the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Mar 2025
Hydroxymethanesulfonate (HMS) Formation under Urban and Marine Atmosphere: role of aerosol ionic strength
Abstract. Hydroxymethanesulfonate (HMS) has emerged as a critical organosulfur species in ambient aerosols under polluted atmosphere, with limited observations conducted in pristine and marine environment. The impact of aerosol properties, particularly ionic strength (IS), on the formation of HMS remains uncertain, especially considering their significant variations that may occur under different atmospheric conditions. Here, HMS levels in a continental city (Nanjing, China) and, for the first time, in marine atmosphere (Yellow Sea and Bohai Sea) were quantified at 0.30±0.10 and 0.05±0.01 μg m–3, respectively. Elevated HMS concentration was recorded during a Nanjing haze event resulted from enhanced HMS formation rates, which can be largely attributed to reduced IS levels on hazy days as the IS-dependent enhancement on HMS formation increased with decreasing IS within the continental IS range (6–20 mol kg–1). This arises from the fact that elevated IS can boost HMS formation rate constants but also hinder the solubility of HMS precursor (SO2) and its further dissociation. Consequently, the IS-dependent enhancement initially rose with increasing IS, peaking at 4 mol kg–1 before declining. This also elucidated the negative correlation between HMS level and IS found in urban Nanjing (R=-0.62). Additionally, marine aerosols with lower IS levels (2.0–6.0 mol kg–1) likely exhibited more pronounced enhancements on marine HMS formation. Furthermore, the study highlights the significant impact of ambient humidity on aerosol IS (R=-0.89), suggesting the integration of ionic strength into chemical models to enhance the representation of particulate sulfur chemistry, particularly in humid environments.
- Preprint
(22458 KB) - Metadata XML
-
Supplement
(2184 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2025-683', Anonymous Referee #1, 23 Apr 2025
This paper investigates HMS formation as a function of particle pH and ionic strength (IS) based on two data sets: an urban site and a marine site. Modeling shows interesting interaction between IS and pH on HMS production rates. The concentrations of HMS are very low and a very small fraction of PM2.5 mass in both locations so it is unclear why the authors are attempting this analysis with this data – the contribution of HMS seems of little importance in these data. I guess the belief is the findings here have broad applicability. This could be made clearer. The analysis assume the HMS was formed under the conditions in which the measurements were made, otherwise the analysis falls apart. Other than low wind speeds, there is little proof offered to support this. Furthermore, how do the authors know that the formation didn’t occur in clouds not the aerosol, which are much more conducive to HMS formation. More evidence of formation in ALWC would be helpful. Aerosol pH is a critical parameter, but no evidence is provided to suggest it is reasonably predicted and assumptions are made due to lack of ammonia data. Extensive modeling analysis is performed, and the results are interesting, but little evidence is given that the data strongly support the model predictions; the main point seems to be that the model and data agree at the level of general trends in pH and IS. A more in-depth comparison might be helpful. A more direct comparison between model and measurements would be useful. Finally, given all the caveats stated in this manuscript relating to uncertainties in rate constants, phase separations, and possible surface vs bulk particle reactions, how does one interpret the modeling results as being reliable? This is why more direct comparisons would be useful. Overall, the modeling results are very interesting but there could be improvements to the manuscript. More details are given below.
Some main overall points.
- HMS was measured by ion chromatography. Regarding the detailed discussion in the supplemental material page 10 and Fig S10. Work by (Wei et al., 2020; Lai et al., 2023) suggest that HMS is actually detected as sulfite since it is converted to other or free S(IV) species in the IC column due to the eluent. To address this, a method of adding H2O2 to a filter sample extract to convert free bisulfite and sulfite to sulfate. This is found to have a minor effect since the untreated and treated samples give similar HMS concentrations. However, a similar treatment in Fairbanks shows that this sample treatment results in significant changes in HMS concentrations and that this is likely associated with other aldehyde-S(IV) adducts See Dingilian et al, ES&T Air 2024 https://doi.org/10.1021/acsestair.4c00012 There is evidence for these other aldehyde-S(IV) adducts in NCP hazes from single particle analysis, see https://doi.org/10.5194/acp-20-5887-2020 and https://doi.org/10.1016/j.envpol.2022.120846, although they may not be a large fraction of S(IV) measured. This may not affect the results of this study, but a brief discussion of this in the main text would be useful – the whole paper depends on the assumption that HMS is what is being measured by the IC.
