A modified travelling wave ion mobility mass spectrometer as a

Subscriber access provided by University of Otago Library

Article

A modified travelling wave ion mobility mass spectrometer as a versatile platform for gas-phase ion–molecule reactions Martin F. Czar, Adrien Marchand, and Renato Zenobi Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b00541 • Publication Date (Web): 22 Apr 2019 Downloaded from http://pubs.acs.org on April 22, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

A modified travelling wave ion mobility mass spectrometer as a versatile platform for gas-phase ion–molecule reactions Martin F. Czar,a Adrien Marchand,a and Renato Zenobia,* a

Laboratory of Organic Chemistry, Department of Chemistry and Applied Biosciences, ETH Zurich, Zurich, Switzerland Corresponding author: Renato Zenobi, Department of Chemistry and Applied Biosciences, ETH Zurich, 8093 Zurich, Switzerland. Email: [email protected]. *

Abstract Taken individually, chemical labeling and mass spectrometry are two well-established tools for the structural characterization of biomolecular complexes. A way to combine their respective advantages is to perform gas-phase ion–molecule reactions (IMRs) inside the mass spectrometer. This is, however, not so well developed because of the limited range of usable chemicals and the lack of commercially available IMR devices. Here, we modified a travelling wave ion mobility mass spectrometer to enable IMRs in the trapping region of the instrument. Only one minor hardware modification is needed to allow vapors of a variety of liquid reagents to be leaked into the trap travelling wave ion guide of the instrument. A diverse set of IMRs can then readily be performed without any loss in instrument performance. We demonstrate the advantages of implementing IMR capabilities in general, and to this Q-IM-TOF mass spectrometer in particular, by exploiting the full functionality of the instrument, including mass-selection, ion mobility separation, and post-mobility fragmentation. The potential to carry out gas-phase IMR kinetics experiments is also illustrated. We demonstrate the versatility of the setup using gas-phase IMRs of established utility for biological mass spectrometry, including hydrogendeuterium exchange, ion– molecule proton transfer reactions, and covalent modification of DNA anions using trimethylsilyl chloride.

Page 1 of 20 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemical labeling performed in solution is a common method to obtain structural information on biomolecules and biomolecular complexes.1–8 For example, hydrogen– deuterium exchange (HDX) reactions can be performed on proteins in solution to give information about the solvent accessible surface area.2–4 The change in deuterium uptake upon binding of a ligand can reveal potential conformational changes. Other types of strategies such as chemical footprinting can also be used to obtain structural insights on solvent accessible functional groups. For example, dimethylsulfate footprinting of DNA or RNA has been used to determine which base within a nucleic acid strand is involved in hydrogen bond formation.7 These reactions are typically performed in solution and often analyzed using mass spectrometry. However, like every condensed-phase approach, chemical labeling suffers from a lack in specificity. Indeed, it is blind to the potential presence of multiple coexisting stoichiometries or conformations of the drug–target complex of interest. The ability to tackle this problem is the main strength of mass spectrometry, in particular when coupled with ion mobility spectrometry.8 A way to combine both the advantages of chemical labeling and mass spectrometry is to perform gas-phase ion–molecule reactions (IMRs) inside the mass spectrometer. Efforts of the past three decades have shown IMRs to be a useful structural tool in the field of mass spectrometry.9 In an IMR, analyte ions, usually generated by electrospray ionization (ESI), are allowed to react with reagent gases in one of the vacuum chamber compartments of a mass spectrometer; since a chemical reaction almost always results in a detectable change in analyte mass and/or charge, a mass spectrometer acts as a convenient reaction vessel for conducting IMRs. Perhaps the best known IMR is gas-phase HDX,10–19 which is done by exposure of ions to deuterated reagent gases such as ammonia (ND3). The degree of deuterium uptake, which is easily discernible from mass spectral shifts, can report on the conformations of gaseous ions. Another class of IMRs are proton transfer reactions,20,21 including those involving cation–anion pairs (i.e., ion–ion reactions),22–24 which have been shown to be analytically useful for the simplification of congested mass spectra,21,24 for elevating sequence coverage in tandem MS experiments,25–28 and which have also been of fundamental interest in order to study the relationship between gas-phase biomolecular structure and charge state.29–33 Dozens of IMR-based methods have emerged, which have found many uses: the positions of unsaturated carbon–carbon bonds in lipids can be localized using reactions with ozone;34 structural isomers that are not differentiable by other means often show differences in gas-phase reactivity,35,36 thus enabling their resolution;


