the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Sep 2024
Analyzing the chemical composition, morphology and size of ice-nucleating particles by coupling a scanning electron microscope to an offline diffusion chamber
Abstract. To understand and predict the formation of clouds and rain and their influence on our climate, it is crucial to know the characteristics and abundance of ice-nucleating particles (INPs) in the atmosphere. As the ice-nucleating efficiency is a result of individual particle properties, a detailed knowledge on these properties is essential. Here, we present an offline method for the comprehensive analysis of ambient INPs that benefits from the combination of two instruments already used for ice nucleation measurements. First, the aerosol is sampled on silicon wafers. INPs are then activated at different temperature and humidity conditions in the deposition nucleation and condensation freezing mode using a static diffusion chamber. Activated INPs are located in a coordinate system, which allows for recovery of the individual particles causing the nucleation in a scanning electron microscope. Here, the size, chemistry and morphology of the particles are identified. Finally, the INPs are classified into categories based on their measured properties. As a result, a size resolved spectrum of the INP classes can be determined.
The performance of this coupling method is investigated in a case study on samples from the high-altitude field side Jungfraujoch (JFJ), Switzerland. 200 individual INPs from 14 samples obtained during a 5-week period were classified. Most deposition nucleation / condensation freezing mode INPs from Jungfraujoch, activated at −30 °C, were of irregular shape and had projected area diameters in the range from 300 nm to 35 µm, with a distinct maximum at 1–2 µm. A major contribution of mineral particles, mainly aluminosilicates / Al-rich particles, but also carbonates and silica, was identified for the entire INP size spectrum at −30°C. Further contributions were from carbon-rich particles, consisting of both smaller soot particles and larger biological particles. Mixed particles, here mostly particles rich in Al and C, were identified in higher abundances primarily between 3 µm and 9 µm. Minor contributions were seen from sulfates and metal oxides, with the latter ones found with increased proportion in the size range below 500 nm.
Such results are useful for evaluating INP type-specific parametrizations, e.g., for use in atmospheric modeling, and in closure studies.
- Preprint
(1322 KB) - Metadata XML
-
Supplement
(295 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-2797', Anonymous Referee #1, 01 Oct 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2797/egusphere-2024-2797-RC1-supplement.pdf
-
RC2: 'Review of “Analyzing the chemical composition, morphology and size of ice-nucleating particles by coupling a scanning electron microscope to an offline diffusion chamber” by Lisa Schneider et al.', Anonymous Referee #2, 24 Oct 2024
The manuscript reviews how ice nucleating particles can be located on a wafer and their composition and size analysed using an electron microscope. The use of the method is exemplified on a set of wafers collected at Jungfraujoch in 2017.
The coupling of the ice nucleation chamber FRIDGE with EM analysis is very suitable for the task of gaining information on the abundance of a specific category of particles active as INP at a certain temperature.However, the methodology and detail of coupling FRIDGE with EM has been discussed in previous papers and it is not made clear what novel information is provided in the current manuscript. Concerning the methodology, it seems not to go beyond what is already published in Schrod et al., 2016 and He et al., 2023.
The JFJ case study is a valuable dataset by itself, but the attempted validation of the FRIDGE-EM coupling by comparing to different techniques that investigated the INP composition at different activation temperatures on JFJ is not convincing. As the authors note themselves at best only a rough comparison can be made.
Listed in the comments below are several inaccuracies and inappropriate references.
The line of explanations should also be structured clearer. On several occasions, statements are made that are not comprehensible and are explained only by information given later in the text. Leading with the necessary information and explanations before a conclusion or result, would make it easier to follow.
It is irritating that the author state on couple occasions that a detailed analysis is not possible or feasible, but the analysis is then done anyway in the following.
Because the manuscript lacks novelty, rigor and structure, I recommend major revisions before consideration for publication.
Specific comments:
Title: The analysis of particle morphology is not discussed in the manuscript. Consider adjusting the title accordingly.
Abstract
Line 29: Specify how the results can be used to evaluate parametrizations.
Introduction
The introduction should be shortened and focussed on motivating the presented methodology and be less of a review of the subject area in general.