- Why does one assume that the conditions in which the aerosol is being measured represents the conditions under which HMS was formed under. Low wind speed seems the only basis. This argument depends on the estimated lifetime of HMS, ie that it has a short lifespan. If the HMS (and possibly sulfate) was formed in clouds all of what is presented in this manuscript is not correct. This is especially true for the marine environment where cloud processing is frequent and expected. It is noted in the manuscript (pg 4 line 3 and on) that only 33% of HMS is thought to be formed in aerosol particles over the NCP, with 66% in cloud and fogs, so why all in aerosols in this study? This is critical and should be given more attention.
- In this study HMS is a very small fraction of sulfate both in the urban and marine locations. Why is it so small compared to other studies in the NCP. Seems like it is not really that important since it is a very small mass component of PM2.5. This is discussed on page 8, but one could add a justification that the analysis of IS effects can still be applied even if the concentrations of HMS are low. Is this why HMS is not shown on Fig 1d that shows the chemical speciation of PM2.5 (it is shown in Fig 1h, which is odd). Maybe the low concentrations reflect that it was not formed in cloud/fog water.
- From the chemical speciation it seems the most important PM2.5 component by far is ammonium nitrate. Maybe that should be pointed out. The interesting thing about the prevalence of nitrate is that it suggests a certain pH range (roughly 3 or higher) https://doi.org/10.5194/acp-18-12241-2018. Might want to note that this is roughly consistent with the calculated pH. Furthermore, it would be worth noting what chemical species drove ALWC, seem to be nitrate (Section 3.3).
- General comparisons between HMS (and sulfate, eg page 8) formation to Fairbanks and the urban study in this case are tenuous at best give the vastly different environmental conditions (Fairbanks winter T down to -35C very little sunlight/photochemistry in contrast to this study). These comparisons do not seem valid.
Specific Comments:
Pg 3 First line, what is meant by “crucial”. Add justification (if there is any) for this characterization? Eg, the term “up to” is not specific (ie, it is the max value). What is the mass fraction of HMS in PM2.5 when HMS is 18 g/m3? Compare this to the typical HMS mass fraction in PM2.5? Campbell should not be in the ref list, that is not data from the N. China Plane, check all references here are NCP.
Line 10 pg 3, I believe the models of Moch et al are for cloud processing in the formation of HMS. Is this true for all the references here. Please specify, the route, ie clouds/fogs vs aerosol particle water – it matters.
Line 20 pg 3. Regarding meteorological effects. A big influence of temperature is the effect on pH. See Campbell Sci Adv 2024, DOI: 10.1126/sciadv.ado4373 Tao and Murphy ACP 2019 https://doi.org/10.5194/acp-19-9309-2019
Line 36-37 pg 3 that states; In controversy,.... Campbell et al. 2022. I do not agree with this sentence. If you read that paper carefully you will see that pH was estimated since total (gas plus particle) ammonium was not measured. For more discussion on pH in Fairbanks and relation to HMS see Campbell Sci Adv 2024, DOI: 10.1126/sciadv.ado4373
I also do not understand the line following the one above, line 37-40 page 3. As noted above, why does one assume that the conditions in which the aerosol is being measured represents the conditions under which HMS was formed under. This depends on the estimated lifetime of HMS. This should be considered in the interpretation of these studies and especially the interpretation of the data in this paper.
Line 13 page 4. To have ALWC of 100’s of ug/m3 in polluted urban areas it must correspond to very poor air quality only found in some countries – ie this is not typical. Be specific what urban areas would have this level of ALWC, ie, in cities in the NCP?
First lines of page 6. It makes no sense using the epsilon of 0.5 for NH3/NH4+ partitioning in the pH analysis based on Fairbanks Alaska data which is an extremely different environment than that of this study. One might consider an iterative approach, see doi:10.2533/chimia.2024.1
R1 and R2 are shown as equilibrium reactions, but then the reaction rate constants are discussed. This is confusing.
Fig 1 caption, typo change penal to panel. T in (a) is in C, t in (e) is in Kelvin, why? Why is HMS not shown in Fig 1d?
Last line page 7, referring to similar results from Moch et al. But wasn’t Moch’s results based on HMS formation in clouds, which is not what is assumed here? Also, next line on top of page 8, comparing HMS concentrations to Fairbanks makes little sense given how different the situation is.
Line 40 page 8, too strong a statement, correlation is not causation. This applies throughout the manuscript.
Fig 2 b (and in other places) specify if ratio of HMS/Sulfate is mass or molar.
