Evaluation of cation exchange membrane performance under exposure to high Hg<sup>0</sup> and HgBr<sub>2</sub> concentrations

Miller, Matthieu B.; Dunham-Cheatham, Sarrah M.; Gustin, Mae Sexauer; Edwards, Grant C.

doi:https://doi.org/10.5194/amt-12-1207-2019

Articles | Volume 12, issue 2

https://doi.org/10.5194/amt-12-1207-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/amt-12-1207-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 12, issue 2

Research article

|

26 Feb 2019

Research article |

| 26 Feb 2019

Evaluation of cation exchange membrane performance under exposure to high Hg⁰ and HgBr₂ concentrations

Matthieu B. Miller, Sarrah M. Dunham-Cheatham, Mae Sexauer Gustin, and Grant C. Edwards

Download

Final revised paper (published on 26 Feb 2019)
Supplement to the final revised paper
Preprint (discussion started on 27 Jun 2018)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review of CEM evaluation under high exposure conditions', Anonymous Referee #2, 04 Jul 2018
- AC1: 'Author’s Response to Referee #2', Matthieu Miller, 05 Aug 2018
RC2: 'Comments on PES CEM performance study', Anonymous Referee #1, 27 Jul 2018
- AC2: 'Author’s Response to Referee #1', Matthieu Miller, 20 Aug 2018

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Matthieu Miller on behalf of the Authors (21 Oct 2018) Author's response

ED: Referee Nomination & Report Request started (02 Nov 2018) by Jonathan Abbatt

RR by Anonymous Referee #2 (08 Nov 2018)

Suggestions for revision or reasons for rejection

Reviewer #2 Response:

Response to Major concerns:
1. In response to the concern that no standard reference material is used during the
acid-digestion analysis:
R2 – The reviewers have adequately addressed this concern

2. In response to concerns about pyrolyzer efficiency:
R2 – The authors response is inadequate in regard to the pyrolyzer. The authors attest to the pyrolyzer working at 500 C for GOM by referencing other papers, but their own data shows that there is a rather large discrepancy between the Tekran concentrations and the GOM concentration in this dataset of which the authors acknowledge. This could be (1) inefficient reduction of all GOM to GEM for subsequent Tekran analysis or (2) some other artefact in the experiments. Using the Tekran was a clever way to QA/QC check the concentrations being sorbed to the CEMs. But there was substantial discrepancies in the recoveries. To simply say “the pyrolyzer was inefficient” and then argue in response to my criticism of this by showing references stating that the pyrolyzer should work under at the tested temperatures is inadequate when clearly something was array in the experiments in terms of these recoveries. The authors could have and should have EASILY re-tested the experiments at the pyrolyzer temperatures they themselves suggest. It is my opinion that this simple, additional experimental treatment needs to be included in the study. This relates to point 5 below.

3. In response to concerns over removal of first in-series CEM filter during GEM permeations:
R2 - Here I feel the authors have adequately addressed the concern by listing both the proportion of sorbed GEM with and without inclusion of the first filter. It is still my preference that the first filters be included and as they authors state this is still a very small amount. This would also reduce the length of the manuscript and discussion on this could easily be moved to the SI without affecting anything, which would help with the suggestion I make in point 5 below.
Nonetheless, one concern that is still yet to be addressed is why is the uptake of GEM on the first filter not linear with increasing GEM concentration? It is exponential. I brought up this concern in the previous round of comments and it was ignored.

4. In response to the concern over unheated Teflon sample tubing length between Line
1 and Line 0:
R2 - Indeed another factor that may be contributing to the poor recovery results and another part of the experiments that could have easily been adjusted and retested. Possibly not a big issue as the authors suggest, but I imagine it also could have been adjusted easily.

5. “The authors agree that more experimentation is both welcome and necessary, and
indeed is now ongoing within the scope of an expanded and fully funded NSF research
project. However, our original and primary concern within the more narrow scope of this
manuscript was with GEM uptake and HgBr2 breakthrough at high loading rates.
These questions were of particular importance to a companion manuscript, also currently
submitted to AMT.”
R2 – What the authors are attempting to do here is manuscript quantity over manuscript quality. These experiments show useful data, but as I have suggested they appear rushed and too many questions and concerns have been left unanswered and unaddressed, when they could have easily been rectified by simply re-running experiments making slight adjustments to the existing conditions. This is THE STUDY addressing “GEM uptake and HgBr2 breakthrough at high loading rates” it should not need “companion” studies or “follow-up" studies to comprehensively answer the research question it set out to do. That is the role of this manuscript. It is my suggestion that this study be combined with this “companion study” to make a more complete paper. There is a lot of over-elaboration in the discussion and many points could easily be moved to an SI without loss of information to combine this study with complimentary work. This study definitely has merit, but it needs to be more polished.

