Characterization of the Newly Designed Wall-Free Particle Evaporator (WALL-E) for Online Measurements of Atmospheric Particles

Gao, Linyu; Zgheib, Imad; Stergiou, Evangelos; Carstens, Cecilie; Sari Doré, Félix; Dupanloup, Michel; Bourgain, Frederic; Perrier, Sebastien; Riva, Matthieu

doi:https://doi.org/10.5194/egusphere-2025-1072

Preprints

https://doi.org/10.5194/egusphere-2025-1072

Preprints

24 Mar 2025

| 24 Mar 2025

Characterization of the Newly Designed Wall-Free Particle Evaporator (WALL-E) for Online Measurements of Atmospheric Particles

Linyu Gao, Imad Zgheib, Evangelos Stergiou, Cecilie Carstens, Félix Sari Doré, Michel Dupanloup, Frederic Bourgain, Sebastien Perrier, and Matthieu Riva

Abstract. Organic aerosols (OA) play a critical role in the atmosphere by directly altering human health and climate. Understanding their formation and evolution as well as their physicochemical properties requires a detailed characterization of their chemical composition. Despite advanced analytical techniques developed within the last decades, real-time online measurement of atmospheric particles remains challenging and suffers from different artifacts. In this work, we introduce the newly designed wall-free particle evaporator (WALL-E) coupled with a chemical ionization mass spectrometer (CIMS) using bromide (Br^–) as the reagent ion. We comprehensively evaluate the performance of the WALL-E system, demonstrating its ability to evaporate particles while maintaining the integrity of the compounds composing the particles (i.e., minimal thermal decomposition). To demonstrate WALL-E’s performance, the composition of aerosol particles formed from α-pinene ozonolysis in the presence of SO₂ is characterized. In addition, by applying the scan declustering method, we can now provide a quantification of the different species present in the condensed phase, e.g., C₁₀H₁₆O₄ 84 ng m^-3, C₁₉H₂₈O₇ 7 ng m^-3for a total SOA mass of 1 µg·m^-3. While dimers exhibit higher sensitivities, they account for only 14–18 % of the total particle masses, which is considerably lower than their signal fractions (23–29 %). This suggests a potential overestimation of the dimer contributions when relying solely on signal fractions. In addition, volatility analysis using thermograms reveals a clear relationship between T₅₀ and compound saturation vapor pressure (C^*), with lower-volatility species desorbing at higher temperatures. In addition, measured T₅₀ for α-pinene-derived SOA products agree well with theoretical volatility estimation models (e.g., SIMPOL). Overall, this study demonstrates that WALL-E system coupled to a CIMS is a promising technique for real-time particle characterization (i.e., composition, quantification, and volatility) of atmospheric aerosols.

Received: 06 Mar 2025 – Discussion started: 24 Mar 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 2258 KB)

Supplement (1753 KB)

Download & links

Linyu Gao, Imad Zgheib, Evangelos Stergiou, Cecilie Carstens, Félix Sari Doré, Michel Dupanloup, Frederic Bourgain, Sebastien Perrier, and Matthieu Riva

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-1072', Anonymous Referee #1, 14 Apr 2025
This article provides a detailed review of existing methods and introduces a new analytical method called WALL-E, which is designed to measure the chemical composition of atmospheric particles in real time. The work is significant and innovative, but the readability of the article needs improvement. Here are the specific details:
Introduction: The Introduction compares many methods, but it doesn’t clearly state that the main contribution is adding a device that integrates TD (thermal desorption) and other functions in front of the Br-CIME. This might be confusing for users of the instrument, especially those who are not developers, as they may struggle to quickly understand the research goal. In addition, the introduction of VIA-NO3-CIMS is too simplistic and needs to be strengthened.

Abstract: The phrase “suffers from different artifacts” in the Abstract is too vague. It should specify what kind of artifacts WALL-E addresses.

Undefined Term in Abstract: T50 is not defined in the Abstract.

Figures 1 and 2: Combining Figures 1 and 2 would make it easier to see both the appearance of the TD and the airflow simulation inside. It would also be helpful to label the “Dilution/Cooling Unit” and “Gas-phase Denuder.”

Lines 346-363: While it’s possible to guess what DC and dV50 mean, it would be clearer if their full names were provided.

Lines 377-381: The explanation here is not clear or intuitive enough in the following.

Lines 407-418, Lines 430-438: This part repeats information from the Introduction. It should be shortened and focus on the core issues. Some part is better for Introduction

Line 474: The statement “The corrected values align well with those reported for VIA and FIGAERO” is too general. It should include references and specific results.