Page 10 top of page, what is the lifetime of HMS, this would support the idea that it is formed locally. Low wind speed does support the idea of local formation but does not prove the conditions measured were those under which HMS was formed.
Fig 3a. I assume this graph is predicted IS vs measured RH; this should be clarified. IS depends on ALWC and ALWC depends on RH and other factors (particle composition). Why not focus on ALWC and IS, that is what matters and what is more generalizable to other locations vs RH.
Fig 4b, the colors for data, urban clean vs polluted are too close to being the same. Typo in caption (radium?). From this plot, does the data fit with the model? Looking at the urban polluted vs clean, it does not seem to fit that well, there are many blue markers on similar isopleths as urban clean and urban polluted. The graph does not actually show data on measured HMS formation rates, or maybe all one can do is show HMS concentration. So how does this plot assess the comparison between these data from this study and the predictions. At the very least state associated HMS concentrations for the groups of data shown on the plot. The data on HMS concentrations and production rates of HMS, eg Fig 2, should be added. Furthermore, the plot is for ALWC = 100 ug/m3, but the data seem to cover a wide range of ALWC. How is that reasonable?
I do not understand the first part of the last paragraph, how does cycling of HMS through formation and loss in the liquid drop make sulfate? Explain more.
Page 17, what is Enteromorpha ?
Citation: https://doi.org/10.5194/egusphere-2025-683-RC1 -
AC3: 'Reply on RC1', Rongshuang Xu, 17 Jun 2025
We thank the reviewer for his/her insightful and constructive comments, which have greatly assisted us in enhancing this paper.
- First of all, we concur with the reviewer that cloud water could serve as a significant HMS formation medium and more declaration should be given to elucidate the contribution of in-cloud HMS formation to our observations. In the wintertime of Nanjing, characterized by weak vertical exchange due to temperature inversions, low wind speeds, and reduced cloud water content, it is plausible that the impact of in-cloud HMS formation on the observed levels of particulate HMS may be insignificant (please see our response to Major Comment #2 for further details). We acknowledge that this study observed lower levels of HMS compared to polluted areas such as Northern China, possibly due to the minor contribution from in-cloud formation. However, our observations in Nanjing could offer valuable insights into the aerosol chemistry related to HMS formation. Furthermore, despite the lower HMS levels, this study has provided supplementary information on ambient HMS abundance in less polluted and even marine pristine environments, enhancing our understanding of the prevalence of particulate sulfur species.
- Regarding the aerosol pH estimations, we acknowledge the need for clearer descriptions. We have monitored the gas-phase NH3 in Nanjing using a Monitor for AeRosols and GAses (MARGA; Metrohm Ltd., Switzerland). And we adopted a particle to particle + gas partitioning fraction of NH4+ (ε) of 0.5 to estimate the marine aerosol pH, drawing from prior observations conducted in the Bohai Sea (refer to our reply to Specific Comment #7).
- Moreover, we have expanded on the discussion comparing our model results with ambient measurements, as outlined in our responses to Specific Comment #15.
Please kindly refer to the Supplement for detailed point-to-point responses to the referee's comments.
-
RC2: 'Comment on egusphere-2025-683', Anonymous Referee #2, 25 Apr 2025
The manuscript presents a detailed study of HMS formation in urban and marine environments, examining the occurrence of a haze event in urban Nanjing. An ion chromatography method is used to analyze collected samples for HMS quantification and ISORROPIA Ⅱ and AIOMFAC were used to estimate ALWQ, pH and ionic strength (IS). An interesting effect of the ionic strength in HMS formation is discussed, providing evidence of HMS enhancement at low ionic strength during high humidity and pollution days. Although the study is well structured it lacks support of some key results, raising questions regarding the pH estimation, and model results vs measurements. The results are interesting, and overall, the work is suitable for publication in “Atmospheric Chemistry and Physics (ACP)” provided that the comments outlined below are adequately addressed.
- The authors use hydrogen peroxide to react the available S(IV) species, since the ion chromatography signal presents HMS as sulfite. Although hydrogen peroxide is not expected to decompose HMS at pH<6, have the authors tested this reaction for the examined conditions? Since filter aliquots are used the pH needs to monitored and test samples need to be examined for potential HMS decomposition (even from formed radicals) in the analyzed sample.
- The study focuses mainly on aerosol HMS presence, without providing evidence that the HMS formation occurred in the aerosol vs the cloud/fog water phase. Since 66% of HMS is formed in cloud and fog water, an analysis and/or discussion on the fate and formation of HMS in cloud and fog water is required and a comparison with this studies finding in the aerosol phase is necessary.