6. In response to concerns about high CEM filter blanks:
R2 – This concern was brought up here because it has been inadequately addressed in previous studies. Take the Huang et al (2017) study for example:

“The bi-weekly MDLs (336 h) for active systems with cation-exchange and nylon membranes were 2–68 pg m−3 (mean: 24 pg m−3) and 0.01–14.6 pg m−3 (mean: 2.1 pg m−3), respectively. Biweekly MDL was calculated from 3 times the standard deviation of bi-weekly blanks. The MDL was calculated for each period of sampling, due to the fact this can vary based on treatment of the membranes, the time samples are prepared for deployment, deployment at the field site, and handling once returned to the laboratory. The membranes may also vary by material lot. All samples were corrected by subtracting the blank for the corresponding 2-week period.”

(a) A 2 week mean detection limit concentration of 24 pg m-3 is at or below typical GOM concentration even in industrial city sites (examples: Lyman and Gustin 2009; Huang et al. 2012; Choi et al. 2013).
(b) These are DETECTION limits. Not QUANTIFICATION limits. Attempting to provide high certainty, quantifiable results at concentrations at or below detection limits is simply erroneous.
(c) The purpose of averaging method detection and/or quantification limits is to prevent exactly what the authors are discussing here by smoothing out inconsistency in variable handling of materials, residual Hg, etc. By subtracting “the blank” (a single blank for a single sampling period; n=1) is again incorrect and based.
(d) This completely contradicts a statement the authors of the current manuscript have introduced in this round of reviews in terms of their attempting to caution the use of the term RM:
“a broad term that favors basic accuracy of measurement over determination of specific compounds”
How can this favour “accuracy of measurement” if attempts at field measurements at typical concentrations encountered in the environment are at or below detection (not quantification) limits. There is no confidence in such results.
As such these concerns do still need to be addressed for this sampling method.

R2 Other comments of this latest version:
Lines 39-40: This “cautionary” description of RM that was suggested by me previously is inadequate as decribed above. This research group attempts to use this same sorbent material to determine specific compounds of GOM in previous work (Huang et al. 2017). Thus, by making this statement here they are saying the results from this previous work are not accurate measurements.

Line 95: Once again here the authors should include a reference to the Marusczak et al. 2017 that describes how adding the zero flushes from GOM analysis to the actual GOM concentrations increases the derived concentrations to agree more closely with alternative measurement techniques and some modelled values. This was paper was specifically mentioned in the previous round of comments to be added to the literature review to balance impartiality but this comment was ignored.

Lines 128-131: Can the authors please reference where exactly in the Gustin et al. 2015 paper is the mention of PTFE/PFA producing zero sorption of GEM. Personally, I could not find this specific point within the reference.

Lines 220-222: Yes, the uptake was linear if the first filter was not included. But when the first filter is included it is exponential. Why? As yet this has not been addressed by the authors anywhere.

Line 232: Approximate not approximately.

Lines 236-237: “but it should be noted that the performance of the CEM filters at low concentrations could be slightly different"
This should be changed to:
“but it should be noted that these concentrations are 50-1000x above typical background concentrations and the performance of the CEM filters at low concentrations could be slightly different”
As was previously suggested.

Lines 263-266: Caution needed here. You would never use such a system that has been used in contaminated environments for background work. Even a small memory of the Hg would overwhelm the background signal. Lines would need to be discarded or thoroughly acid cleaned for background work. Please make a cautionary note on this here. A comment was made about this in the previous revisions, but ignored.

Lines 284-285: Oagain I ask is "photochemistry driven by room fluorescent
285 lighting." a process reported in the literature? If so please reference, if not this is just pure speculation.

Lines 340-343: Again, why speculate when this could have easily been determined by repeating the experiments at higher pyrolyzer temperatures?

