Coupling a gas chromatograph simultaneously to a flame ionization detector and chemical ionization mass spectrometer for isomer-resolved measurements of particle-phase organic compounds

Bi, Chenyang; Krechmer, Jordan E.; Frazier, Graham O.; Xu, Wen; Lambe, Andrew T.; Claflin, Megan S.; Lerner, Brian M.; Jayne, John T.; Worsnop, Douglas R.; Canagaratna, Manjula R.; Isaacman-VanWertz, Gabriel

doi:https://doi.org/10.5194/amt-14-3895-2021

Articles | Volume 14, issue 5

https://doi.org/10.5194/amt-14-3895-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/amt-14-3895-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 14, issue 5

Research article

|

27 May 2021

Research article |

| 27 May 2021

Coupling a gas chromatograph simultaneously to a flame ionization detector and chemical ionization mass spectrometer for isomer-resolved measurements of particle-phase organic compounds

Chenyang Bi, Jordan E. Krechmer, Graham O. Frazier, Wen Xu, Andrew T. Lambe, Megan S. Claflin, Brian M. Lerner, John T. Jayne, Douglas R. Worsnop, Manjula R. Canagaratna, and Gabriel Isaacman-VanWertz

Download

Final revised paper (published on 27 May 2021)
Supplement to the final revised paper
Preprint (discussion started on 12 Aug 2020)
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 AMT 2020 264', Anonymous Referee #1, 08 Sep 2020
- AC1: 'Response to reviewers', Chenyang Bi, 05 Nov 2020
RC2: 'Review of Bi et al.', Anonymous Referee #2, 03 Oct 2020
- AC2: 'Response to reviewers', Chenyang Bi, 05 Nov 2020

Peer-review completion

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

AR by Chenyang Bi on behalf of the Authors (05 Nov 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (06 Nov 2020) by Mingjin Tang

RR by Anonymous Referee #1 (23 Nov 2020)

RR by Anonymous Referee #2 (25 Nov 2020)

Suggestions for revision or reasons for rejection

Summary:

I thank the authors for their responses to my comments as “reviewer 2”. However, I still find that the there is a clear issue with the Fig. 4a inset, which prevents me from having full confidence that this new instrument is functioning in the way the authors are describing it should. I understand why the authors prefer to keep their main data analysis for a subsequent manuscript, but I think that this manuscript should still be able to show consistency with typical iodide CIMS instruments, and thus give confidence that the subsequent data analysis will be as interpretable and useful as suggested.

Comments:

The authors’ reply to my comments about vanillin ionization are not entirely satisfying. There are numerous parts of the reply that are inaccurate or irrelevant. For instance, the authors wrote:
“A critical difference between this instrument and direct-air sampling instrumentation is the ability to collect “clean” mass spectra of individual analytes,…”
It is indeed straightforward to get ‘clean’ spectra with a direct-air sampling iodide CIMS, by just wafting a bottle of vanillin or any other single analyte in front of a sampling inlet (impurities in the commercially available vanillin are irrelevant to this discussion). This new instrument is novel because it can produce ‘clean’ spectra from complex mixtures which is a nice advance (though it should still produce the same spectra as when sampling a single analyte), but that’s not what my comment was about. Then the authors say:
“…we believe that a lot of the apparent discrepancy comes from the fact that this instrument specifically provides an ability to see and explore the non-adduct ions, while a typical CIMS does not straightforwardly relate adduct ions to potential non-adduct counterparts.”
When sampling just vanillin from a bottle directly into the iodide CIMS like I described, you do get all of the adduct and non-adduct ions, where you can directly explore which non-adduct ions form from ionization of vanillin. Therefore the authors’ statement is incorrect (while the new instrument could be useful for identifying the parents of non-adduct ions in a complex mixture, again a nice advance, that’s not the issue I am addressing).

