the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 May 2026
Co-condensation and co-evaporation of levoglucosan onto and from deliquesced ammonium sulfate particles – influence of relative humidity, particle mass and size, and presence of a surfactant
Abstract. Co-condensation is the process by which condensable vapors condense alongside water vapor onto growing aerosols, and it can significantly alter the properties of clouds. Semi-volatile species partition dynamically between the gas and the condensed phases, and their co-condensation together with water vapor may increase the condensed mass and amplify water uptake. Although this process is based on thermodynamics and has been simulated in models, it has been scarcely investigated experimentally. In this study, the ability of levoglucosan to co-condense and co-evaporate from inorganic particles together with water was investigated by monitoring, under ambient-like conditions, its gas-particle partitioning on ammonium sulfate (AS) monodispersed particles in the Experimental Multiphasic Atmospheric Simulation Chamber (CESAM). The net evaporation flux of levoglucosan depending on particle size and the presence of a surfactant was explored for relative humidities (RH) from 100 % to dry conditions as particle concentration in the chamber was reduced. Due to the high deliquescence point of AS, wet experiments were initialized at RH above 80 %. Co-condensation of levoglucosan was observed when RH increased up to 100 %, while co-evaporation of levoglucosan occurred when RH decreased. It was shown that gas-particle partitioning of levoglucosan was sensitive to all the investigated parameters, but the main drivers were the levoglucosan-to-sulfate mass ratios and RH. The effect of the levoglucosan-to-sulfate mass ratio was interpreted as salting-out, and the significant influence of RH provides experimental proof of co-evaporation or co-condensation of levoglucosan from or onto AS particles.
- Preprint
(1754 KB) - Metadata XML
-
Supplement
(802 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on ar-2026-14', Anonymous Referee #1, 01 Jul 2026
-
RC2: 'Comment on ar-2026-14', Anonymous Referee #2, 13 Jul 2026
This manuscript describes a set of experiments conducted in the CESAM chamber to study evaporation and condensation of levoglucosan to/from aqueous solutions droplets under varying conditions of relative humidity and initial dry particle size.
The topic is relevant and the experimental idea interesting. However, I find however that the manuscript in its current form is very difficult to read and that it needs to be reconsidered after clarifications and major revision.
First, I find that the description of the experiments and the corrections need clarification, see further comments and some suggestions for improvement below.
Second, in some places, statements are given without literature references, clear description of assumptions, or without support by thermodynamic calculations which could be done using online-tools (for example using AIOMFAC or E-AIM). The authors could attempt to constrain or model droplet and gas phase concentrations of levoglucosan from known properties. I suggest it is better explained in the context of existing knowledge how the experiments contribute new insights.
Third, I find that the authors make quite firm conclusions, and in the current form of the manuscript it is hard to see that they are all fully justified. For example, as I understand, the conclusion on co-condensation is based entirely on one experiment (3d) and significant corrections were done to arrive at Dve due to technical issues with an RH censor. In the conclusion section, the authors provide numbers on atmospheric lifetimes which I do not see are explained. Some suggestions for improvement are given below.
Methods
Introduction of dry monodisperse particles into the chamber: If I understand correctly, dry particle diameter was selected by AAC and converted to electrical mobility diameter. This was checked by SMPS – how well did the two methods correspond -what were actual numbers in terms of e.g. geometric mean diameters obtained? Most places it says “~” when a diameter is given – what were the actual values measured?
The dry size distribution was measured using the MPSS system. Due to a problem in the drying system, a correction was used to get the dry size distribution. The description of the correction and the related uncertainties is substantial, but difficult to follow and could be explained more clearly. I miss figures showing the actual and corrected particle size distributions.
If the authors have the possibility to repeat some of the experiments with correct drying (in particular exp. 3d) and thus avoid the corrections it would significantly strengthen the discussion and conclusions that can be made.
Wet particle size during experiments were monitored using an SMPS system. Was the relative humidity of the sheath air controlled or could it be different than the relative humidity in the aerosol flow coming in ? If so – what uncertainties does that introduce on derived growth factors?
I suggest a schematic (could be in supporting material) showing all the instruments and where and when they were connected with flow rates given as well as indications for RH in different parts of the system and where it was measured.
I miss a clear timeline for each type of experiment. What was the conditions during exp 1a,b? at what times was the RH varied and how? Similar, please also give an explanation of the timeline in other experiments where RH was varied.
I suggest table 1 includes : 1) the size aerodynamic size selected by the AAC?, 2) the corresponding calculated mobility size, and the volume equivalent size. Please include a column stating whether RH was constant or varied during a given experiment (and if it was varied, give the ranges).
Section 2.6.
The assumptions behind equations (3) and (4) should be stated.