Instead of repeatedly citing chapter 1 and 8 from the 2017 AMS reviews of Kanji et al. and Cziczo et al., it would be more helpful to cite specific references to the individual topics.Line 64: clarify what is meant by "the efficiency of metal oxides to activate as INP depends on the type of metallic particle". Do you mean the type of metal cation?
Line 83: All the references provided here seem to be for measurements of INP concentrations only. Add references specific for the mentioned identification of particle type and size.
Line 85-89: The logic is not clear in these sentences. Clarify if it is the aerosol composition, particle class, main type, or specific type that should be related to INP to improve simulations.
Line 101-103: If for IR, scavenged particles cannot be distinguished from INP, there is no information on the INP.
Line 103-104: Clarify how INP can be identified using a CFDC by activation of sampled non-activated aerosol.
Line 106-107: specify what is meant by the “appearances of particles” and explain what observable differences can result from the evaporation process.
Line 111: specify what problems online methods can run into with low INP concentrations.
Line 129-130: Clarify how the property influencing ice nucleation can be isolated using SEM. It would be more precise to state that the chemistry, shape and size of INP can be obtained.
Fig.4: The perspective of the SEM image on top of the EDX spectra is confusing. Why is there a large shadow?
Line 151: Schrod et al., 2017 state that: “From the present SEM analysis we cannot draw conclusions on the chemical composition and nature of INPs, which make only a 10^-3 to 10^-5 fraction of the randomly selected particles on a wafer.” This was obviously a different FRIDGE-SEM-coupling, and the novelty of the method presented here should be highlighted. However, the description of chemical analysis of wafers provided in He et al., 2023 seems to already describe the current method.
Methodology
This section resembles an operation manual, and it doesn’t substantially go beyond Schrod et al., 2016 and He et al., 2023.
I’m missing an explanation on how the ice is evaporated between the FRIDGE experiment and the SEM, and an analysis if IR are moved during the process.Line 267: quantify the coordinate uncertainties and discuss where the uncertainties come from. Based on the pixel size and the criteria of 30 pixels to identify an ice crystal location, the INP could be up to 300um away from the centre if the crystal grows as needle. It could be explained in more detail why 50um is a good value. Is it because at the investigated conditions the ice growth regime is plate like?
Line 276: please define refractory particles in this context.
Line 277: it is unclear what is meant by “with respect to the analysed particles” here.
Line 279: Explain why adjusting the scanning radius optimizes the analysis. It could be quantified on a lightly loaded wafer what the distance of INP from the coordinate usually is to exclude particles outside the range.
Line 285: Explain how it can be known if a feature is relevant for ice formation.
Line 299-302: Describe the stage where ice is evaporated before the SEM analysis.
2.6 Chemical classification: Fig. S1 could be shown here and referred to, to guide the reader and help to follow the descriptions.
Line 307: Again, how can be identified if a certain surface property is relevant for ice nucleation?
Line 315: Explain, based on what information the classification scheme is modified. Does this make the scheme subjective?
Fig. 3: x-, y-axis scale are too small to read. Also, axis labels should be added.
Case Study
Clarify if sampling was conducted downstream of an inlet or in the open.
Line 391: Define cINP as INP concentration.
Fig.4: increase the contrast of the figure. Corresponding sample numbers are not shown. In the caption, do you mean adapted from Weber (2019) instead of modified according to Weber (2019)? Specify that the 5-day average is a running average.
Line 402: Provide a reference for Saharan dust being active below -20°C.
Line 402: From Fig. 4 it is not clear when the 14 samples were taken. 31 cINP on 11 days are marked with triangles.
Fig. 5: What’s the point of this figure? As shown in Fig. 2 and explained in Sec. 2.4. the edge region ice crystals were excluded for SEM. There should therefore be clearly more ice crystals detected by FRIDGE than SEM positions.
In the caption, mention that the 1:1 line is shown in red.Line 420: Clarify why the algorithm is inconsistent with excluding the temperature sensor area.
Line 437: Explain why a meteorological interpretation is not feasible for the current manuscript. Chapter 6 in the thesis of Weber 2019 contains a meteorological interpretation of the JFJ results and Fig.6.18 therein shows a comparison during and outside SDE.