Page 2 of 20

Page 3 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and the presence of specific functional groups,37–41 including those in post-translationally modified peptides,42–44 can be unambiguously detected. While the utility of IMRs is well-established, relatively few groups have harnessed their potential, and descriptions of straightforward setups for their implementation on commercially available instruments, though existing,40,45,46 are relatively scarce. Aside from the inherent difficulties in finding viable chemical reactions for gas-phase work, progress in this area is hindered by the requirement to significantly alter instrument hardware. Some applications also require the purchase of pure reagent gases, which can be prohibitively expensive. To date, most IMR work has been carried out on ion trapping instruments,9 since the exposure time of ions to reagent gases can be tuned to great effect on such instruments (typically tens of milliseconds to seconds), and since they offer the option to perform multiple steps of ion isolation and fragmentation. The development of some modern mass spectrometers, such as the hybrid quadrupole–ion mobility–time-of-flight (Q-IM-TOF) instrument used in this work (see Figure 1), enable the hyphenation of multiple steps of mass discrimination, ion fragmentation, and, sometimes, ion mobility separation in a single experiment. The integration of IMRs with such instruments has provided interesting new opportunities.16,17,47– 50

For example, the Rand group recently showed49,50 that gas-phase HDX can be effected in

the “cone entry” or “cone exit” (source) regions of such an instrument using vapors of liquid deuterated exchange reagents. In related work, Nagy et al.48 used the ion mobility cell of such an instrument as a reaction vessel for HDX. A particularly appealing aspect of these setups is that they are relatively simple to implement, as only very minor and easily reversible modifications to the instrument were required in both cases. Furthermore, the ability to use liquid reagents, aside from the lower cost entailed, provides considerable flexibility in terms of reagent choice. However, unlike the previous works using purified gaseous reagents,16,17 neither of these setups take full advantage of the instrument’s capabilities. Chemical labeling in the source region, while convenient for on-line analysis, precludes the possibility of performing IMRs on mass-selected species. This sacrifice in chemical control can be a detriment upon analysis of more complex samples. Likewise, infusion of reagents into the mobility cell precludes the possibility for straightforward ion mobility separation. To the best of our knowledge, a generalizable setup that allows IMRs to be performed using liquid reagents in the low-pressure regions of such an instrument has never been reported in the literature.



Here, we present a versatile setup for performing ion–molecule reactions in the trapping region of a hybrid Q-IM-TOF instrument using vapors of various liquid reagents. We show that, with only one simple modification to the instrument, ion–molecule reactions can be conducted on mass-selected species, which can then be further structurally characterized by ion mobility and/or collision-induced dissociation. As prototypical reactions, we considered gas-phase HDX, proton transfer reactions, and covalent modification reactions using trimethylsilyl chloride (TMSCl). We also show that the timescales available for IMRs on this platform rival those accessible using ion trap instruments.

Experimental Chemicals Leucine-enkephalin acetate salt hydrate, bovine ubiquitin, LCMS-grade water, glacial acetic acid, formic acid, methanol, ammonium acetate salt, piperidine, and aqueous ammonium hydroxide (28% NH3 in H2O) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Deuterated water (D2O, 99.9 atom % D), methanol-OD (MeOD, CH3OD, 99.5 atom % D), ethanol-OD (EtOD, CH3CH2OD, 99.5 atom % D), and deuterated aqueous ammonium hydroxide solution (25% ND3 in D2O, 99 atom % D) were purchased from Cambridge Isotope Laboratories, Inc. (Andover, MA, USA). The DNA sequences T6 (d(TTT-TTT)) and ds26 (d(CAA-TCG-GAT-CGA-ATT-CGA-TCC-GAT-TG)) were purchased from Eurogentec (Seraing, Belgium) in RP‐cartridge purified and lyophilized form. All chemicals were used as received. Instrument modifications The modified instrument (Figure 1) is a commercially available Synapt G2-S mass spectrometer (Waters, Wilmslow, UK), equipped with a Z-spray electrospray ionization source. Under conditions of normal instrument operation, the trap TWIG is fed continuously with Ar collision gas at flow rates on the order of 1–10 mL/min1, to maintain a pressure reading of ~102 mbar. This is achieved by a mass flow controller and solenoid valve, which are connected in series upstream of the trap TWIG gas inlet, and which are both located inside the instrument housing. To allow IMRs, the only required modification to the instrument hardware was the removal of the gas connection between the trap TWIG’s solenoid valve and its vacuum chamber compartment gas inlet. This gas connection was replaced with a modified connection containing a 1/8-inch Swagelok tee, which allows reagent gas to be mixed in with the Ar gas (see the SI for additional details). Page 4 of 20 ACS Paragon Plus Environment

