Design and evaluation of a catalytic stripper with a plate electrical aerosol classifier

Guo, Chengxiang; Yu, Tongzhu; Yang, Yixin; Gui, Huaqiao; Ji, Zhe; Wang, Jian; Liu, Jianguo; Liu, Wenqing

doi:10.5194/egusphere-2026-565

Preprints

https://doi.org/10.5194/egusphere-2026-565

Preprints

05 Mar 2026

| 05 Mar 2026

Design and evaluation of a catalytic stripper with a plate electrical aerosol classifier

Chengxiang Guo, Tongzhu Yu, Yixin Yang, Huaqiao Gui, Zhe Ji, Jian Wang, Jianguo Liu, and Wenqing Liu

Abstract. The catalytic stripper (CS) for removing volatile particles is a critical unit within the measurement system. However, the penetration efficiency of small size particles is currently significantly lower than large particles in CS. Therefore, further improving the penetration efficiency of small size particles is of significant research interest. This study aims to enhance the penetration efficiency of small size particles by reducing the thermophoretic loss. For this purpose, a CS equipped with a plate-type electrical aerosol classifier (EAC) was designed and developed, and its performance was evaluated. The particles are prevented from depositing on the tube wall by applying an electric field force to them in the opposite direction to the thermophoretic force they are subjected to, which ultimately serves the purpose of further improving the particle penetration efficiency. The experimental results demonstrated that the CS achieved a removal efficiency (RE) higher than 99.9 % at a flow rate of 1.5 L/min or lower. At a sample flow of 0.3 L/min and a temperature of 350 °C, the penetration efficiency of CS+EAC without voltage was evaluated. Combined with the CS+EAC voltage-penetration efficiency curve, applying -112 V on the EAC, the penetration efficiency was further improved under the same experimental conditions, and the smaller the particle size, the greater the improvement. Compared to the 0 V, the improvement rate for 15 nm at -112 V was 24.4 %, while that for 23 nm was 18.9 %. Further experimental results show that the EAC can remove particles smaller than 10 or 23 nm by further increasing the voltage. This capability enables rapid particle classification and facilitates high temporal resolution measurements of particle number concentrations across different size intervals.

Received: 30 Jan 2026 – Discussion started: 05 Mar 2026

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 paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 1635 KB)

Supplement (808 KB)

Download & links

Chengxiang Guo, Tongzhu Yu, Yixin Yang, Huaqiao Gui, Zhe Ji, Jian Wang, Jianguo Liu, and Wenqing Liu

Status: final response (author comments only)

RC1:
'Comment on egusphere-2026-565', Anonymous Referee #1, 29 Mar 2026
This study utilizes a combined CS+EAC design, aiming to further improve the penetration efficiency of particles by applying an electric field force counteracting the thermophoretic force they experience after CS treatment. The manuscript articulates the feasibility of this method through structural design and theoretical analysis, and demonstrates its effectiveness through experimental design and data analysis. This is a valuable and innovative contribution to aerosol measurement techniques.
However, there are the following issues that need to be clarified:
In section 2.1, the electric field width is the same as the aerosol slit width, which will cause the particles entering from the edge of the aerosol slit to be subjected to a distorted electric field force. This will affect the trajectory of these particles, and consequently, there is a possibility that the improvement in particle penetration efficiency will be limited.

In section 2.3.3, the control information for the linear scanning of the EAC electric field is inadequately described. Table S2 does not list the specific particle size range and its corresponding voltage range and step voltage.

In section 2.3.4, when verifying the ability of CS+EAC to roughly estimate the peak particle size, the manuscript does not explain the impact of multiple charging on this method, nor does it state how to eliminate or ignore this impact.

In section 2.4.5, there is a descriptive inconsistency: the original text states that a square wave voltage of 0V and -200V was used for the experiment, but the red part of the experimental setup diagram S3 describes it as 0/-160V.

The parameter 𝛻 T in Equation (1) and the variable V𝑡ℎ in Equation (2) are undefined.

In section 3.2.2, the peak of the 15nm curve for DR3 in Figure a is significantly higher than other results, and the same applies to the 15nm curve peak of DR3 in Figure b. It is recommended to add an explanation for this phenomenon.