- The estimation of pH is achieved without ammonia measurements. How accurate is this estimation? More information is needed.
- The model results are interesting, providing new insights on HMS fate. However, with the exception of correlations, how do these results compare with the measurements? What are the most sensitive parameters of the model and what is the deviation between measurements and model results. How does this affect the conclusions?
- The HMS to sulfate ratio is mass or molar ratio?
- Page 6: Why is an epsilon of 5 for NH3/NH4+ partitioning used for the present study? It is stated that is based on Fairbanks Alaska, but this works investigated environment is significantly different and with higher pollution.
- Page 8 Lines 18-21: This statement is a bit confusing. Since HCHO is a precursor of HMS, wouldn’t it be expected to observe a decrease of HCHO and an increase of HMS rather than simultaneous peaks of the two species?
Citation: https://doi.org/10.5194/egusphere-2025-683-RC2 - AC1: 'Reply on RC2', Rongshuang Xu, 17 Jun 2025
-
RC3: 'Comment on egusphere-2025-683', Anonymous Referee #3, 05 May 2025
The authors present a study that uses filter measurements of hydroxymethanesulfonate (HMS), together with sulfate and formaldehyde (HCHO) data, to investigate the role of ionic strength in HMS formation in both Nanjing China and a marine cruise. They find that HMS concentrations are elevated during hazy days, which the authors attribute to lower aerosol ionic strength due to increased relative humidity (RH). While the topic is of broad interest to the atmospheric chemistry community, the manuscript currently lacks sufficient detail, clarity, and robust evidence to support its central conclusions. I have several major concerns below.
1.HMS Formation Pathways: Cloud Droplets vs. Aerosols
The authors seem to assume that HMS forms predominantly in aerosols in Nanjing. However, previous studies-particularly from China-have shown that HMS mainly forms in cloud or fog droplets (e.g. Moch et al., 2018), with little evidence of forming in aerosols. The only exception is wintertime Fairbanks in Alaska where there is no liquid cloud water (Campbell et al., 2024). This manuscript does not convincingly rule out the possibility of formation in cloud droplets. The argument of fog events shorter than two hours (Page 10, Line 9) is not sufficient, as HMS can form in cloud droplets above surface and be brought down to surface afterwards. Additionally, there is no discussion of HMS formation in marine aerosols, which further weakens the generalizability of the conclusions.
2. Section 3.3 lacks clarity. In Page 10 Line 41, the authors suggest a correlation between HMS and ALWC. This should not be surprising, as sulfate is highly hygroscopic. This correlation could simply reflect the high correlation between HMS and sulfate (Page 8, Line 26).
For Figure 3, it is expected that ionic strength inversely correlates with RH due to dilution effects at high humidity (high RH leads to high ALWC and therefore diluted solutions). Figure 3 does not make a good argument for ionic strength as a controlling factor in HMS formation. The right panel relies entirely on model output without supporting observational data, making the conclusion speculative.
There is simply no direct evidence that ionic strength itself is influencing HMS formation. The observed increase in HMS during hazy days could stem from several other factors: (1) Increased ALWC (Figure 2e), which enhances the dissolution of SO₂ and HCHO, potentially accelerating HMS production. (2) Elevated HCHO (Figure 2d), which increases by a factor of three during haze events and could drive more HMS formation. These alternative explanations are not adequately addressed in the manuscript.
3. Ionic Strength and pH Calculations. The calculation of ionic strength lacks details. Although ISORROPIA II can calculate ionic strength, the authors use the AIOMFAC model without discussing its added value. In particular, how important are organic species to ionic strength and pH?
4. The pH calculation (Page 5, Line 39) references the use of gas-phase species such as NH3 and HNO3, but these concentrations are not shown. Providing this information would help understand the pH and ionic strength estimates.
Minor Comments
5. Figure 1. The authors might consider adding panel titles to clearly distinguish between two different measurements (left vs. right panels).
References
Moch, J. M., Dovrou, E., Mickley, L. J., Keutsch, F. N., Cheng, Y., Jacob, D. J., Jiang, J., Li, M., Munger, J. W., Qiao, X., and Zhang, Q.: Contribution of Hydroxymethane Sulfonate to Ambient Particulate Matter: A Potential Explanation for High Particulate Sulfur During Severe Winter Haze in Beijing, Geophys. Res. Lett., 45, https://doi.org/10.1029/2018GL079309, 2018.