The main issue that I still have is that when you directly sample vanillin like this with typical iodide ionization, the [M+I]- signal is 100x higher than the [M-H]- signal (I’ve measured this myself before), contrary to what is shown in Fig. 4a. The authors describe various things that can affect sensitivity and declustering in a TOF mass spectrometer, e.g. voltages and IMR pressure, but mostly that will only affect whether an adduct stays an adduct or declusters back to a neutral analyte and iodide anion (or further fragmentation e.g. loss of –CO2 or –H2O or –NO2 for some compounds), it won’t change the ionization pathway to enhance [M-H]-. In other words, any properly operating iodide CIMS should always sample vanillin with [M+H]- at much higher signal than [M-H]-, and never with those reversed. Because of this, I still have to conclude that something is amiss with the ionization in your vanillin example in Fig 4. The levoglucosan and undecanoic acid examples given in Fig. R3 do look fine and are as expected, but why is vanillin different? If the answer is some sort of artifact with your instrument design, then I would say there is a potentially major issue with the utility of this instrument, because your iodide CIMS data will not be comparable to other instruments for at least a subset of compounds. If it really is that you’re operating your CIMS in some atypical configuration that gives this spectrum, then I think you need to figure out why and reconfigure to something more standard, otherwise you’re negating the benefits of using an iodide CIMS by unnecessarily complicating the ionization chemistry and making it incomparable to other iodide CIMS data. I understand you’re arguing that your goal in this manuscript is not to fully understand all of the ionization chemistry (that’s the next paper) or to have all the answers, but I would much prefer you don’t publish your paper with a Fig 4a inset that is labeled as iodide ionization but is definitely strongly influenced by something else. The purpose of this paper is to show that your new instrument works and briefly show its benefits, but Fig 4a inset tells me something is not working as intended.

That said, it could still be a simple answer. I suggested that the ionization could have changed due to the IMR temperature, but the authors have pointed out that the IMR itself is not heated. It sounds like there could possibly be a surface where the transmission line mates with the IMR that could be at least somewhat heated, but barring this, there are other options. The CIMS could be operating in some strange configuration of voltages or pressures, etc, but I find this unlikely. Is it possible that you just have a lot more O2- impurity in your iodide-only mode than is typical? Is O2 diffusing in or leaking in from your zero air source or even from room air through a leaky fitting? If 225C temps are ruled out, then an O2 leak seems most likely. I’m not sure why this wouldn’t show up in your Fig R3 of levoglucosan and undecanoic acid; either O2- is very insensitive to those compounds while being very sensitive to vanillin, or the potential O2 leak was occurring only in your vanillin experiment but not your Fig R3 experiments. The last thing I can think of is that in your Fig4a, that main elution peak for which you’re showing the mass spectrum is not actually only vanillin, but may be dominated by some other compound that predominantly forms that non-adduct, which seems unlikely.

In summary, the Fig. 4a inset is definitely not normal iodide ionization of vanillin, and that doesn’t convince me as a reader that you have shown this instrument to be sufficiently described for this publication and subsequent use. I don’t know the answer, but I suggest the authors start by sampling vanillin directly with the iodide CIMS to verify they measure mostly [M+I]-.

Additionally, about my comment 2 on decomposition products, I thank the authors for the thorough description. I intended to refer to ‘upstream decomposition’, where heating for desporption (and in the column) could fragment to form smaller product molecules that get transmitted to both the CIMS and the FID. Then what I’m suggesting is that the iodide CIMS is simply insensitive to these smaller fragmentation products generally, while they do get sampled in the FID. It would be interesting (and valuable for interpretation) to know how much of this signal that gets measured in the FID but not CIMS is due to fragmentation (versus compounds that don’t fragment but are not sensitive to iodide). Thank you for showing the FID signal in Fig. R1, but in hindsight that’s not actually very useful because that is a mix of several/many known and unknown compounds, so there’s still no way of separating out the fragmentation pathway. To answer that question, you’d have to reproduce Fig. R1, but with injections of only a single known compound/isomer at a time. Then, you should have only a single eluting peak that is the known compound/isomer, and any other peaks are fragmentation products that may well show up in just the FID and not the iodide CIMS. I think that including this type of analysis would greatly help the quality of this manuscript, but perhaps it could be included in a future manuscript instead.

Hide

ED: Reconsider after major revisions (25 Nov 2020) by Mingjin Tang

AR by Chenyang Bi on behalf of the Authors (31 Mar 2021) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (01 Apr 2021) by Mingjin Tang

RR by Anonymous Referee #1 (14 Apr 2021)

RR by Anonymous Referee #2 (19 Apr 2021)

ED: Publish as is (20 Apr 2021) by Mingjin Tang

AR by Chenyang Bi on behalf of the Authors (26 Apr 2021)

Short summary

Measurement techniques that can achieve molecular characterizations are necessary to understand the differences of fate and transport within isomers produced in the atmospheric oxidation process. In this work, we develop an instrument to conduct isomer-resolved measurements of particle-phase organics. We assess the number of isomers per chemical formula in atmospherically relevant samples and examine the feasibility of extending the use of an existing instrument to a broader range of analytes.