REFERENCES:
Choi, H. D., Huang, J., Mondal, S., & Holsen, T. M. (2013). Variation in concentrations of three mercury (Hg) forms at a rural and a suburban site in New York State. Science of the Total Environment, 448, 96-106.

Huang, J., Miller, M. B., Edgerton, E., & Sexauer Gustin, M. (2017). Deciphering potential chemical compounds of gaseous oxidized mercury in Florida, USA. Atmospheric Chemistry and Physics, 17(3), 1689-1698.

Huang, J., Liu, C. K., Huang, C. S., & Fang, G. C. (2012). Atmospheric mercury pollution at an urban site in central Taiwan: Mercury emission sources at ground level. Chemosphere, 87(5), 579-585.

Lyman, S. N., & Gustin, M. S. (2009). Determinants of atmospheric mercury concentrations in Reno, Nevada, USA. Science of the total environment, 408(2), 431-438.

Marusczak, N., Sonke, J. E., Fu, X. and Jiskra, M. (2017). Tropospheric GOM at the Pic du Midi Observatory—Correcting Bias in Denuder Based Observations. Environmental Science & Technology 51 (2), 863-869

Hide

RR by Anonymous Referee #3 (23 Nov 2018)

Suggestions for revision or reasons for rejection

Evaluation of CEMs for GEM uptake and GOM capture and retention for use in a flow-based technique is an important step towards obtaining realistic measurements of reactive mercury. This study presents data that shows promising results for laboratory air. Of course, further work using an ambient air matrix and other species of GOM is need.

I am concerned there was no mixing volume downstream of the bromide permeation tube. In trying to keep line lengths short to minimize wall losses there may not be adequate mixing to insure equal concentrations travel through Line 0 and Line 1. This may bias uptake and capture measurements and may have contributed to the mass balance problems. It may be one of the causes of the large amount of noise in the pyrolyzed signal shown in Figure 3. If mixing volume use was not possible then only running with a single Line should have been considered.

I am also concerned about the poor performance of the pyrolyzer. Inadequate conversion due to residence time and temperature problems is fairly easy to address. It would have been a much better paper if simple tests of increasing the pyrolyzer temperature had been done. Incomplete reduction or varying efficiency may be another cause of the noisey signal in Figure 3.

Another issue that was not raised in the paper is the effect of back-reactions downstream of the pyrolyzer to reform HgBr2. What is the fate of the bromine?

Other Comments:

Line 43: Not ppt but parts per quadrillion

Line 145: Locating the AC scrubber downstream of the humidifier could cause problems: the carbon’s ability to uptake GEM is reduced at higher humidity. Some AC scrubbers contain iodine which can be released and migrate with humidity, which can then trap Hg downstream.

Line 158: Was the dry N2 flow included when calculating the humidity values?

Line 165: The volume of pyrolyzer seems small. What is residence time? Were any tests done to insure that reduction was quantitative?

Line 327: How did you estimate line losses?

Line 330: I can believe that the pyrolyzer was not 100% efficient, but did you try to increase the temperature and look for increased conversion?

Figure 1: It would be nice to show humidifier in this figure. I am concerned about the lack of mixing volume downstream of the permeation gas TEE. Without adequate mixing there may not be equal concentration of HgBr2 going to Line 0 and Line 1. This may be one of the reasons that there is so much noise of the red trace of Figure 3 and could throw off your mass balance calculations.

SI Figure 1: The gaps between 2537 calibrations is about a month. Why wasn’t it done on a regular basis or at the end of each test?

Hide

ED: Reconsider after major revisions (26 Nov 2018) by Jonathan Abbatt

AR by Matthieu Miller on behalf of the Authors (06 Jan 2019) Author's response Manuscript

ED: Referee Nomination & Report Request started (07 Jan 2019) by Jonathan Abbatt

RR by Anonymous Referee #2 (08 Jan 2019)

Suggestions for revision or reasons for rejection

GENERAL COMMENTS:
I believe the supplementary experiments that tested the pyrolyzer up to 1000 °C are of great benefit to this manuscript as it provide the necessary evidence that it is likely the responsible party for the higher recoveries. While a safe and operational 1000 °C pyrolyzer would have been ideal for all the experiments I can understand concerns if it was not entirely stable or safe for on-going operation.

