Rapid Determination of Ethylene Oxide and 75 VOCs in Ambient Air with Canister Sampling and Associated Growth Issues

Round 1

Reviewer 1 Report

Hoisington and Herrington presented an interesting study for the GCMS detection of ethylene oxide (EtO) without derivatization from the canisters with the help of preconcentration and cryogenic cooling. The method is based on US EPA Method TO-15A and a method validation was performed to evaluate the MDL, linearity, accuracy and precision. Additionally, the cleaning procedure for EtO in canister and potential EtO growth mechanism was investigated. The study was well-designed and the manuscript is well-written. It is therefore recommended for this manuscript to be accepted for publication after minor revisions as indicated below:

Minor comments:

1. Please define NATTS on Page 2 Line 51.
2. In Table 2 on Page 4, the % area increase from 0 °C to -30 °C were calculated incorrectly: the increase should be 58% and 74% respectively. The calculated values are the % area, not % area increase
3. Figure 1 on Page 4 needs error bars incorporated in the plots.
4. Could the author provide legends for chromatograms shown in Figure 3 on Page 6?
5. In Figure 4 on Page 7, it seems like the chromatograms shown were acquired with a combination of scan method (top) and SIM methods (mid and bottom), if this true, can the author add legends or captions to them? If this is not true, can the author explain why there are so many traces in the mid plot?

        Also, are compounds 5 (n-Butane) and 6 (vinyl chloride) successfully resolve? This is not clearly indicated in Figure 4.

6. As part of the method validation for linearity, the calculation results shown in Table 5 used ISTD and plotted the relative response of EtO against ISTD, this is great considering that ISTD can correct potential response variation from the instrument due to different factors. such as different injection volume. It was not clear whether the other calculations regarding MDL, accuracy, and precision also used the relative response against ISTD. Can the author clarify that?
7. Can the author provide more significant figures (SF) for the MDL calculation shown on Page 11? The current calculation does not seem to have sufficient SF for the calculation of SD and MDL.
8. Is there any requirement/criteria for the determination if EtO is stable in a canister? As shown in Table 9 on Page 12, it might be better to show the full stability test results up to 14 days for comparison purposes.
9. In Table 10 on Page 12, can the author report “< LOQ” instead of “ND”? Also applies to Table 11 on Page 13.
10. In order to show that the EtO is below the MDL, if might be better to show an GCMS overlay of the post-proprietary cleaning canister, with an EtO standard injection at the MDL to replace Figure 9 on Page 14.
11. For Table 12 on Page 14, it is evident that the proprietary cleaning is capable of reducing the amount of EtO to a certain level at least. And it is reasonable to explain the increase of EtO level following Proprietary cleaning 1 with potential breakdown of larger carbon containing compounds, assuming that the proprietary cleaning was exclusively targeting EtO and may not be that effective in cleaning other chemicals in the canisters. Still, for canister 2, 3, and 4, it remains unclear why the level of EtO decrease or stay the same. Is it possible to rule out the possibility that the “ND” might be related to the current method not being able to accurately measure low concentration?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Overall comment:

This manuscript by Hoisington and Herrington describes a GC-MS method which is claimed to be an improved version of the US EPA TO-15 compendium method with particular attention on the analysis of EtO, along with other 75 VOCs. Since EtO is easily co-eluted with compounds like methanol and acetaldehyde, both of which are commonly found in ambient air either via primary emission or via photochemistry. Proper selection of a column and a lower starting oven temperature for the temperature program are two key approaches to allow successful separation of EtO from other interfering species. Some key aspects of quality assurance, e.g., recovery, linearity, reproducibility (RSD), MDL, etc. were performed to support the robustness of their proposed method. It is basically a re-validation of TO-15 with new instrumentation; therefore, little innovation can be sighted in the article. In the early 90’s people would use home-built trap with liquid nitrogen to enrich target VOCs and hot water to desorb them to GC. With commercial thermal desorption (TD) units available nowadays, automated analysis of VOCs by GC-MS becomes relatively easy and straightforward. The TD unit chosen by the authors was made by Markes which uses chemical sorbents plus thermal electric cooling to trap VOCs at sub-ambient temperatures. In this work the trapping temperature is set at -30°C and the desorption temperature is 300°C. Several other GC-MS studies in the literature used home-built TDs or chromatographic techniques and showed satisfactory results as well, which are considered to be more innovated than simply using commercial units to perform an existing EPA method.