Page 4 of 20

Page 5 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The components used for preparing reagent vapors for IMRs are shown on the lefthand-side of the mixing tee in Figure 1. A KF16 cross is used to connect all components to a common compartment containing the reagent vapor and liquid. A 100 mL glass media bottle, filled with ~20 mL of reagent liquid, is connected to the bottom of the KF16 cross through a KF16–KF25 reducing union. A centering O-ring is sandwiched between the glass bottle and the reducing union, and is held in place using the bottle’s threaded cap, which has a hole drilled through it. The reagent bottle is immersed in a thermal equilibration bath containing 2 L of room-temperature water. A Pirani gauge is connected to the top of the KF16 cross, which is used to monitor the relative pressure in the reagent compartment, and to ensure that liquid–vapor equilibrium has been reached. A gate valve is placed between the reagent compartment and a roughing pump, which is regulated using a foreline angle valve; the pump serves to degas the reagent liquid, and to remove ambient air from the reagent compartment. To protect the roughing pump from reagent gases, two 78 C cold traps (using acetone/dry ice mixtures) are connected in series between the pump and the gate valve. A leak valve (SSSS6MM series from Swagelok) and ball valve are placed between the KF16 cross and the mixing tee, to control the flow of reagent gas into the trap TWIG vacuum chamber compartment. General workflow for IMRs For experiments involving Leu-enkephalin, pneumatically-assisted ESI was used to generate singly-protonated ions, and typical capillary voltages were ~2 kV. For all other analytes, nano-ESI was used. For this, solutions were sprayed from borosilicate glass capillaries with inner diameters of 0.78 mm and outer diameters of 1.00 mm, which were pulled to an opening size of ~5 μm using a P-1000 micropipette puller (Sutter Instrument Co., Novato, CA, USA). A platinum wire was inserted through the back end of the capillaries to act as an ESI electrode. Typical capillary voltages were ~1 kV. Reference ESI mass spectra were recorded by returning the instrument to normal operating conditions; all valves connected to the reagent compartment were closed, and the pressure reading in the trap was recorded, along with all gas flow parameters (Ar, He, N2). Instrument voltages were optimized using the instrument’s MassLynx software (Waters) so as to give satisfactory signal intensity for the analyte, and a reference mass spectrum was recorded. For IMRs, ~20 mL of liquid reagent was poured into a clean, 100 mL media bottle. The bottle was connected to the reagent compartment and then degassed and pumped for several seconds so as to achieve a stable reading on the reagent compartment Pirani gauge. Page 5 of 20 ACS Paragon Plus Environment


The gate valve was closed, and the reagent compartment was allowed to re-equilibrate for several minutes. The ball valve was opened to evacuate the section of the gas line between the ball valve and the leak valve. The leak valve was gradually opened until evidence for the desired IMR was observable in the mass spectrum. The trap pressure reading was allowed to stabilize and then recorded again. Finally, a mass spectrum for the target IMR was measured. The mass spectra shown here are typically the average of 100–300 scans, recorded in resolution mode with IMS enabled, with a scanning range of 100–3000 m/z, and with a scan time of 1 second. Under these conditions, the average exposure time of ions to reagent gases is approximately 11 ms, unless otherwise stated. We note that the partial pressure of reagent gas in the trap TWIG is defined here as the difference in the trap TWIG pressure reading with and without reagent being leaked into the trap (Ptrap). However, the pressure gauge for the trap TWIG is located on the opposite end of the vacuum chamber compartment relative to the gas inlet; thus, the actual pressures in the trap may be higher than indicated here. Furthermore, the gas pressure readings reported here do not take into account differences in effusion rates and detection efficiencies between the different gases. We note that the effusion of reagent gases from the trap TWIG to the adjacent TWIG compartments is likely negligible in the experiments presented here. This is firstly because the pressure in the IMS TWIG (~2.5 mbar) is much higher than in the trap TWIG ( MeOD > EtOD  D2O. This trend agrees with the results of Mistarz et al.,49 who carried out HDX in the cone exit region of a similar instrument using different reagent gases, and can be rationalized in terms of the differences in proton affinities (and the different reaction mechanisms possible) for the different reagents.12,13 We next examined ion–molecule proton transfer reactions between ubiquitin cations and three bases (water, ammonia, and piperidine), and also those between ds26 DNA anions and three acids (water, acetic acid, and formic acid). Figure 3a compares positive ESI mass spectra of ubiquitin from native solution conditions without (grey trace) and with (red trace) piperidine base being leaked into the trap, whereas Figure 3b compares negative ESI mass spectra of the DNA sequence ds26 sprayed from pure water without (grey trace) and with (dark red) formic acid present in the trap. In both cases, the addition of proton transfer reagent to the trap results in a marked decrease in analyte charge state, suggesting the occurrence of proton transfer reactions. The products of proton transfer reactions always possess fewer charges than the precursor ion(s). We note that no higher charge state was observed when an acid was reacted with positive ions or a base with negative ions. We note further that no charge state shift was observed when operating at elevated Ar pressures with