I miss the explanation for how to come from the change in particle mass with time due to evaporation or condensation for a particle in the transition regime to a first order loss equation. What are the assumptions, is the Kelvin effect ignored?
Regarding figure 2 – in the text the steps of increasing RH are stated as 13.45 and 16.15, but the timescale in the figure has been normalized and it is not clear what time zero is.
It says that the mass shows an exponential decay – this could be supported by showing an exponential fit.
Why were some data removed from the figures and what is “the final corrected data” ?
Page 6, line 186 – which shape factor was used? Is there a reference?
Is the HUMANS system new – or why is it described in so much detail?
Line 152: “typically” – which experiments were “typical” and which not?
Concentrations of levoglucosan in particle, droplet and gas phase:
I miss some even simple calculations to constrain the budgets of gas phase and particle/droplet phase levoglucosan. For example, knowing the number concentration of particles one could estimate the maximum gas phase concentration of levoglucosan in the chamber if all evaporated and at different stages on the way to this point.
In the atomizer bottle the mass concentrations of ammonium sulfate and levoglucosan are 1:1. Do the authors assume that the composition of the particles coming out of the atomizer are also 1:1? If so, what is the initial value of fo/s? (based on line 215 I get fo/s ~1.19 initially). If the value of 1.19 is correct for the initial composition, how is it possible that fo/s is higher than 1.19 at the start of the experiments if levoglucosan evaporated from the particles during the injection? (which would mean that fo/s should go down)?
Above figure 3 it says that the initial particle size varied between experiments due to the different times of injection and evaporation during this time. How come that before time zero (time of injection) particle diameter Dve goes up and fo/s goes up? If particles were evaporating, I would expect the diameter to decrease and also the mass of organics? (if levoglucosan is evaporating)?
Similarly, why in figure 4 is fa/s going up in the period of injection when RH is constant at around 85% ? This would suggest that levoglucosan condenses during the injection period?
In figure 3 panel 3h the fraction of organics in the particles is changing from ¨2 to ~0.4 (a factor of five) - - what change in volume (Dve) does that correspond to based on simple estimations using mass and density of AMS and LG? Does it correspond to the measured change in Dve and thus particle volume?
What is assumed about the wall loss rate of levoglucosan in the gas phase relative to wall loss rate of the particles? Is it used to explain some of the observed trends?
General
I find that the title should revised - co-condensation is not studied in the presence of a surfactant for example.
Please give the values used in various calculations and discussions (e.g. values for ERH and DRF points etc. with literature references.
The word co-condensation was originally for condensation of soluble gases alongside water onto growing cloud droplet at supersaturated conditions. The current study is for RH<100%. I suggest to explicitly state that the word is used here before activation into a cloud droplet.
Line 48: “Although this process is based on thermodynamics and has been simulated in models “ – please add references for this statement.
As already pointed out by reviewer 1, the gas-particle partitioning of organics to/from aqueous solution droplet containing salt has been previously investigated and it would be natural to use the results from that work as perspective (Yli-Juuti et al. EST 2013, DOI: 10.1021/es401233c).
Page 2 line 55. The cloud experiments referred to was conducted at low temperature (243 or 263 K) (Lopez et al. 2025) – how does it relate to the conditions in the current experiments? What is the importance of the diagonal basis-set for the current work?
Line 60: Levoglucosan is a solid at room temperature – the term miscible applies to mixtures of liquids? Should it perhaps say "soluble" instead?
Later it says: semi-volatile and water-soluble – please provide values and references for water solubility and vapor pressure.
Pag2 line 63 “high emission factors” – please be quantitative and provide references
Figure 4: 60 minutes is highlighted in the text – it is difficult to see in the figure what happens at 60 minutes – the second fit is started later. It says that RH stabilizes after 60 minutes but from panel a RH seems to decrease steadily after 60 min?
Regarding the SDS experiments: I miss figures with data (evaporation rates, relative humidities etc.) for the SDS experiments, and the table values do not include uncertainty estimates – I suggest these elements are included and discussed in a revised manuscript.
Regarding the photochemical degradation rates of the particle phase of approximately 17 hours (line 592) – where does this number come from? It would depend on the OH concentration in the droplets? How is the atmospheric lifetime calculated?
