Comparison of Two Derivative Methods for the Quantification of Amino Acids in PM2.5 Using GC-MS/MS
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsDear authors,
Your manuscript is interesting, well written and organised. The comparison of the two derivatisation method would be helpful for further studies. However, you should discuss in more details the advantages and inconvenience of each derivatisation methods. You states that there are no statistical differences between the two methods but regarding results MtBSTFA should be better. that's different in your conclusion. In addidtion, the protocol for assessing the two derivatisation techniques should also discuss in terms of feasability and simplicity. MtBSTFA derivatisation have to be done in strict dry conditions. How this conditions was tested in your experiments as you write that your evaporate extracts to dryness but no more infomations about the potential remaining of water traces which could have an effect on the efficiency of the derivatisation.
Author Response
Your manuscript is interesting, well written and organised. The comparison of the two derivatisation method would be helpful for further studies. However, you should discuss in more details the advantages and inconvenience of each derivatisation methods. You states that there are no statistical differences between the two methods but regarding results MtBSTFA should be better. that's different in your conclusion. In addidtion, the protocol for assessing the two derivatisation techniques should also discuss in terms of feasability and simplicity. MtBSTFA derivatisation have to be done in strict dry conditions. How this conditions was tested in your experiments as you write that your evaporate extracts to dryness but no more infomations about the potential remaining of water traces which could have an effect on the efficiency of the derivatisation.
Response
I greatly appreciate the insightful comment. Based on your comments, I have revised the manuscript and summarized three categories as below. Please refer to the revised manuscript prepared as a PDF file with traceable changes.The response letter can also be found in the attached file.
Comment 1
Your manuscript is interesting, well written and organised. The comparison of the two derivatisation method would be helpful for further studies. However, you should discuss in more details the advantages and inconvenience of each derivatisation methods. You states that there are no statistical differences between the two methods but regarding results MtBSTFA should be better. that's different in your conclusion. In addidtion, the protocol for assessing the two derivatisation techniques should also discuss in terms of feasability and simplicity.
Response
It is known that there are over 500 amino acids containing a basic amino group (-NH2) and a carboxyl group (-COOH) exist in nature, they have diverse structures and polarities, making it difficult to specify a single advantageous derivative method. The targeted 13 amino acids in PM2.5 in this study are also compounds with diverse structures and polarities, as shown in Table S1. The only derivatization used for amino acid analysis in atmospheric samples so far has been the MTBSTFA derivative method, but the disadvantage of this method is that it takes a lot of time to dry the aqueous extract because it should be performed under strict drying conditions as you pointed out. Therefore, the research that require monitoring a large number of samples are subject to time constraints. From this perspective, a rapid derivative method using ethyl chloroformate that does not require strict drying conditions was attempted for amino acid analysis in this study. In response to your comments, the pros and cons of both derivative methods have been added to the introduction, including aspects of feasibility and simplicity.
Comment 2
MTBSTFA derivatisation have to be done in strict dry conditions. How this conditions was tested in your experiments as you write that your evaporate extracts to dryness but no more infomations about the potential remaining of water traces which could have an effect on the efficiency of the derivatisation.
Response
We regret that the concentration process of sample extracts was not described in detail, which may have caused confusion. MTBSTFA method requires complete drying as you pointed out, as the derivative reaction of amino acids is interfered by the presence of moisture. Therefore, it was completely dried to remove any residual moisture under nitrogen stream, and was stated in the text.
Comment 3
Figures and Tables are not clear and well-presented.
Response
Figures and Tables have been edited to make them clearly visible in the text and we have also attached them as separate files just in case.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsAmino acids (AAs) inPM2.5 are directly and indirectly related to climate change, and to allergic diseases. In this work, two methods, MTBSTFA w/1% t-BDMCS for silylation, and ethylchloroformate (ECF) with methanol for chloroformate derivatization, were used to compare and evaluate 13 AA derivatives in PM2.5 samples. The work is very interesting. However, the innovation points were not sufficiently emphasized. I think that a major reversion is needed before acceptation.
The following issues should be also considered:
- In lines of 20, 213, 220, 296, “N”and “tert” in “N-tert-butyldimethylsilyl- N-methyltrifluoroacetamide” should be italic.
- In Figure 2, it seems that the reactions were not complete at the time of 2 h. Please add the derivatization experiment at 4 h.
- In table 1, please the concentration of AAs where RSD was measured. The max RSD (%) reached 24.8, indicating unacceptable results
- In table 2, Relative difference of Asn and Ser is too large between the two methods, while both had similar recoveries. Why?