Campbell, J. R., Michael, Dingilian, K. K., Cesler-Maloney, M., Simpson, W. R., Robinson, E. S., DeCarlo, P. F., Temime-Roussel, B., D’Anna, B., Holen, A. L., Wu, J., Pratt, K. A., Dibb, J. E., Nenes, A., Weber, R. J., and Mao, J.: Enhanced aqueous formation and neutralization of fine atmospheric particles driven by extreme cold, Science Advances, 10, eado4373, https://doi.org/10.1126/sciadv.ado4373, 2024.
Citation: https://doi.org/10.5194/egusphere-2025-683-RC3 - AC2: 'Reply on RC3', Rongshuang Xu, 17 Jun 2025
Status: closed
-
RC1: 'Comment on egusphere-2025-683', Anonymous Referee #1, 23 Apr 2025
This paper investigates HMS formation as a function of particle pH and ionic strength (IS) based on two data sets: an urban site and a marine site. Modeling shows interesting interaction between IS and pH on HMS production rates. The concentrations of HMS are very low and a very small fraction of PM2.5 mass in both locations so it is unclear why the authors are attempting this analysis with this data – the contribution of HMS seems of little importance in these data. I guess the belief is the findings here have broad applicability. This could be made clearer. The analysis assume the HMS was formed under the conditions in which the measurements were made, otherwise the analysis falls apart. Other than low wind speeds, there is little proof offered to support this. Furthermore, how do the authors know that the formation didn’t occur in clouds not the aerosol, which are much more conducive to HMS formation. More evidence of formation in ALWC would be helpful. Aerosol pH is a critical parameter, but no evidence is provided to suggest it is reasonably predicted and assumptions are made due to lack of ammonia data. Extensive modeling analysis is performed, and the results are interesting, but little evidence is given that the data strongly support the model predictions; the main point seems to be that the model and data agree at the level of general trends in pH and IS. A more in-depth comparison might be helpful. A more direct comparison between model and measurements would be useful. Finally, given all the caveats stated in this manuscript relating to uncertainties in rate constants, phase separations, and possible surface vs bulk particle reactions, how does one interpret the modeling results as being reliable? This is why more direct comparisons would be useful. Overall, the modeling results are very interesting but there could be improvements to the manuscript. More details are given below.
Some main overall points.
- HMS was measured by ion chromatography. Regarding the detailed discussion in the supplemental material page 10 and Fig S10. Work by (Wei et al., 2020; Lai et al., 2023) suggest that HMS is actually detected as sulfite since it is converted to other or free S(IV) species in the IC column due to the eluent. To address this, a method of adding H2O2 to a filter sample extract to convert free bisulfite and sulfite to sulfate. This is found to have a minor effect since the untreated and treated samples give similar HMS concentrations. However, a similar treatment in Fairbanks shows that this sample treatment results in significant changes in HMS concentrations and that this is likely associated with other aldehyde-S(IV) adducts See Dingilian et al, ES&T Air 2024 https://doi.org/10.1021/acsestair.4c00012 There is evidence for these other aldehyde-S(IV) adducts in NCP hazes from single particle analysis, see https://doi.org/10.5194/acp-20-5887-2020 and https://doi.org/10.1016/j.envpol.2022.120846, although they may not be a large fraction of S(IV) measured. This may not affect the results of this study, but a brief discussion of this in the main text would be useful – the whole paper depends on the assumption that HMS is what is being measured by the IC.
- Why does one assume that the conditions in which the aerosol is being measured represents the conditions under which HMS was formed under. Low wind speed seems the only basis. This argument depends on the estimated lifetime of HMS, ie that it has a short lifespan. If the HMS (and possibly sulfate) was formed in clouds all of what is presented in this manuscript is not correct. This is especially true for the marine environment where cloud processing is frequent and expected. It is noted in the manuscript (pg 4 line 3 and on) that only 33% of HMS is thought to be formed in aerosol particles over the NCP, with 66% in cloud and fogs, so why all in aerosols in this study? This is critical and should be given more attention.
- In this study HMS is a very small fraction of sulfate both in the urban and marine locations. Why is it so small compared to other studies in the NCP. Seems like it is not really that important since it is a very small mass component of PM2.5. This is discussed on page 8, but one could add a justification that the analysis of IS effects can still be applied even if the concentrations of HMS are low. Is this why HMS is not shown on Fig 1d that shows the chemical speciation of PM2.5 (it is shown in Fig 1h, which is odd). Maybe the low concentrations reflect that it was not formed in cloud/fog water.