the leak valve closed. The pressure-dependence profiles using three different proton transfer reagents for both ubiquitin (Figure 3c) and ds26 (Figure 3d) show almost exclusively a monotonic decrease in charge state with increasing reagent pressure. As may be expected, deprotonation of ubiquitin cations is most efficient with the strongest base tested (piperidine), whereas protonation of ds26 anions is most efficient with the strongest acid (formic acid). Structural characterization of IMR product ions by ion mobility spectrometry To leverage both the tandem MS and ion mobility capabilities of the instrument, we have extended these proton transfer reaction studies to those on mass-selected ubiquitin cations, which were subsequently structurally characterized by IMS. To produce the higher charge states of ubiquitin, ESI was done from denaturing solution conditions (Figure 4a). The 10+ through 5+ charge states were individually mass-selected and exposed to water vapor during their accumulation period prior to IMS-MS analysis. This is depicted in Figure 4b for the 10+ charge state, for which the 9+ through 5+ charge states are observed upon exposure to water vapor. By adjusting pressure of water vapor in the trap, vastly different product ion charge state distributions could be obtained (see Figure S3). Here, for ease of comparison of the different precursor charge states, the pressure of water vapor was maintained at 4.9102 mbar, so as to ensure comparable levels of collisional activation and/or cooling of ions during their accumulation in the trap. Ion mobility arrival time distributions for 6+ and 5+ ions produced from different precursor charge states are compared in Figures 4c,d. Broadly speaking, for both the 6+ and 5+ ions, the arrival time distributions exhibit a shift towards shorter average arrival times with decreasing precursor charge state. For example, 6+ and 5+ ions produced from the more highly charged precursors (9+,10+) occupy similar conformational ensembles that are relatively expanded (with longer arrival times). In contrast, 6+ and 5+ ions produced from precursor ions bearing fewer charges progressively occupy more compact conformers (with shorter arrival times). These findings, which are consistent with previous results,31–33,51 are in accordance with the view that the conformations adopted by a charged, gaseous protein are not solely determined by its charge state. Other factors, including the degree of collisional activation, the timescales for structural rearrangement in the gas phase, and (as is most relevant presently) the progenitor ion’s charge state and conformation can all have a significant effect. The simple modifications described herein enable investigations into these facets of protein folding in the absence of solvent.


Page 8 of 20

Page 9 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Collision-induced dissociation of mobility-separated IMR product ions To extend the capabilities of the setup, we showcase results from experiments in which a mass-selected ion is subjected to IMRs to produce a variety of second-generation product ions (covalent adducts), which are subsequently separated by IMS, and finally subjected to CID in the transfer cell prior to MS analysis. The reaction of triply deprotonated hexathymine (T63) with trimethylsilyl chloride (TMSCl) was chosen as a test system, as TMSCl is known to form covalent TMS adducts when reacted with the phosphates of DNA anions, either through nucleophilic substitution (leading to the loss of Cl) or addition–elimination (leading to the loss of HCl), depending on the charge state of the DNA.52 Comparison of mass spectra produced by isolation of T63 with and without TMSCl present in the trap (Figure 5a,b) shows the efficient production four second-generation ions of the form [T6+nTMS]2 (n = 0, 1, 2, 3) from the precursor T63. These product ions are likely formed by proton transfer, nucleophilic substitution, and sequential addition–elimination reactions, respectively. Figure 5c shows the ion mobility separation of the four second-generation product ions (shown in Figure 5b) followed by CID in the transfer cell. Comparison of the arrival time distributions of the four second-generation ions (Figure 5c, left panel) reveals that the four second-generation ions are sufficiently well-resolved by IMS so that their respective CID spectra can be extracted from the data by integration of mass spectra over their respective arrival time intervals. This procedure is depicted in Figure 5c as the shaded regions with different colors, and the resultant CID spectra of individual second-generation ions are shown in Figure 5d–g. Fragment ions are annotated according to the nomenclature proposed by McLuckey et al.53 Chiefly, all CID spectra show the full gamut of w-type sequence ions, which result from cleavage of 3' C–O bonds between the sugar moieties and their adjacent phosphate groups. Furthermore, the number of TMS adducts present on the sequence ions (or the distribution thereof) depends strongly on the number of TMS present on the individual second-generation (pseudo-precursor) ion. For instance, the CID spectrum of [T6+3TMS]2 (Figure 5g) features w5 ions containing primarily three TMS adducts, consistent with the view that TMSCl reacts exclusively with the phosphate groups of the DNA. In a similar vein, the smaller w-type fragment ions produced from [T6+3TMS]2 can contain two, one, or no TMS adducts. While an exhaustive discussion of the spectral features present in the CID spectra is beyond the scope of this work, these results do show clearly that mixtures of IMR product ions can be further structurally characterized individually via post-mobility