Fig.6: Fig. S2 implies that the composition of the INP population active at -30°C can substantially vary from day to day. E.g., looking at W2, W7, W11 and W34 where a similar number of particles were analysed, the abundance of components is never similar. Please analyse and discuss the implications on sampling statistics and what the total chemical composition in Fig.6 represents.
In the caption, what artifact is excluded?Line 487-488: It is mentioned that a comparison of abundance is not possible. Clarify the purpose of doing a comparison if the abundance, which is the main result, cannot be compared.
Line 505: Clarify how the agreement can be considered good if only a rough comparison can be made. The comparison suffers from the previous mentioned differences in activation temperature and overall technique, making a direct comparison questionable.
Line 537: Specify the role of volatility.
Fig. 7: Clarify if analysed particles activated at all of the listed RH’s or at least one.
Line 549: This is a misunderstanding. DeMott et al. 2010 found that the concentration of INP correlates to the concentration of particles >500nm. Not that they are >500nm.
Line 569: Quantify the statistical uncertainty.
Line 573: Looking at Fig.7 in Lacher et al., 2021 the second maximum in the OPC data appears between 0.5.1um. Clarify how the IR OPC data is compared to the current results.
Line 574: Is a comparison just difficult or not possible?
Line 586: Explain how the size of ice crystals is linked to the size of INP/IR.
Line 589: Better references for the importance of immersion freezing are Ansmann et al. 2009 or Westbrook and Illingworth, 2011.
Line 591: It needs to be pointed out clearly what part of the presented method is novel.
Line 594: The analysis of morphology and surface properties has not been demonstrated in this work.
Line 606: Specify what improvements are necessary.
Line 625-627: It has not been demonstrated in this work that meaningful structural information can be obtained, and it is unclear how information on the relevance of a property for ice nucleation can be gained with this method.
Line 631-632: Clarify what element of the method need adaptation. It can be assumed that only the EAC would be flown, and the method of analysis remains the same. Also, is the EAC not already usable in an aircraft setting?
Technical corrections:
Line 48: I can’t find information about hydrogen-bridging functional groups in Kanji et al., 2008. Double check the reference.
Line 51: Replace Hoose & Möhler, 2012 with a more specific reference about the influence of particle size on ice nucleation.
Line 52: Marcolli 2017 would be a more specific reference for pre-activation than Abdelmonem et al., 2020.
Line 174: I can only find information on the EAC in the Supplement of DeMott et al., 2018 and it is not clear how the EAC was modified compared to the description in Klein et al., 2010.
Line 175: Lacher et al., 2024 report FRIDGE measurements using filter samples, the EAC seems not to have been used.
Line 270: “two effects” instead of “to effects”
Line 496: superfluous )
References:
Abdelmonem, A., Ratnayake, S., Toner, J. D., Lützenkirchen, J.: Cloud history can change water-ice-surface interactions of oxide mineral aerosols: a case study on silica, Atmos. Chem. Phys., 20, 1075-1087, doi:10.5194/acp-20-1075-2020, 2020
Ansmann, A.; Tesche, M.; Seifert, P.; Althausen, D.; Engelmann, R.; Fruntke, J.; Wandinger, U.; Mattis, I.; Müller, D. Evolution of the ice phase in tropical altocumulus: SAMUM lidar observations over Cape Verde, J. Geophys. Res., 2009, 114, D17208, doi:10.1029/2008JD011659