- From the chemical speciation it seems the most important PM2.5 component by far is ammonium nitrate. Maybe that should be pointed out. The interesting thing about the prevalence of nitrate is that it suggests a certain pH range (roughly 3 or higher) https://doi.org/10.5194/acp-18-12241-2018. Might want to note that this is roughly consistent with the calculated pH. Furthermore, it would be worth noting what chemical species drove ALWC, seem to be nitrate (Section 3.3).
- General comparisons between HMS (and sulfate, eg page 8) formation to Fairbanks and the urban study in this case are tenuous at best give the vastly different environmental conditions (Fairbanks winter T down to -35C very little sunlight/photochemistry in contrast to this study). These comparisons do not seem valid.
Specific Comments:
Pg 3 First line, what is meant by “crucial”. Add justification (if there is any) for this characterization? Eg, the term “up to” is not specific (ie, it is the max value). What is the mass fraction of HMS in PM2.5 when HMS is 18 g/m3? Compare this to the typical HMS mass fraction in PM2.5? Campbell should not be in the ref list, that is not data from the N. China Plane, check all references here are NCP.
Line 10 pg 3, I believe the models of Moch et al are for cloud processing in the formation of HMS. Is this true for all the references here. Please specify, the route, ie clouds/fogs vs aerosol particle water – it matters.
Line 20 pg 3. Regarding meteorological effects. A big influence of temperature is the effect on pH. See Campbell Sci Adv 2024, DOI: 10.1126/sciadv.ado4373 Tao and Murphy ACP 2019 https://doi.org/10.5194/acp-19-9309-2019
Line 36-37 pg 3 that states; In controversy,.... Campbell et al. 2022. I do not agree with this sentence. If you read that paper carefully you will see that pH was estimated since total (gas plus particle) ammonium was not measured. For more discussion on pH in Fairbanks and relation to HMS see Campbell Sci Adv 2024, DOI: 10.1126/sciadv.ado4373
I also do not understand the line following the one above, line 37-40 page 3. As noted above, why does one assume that the conditions in which the aerosol is being measured represents the conditions under which HMS was formed under. This depends on the estimated lifetime of HMS. This should be considered in the interpretation of these studies and especially the interpretation of the data in this paper.
Line 13 page 4. To have ALWC of 100’s of ug/m3 in polluted urban areas it must correspond to very poor air quality only found in some countries – ie this is not typical. Be specific what urban areas would have this level of ALWC, ie, in cities in the NCP?
First lines of page 6. It makes no sense using the epsilon of 0.5 for NH3/NH4+ partitioning in the pH analysis based on Fairbanks Alaska data which is an extremely different environment than that of this study. One might consider an iterative approach, see doi:10.2533/chimia.2024.1
R1 and R2 are shown as equilibrium reactions, but then the reaction rate constants are discussed. This is confusing.
Fig 1 caption, typo change penal to panel. T in (a) is in C, t in (e) is in Kelvin, why? Why is HMS not shown in Fig 1d?
Last line page 7, referring to similar results from Moch et al. But wasn’t Moch’s results based on HMS formation in clouds, which is not what is assumed here? Also, next line on top of page 8, comparing HMS concentrations to Fairbanks makes little sense given how different the situation is.
Line 40 page 8, too strong a statement, correlation is not causation. This applies throughout the manuscript.
Fig 2 b (and in other places) specify if ratio of HMS/Sulfate is mass or molar.
Page 10 top of page, what is the lifetime of HMS, this would support the idea that it is formed locally. Low wind speed does support the idea of local formation but does not prove the conditions measured were those under which HMS was formed.
Fig 3a. I assume this graph is predicted IS vs measured RH; this should be clarified. IS depends on ALWC and ALWC depends on RH and other factors (particle composition). Why not focus on ALWC and IS, that is what matters and what is more generalizable to other locations vs RH.
Fig 4b, the colors for data, urban clean vs polluted are too close to being the same. Typo in caption (radium?). From this plot, does the data fit with the model? Looking at the urban polluted vs clean, it does not seem to fit that well, there are many blue markers on similar isopleths as urban clean and urban polluted. The graph does not actually show data on measured HMS formation rates, or maybe all one can do is show HMS concentration. So how does this plot assess the comparison between these data from this study and the predictions. At the very least state associated HMS concentrations for the groups of data shown on the plot. The data on HMS concentrations and production rates of HMS, eg Fig 2, should be added. Furthermore, the plot is for ALWC = 100 ug/m3, but the data seem to cover a wide range of ALWC. How is that reasonable?