I will now accept the manuscript for publication if the following minor corrections and updates are made:

SPECIFIC COMMENTS:
Line 29: Please write the actual value (127%) and not >100%

Lines 30-32: Please rephrase this sentence to:
"The low HgBr2 breakthrough on the downstream CEMs (<1%), suggest that the elevated recoveries are more likely related to sub-optimal pyrolyzer conditions or inefficient collection on the Tekran 2537 gold trap."
The low breakthrough is what suggests the pyrolyzer likely being responsible for the elevated recoveries, not the elevated recoveries suggesting poor pyrolyzer conditions.

Lines 38-39: The authors have now removed any and all discussion on the concerns of using the more general terminology RM, which was my concern in the original round of reviews “More caution should be used in the definition of the term reactive mercury (RM).”
The “Often” (or perhaps sometimes) use of RM is by one or maybe two research groups. This does not qualify as often.
Since the authors are struggling for a descriptive terminology due to their own interchanging use of GOM and RM in their literature I suggest the following simple and clear addition:
“While the term RM does dilute some specific information in regard to the state of oxidized Hg in the atmosphere it does remove some of uncertainty as whether or not PBM contributes to the Hg collected by the CEMs.”

Lines 117-118: Should be “We attempted to explain…”

Lines 200-202: Please do not use "pretty" in a scientific paper. It is a colloquial and very unspecific term. Can the authors be more quantitative here? Use the data form SI Figure 4. What was the collection efficiency compared to 1000 °C? State this. This doesn’t mean this value describes the exact inefficiency of the existing system (which should be mentioned), but it does a better job than “pretty”.

Lines 276-280: Linearity of the first filters remains poorly addressed. Fit both a linear and exponential curve to the data with equations and r2 values (as suggested below). The authors say in their response:
“There are only five data points, so we feel any assertion about this relationship being linear or exponential is not robust.”
The authors seems to be using this as some form excuse as to why this should not be discussed, but on the contrary why did this again not sound some alarm bells and why was this set-up not re-tested? This goes specifically with the sentiments of myself and one other reviewer that experiments were rushed and less conclusive than they should have been. The authors are happy to conclude the linearity of the relationship based on limited data, but seem to not want to make conclusions if they pose a problem. This is and has been a concern of mine because if it is a exponential relationship then it would pose a problem for lower concentrations. Some form of description or discussion on this point IS necessary.

Line 369-371: This sentence should be changed to:
“Although further work is required to more definitively determine detection and quantification limits of the various CEM methodologies, based on the mean total Hg mass of 50 ± 20 pg observed in this study, the artifact of GEM uptake to the CEMs would be below the detection limit observed here.”
This sentence finally provides an (albeit subtle) mention that work is still required on defining detection and quantification limits of the CEM methodologies that I have suggested in all rounds of review.

Line 406: Please add the following after the sentence concluding on line 406:
“Evidence for this conclusion can be seen in by the XX % increase in Hg collection by the pyrolyzer at 1000 °C (see SI Figure 4) in supplementary experiments.”

SUPPLEMENTAL
In the current supplemental information that caption to SI Figure 5 has a caption. This caption should obviously be removed for publication.
BUT, it makes a perfect point. Add both linear and exponential curves with their appropriate equations and r2 values. This is something I have been inquiring about throughout the reviews and even co-authors have asked for its inclusion.

Hide

ED: Publish subject to minor revisions (review by editor) (22 Jan 2019) by Jonathan Abbatt

AR by Matthieu Miller on behalf of the Authors (27 Jan 2019) Author's response Manuscript

ED: Publish as is (29 Jan 2019) by Jonathan Abbatt

Short summary

This study was undertaken to demonstrate that a cation exchange membrane (CEM) material used for sampling reactive mercury (RM) does not possess an inherent tendency to collect gaseous elemental mercury (GEM). Using a custom-built mercury vapor permeation system, we found that the CEM material has a very small GEM uptake of approximately 0.004 %, too small to create a significant artifact. We also found that a representative RM compound was collected by the CEM material with high efficiency.

Evaluation of cation exchange membrane performance under exposure to high Hg0 and HgBr2 concentrations

Download

Interactive discussion

Peer-review completion

Evaluation of cation exchange membrane performance under exposure to high Hg⁰ and HgBr₂ concentrations