In section 3.4.1, when soot particles with a peak size of 35nm pass through the CS+EAC at 0V, their peak size changes. This phenomenon will inevitably affect the subsequent use of CS+EAC to retrieve the particle size distribution; therefore, it is suggested to supplement methods to avoid this influence.

In section 3.4.1, the description of the relationship between the peak particle size and the step voltage is unclear. It is recommended to add specific mathematical formulas and describe them in combination with the information in the figures.

In sections 3.4.1, the significance of the annotations for -100V and 0V in Figure 9b is unclear, making it impossible to understand their specific meanings at a glance. It is recommended to change the expression format.

In section 3.4.2, The meanings of 0V and -200V are unclear.
Citation: https://doi.org/10.5194/egusphere-2026-565-RC1
- AC1: 'Reply on RC1', Tongzhu Yu, 29 Apr 2026
  
  Tongzhu Yu
  
  Key Laboratory of Environmental Optics and Technology, Anhui Institute of Optics and Fine Mechanics, Hefei Institutes of Physical Science, Chinese Academy of Sciences,
  
  Hefei 230031, China
  
  April, 2026
  
  Dear Editor,
  Thank you for the relevant revision comments. We have now checked and corrected our manuscript, and the detailed reply to each comment is as follows:
  
  Reviewer #1:
  
  Overall Comments:
  
  Comment 1: In section 2.1, the electric field width is the same as the aerosol slit width, which will cause the particles entering from the edge of the aerosol slit to be subjected to a distorted electric field force. This will affect the trajectory of these particles, and consequently, there is a possibility that the improvement in particle penetration efficiency will be limited. Since the high-voltage electrode of the EAC is a circuit board, the edges of the board require insulation, which results in a narrower electric field width.
  
  Reply: Thank you for your comment. As shown in Figure 1, the width of the aerosol slit is 20 mm, while the width of the classification zone is 30 mm. Based on this, the sheath flow will have a certain inhibitory effect on the diffusion of particles in the sample flow. We will incorporate your suggestions to improve the design and use them in future optimizations, replacing the circuit board with stainless steel for the high-voltage electrodes.
  Comment 2: In section 2.3.3, the control information for the linear scanning of the EAC electric field is inadequately described. Table S2 does not list the specific particle size range and its corresponding voltage range and step voltage.
  
  Reply: Thank you for your suggestion. We have added the voltage adjustment range and step voltage parameters for each test particle size to Table S3. In addition, this table is also provided in the attachment to the response.
  Comment 3: In section 2.3.4, when verifying the ability of CS+EAC to roughly estimate the peak particle size, the manuscript does not explain the impact of multiple charging on this method, nor does it state how to eliminate or ignore this impact.
  
  Reply: Thank you for your suggestion. The current tests merely demonstrate that CS+EAC can provide a rough estimate of the peak particle size. Since the experimental data were provided by the SMPS, they already include multi-charge corrections for soft X-rays. The final design of the CS+EAC includes a unipolar charging device for charging particles and an electrometer for measuring particle concentration. In our subsequent work, we will use the designed unipolar charge device as a basis to systematically develop and optimize a particle size distribution scanning algorithm, including multi-charge corrections.
  Comment 4: In section 2.4.5, there is a descriptive inconsistency: the original text states that a square wave voltage of 0V and -200V was used for the experiment, but the red part of the experimental setup diagram S3 describes it as 0/-160V.
  
  Reply: Thank you for pointing that out; we have corrected the error. The actual voltage is -200V. The revised Figure S3 has been provided in the attachment to this response.
  Comment 5: The parameter ∇T in Equation (1) and the variable V_th in Equation (2) are undefined.
  
  Reply: Thank you for your valuable suggestion. An explanation of ∇T has been added in line 147 to clarify its meaning in the revised manuscript. Since V_th is already explained on line 149, we will not make any changes there.
  Comment 6: In section 3.2.2, the peak of the 15nm curve for DR3 in Figure a is significantly higher than other results, and the same applies to the 15nm curve peak of DR3 in Figure b. It is recommended to add an explanation for this phenomenon.
  