I do not understand the first part of the last paragraph, how does cycling of HMS through formation and loss in the liquid drop make sulfate? Explain more.
Page 17, what is Enteromorpha ?
Citation: https://doi.org/10.5194/egusphere-2025-683-RC1 -
AC3: 'Reply on RC1', Rongshuang Xu, 17 Jun 2025
We thank the reviewer for his/her insightful and constructive comments, which have greatly assisted us in enhancing this paper.
- First of all, we concur with the reviewer that cloud water could serve as a significant HMS formation medium and more declaration should be given to elucidate the contribution of in-cloud HMS formation to our observations. In the wintertime of Nanjing, characterized by weak vertical exchange due to temperature inversions, low wind speeds, and reduced cloud water content, it is plausible that the impact of in-cloud HMS formation on the observed levels of particulate HMS may be insignificant (please see our response to Major Comment #2 for further details). We acknowledge that this study observed lower levels of HMS compared to polluted areas such as Northern China, possibly due to the minor contribution from in-cloud formation. However, our observations in Nanjing could offer valuable insights into the aerosol chemistry related to HMS formation. Furthermore, despite the lower HMS levels, this study has provided supplementary information on ambient HMS abundance in less polluted and even marine pristine environments, enhancing our understanding of the prevalence of particulate sulfur species.
- Regarding the aerosol pH estimations, we acknowledge the need for clearer descriptions. We have monitored the gas-phase NH3 in Nanjing using a Monitor for AeRosols and GAses (MARGA; Metrohm Ltd., Switzerland). And we adopted a particle to particle + gas partitioning fraction of NH4+ (ε) of 0.5 to estimate the marine aerosol pH, drawing from prior observations conducted in the Bohai Sea (refer to our reply to Specific Comment #7).
- Moreover, we have expanded on the discussion comparing our model results with ambient measurements, as outlined in our responses to Specific Comment #15.
Please kindly refer to the Supplement for detailed point-to-point responses to the referee's comments.
-
RC2: 'Comment on egusphere-2025-683', Anonymous Referee #2, 25 Apr 2025
The manuscript presents a detailed study of HMS formation in urban and marine environments, examining the occurrence of a haze event in urban Nanjing. An ion chromatography method is used to analyze collected samples for HMS quantification and ISORROPIA Ⅱ and AIOMFAC were used to estimate ALWQ, pH and ionic strength (IS). An interesting effect of the ionic strength in HMS formation is discussed, providing evidence of HMS enhancement at low ionic strength during high humidity and pollution days. Although the study is well structured it lacks support of some key results, raising questions regarding the pH estimation, and model results vs measurements. The results are interesting, and overall, the work is suitable for publication in “Atmospheric Chemistry and Physics (ACP)” provided that the comments outlined below are adequately addressed.
- The authors use hydrogen peroxide to react the available S(IV) species, since the ion chromatography signal presents HMS as sulfite. Although hydrogen peroxide is not expected to decompose HMS at pH<6, have the authors tested this reaction for the examined conditions? Since filter aliquots are used the pH needs to monitored and test samples need to be examined for potential HMS decomposition (even from formed radicals) in the analyzed sample.
- The study focuses mainly on aerosol HMS presence, without providing evidence that the HMS formation occurred in the aerosol vs the cloud/fog water phase. Since 66% of HMS is formed in cloud and fog water, an analysis and/or discussion on the fate and formation of HMS in cloud and fog water is required and a comparison with this studies finding in the aerosol phase is necessary.
- The estimation of pH is achieved without ammonia measurements. How accurate is this estimation? More information is needed.
- The model results are interesting, providing new insights on HMS fate. However, with the exception of correlations, how do these results compare with the measurements? What are the most sensitive parameters of the model and what is the deviation between measurements and model results. How does this affect the conclusions?
- The HMS to sulfate ratio is mass or molar ratio?
- Page 6: Why is an epsilon of 5 for NH3/NH4+ partitioning used for the present study? It is stated that is based on Fairbanks Alaska, but this works investigated environment is significantly different and with higher pollution.
- Page 8 Lines 18-21: This statement is a bit confusing. Since HCHO is a precursor of HMS, wouldn’t it be expected to observe a decrease of HCHO and an increase of HMS rather than simultaneous peaks of the two species?