Author Response
Amino acids (AAs) in PM2.5 are directly and indirectly related to climate change, and to allergic diseases. In this work, two methods, MTBSTFA w/1% t-BDMCS for silylation, and ethylchloroformate (ECF) with methanol for chloroformate derivatization, were used to compare and evaluate 13 AA derivatives in PM2.5 samples. The work is very interesting. However, the innovation points were not sufficiently emphasized. I think that a major reversion is needed before acceptation.
Response
Thanks for your valuable suggestions. We have revised the manuscript to highlight the innovative aspects of this study as your comments. Please refer to the revised manuscript prepared as a PDF file with traceable changes. The response letter can also be found in the attached file.
Comment 1
In lines of 20, 213, 220, 296, “N”and “tert” in “N-tert-butyldimethylsilyl- N-methyltrifluoroacetamide” should be italic.
Response
The entire text has been reviewed and rewritten in italics.
Comment 2
In Figure 2, it seems that the reactions were not complete at the time of 2 h. Please add the derivatization experiment at 4 h.
Response
Amino acids containing the functional amino group (-NH2) and carboxyl group (-COOH) have different derivatization reaction rates and the types of derivative forms(M+1TBDM or M+2TBDM or M+3TBDM or M+4TBDM) depending on their molecular structures. The targeted 13 amino acids in PM2.5 in this study are also compounds with diverse structures and polarities, as shown in Table S1. Considering the target derivative products of individual amino acids (M+2TBDM or M+3TBDM) as a whole as described in Table S2, 2 h was found to be an appropriate derivatization time, and please note that each derivative products may change different forms when it exceeds 2 h. Also, please understand that 4 hours for a derivatization experiment is too long from a sample preparation perspective, as it also involves a procedure of concentrating long aqueous sample extracts.
Comment 3
In table 1, please the concentration of AAs where RSD was measured. The max RSD (%) reached 24.8, indicating unacceptable results.
Response
As you pointed out, 24.8% RSD is high as the RSD for the RRF of the calibration curve. However, please consider that RSDs may vary depending on the calibration ranges of AAs and they were evaluated as derivatives (less stable), not the original forms of AAs. Also, the RSD tolerance of the RRF we confirmed was within 25% as shown below. Please refer to the table below which summarizes EPA and authoritative references.
|
Reference |
Acceptance Criteria |
Description |
|
U.S. EPA. Guide to Method Flexibility and Approval of EPA Water Methods 1) |
RSD < 25% |
Traditional performance specifications considered any regression line with a correlation coefficient (r) of 0.995 or greater as linear. Also, for organic analytes, a relative standard deviation (RSD) of 25% or less is considered linear. |
|
McClenny, W., & Holdren, M. (1999), Method TO-15A VOCs, EPA 2) |
RRF ≤ 30% RSD |
Average RRF ≤ 30% RSD and each calibration level within ±30% of theoretical concentration; for quadratic or linear curves, coefficient of determination ≥ 0.995, and each calibration level within ±30% of theoretical concentration |
|
Ma, T. et al.3) |
RSD < 25% |
The relative standard deviation (RSD) of the relative response factors (RRF) of each substance at different concentrations must be less than 25%. |
|
Camino-Sánche et al. 4) |
RSD < 25% (ISO and IUPAC guidelines) |
The method performance requirements were established as follows: (a) Linearity, the determination coefficient (R2) must be equal or greater than 0.990 and maximum residual deviation must be less than 25 %. |
|
Rozio et al. 5) |
RSD < 30% |
Calculate the RRF at each concentration level over the tested linearity range. The RSD on the obtained RRF values must be not more than 30% |
1) U.S. EPA (1996). Guide to Method Flexibility and Approval of EPA Water Methods. EPA 821-D-96-004.
2) McClenny, W. A., & Holdren, M. W. (1999). Method TO-15A: Determination of Volatile Organic Compounds (VOCs) in Air. U.S. EPA.
3) Ma, T., Teng, Y., Christie, P., Luo, Y., Chen, Y., Ye, M., & Huang, Y. (2013). A new procedure combining GC-MS with accelerated solvent extraction for the analysis of phthalic acid esters in contaminated soils. Frontiers of Environmental Science & Engineering, 7(1), 31-42. https://doi.org/10.1007/s11783-012-0463-2
4) Camino-Sánchez, F. J., Zafra-Gómez, A., Ruiz-García, J., & Vílchez, J. L. (2013). Screening and quantification of 65 organic pollutants in drinking water by stir bar sorptive extraction-gas chromatography-triple quadrupole mass spectrometry. Food Analytical Methods, 6(3), 854-867. https://doi.org/10.1007/s12161-012-9495-2
5) Rozio, M. G., Angelini, D., & Carrara, S. (2025). Uncertainty factors and relative response factors: correcting detection and quantitation bias in extractables and leachables studies. Analytical and Bioanalytical Chemistry, 1-19. https://doi.org/10.1007/s00216-025-05946-5
We have provided tolerances based on references of RSD in the text (page 5) and have cited additional references as new.