  Reply: Thank you for your comment. First, it should be noted that the two figures essentially represent the same dataset, with only the x-axis differing. The higher peak normalized penetration efficiency compared to other particle sizes is due to data fluctuations at low voltages for the 15 nm particles. To minimize the influence of data fluctuations at low voltages, the normalized penetration efficiency was averaged for the data points with x-axis values ranging from 0 to 0.32593, thereby reducing the error in this dataset. The revised Figure 6 has been provided separately in the Supplementary Materials.
  Comment 7: In section 3.4.1, when soot particles with a peak size of 35nm pass through the CS+EAC at 0V, their peak size changes. This phenomenon will inevitably affect the subsequent use of CS+EAC to retrieve the particle size distribution; therefore, it is suggested to supplement methods to avoid this influence.
  
  Reply: Thank you for your suggestion. We have adjusted the number concentrations measured at different voltages based on the penetration efficiency of CS+EAC at 0 V. In addition, the V50 values were corrected based on the rate of decline in total concentration or 0.5 times the total concentration. The error in the corrected peak particle size was further reduced. The detailed revisions are provided in Section 3.4.1 and Appendix D of the Supplementary Materials.
  Comment 8: In section 3.4.1, the description of the relationship between the peak particle size and the step voltage is unclear. It is recommended to add specific mathematical formulas and describe them in combination with the information in the figures.
  Reply: Thank you for your suggestion. In this study, particles smaller than the voltage-based cut-off size are removed. By gradually increasing the voltage from 0 V, the particle size distribution can be progressively reduced to zero. The particle size distribution will be vertically cut off when the transfer function of the EAC is sufficiently ideal. However, since the transfer function is not ideal and has a certain slope, some error is introduced during the cutting process. With the appropriate step voltage setting, this error is negligible. When the voltage corresponding to a 50% reduction in total concentration is substituted into the curve fitted in Figure 7, the resulting particle size is essentially consistent with the actual peak particle size. On the one hand, from the perspective of the rate of concentration decrease, when the D50 corresponding to the voltage is the peak particle size, the rate of total concentration decrease measured by the instrument should be at its maximum. Therefore, this study uses either half of the total concentration or the peak rate of decrease in total concentration to roughly estimate the peak particle size of the particle size distribution. By further adjusting the results based on the penetration efficiency of the different voltage-cut sections, the measurement error in the peak particle size can be further reduced.
  Comment 9: In sections 3.4.1, the significance of the annotations for -100V and 0V in Figure 9b is unclear, making it impossible to understand their specific meanings at a glance. It is recommended to change the expression format.
  Reply: Thank you for your comment. Because the application of an electric field improves particle penetration efficiency, the particle size distribution shows that the peak of the distribution rises when the voltage is between -100 V and -300 V. Therefore, the labels “0 V” and “−100 V” have been added to more clearly indicate the particle size distributions corresponding to different voltages. In addition, we have included a detailed description of Figure 9.B in the supplementary materials. This figure has been included in the attachment to the response to the reviewers.
  Comment 10: In section 3.4.2, The meanings of 0V and -200V are unclear.
  Reply: Thank you for pointing this out. An explanation of the basis for voltage selection has been added in lines 414–416 of the revised manuscript.
  I am looking forward to hearing from your office soon!
  Sincerely,
  
  Tongzhu Yu
  
  Key Laboratory of Environmental Optics and Technology, Anhui Institute of Optics and Fine Mechanics, Hefei Institutes of Physical Science, Chinese Academy of Sciences,
  
  Hefei 230031, China
  Email: tzyu@aiofm.ac.cn
  
  Citation: https://doi.org/10.5194/egusphere-2026-565-AC1
RC2:
'Comment on egusphere-2026-565', Anonymous Referee #2, 06 Apr 2026
This manuscript presents a study on how to improve the penetration efficiency of small solid particles through a catalytic stripper by integrating a plate-type electrical aerosol classifier to counteract thermophoretic losses. The experimental setup is detailed and the data are comprehensive; however, I have serious concerns regarding the design of the instrument. Comments below need to be addressed before a second round of review:
Major comments:
If the primary goal is to reduce thermophoretic losses, why were simpler design approaches not considered? For instance, adding a concentric tube at the CS outlet to supply a sheath flow around the sample flow could keep particles away from the wall and reduce the thermophoretic loss. Have the authors explored or compared such methods? If not, a justification for favoring the active electric-field approach is needed.