Citation: https://doi.org/10.5194/egusphere-2025-683-RC2 - AC1: 'Reply on RC2', Rongshuang Xu, 17 Jun 2025
-
RC3: 'Comment on egusphere-2025-683', Anonymous Referee #3, 05 May 2025
The authors present a study that uses filter measurements of hydroxymethanesulfonate (HMS), together with sulfate and formaldehyde (HCHO) data, to investigate the role of ionic strength in HMS formation in both Nanjing China and a marine cruise. They find that HMS concentrations are elevated during hazy days, which the authors attribute to lower aerosol ionic strength due to increased relative humidity (RH). While the topic is of broad interest to the atmospheric chemistry community, the manuscript currently lacks sufficient detail, clarity, and robust evidence to support its central conclusions. I have several major concerns below.
1.HMS Formation Pathways: Cloud Droplets vs. Aerosols
The authors seem to assume that HMS forms predominantly in aerosols in Nanjing. However, previous studies-particularly from China-have shown that HMS mainly forms in cloud or fog droplets (e.g. Moch et al., 2018), with little evidence of forming in aerosols. The only exception is wintertime Fairbanks in Alaska where there is no liquid cloud water (Campbell et al., 2024). This manuscript does not convincingly rule out the possibility of formation in cloud droplets. The argument of fog events shorter than two hours (Page 10, Line 9) is not sufficient, as HMS can form in cloud droplets above surface and be brought down to surface afterwards. Additionally, there is no discussion of HMS formation in marine aerosols, which further weakens the generalizability of the conclusions.
2. Section 3.3 lacks clarity. In Page 10 Line 41, the authors suggest a correlation between HMS and ALWC. This should not be surprising, as sulfate is highly hygroscopic. This correlation could simply reflect the high correlation between HMS and sulfate (Page 8, Line 26).
For Figure 3, it is expected that ionic strength inversely correlates with RH due to dilution effects at high humidity (high RH leads to high ALWC and therefore diluted solutions). Figure 3 does not make a good argument for ionic strength as a controlling factor in HMS formation. The right panel relies entirely on model output without supporting observational data, making the conclusion speculative.
There is simply no direct evidence that ionic strength itself is influencing HMS formation. The observed increase in HMS during hazy days could stem from several other factors: (1) Increased ALWC (Figure 2e), which enhances the dissolution of SO₂ and HCHO, potentially accelerating HMS production. (2) Elevated HCHO (Figure 2d), which increases by a factor of three during haze events and could drive more HMS formation. These alternative explanations are not adequately addressed in the manuscript.
3. Ionic Strength and pH Calculations. The calculation of ionic strength lacks details. Although ISORROPIA II can calculate ionic strength, the authors use the AIOMFAC model without discussing its added value. In particular, how important are organic species to ionic strength and pH?
4. The pH calculation (Page 5, Line 39) references the use of gas-phase species such as NH3 and HNO3, but these concentrations are not shown. Providing this information would help understand the pH and ionic strength estimates.
Minor Comments
5. Figure 1. The authors might consider adding panel titles to clearly distinguish between two different measurements (left vs. right panels).
References
Moch, J. M., Dovrou, E., Mickley, L. J., Keutsch, F. N., Cheng, Y., Jacob, D. J., Jiang, J., Li, M., Munger, J. W., Qiao, X., and Zhang, Q.: Contribution of Hydroxymethane Sulfonate to Ambient Particulate Matter: A Potential Explanation for High Particulate Sulfur During Severe Winter Haze in Beijing, Geophys. Res. Lett., 45, https://doi.org/10.1029/2018GL079309, 2018.
Campbell, J. R., Michael, Dingilian, K. K., Cesler-Maloney, M., Simpson, W. R., Robinson, E. S., DeCarlo, P. F., Temime-Roussel, B., D’Anna, B., Holen, A. L., Wu, J., Pratt, K. A., Dibb, J. E., Nenes, A., Weber, R. J., and Mao, J.: Enhanced aqueous formation and neutralization of fine atmospheric particles driven by extreme cold, Science Advances, 10, eado4373, https://doi.org/10.1126/sciadv.ado4373, 2024.
Citation: https://doi.org/10.5194/egusphere-2025-683-RC3 - AC2: 'Reply on RC3', Rongshuang Xu, 17 Jun 2025
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 373 | 69 | 22 | 464 | 48 | 16 | 34 |
- HTML: 373
- PDF: 69
- XML: 22
- Total: 464
- Supplement: 48
- BibTeX: 16
- EndNote: 34
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1