Cziczo, D. J., Ladino, L., Boose, Y., Kanji, Z. A., Kupiszewski, P., Lance, S., Mertes, S., Wex, H.: Measurements of Ice Nucleating Particles and Ice Residuals, Meteor. Mon., 58, 1-13, doi: 10.1175/AMSMONOGRAPHS-D-16-0008.1, 2017
DeMott, P. J., Prenni, A. J., Liu, X., Kreidenweis, S. M., Petters, M. D., Twohy, C. H., Richardson, M. S., Eidhammer, T., Rogers, D. C.: Predicting global atmospheric ice nuclei distributions and their impacts on climate, P. Natl. A. Sci, 107, no. 25, 11217-11222, doi: 10.1073/pnas.0910818107, 2010
DeMott, P. J., et al.: The Fifth International Workshop on Ice Nucleation phase 2 (FIN-02): laboratory intercomparison of ice nucleation measurements, Atmos. Meas. Tech., 11, 6231-6257, doi: 10.5194/amt-11-6231-2018, 2018
He, C., Yin, Y., Huang, Y., Kuang, X., Cui, Y., Chen, K., Jiang, H., Kiselev, A., Möhler, O., Schrod, J.: The Vertical Distribution of Ice-Nucleating Particles over the North China Plain: A Case of Cold Front Passage, Remote Sens., 15, 4989, doi: 10.3390/rs15204989, 2023
Hoose, C. and Möhler, O.: Heterogeneous ice nucleation on atmospheric aerosols: a review of results from laboratory experiments, Atmos. Chem. Phys., 12, 9817-9854, doi: 10.5194/acp-12-9817-2012, 2012
Kanji, Z. A., Florea, O., Abbatt, J. P. D.: Ice formation via deposition nucleation on mineral dust and organics: dependence of onset relative humidity on total particulate surface area, Environ. Res. Lett., 3, 025004, doi: 10.1088/1748-9326/3/2/025004, 2008
Kanji, Z. A., Ladino, L. A., Wex, H., Boose, Y., Burkert-Kohn, M., Cziczo, D. J., Krämer, M.: Overview of Ice Nucleating Particles, Meteor. Mon., Vol. 58, doi: 10.1175/AMSMONOGRAPHS-D-16-0006.1, 2017
Klein, H., Haunold, W., Bundke, U., Nillius, B., Wetter, T., Schallenberg, S., Bingemer, H.: A new method for sampling of atmospheric ice nuclei with subsequent analysis in a static diffusion chamber, Atmos. Res., 96, 218-224, doi: 10.1016/j.atmores.2009.08.002, 2010
Lacher, L., Clemen H.-C., Shen, X., Mertes, S., Gysel-Beer, M., Moallemi, A., Steinbacher, M., Henne, S., Saathoff, H., Möhler, O., Höhler, K., Schiebel, T., Weber, D., Schrod, J., Schneider, J., Kanji, Z. A.: Sources and nature of ice nucleating particles in the free troposphere at Jungfraujoch in winter 2017, Atmos. Chem. Phys., 21, 16925-16953, doi: 10.5194/acp-21-16925-2021, 2021
Lacher et al.: The Puy de Dôme ICe Nucleation Intercomparison Campaign (PICNIC): comparison between online and offline methods in ambient air, Atmos. Chem. Phys., 24, 2651-2678, doi: 10.5194/ACP-24-2651-2024, 2024
Marcolli, C.: Pre-activation of aerosol particles by ice preserved in pores, Atmos. Chem. Phys., 17, 1595–1622, https://doi.org/10.5194/acp-17-1595-2017, 2017.
Schrod, J., Danielczok, A., Weber, D., Ebert, M., Thomson E. S., Bingemer H. G.: Re-evaluating the Frankfurt isothermal static diffusion chamber for ice nucleation, Atmos. Meas. Tech., 9, 1313-1324, doi: 10.5194/amt-9-1313-2016, 2016
Schrod, J., Weber, D., Drücke, J., Keleshis, C., Pikridas, M., Ebert, M., Cvetkovic, B., Nickovic, S., Marinou, E., Baars, H., Ansmann, A., Vrekoussis, M., Mihalopoulos, N., Sciare, J., Curtius, J., Bingemer, H. G.: Ice nucleating particles over the Eastern Mediterranean measured by unmanned aircraft systems, Atmos. Chem. Phys., 17, 4817-4835, doi: 10.5194/acp-17-4817-2017, 2017
Weber, D.: Eisnukleation von Aerosolen: Laborexperimente und Messungen im Feld, Ph.D. thesis, Goethe Universität Frankfurt, Germany, 2019
Westbrook, C. D., and Illingworth, A. J.; Evidence that ice forms primarily in supercooled liquid clouds at temperatures > −27°C, Geophys. Res. Lett., 2011, 38, L14808, doi:10.1029/2011GL048021.