It is unclear how the instrument is used when measuring real motor emitted particles. Is an X-ray neutralizer installed upstream the instrument so that the EAC can function? If so, it must be noted that the charging efficiency of small particles (10 nm, 23 nm) are very low. In this case, the EAC does not exert electric forces to the neutral particles and only contribute to their losses.

For particles of one polarity, the applied electric field force opposes the thermophoretic force, thereby reducing wall deposition. However, for particles of the opposite polarity, the electric field force aligns with the thermophoretic force, potentially increasing particle losses. This polarity-dependent effect should be explicitly acknowledged and discussed in the manuscript

Based on the results shown in Figure 9, it is not evident that the CS+EAC+particle counter combination can reliably determine the peak particle size of a polydisperse aerosol. In particular, the scatter plot in Figure 9a shows a relatively constant slope region in the middle, making it difficult to identify the voltage at which the particle number concentration declines most rapidly. To improve clarity, I suggest color-coding the scatter plot markers by applied voltage and also plot the derivatives of particle number concnentration.

Minor comments:
Line 383: ‘is gradually removed’ -> ‘gradually moves’

Page 7: the numbers in Psi 1,2,3 should be in subscript

Figure 7b: ‘volatility range’?

Figure 8: the numbers can be put on the color bars for the 15 nm data.
Citation: https://doi.org/10.5194/egusphere-2026-565-RC2
- AC2: 'Reply on RC2', Tongzhu Yu, 29 Apr 2026
  
  Tongzhu Yu
  
  Key Laboratory of Environmental Optics and Technology, Anhui Institute of Optics and Fine Mechanics, Hefei Institutes of Physical Science, Chinese Academy of Sciences,
  