Specific comments:

1. Since this manuscript is based on the existing TO-15 method. The authors are expected to overview the common problems encountered when using the “old” TO-15 method, and also to point out why there are needs for an improved method for better results of EtO analysis. What are the most noticeable improvements resulted from their instrumentation and approaches compared ot the old method? This part can be added in the last paragraph of their introduction section and also the right places in the main text.
2. Figure 1 is about the determination of breakthrough, as written in the caption, but the breakthrough volumes are not reported. If they intend to perform breakthrough in addition to the property of linearity, then more data points are probably necessary; that is to add more data points until the straight lines start to bend over. If their intention is not to report breakthrough but linearity only, then drop the word “breakthrough” from the text and figure.
3. Figure 3 is terribly made and gives a lot of confusion. There are so many traces in different colors but no or little labeling and illustration is given. I tried very hard to understand these traces but gave up in the end. This figure is considered as the most important result of their work, because separation of EtO from MeOH is the paramount issue they attempted to address.
4. Some problem with Fig. 4 as with Fig. 3, no labeling or sufficient illustration is given either in the text or caption. In other words, very poor figure quality!
5. The title says EtO plus 75 other VOCs - that is 76 species, but Fig. 4 lists 80 VOCs. Why is there an inconsistency?
6. 6, is the table next to the graph necessary? Should it be discarded?
7. “The results of the MDL study are shown in Table 7. The low calibration point of 34 pptv was determined as the spiking level of the MDL study due to its S/N ratio of 5.9.” Why is the S/N = 5.9 is the criteria for the lowest concentration to determine MDL? The finding of the lowest concentration for DL determination requires analysis of 7 aliquots of a sample within 5 fold of the presumed DL as the first step. Further dilution of the sample is required until finding the right level of the sample for DL determination. Although it is a lengthy process, it should be performed and described as required, since DL of EtO is an important piece of information particularly for this article.
8. What is the huge peak in the early segment of the chromatograms in Fig. 8 and 9, retention time around 4 min? Is it CO2? Would it burry peaks of EtO and other highly volatile species? Is SIM mode used to detect these compounds? What are the ions used for detecting EtO to avoid other interfering species? In fact, to make Table 4 more useful, the target ions for each VOC and its DL, RSD, R2, recovery, etc. can be combined within one table.
9. L299-300. The two proposed reactions are not drawn in a correct manner. Stepwise reactions to form final products should be given if reactions are proposed.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

1. 3. The same problem that has been pointed out in my first round of review is still there; that is, the colored traces in the figure are not labeled in the plot and explained in the caption. It is annoying to have to figure out things by themselves, things that are crucial to the clarity of the technique.
2. Same comment goes to Fig. 4. There are colored traces without explanation.

3. Table 5. This table is not made in a proper way.

4. 6. Is the table to the right necessary? If so, then explain the content explicitly. Otherwise, they should consider to drop it. This point has been raised in my last review but without correction.

5. If Table 9 is about showing stability in recovery%, then why the values are not shown as such? It would be a lot easier for readers to read percent recovery than the absolute concentrations.

6. 7 and Table 11 do not state whether humid air, dry air or inert gas is used. Readers would have to guess that humid air was used for this particular test, correct?

7. 2, if all these mass spectra are taken from the MS library and not the real experimental data in the study, it is suggested that these spectra are placed in supporting information instead of the main text.

8. L168-: “s. The Rtx-VMS column…..Initial testing using ambient starting temperatures failed to fully resolve the MeOH/EtO 168 coelution, but the Rxi-624Sil MS column was able to separate the pair at lower temperatures using cryogenic cooling…”

Comment: The Rxi-VMS column was not able to resolve these two peaks at ambient temperatures, but what if it was done at sub-ambient temperatures like the 624 column, would the two compounds be resolved then? Would there be other types of columns that can separate these two compounds at ambient temperatures? Obviously only two columns have been tested in these study. It may be worth trying to find a column that has the desired separation ability without having to lower the GC oven temperature to sub-ambient using cryogens.

This manuscript although has improved quite a bit from its original version, it still lacks the required standard to be published in a high-quality journal. Again, the technique is based on TO-15, which is an established method since 90’s. Novelty is not seen in this paper and the way the results is presented is still rough and requires some substantial polishing and modification. Figures and Tables are not made in a professional manner. I would suggest somewhere between rejection and major revision in view of the existing weaknesses, even though the manuscript has gone through the first round of revision.    

Comments for author File: Comments.pdf

Author Response

1. The same problem that has been pointed out in my first round of review is still there; that is, the colored traces in the figure are not labeled in the plot and explained in the caption. It is annoying to have to figure out things by themselves, things that are crucial to the clarity of the technique.

The caption states what each trace is. However, we have added notes on figure ID’ing the traces and improved picture quality to help clarify it.

2. Same comment goes to Fig. 4. There are colored traces without explanation.

The caption states what the traces are. However, we have added notes to each trace labeling the TIC, EIC, and SIM traces to help clarify.

3. Table 5. This table is not made in a proper way.

We are unsure what is meant by this. If a clearer idea of the issue is given the table can be modified as needed.