Section 3.3: The main innovation is the use of T50, but the subsection titles are confusing. Tmax is discussed in Sections 3.3.1 and 3.3.3, while T50 is in Sections 3.3.2 and 3.3.4. This arrangement makes it hard for readers to follow.
Citation: https://doi.org/10.5194/egusphere-2025-1072-RC1
RC2:
'Comment on egusphere-2025-1072', Anonymous Referee #2, 17 Apr 2025
Gao et al. presented a newly designed Wall-Free Particle Evaporator (WALL-E) inlet for online chemical measurements of atmospheric particles using a chemical ionization mass spectrometer (CIMS). The authors claim WALL-E can efficiently evaporate organic particles with minimum thermal decomposition, which can be a good technique for real-time particle characterization. However, some sections are not very clear or convincing, which makes me unsure about the reliability of WALL-E. Please see my comments below for more details. Thus, I suggest major revision before publishing this paper.
General comments:
I have several questions about the thermal profile of WALL-E system:
The temperature profile in Figure 2 is based on the COMSOL simulation. Although COMSOL is a powerful tool to understand the system's thermodynamics, which is good for system design, the simulation results might not match the real results. Please comment on this by discussing the difference between the simulation results and real conditions. I suggest adding some temperature measurements to confirm the temperature profile. You can add a thermocouple at the center of the tubes. This is also important for your temperature correction.

Are there any reasons that you don’t put any insulation outside the WALL-E system to prevent the thermal exchange with the ambient?

In the Dilution/cooling unit, I do not understand how this can prevent the re-condensation of vaporized species. I expect less volatile species could condense on the wall. I think you need to do more tests with different ambient relevant species to investigate this.

Also, the thermal decomposition was only checked for one species. I suggest adding more standards to validate that, especially those relevant to ambient organics.

I also have a hard time following the SOA sections.
Why do you add SO₂ in the SOA generation? It seems very unusual in the literature. You will generate organic sulfate, which can lead to more complicated properties. This also lead to challenges to compare with any literature values (e.g., chemical composition, volatility, thermal properties, etc.)

I do not understand how you did sensitivity corrections. Please explain that a little bit more.

It is very surprised to me that your SOA mass derived from CIMS is much higher than that from SMPS. Those results do not convince me since I expect the mass to be underestimated by CIMS due to evaporation and transport efficiency, re-condensation on the wall, and loss of volatile species in the activated carbon denuder in WALL-E. I do not really understand why WALL-E leads to overestimation.

I do not find discussion about volatility based on their measurements.

Are there any size dependencies in your results?

What is the sensitivity and detection limitation of using WALL-E?

Specific comments:
1, could you label each part and the flow direction? I also do not understand why there are two tees for the sheath flow. Is WALL-E like a distillation tube where a sample tube is inserted inside a big tube? Then why are these tees in the same position?

Section 2.2.1. It is unclear to me how you mixed these solutions. What is the fraction of each chemical?

L330=331, “Without sensitivity … ug m-3.” The signal is only 0.04 ncps, which is the same as the signal that no VOC was injected (L326). Therefore, the SOA signal could just be the background noise.

Figure S6. I suggest using scattering plots with fittings.

Figure numbers in SI need to be correct.

dV50 is not well defined.

L387-389, “Using the dV50 … 8-compound fitting).” Do you expect that high amount of H2SO4? How much SO2 was added to the system?

Section 3.3.2. I think either Tmax or T50 works for discussing the volatility, but you are mixing them up and making it hard for me to follow. I suggest either picking one or separating them into two sections.

L454-455, “For comparison … of 98%.” Why do you use 98%, not 99.5% or 99.9%, as other studies you mentioned before?

L457-458, “In addition … fewer data points.” Do you mean fewer data points at higher temperatures? Overall, you have lots of data points, and it seems that only PEG-17 do not have enough data points after reaching the Tmax.

L460-463, “The signals … TD techniques.” How much improvement compared to other techniques? Did other techniques also use PEG?

Figure S14, please provide references for the FIGAERO data.

L485-486, “Longer residence … values.” I think this should be the opposite. Shorter residence time particles might not reach thermal equilibrium, so they need a higher temperature to evaporate completely.

Figure 7: I don't understand the purpose of this figure since you did not show any of your data.

Figure 8: Is the mean line showing the average of heat and cool? It is not clear to me how you used the data from the fast cool ramp.

L535-537, “By using … 1 ug m-3).” I am not fully getting this. Where did you show these results?
Citation: https://doi.org/10.5194/egusphere-2025-1072-RC2
RC3:
'Comment on egusphere-2025-1072', Anonymous Referee #3, 17 Apr 2025
Gao et al. presented a novel technique WALL-E, which is a thermal desorption unit coupled with CIMS. It can detect and quantify the chemical composition of aerosol particles with CIMS in real-time. This study is very interesting and innovative. However, the manuscript’s readability could be improved. I recommend a major revision prior to publication. Please see my detailed comments below:
General comments:
Stainless steel tubing was used for WALL-E in this study. Uncoated stainless steel tubing can adsorb semi-volatile and polar organics at elevated temperature, leading to sample loss and memory effects. Did you use any inert-coating on these tubes? if not, I recommend using inert-coated tubing (e.g. sulfinert coated stainless steel tubing) in the future, and add discussion of the caveat of using uncoated stainless steel tubing

SMPS was used in the experiments, and I wonder if the evaporation efficiency, T₅₀, and volatility characterization are particle size and mass loading dependent? Please include the number and size distribution for standards and SOA. Are SMPS and CIMS measure the same particles?