Citation: https://doi.org/10.5194/egusphere-2024-2797-RC2 -
RC3: 'Comment on egusphere-2024-2797', Anonymous Referee #3, 07 Nov 2024
In this manuscript, the authors present an offline method that combines an ice nucleation counter, the FRankfurt Ice nucleation Deposition freezinG Experiment (FRIDGE), with Scanning Electron Microscopy (SEM) to analyze the chemical composition, size, and morphology of Ice Nucleating Particles (INPs) collected from ambient air. The authors begin by providing an overview of the methodology, followed by a case study demonstrating its application to ambient aerosols collected during the 2017 CLACE/INUIT campaign at the Jungfraujoch station. The methodological section appears unfinished, as it fails to demonstrate all the potential features the authors claim the method can analyze (e.g., morphology, pores). Additionally, the lack of standards in this section raises concerns about evaluating the performance of the technique and estimating statistical error, particularly for size measurements and the coupling of particle size with ice nucleation efficiency. The case study lacks sufficient statistical analysis and fails to provide a clear connection to other parameters measured during the campaign (e.g., aerosol size distribution and number concentration). It appears the authors did not clearly decide whether to (A) describe and evaluate the method in detail, suitable for AMT journal or (B) focus on the CLACE/INUIT campaign with further analysis. As a result, the manuscript presents two incomplete studies that are not well connected.
Specific comments
- A clearer explanation is needed as to why deposition and condensation freezing modes are grouped together as the two primary ice nucleation modes. Specifically, the cycles with FRIDGE include measurements taken below water saturation at RH=100%, which would typically prevent condensation freezing. Is this grouping due to uncertainty in the RH measurements, which could allow for RH to exceed 100%?
- The authors argue that volatile compounds are not detected and that these compounds are generally not known to be efficient INPs. Is there any estimation of which type of volatile material is lost, a lower estimation of vapor pressure? How does this affect SEM measurements, notably for carbon? Could there be potential effect of freezing point depression, such as the competition for adsorption on active ice nucleating sites between water and other volatile compounds?
- The case study part focuses primarily on chemical composition and size analysis, but the title of the paper also mentions morphology. The conclusion, line 594, states, " This coupling allows for detailed analysis of various INP properties, such as chemical composition, mixing state, size, morphology, and surface properties like cracks or pores." Yet, no information is provided regarding the mixing state, morphology or surface properties.
- Figure 3: The x and y axis are not readable. Only few EDX spectra are provided in figure 3 for the case study analysis. How can we evaluate the reliability of the chemical composition on all particles?
- Figure 4: How is the 5-day average calculated and plotted? Why are there no error bars? What is the background level? How many measurements were performed per wafer, 1 only? Schrod, J. et al. (2016) provided statistical analysis for FRIDGE.
- Figure 5: What is the red curve, is this the ideal version with 1:1 ratio?
- Figure 6: Why is n=199, shouldn’t it be 200? Are the same particles active at all RH values (95%, 97%, 99%, 101%)? The percentages should be linked to the previous figure, as you mention the total chemical INP composition, but only 15% of all INPs are considered. Additionally, there are no error bars. Soot accounts for 1% of the total INP, does this fall within the uncertainty range?
- Figure 7 and discussion: No details comparison is made between initial size distribution of particles and size of INP, which do not enable correct estimation how size affect ice nucleation. In lines 556-566, the range referred here may be biased by simply higher initial concentrations. You need to better connect this part to size distribution measurements in lines 571-584, with statistical analysis.
Technical corrections
- Line 350 “natural mineral durst” correct to dust
-Line 401: “This was not the case for -20°C, because Saharan primarily activate as temperatures below - 20°C.” reference?
Citation: https://doi.org/10.5194/egusphere-2024-2797-RC3
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 160 | 48 | 51 | 259 | 14 | 2 | 2 |
- HTML: 160
- PDF: 48
- XML: 51
- Total: 259
- Supplement: 14
- BibTeX: 2
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1