4. Is the table to the right necessary? If so, then explain the content explicitly. Otherwise, they should consider to drop it. This point has been raised in my last review but without correction.

Our last response made a case for keeping the table, as it included the %RSD and calibration equation and we felt that information was important to readers. However, we have removed the table from the figure and provided the information in the lines below the figure.

5. If Table 9 is about showing stability in recovery%, then why the values are not shown as such? It would be a lot easier for readers to read percent recovery than the absolute concentrations.

This has been changed to recovery %.

6. 7 and Table 11 do not state whether humid air, dry air or inert gas is used. Readers would have to guess that humid air was used for this particular test, correct?

We cover this in Section 2.1 – “Unless otherwise noted, all standards and blanks were made in 6L SilcoCan air canisters (Restek Corporation, Bellefonte, PA, USA) using zero air humidified to 50% relative humidity (RH) using deionized water, matching the suggested fill gas and humidification levels suggested in TO-15A.”

However, we have also updated all tables to explicitly state the use of 50% humid air when used to add clarity.

7. 2, if all these mass spectra are taken from the MS library and not the real experimental data in the study, it is suggested that these spectra are placed in supporting information instead of the main text.

Moved to supporting information

8. L168-: “s. The Rtx-VMS column…..Initial testing using ambient starting temperatures failed to fully resolve the MeOH/EtO 168 coelution, but the Rxi-624Sil MS column was able to separate the pair at lower temperatures using cryogenic cooling…”

Comment: The Rxi-VMS column was not able to resolve these two peaks at ambient temperatures, but what if it was done at sub-ambient temperatures like the 624 column, would the two compounds be resolved then? Would there be other types of columns that can separate these two compounds at ambient temperatures? Obviously only two columns have been tested in these study. It may be worth trying to find a column that has the desired separation ability without having to lower the GC oven temperature to sub-ambient using cryogens.

Modified lines 166-169 - The Rtx-VMS column was unable to provide separation for EtO and MeOH at ambient or sub-ambient temperatures. Initial testing using the Rxi-624Sil MS column at ambient starting temperatures failed to fully resolve the MeOH/EtO coelution, but it was able to separate the pair at lower temperatures using cryogenic cooling.

Added lines 179-181 - Other column phases were not investigated, as the Pro EZGC Modeler showed that the Rtx-VMS and Rxi-624Sil MS provided the most efficient separation for the TO-15 compounds. In addition, we know from our collaborations that no one has successfully provided a single column solution for EtO and the other TO-15A VOCs, without the use of on column cooling. We know this to be true for 1-type and 5-type columns. The only potentially viable ambient approach would be the use of plot columns and/or with the use of dean switches; however, these two options represent their own set of hurdles when it comes to analyzing EtO and the TO-15A VOCs in one analysis, and with MS confirmation.

This manuscript although has improved quite a bit from its original version, it still lacks the required standard to be published in a high-quality journal. Again, the technique is based on TO-15, which is an established method since 90’s. Novelty is not seen in this paper and the way the results is presented is still rough and requires some substantial polishing and modification. Figures and Tables are not made in a professional manner. I would suggest somewhere between rejection and major revision in view of the existing weaknesses, even though the manuscript has gone through the first round of revision. 

We have had several discussions with commercial laboratories and the US EPA regarding the analysis of ethylene oxide in ambient air, and the labs performing the analysis expressed concerns that some of their canisters were checking out as clean in the lab blanks, but showing abnormally high levels of ethylene oxide when used in the field. Therefore, we started by sorting out all of the chromatography, which was and remains a confounding variable for laboratories, hence this is a significant contribution to the scientific community. Next,  our investigations led to showing the link between canister cleanliness and humid air in ethylene oxide growth in canisters, which labs were and are missing due to using nitrogen blanks and doing single target SIM analysis for ethylene oxide. This issue is wholly absent in published literature, and therefore, most labs remain blind to all the aforementioned. We feel the novelty of this paper comes from providing a complete solution that addresses the sample collection and storage issues, as well as analysis. Our talks with labs and regulatory agencies show that there is a need for this knowledge as well, as it is absent from existing literature. It is important to note that the US EPA considers ethylene oxide to be a priority topic given its toxicity and recent flurry of media attention. Hopefully, you too will see the timely value of our manuscript, based on all of the aforementioned.

Some of the figures have been redone to improve their quality, and the specific comments aimed at the lack of clarity have been addressed by adding labels in the figures themselves in addition to the captions. The specific comments regarding tables have all been addressed, however the comment for table 5, and the general comment that tables and figures re not made professionally, provides no constructive criticism we can  respond to. If more clarity is given on this we will be happy to address the specific issues.

Round 3

Reviewer 2 Report

This third revision seems to have made corrections or modification in accordance with the review comments. The quality of this manuscript has greatly improved. Acceptance is recommended.