There are two figure S5 and two figure S6 in the SI. It’s very hard to follow which plot is being referred to. Please correct.

For the sensitivity calibration, the unit of ncps/(ug/m³) was used. When comparing the sensitivity among different compounds and calibration (second figure S6 and figure S8), the unit of ncps/ppm should be used to eliminate the influence of molecular weight.

In section 3.3.2, the usage of T_max and T₅₀ is confusing. The Tmax in FIGAERO usually refer to the temperature at which the signal intensity is maximum. For VIA, people usually use T₅₀ as well instead of Tmax, as VIA thermogram also shows a sigmoid curve instead of a near-Gaussian shape. The WALL-E and VIA both have continuous aerosol flow into the TD, therefore they have similar thermograms. The Tmax for FIGAERO and T₅₀ for VIA and WALL-E both represent the temperature when the desorption rate is maximum.

Are all the standards and SOA particles fully evaporated after passing through the WALL-E?

Specific comments:
Line 38: T₅₀ is not defined in the abstract.

Line 157: what is the residence time in the TD with sample flow rate at 1 SLPM?

Line 217: what are the densities used for the mass concentration calculated by SMPS for each compounds? Consider include into table S2.

Line 233: how is OH radical generated in the OFR? Please include the ozone concentration, relative humidity, and OH exposure in the table S3, as O₃/OH initiated oxidation was mentioned.

Line 294: any results indicating the better flow stability with the presence of a HF?

Line 317-318: I wonder if you observe any trimers? How about the sulfur containing compounds shown in figure 5B, as SO₂ was added into the system.

Line 326: what is ncps? Normalized count per seconds? How are they normalized?

Line 330-331: “without sensitivity correction, …, mass is 1.0 ± 0.1 μg/m³” 0.04 ncps is the same as the background as mentioned in line 326. I’m not sure if I fully understand this sentence here.

Line 356: Figure S8, why shikimic acid and glucose were excluded? Any criteria?

Line 388-390: how is sulfuric acid formed in the OFR? The particle phase H₂SO₄ mass loading seems to be pretty high. Any further discussion regarding the H₂SO₄? “a good agreement is retrieved” I’m not sure if I fully understand this. Were you saying the SMPS results match the CIMS results? But how can SPMS differentiate organic and inorganic? The mass loading reported with 10 compounds fitting (8.7 μg/m³) is higher than that with 8 compounds fitting (6.4 μg/m³), but the 10-compounds lead to underestimation, and 8 compounds lead to overestimation. Please correct.

Line 454-455: as mentioned in the general comment 4, I’m not sure if you should use T_max from WALL-E to compare with other particle evaporators, or use T₅₀, which is the temperature when the desorption rate is maximum.

Line 461-463: any explanation for the signal decrease after they reach the plateau? Also why this is an improvement compared to other online TD techniques?

Line 479-487: when you mention the T_max for VIA, do you mean the gas temperature or measured temperature? As far as I know, VIA does not measure the gas temperature directly.

Figure S1: chemical ionization section is not mentioned in the plot.

Figure S4: for the lower plot, what does the time in x-axis mean? What different conditions are they corresponding to?

Figure S12: please specify the legend. Also from the main text, I understand the reason to include the comparison between 99.5% and 98%. But why include 40% and 60%? T₅₀should be from your sigmoid fitting results as mentioned in the main text.

Figure S15: do you mean T_max or T₅₀? T₅₀ in the plot, but T_max in the caption and main text.
Citation: https://doi.org/10.5194/egusphere-2025-1072-RC3

Linyu Gao, Imad Zgheib, Evangelos Stergiou, Cecilie Carstens, Félix Sari Doré, Michel Dupanloup, Frederic Bourgain, Sebastien Perrier, and Matthieu Riva

Supplement

https://doi.org/10.5194/egusphere-2025-1072-supplement

Linyu Gao, Imad Zgheib, Evangelos Stergiou, Cecilie Carstens, Félix Sari Doré, Michel Dupanloup, Frederic Bourgain, Sebastien Perrier, and Matthieu Riva

Viewed

Total article views: 237 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
189	38	10	237	19	5	5

HTML: 189
PDF: 38
XML: 10
Total: 237
Supplement: 19
BibTeX: 5
EndNote: 5

Views and downloads (calculated since 24 Mar 2025)

Month	HTML	PDF	XML	Total
Mar 2025	102	23	4	129
Apr 2025	87	15	6	108

Cumulative views and downloads (calculated since 24 Mar 2025)

Month	HTML	PDF	XML	Total
Mar 2025	102	23	4	129
Apr 2025	87	15	6	108

Viewed (geographical distribution)

Total article views: 235 (including HTML, PDF, and XML) Thereof 235 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 29 Apr 2025

Short summary

In this work, we introduce WALL-E, a newly designed Wall-Free Particle Evaporator that enables real-time, online aerosol analysis while minimizing wall interactions and fragmentation, offering a more accurate and efficient approach to studying aerosol properties. WALL-E provides molecular-level insights into aerosol composition, quantification, and volatility distributions.


Total:	0
HTML:	0
PDF:	0
XML:	0