  Hefei 230031, China
  
  April, 2026
  
  Dear Editor,
  Thank you for the relevant revision comments. We have now checked and corrected our manuscript, and the detailed reply to each comment is as follows:
  Major comments:
  Comment 1: If the primary goal is to reduce thermophoretic losses, why were simpler design approaches not considered? For instance, adding a concentric tube at the CS outlet to supply a sheath flow around the sample flow could keep particles away from the wall and reduce the thermophoretic loss. Have the authors explored or compared such methods? If not, a justification for favoring the active electric-field approach is needed.
  Reply: Thank you for your comment. Some commercial diluters do indeed use the method you described to minimize particle loss. However, there are additional considerations behind this study’s decision to use an electric field to enhance penetration efficiency. Ultimately, the CS+EAC in this study will be coupled with an electrometer and a unipolar charger. In addition to improving the penetration efficiency for small-sized particles, this study aims to alter the lower limit of particle size detection (D₅₀) in the measurement system by applying an electric field, thereby enabling a rough estimation of the peak particle size. This is the rationale for selecting this technical approach in this study.
  Comment 2: It is unclear how the instrument is used when measuring real motor emitted particles. Is an X-ray neutralizer installed upstream the instrument so that the EAC can function? If so, it must be noted that the charging efficiency of small particles (10 nm, 23 nm) are very low. In this case, the EAC does not exert electric forces to the neutral particles and only contribute to their losses.
  Reply: Thank you for your comment. Ultimately, the CS+EAC system will be used in conjunction with a unipolar charger and an electrometer, as the unipolar charger is more effective at charging small size particles. However, CS+EAC has not yet been tested in combination with unipolar charger, as this would require a systematic evaluation of the chargers. According to theoretical predictions, CS+EAC can still improve the penetration efficiency of small-particle-sized particles. Although the improvement rate will be lower than the results presented in this paper due to charging efficiency, this still represents a significant enhancement in penetration efficiency for DC equipment. This is because DC equipment must be equipped with unipolar charging devices.
  Comment 3: For particles of one polarity, the applied electric field force opposes the thermophoretic force, thereby reducing wall deposition. However, for particles of the opposite polarity, the electric field force aligns with the thermophoretic force, potentially increasing particle losses. This polarity-dependent effect should be explicitly acknowledged and discussed in the manuscript.
  Reply: Thank you for your suggestion. As noted in the conclusion at the end of the article, this study still requires integration with a unipolar charger. Since the design of the corresponding unipolar charger has not yet been completed, we plan to integrate it and conduct performance testing in future work. Ultimately, efficient particle charging is achieved by the unipolar charger, volatile particles are removed and size-classified by the CS+EAC, and the total number concentration is measured by the electrometer.
  Comment 4: Based on the results shown in Figure 9, it is not evident that the CS+EAC+particle counter combination can reliably determine the peak particle size of a polydisperse aerosol. In particular, the scatter plot in Figure 9a shows a relatively constant slope region in the middle, making it difficult to identify the voltage at which the particle number concentration declines most rapidly. To improve clarity, I suggest color-coding the scatter plot markers by applied voltage and also plot the derivatives of particle number concentration.
  Reply: Thank you for your suggestion. We have revised the data graph and included the derivatives. At the same time, based on the percentage of the concentration values at different voltages relative to the maximum concentration, we fitted the voltage values corresponding to a 50% reduction in concentration and substituted these into the fitted voltages in Figure 7 to obtain the peak particle size. The comparison results indicate that the peak particle size obtained using the voltage value corresponding to a 50% reduction in total concentration is more accurate. To increase the scanning speed, the difference between step voltages is relatively large, which results in significant errors in the derivatives calculated from adjacent voltage values, as it is difficult to determine the voltage values through further fitting. Therefore, this study uses either half of the total concentration to roughly estimate the peak particle size of the particle size distribution. By further adjusting the results based on the penetration efficiency of the different voltage-cut sections, the measurement error in the peak particle size can be further reduced. The detailed revisions are provided in Section 3.4.1 and Appendix D of the Supplementary Materials.
  Minor comments:
  Comment 5: Line 383: ‘is gradually removed’ -> ‘gradually moves’
  Reply: Thank you for your suggestion. I’ve made the changes based on your suggestion. Furthermore, we have added a note stating that the concentration gradually decreases. The corresponding changes have been made in lines 384–385 of the revised manuscript.
  Comment 6: Page 7: the numbers in Psi 1,2,3 should be in subscript
  Reply: Thank you very much for pointing that out; we have made the necessary corrections. The corresponding changes have been made in lines 180–190 of the revised manuscript.
  Comment7: Figure 7b: ‘volatility range’?
  Reply: Thank you for your suggestion. The term “volatility range” has been changed to “fluctuation range” to avoid ambiguity.
  Comment 8: Figure 8: the numbers can be put on the color bars for the 15 nm data.
  Reply: Thank you for your suggestion. The changes have been made accordingly, and the obstructed areas in all data figures have also been optimized to improve readability.
  I am looking forward to hearing from your office soon!
  Sincerely,
  Tongzhu Yu
  Key Laboratory of Environmental Optics and Technology, Anhui Institute of Optics and Fine Mechanics, Hefei Institutes of Physical Science, Chinese Academy of Sciences,
  Hefei 230031, China
  Email: tzyu@aiofm.ac.cn
  
  Citation: https://doi.org/10.5194/egusphere-2026-565-AC2

Chengxiang Guo, Tongzhu Yu, Yixin Yang, Huaqiao Gui, Zhe Ji, Jian Wang, Jianguo Liu, and Wenqing Liu

Supplement

https://doi.org/10.5194/egusphere-2026-565-supplement

Chengxiang Guo, Tongzhu Yu, Yixin Yang, Huaqiao Gui, Zhe Ji, Jian Wang, Jianguo Liu, and Wenqing Liu

Viewed

Total article views: 925 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
611	230	84	925	157	81	77

HTML: 611
PDF: 230
XML: 84
Total: 925
Supplement: 157
BibTeX: 81
EndNote: 77

Views and downloads (calculated since 05 Mar 2026)

Month	HTML	PDF	XML	Total
Mar 2026	541	186	80	807
Apr 2026	69	42	4	115
May 2026	1	2	0	3

Cumulative views and downloads (calculated since 05 Mar 2026)

Month	HTML	PDF	XML	Total
Mar 2026	541	186	80	807
Apr 2026	69	42	4	115
May 2026	1	2	0	3

Viewed (geographical distribution)

Total article views: 926 (including HTML, PDF, and XML) Thereof 926 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 07 May 2026

Short summary

Accurately measuring ultrafine particle number concentration (PN) from motor vehicle emissions is essential for the prevention and control of atmospheric particulate pollution. The latest Euro 7 emission regulations have lowered the particle size detection threshold from 23 to 10 nm and introduced updated standards for exhaust gas pretreatment and particle counting devices.


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