Comment 3
In table 2, Relative difference of Asn and Ser is too large between the two methods, while both had similar recoveries. Why?
Response
Recovery tests were performed by spiking target analytes into blank quartz fiber filters, which does not reflect the complexity of environmental matrices. However, the concentrations of the analytes may differ between the two methods due to the presence of interfering substances or matrix effects in real samples, The post-extraction cleanup procedures to exclude interfering substances could be considered, but further research is needed. We hope this clears things up.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe present study aimed to establish sample preparation procedures and GC-MS/MS conditions for two derivatization methods, and to validate their applicability for the quantitative analysis of 13 amino acids of interest in atmospheric PM2.5 samples. Before the manuscript can be considered for publication, the following comments and suggestions are respectfully offered to improve the clarity, scientific rigor, and overall quality of the work.
1. Keywords: Improve your Keywords by avoiding long compound words.
2. Introduction: The rationale for comparing the two derivatization methods, MTBSTFA and ECF-MeOH, is well justified. However, it would be beneficial to provide a more comprehensive discussion of the specific limitations associated with each method, as reported in previous studies, particularly when applied to complex atmospheric matrices such as PM2.5.
3. Serine was the only AA that exhibited a statistically significant difference between the two derivatization methods. This discrepancy was attributed to possible co-elution or matrix interference; however, no experimental confirmation was performed (e.g., by spiking with pure standards or conducting clean matrix recovery experiments). Incorporating such an evaluation would strengthen the interpretation and reliability of the observed differences and is recommended for inclusion in the study.
4. The authors compared the results obtained from the two proposed methods; however, a comparison with previously reported methods in the literature is also necessary. Such a comparison would allow for a clearer assessment of the analytical performance and practical advantages of the proposed approaches, thereby highlighting their potential improvements in sensitivity, efficiency, or applicability for amino acid analysis in atmospheric PM2.5 samples.
Author Response
The present study aimed to establish sample preparation procedures and GC-MS/MS conditions for two derivatization methods, and to validate their applicability for the quantitative analysis of 13 amino acids of interest in atmospheric PM2.5 samples. Before the manuscript can be considered for publication, the following comments and suggestions are respectfully offered to improve the clarity, scientific rigor, and overall quality of the work.
Response:
We wish to thank the reviewers for the helpful comments. We have adopted most comments and suggestions and revised the paper accordingly. Please refer to the revised manuscript prepared as a PDF file with traceable changes. The response includes some tables, so please see the attached file for convenience.
Comment 1
Keywords: Improve your Keywords by avoiding long compound words.
Response:
We have revised the keywords to avoid long compound terms and improve clarity. The updated list is as follows:
Keywords:
Amino acids; Fine particulate matter; MTBSTFA; Ethyl chloroformate; Derivatization; GC-MS/MS
Comment 2
Introduction: The rationale for comparing the two derivatization methods, MTBSTFA and ECF-MeOH, is well justified. However, it would be beneficial to provide a more comprehensive discussion of the specific limitations associated with each method, as reported in previous studies, particularly when applied to complex atmospheric matrices such as PM2.5.
Response:
Following your point, we have covered the specific limitations associated with each methods more comprehensively in the introduction. As seen in the recently published review paper (2025, https://doi.org/10.1007/s41810-025-00299-z), all the reviewed studies were performed using only silylation methods (especially MTBSTFA derivatization) to determine amino acids in fine aerosol particles in GC/MS analysis. We can say that ECF-MeOH derivatization was first attempted in our study because we were also unable to find any research cases in which this method was utilized for the AAs analysis in fine aerosol particles. Please refer to the introduction for the reasons why we tried the new derivative method and the pros and cons of both methods.
Comment 3
Serine was the only AA that exhibited a statistically significant difference between the two derivatization methods. This discrepancy was attributed to possible co-elution or matrix interference; however, no experimental confirmation was performed (e.g., by spiking with pure standards or conducting clean matrix recovery experiments). Incorporating such an evaluation would strengthen the interpretation and reliability of the observed differences and is recommended for inclusion in the study.
Response:
We agree that your comments are meaningful in understanding the differences in results between the two derivative methods for serine and have performed additional spiking experiments. Since the analysis of actual samples was completed, we did our best to conduct additional experiments using the remaining sample extracts. The procedures and results were presented in Supplementary Table S4. The evaluation of matrix effect was described based on references.
