher surfactant surface content (by a factor 2), therefore a higher surfactant surface concentration (by a factor of 2^1/3), before diffusion of surfactants from the surface to the bulk. In the end, this could lead to a systematic underestimation of droplet surface tension.

l.200 you could also cite the work of (El Haber et al. 2023) in which you can find a lot of complex mixtures of surfactants, salts, and organics.

Figure S4: Could you be more explicit on how you plot this figure? Could you comment on the meaning of ζ? Why did you choose to plot against ζ instead of plotting against the droplet radius?

l.218: “For SDS with glutaric acid cosolute (Fig. 1A), the macroscopic CMC and effective CMC are in close agreement.” It does not seem to me that the droplet surface tension values reach a plateau here. This sentence is a little too affirmative.

Overall, I think that the end of the article lacks a few sentences about the implications for aerosol-cloud interactions. In light of your results on both non-ionic and ionic surfactants, do you think that surfactants could likely play a role in cloud droplet activation? For example, what would be the concentration of surfactants in the droplet required to reach a 40 mN/m surface tension lowering, with the models and the compounds you studied?

Technical corrections

l.9: Do you oppose “soluble” to “surface-active”? The term “fully dissolved” would be more appropriate.

l.28: “jet drop aerosol generation pathways have been shown to produce droplets with enriched in surfactant” typo.

l.54-55: “the addition of NaCl to SDS or CTAB solutions […] enhances the equilibrium and surface concentration”, do you mean the surface concentration at equilibrium?

l.286: “The Simple Kinetic Model, which accounts for the partitioning of surfactant in droplets with high SA-V ratios and agrees well with experimental measurements, is used to calculate the surface tension and fractional surface coverage in a 10 µm radius droplet” is redundant with the first sentence of the same paragraph.

l.307: “The macroscopic CMCs of ionic surfactants are greatly impacted by the presence of salt” change “impacted by” by “reduced in”.

References

Bzdek, Bryan R., Jonathan P. Reid, Jussi Malila, et Nønne L. Prisle. 2020. « The surface tension of surfactant-containing, finite volume droplets ». Proceedings of the National Academy of Sciences 117 (15): 8335‑43. https://doi.org/10.1073/pnas.1915660117.

El Haber, Manuella, Corinne Ferronato, Anne Giroir-Fendler, Ludovic Fine, et Barbara Nozière. 2023. « Salting out, Non-Ideality and Synergism Enhance Surfactant Efficiency in Atmospheric Aerosols ». Scientific Reports 13 (1): 20672. https://doi.org/10.1038/s41598-023-48040-5.

Faust, Jennifer A., et Jonathan P. D. Abbatt. 2019. « Organic Surfactants Protect Dissolved Aerosol Components against Heterogeneous Oxidation ». The Journal of Physical Chemistry A 123 (10): 2114‑24. https://doi.org/10.1021/acs.jpca.9b00167.

Gérard, Violaine, Barbara Nozière, Christine Baduel, Ludovic Fine, Amanda A. Frossard, et Ronald C. Cohen. 2016. « Anionic, Cationic, and Nonionic Surfactants in Atmospheric Aerosols from the Baltic Coast at Askö, Sweden: Implications for Cloud Droplet Activation ». Environmental Science & Technology 50 (6): 2974‑82. https://doi.org/10.1021/acs.est.5b05809.

Gérard, Violaine, Barbara Noziere, Ludovic Fine, Corinne Ferronato, Dharmendra Kumar Singh, Amanda A. Frossard, Ronald C. Cohen, et al. 2019. « Concentrations and Adsorption Isotherms for Amphiphilic Surfactants in PM1 Aerosols from Different Regions of Europe ». Environmental Science & Technology 53 (21): 12379‑88. https://doi.org/10.1021/acs.est.9b03386.

Hyde, Alan M., Susan L. Zultanski, Jacob H. Waldman, Yong-Li Zhong, Michael Shevlin, et Feng Peng. 2017. « General Principles and Strategies for Salting-Out Informed by the Hofmeister Series ». Organic Process Research & Development 21 (9): 1355‑70. https://doi.org/10.1021/acs.oprd.7b00197.

Jacobs, Michael I., Madelyn N. Johnston, et Shahriar Mahmud. 2024. « Exploring How the Surface-Area-to-Volume Ratio Influences the Partitioning of Surfactants to the Air–Water Interface in Levitated Microdroplets ». The Journal of Physical Chemistry A 128 (46): 9986‑97. https://doi.org/10.1021/acs.jpca.4c06210.