Citation: https://doi.org/10.5194/ar-2026-14-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 152 | 59 | 13 | 224 | 48 | 17 | 18 |
- HTML: 152
- PDF: 59
- XML: 13
- Total: 224
- Supplement: 48
- BibTeX: 17
- EndNote: 18
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This manuscript presents measurements of levoglucosan condensation and evaporation to and from aqueous droplets containing inorganic salts, collected with a laboratory chamber setup. While the results are potentially an interesting addition to the relevant literature, the authors should do a better job in relating this work to literature and state-of-the-art knowledge. They should also add additional quantitative analysis to raise to the level where this work can fit the scope Aerosol Research being "focused on studies with general implications for any of the subject areas rather than investigations that are primarily of local interest or of descriptive nature. " At present, I do not think that the authors use what is already known about aerosol thermodynamics and kinetics to say something novel enough about the present system. Furthermore, I have an important concern regarding the appropriateness of the experimental approach for measuring the phenomenon that the authors aim to target - i.e. the use of a dry instead of humidified DMA measurement and the associated evaporation of some of the key species (please see specific comments #8 below). Here are some suggestions on how to improve the manuscript:
General comments:
1. The results should be interpreted more quantitatively to provide useful and specific constraints for understanding the relevant atmospheric processes. The theoretical frameworks for both the relevant thermodynamics (e.g. E-AIM or similar thermodynamic models) as well as the kinetics (condensation and evaporation) exist, and the authors should at least attempt to use these frameworks to quantitatively answer the following questions: Can the behavior of levoglucosan be explained with the present knowledge on thermodynamics of the relevant systems? If not, what do the results imply about the activities of the relevant species in the condensed phase? For some inspiration (as well as potentially interesting older literature) on such quantitative analysis, please look at e.g. the work by Yli-Juuti et al. 2013.
2. The authors motivate the study by the impact that aqueous phase thermodynamics have on cloud droplet activation in supersaturation conditions. This is indeed also the conditions to which the term "co-condensation" usually refers to in the literature. The results presented in the manuscript have, however, been collected in water activities up to 100% RH, without observing the actual cloud droplet activation but rather staying in the metastable hydrated regime. This needs to be clarified throughout the manuscript to avoid misunderstandings, as the water content and condensation kinetics close to the activation point can be rather different from those at lower levels of water vapor saturation.
3. Along the lines of comment #1 above, I encourage the authors to investigate the sensitivity of the theoretical evaporation and condensation rates to e.g. a) the RH in the chamber; b) the concentration of the particles (which links to the gas-phase saturation, but also coagulation); c) assumptions about the molecular properties of levoglucosan and it's interactions with water and AS. Such sensitivity analyses would help you to understand what really governs the behavior of your system and constrain key parameters associated with it.
Specific comments:
4. Title: As mentioned above, I believe the term "co-condensation" is often used in the context of cloud droplet activation (i.e. water vapor supersaturations above the the critical supersaturation), while the results presented in this manuscript are obtained from lower water vapor activities. Please consider revising the title to better reflect the content of the manuscript.
5. Introduction, lines 55-56: The authors state "this process has never been experimentally investigated on a single semi-volatile compound under various controlled conditions." I do not think this statement is true. In minimum, the article by Yli-Juuti et al. and the references therein is evidence of that, and I strongly encourage the authors to do a thorough review of the literature and modify this paragraph accordingly. Secondly, the authors themselves do not directly observe the co-condensation of organic species upon cloud droplet activation (to which many of the references in this paragraph refer to). Please therefore revise this paragraph to better reflect the content of the study and the scientific literature.
6. Introduction, lines 59-60 (and later in the mansucript). Please quantify what you mean by "semi-volatile" and "soluble", i.e. give the relevant values of saturation vapor pressures, aqueous-phase activities and water solubilities available in the literature. You will later on need these to make quantitative estimates on the relevant condensation and evaporation rates.
7. Table 1 and the associated discussion: The number concentrations of particles in the chamber are rather high. What does this mean for the importance of a) wall-losses; b) coagulation; c) gas-phase saturation wrt levoglucosan in interpreting your results? Please add a discussion of this.
8. Methods, lines 237-240: Why did you dry the particles prior to the measurement - isn't there a risk of losing some of the semi-volatile material? Why did you not use a humidified DMA system? As is pointed out in the manuscript, the humidity control is a critical element for measuring what the authors intend to measure. Unless I have severely misunderstood something, I would strongly encourage the authors to consider repeating the experiments with a better RH control throughout the measurements, including the sizing. This is in my view a critical concern for the validity of the results.
9. Conclusions, lines 570-572: The authors state "The results showed that the net evaporation rate of levoglucosan was sensitive to all the investigated parameters, but the main drivers under our experimental conditions were the levoglucosan-to-sulfate mass ratios and RH." Unfortunately, I find this statement rather uninformative, given the potential of the data collected (especially if the RH control is taken care of). However, with the appropriate quantitative frameworks and potential further experimental analysis with a more careful humidity control, I believe the authors can do much better than this to find clearer and more impactful contributions to the present knowledge about atmospheric aerosols and their thermodynamics. I hope that the suggestions given above will help with this task.
10. Conclusions, line 596: I find that the outlook statement about the "modeling study" does not add much without some further elaboration. What kind of modeling do the authors mean, exactly? Please clarify.
References:
DOI: 10.1021/es401233c