Comment 4
The authors compared the results obtained from the two proposed methods; however, a comparison with previously reported methods in the literature is also necessary. Such a comparison would allow for a clearer assessment of the analytical performance and practical advantages of the proposed approaches, thereby highlighting their potential improvements in sensitivity, efficiency, or applicability for amino acid analysis in atmospheric PM2.5 samples.
Response:
Since there are no previous reports regarding ECF-MeOH derivatization to determine amino acids in fine aerosol particles by GC/MS analysis, comparison with previous literature could only be limited to MTBSTFA derivatization for AAs analysis. Even in comparison with previous literature on the results of MTBSTFA derivatization, it is difficult to make an accurate comparison due to the different types of amino acids, different derivatization conditions, and differences in measurement methods. However, when compared with the evaluation results of the representative previous study (https://doi.org/10.1016/j.chroma.2009.11.021) cited in many literatures related to quantitative analysis of AAs in atmospheric PM2.5 samples using GC/MS, the results in our study were superior in terms of MDLs and recoveries of AAs as below.
|
AAs |
This study |
Mandalakis et al. 2010 |
|
|
MTBSTFA |
ECF-MeOH |
MTBSTFA |
|
|
MDL (ng) |
MDL (ng) |
MDL (ng) |
|
|
Ala |
7.2 |
3.5 |
43 |
|
Asn |
5.9 |
5.6 |
5.3 |
|
Asp |
8.1 |
2.8 |
196 |
|
Gly |
72.9 |
10.6 |
108 |
|
Ile |
10.5 |
3.0 |
19 |
|
Leu |
8.2 |
3.7 |
17 |
|
Met |
2.9 |
3.3 |
4.2 |
|
Phe |
3.1 |
0.5 |
21 |
|
Ser |
24.3 |
4.5 |
57 |
|
Thr |
9.2 |
7.0 |
34 |
|
Tyr |
5.8 |
12.8 |
21 |
|
Val |
9.8 |
2.7 |
16 |
|
AAs |
This study |
Mandalakis et al. 2010 |
|
|
MTBSTFA |
ECF-MeOH |
MTBSTFA |
|
|
% Recovery (low / High ) |
% Recovery (low / High ) |
% Recovery |
|
|
Ala |
107.0 ± 5.8 / 104.2 ± 4.3 |
115.4 ± 1.3 / 115.8 ± 2.6 |
105 ± 9 |
|
Asn |
93.2 ± 11.9 / 102.7 ± 23.5 |
116.6 ± 1.4 / 109.6 ± 1.8 |
64 ± 4 |
|
Asp |
98.5 ± 2.7 / 105.4 ± 16.6 |
118.7 ± 10.5 / 113.7 ± 13.8 |
84 ± 13 |
|
Gly |
114.8 ± 15.1 / 103.7 ± 14.9 |
117.2 ± 4.6 / 115.5 ± 4.1 |
47 ± 3 |
|
Ile |
85.6 ± 5.2 / 93.6 ± 5.1 |
81.5 ± 3.8 / 113.6 ± 15.1 |
96 ± 11 |
|
Leu |
84.5 ± 7.1 / 109.2 ± 7.4 |
80.0 ± 1.8 / 97.5 ± 12.9 |
90 ± 8 |
|
Met |
90.5 ± 2.0 / 100.4 ± 4.7 |
90.5 ± 2.0 / 100.4 ± 4.7 |
87 ± 7 |
|
Phe |
98.5 ± 2.7 / 104.6 ± 3.3 |
115.3 ± 9.7 / 114.2 ± 3.6 |
77 ± 5 |
|
Ser |
105.3 ± 4.7 / 108.4 ± 3.6 |
103.1 ± 21.2 / 103.7 ± 13.6 |
66 ± 8 |
|
Thr |
106.3 ± 10.0 / 119.7 ± 11.8 |
98.2 ± 16.8 / 119.4 ± 13.2 |
32 ± 3 |
|
Tyr |
90.7 ± 12.7 / 109.2 ± 15.5 |
94.1 ± 11.2 / 95.6 ± 1.3 |
83 ± 6 |
|
Val |
90.2 ± 9.1 / 119.3 ± 4.4 |
110.3 ± 13.2 / 99.4 ± 4.5 |
97 ± 4 |
The comparative analytical method (Mandalakis et al. 2010) used GC-(Single Quadrupole) MS for the detection, while our study used GC-(QqQ) MS/MS, and there were also differences in the conditions of extraction and derivatization. Please understand that since the main interest of this study was the comparison between a traditional MTBSTFA derivatization and a newly attempted ECF-MeOH derivatization, and so comparison with previously reported methods were not specifically mentioned in the text.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe current manuscript is acceptable for Chemosensors.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors addressed the suggestions, and I deem the manuscript suitable for publication.