fragmentation; as in the present case, this remains true even when covalent modification results in a mixture of product ions differing in mass by only a few percent. Expanding the timescales for ion–molecule reactions For the setup described here, the exposure time of ions to reagent gases is determined by the residence time of ions in the trap TWIG, which under normal operating conditions lie in the range 0.1–50 ms. Exposure times can be kept to a minimum when operating the instrument in TOF mode, as ions are simply transmitted through the trap TWIG under these conditions, and their exposure time to reagent vapors is limited by the range of usable wave velocities. Typical wave velocities lie on the range 10–2000 m/s, which, if one neglects ion roll-over effects, correspond to exposure times in the range 0.1–10 ms. Slightly longer ion exposure times can be achieved when operating the instrument in ion mobility mode. In this configuration, ions are accumulated in the trap TWIG for 8–100 ms prior to being pulsed into the IMS TWIG. This results in average exposure times on the range 4–50 ms. We note that the ion accumulation time is determined by the pusher period, which is in turn linked to the scanning range specified. However, it should be noted that, in ion mobility mode, not all ions are exposed to reagent vapors for the same amount of time. Furthermore, the upper limit of this time range is still relatively low as compared to those accessible with ion trap instruments. In general, for some applications, it may be desirable to have longer and controllable reaction times, for example if the observed rate of a reaction is too low, or if IMR kinetics experiments are of interest. To access longer reaction timescales for IMRs, time-programmable sequences of DC potentials were applied to the entrance and exit electrodes of the trap TWIG assembly. This was done using scripts which were piloted using Waters Research Enabled Software (WREnS).54,55 Briefly, this allows the instrument to be cycled between four states of programmable duration over the course of the experimental sequence: (i) ion accumulation, (ii) ion storage, (iii) ion extraction and analysis, (iv) trap clearing. The ion storage period defines the interaction time of ions with reagent vapor. As a model system, we considered gas-phase HDX of singly protonated Leuenkephalin with D2O (see the SI, Figures S4 and S5 for additional HDX kinetics experiments on ubiquitin cations). A recent experimental and computational study by Chen et al.56 on singly-protonated Leu-enkephalin has shown that six out of nine of its labile hydrogen atoms exchange upon exposure to D2O on the timescale of 10 s (using D2O pressures of 6.7103 mbar). However, here, under conditions of normal instrument operation (with reaction times Page 10 of 20 ACS Paragon Plus Environment

Page 10 of 20

Page 11 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of 11 ms), deuterium uptake levels were not found to exceed ~2 over the range of usable pressures (Figure 2b). Figure 6 shows the deuterium uptake of Leu-enkephalin as a function of ion storage time in the trap TWIG, along with a representative mass spectrum measured with a reaction time of 8 seconds. The deuterium uptake approaches six exchanges with longer ion storage times, suggesting the successful exchange of its six most labile hydrogen atoms. This simple experiment demonstrates that the ion storage period rivals those that can be accessed using ion traps. Discussion and Conclusion We have presented a simple setup for performing ion–molecule reactions in the trapping region of a hybrid Q-IM-TOF instrument using vapors of various liquid reagents. The main advantage of the setup is its versatility, as many reagents can be used. Ten liquid reagents were tested here, with vapor pressures covering a wide range (5–500 mbar). However, comparable ranges of reagent pressures in the trap, and reproducible reaction extents, were attainable by adjustment of the leak valve (see Figure S6 for an example of the day-to-day variability). We note that the total ion current was preserved for reagent pressures below 5102 mbar, which compares favorably with the reagent pressures typically used when conducting IMRs on linear ion traps. The current limitation appears to be the reagent vapor pressure, the lower limit for which is ~5 mbar (at 298 K). By heating the vial and the line to the ion trap of the mass spectrometer, it could be possible to extend the list of reagents. Caution should be exercised to ensure chemical compatibility of the reagents with standard vacuum components. In particular, chemical compatibilities with the O-rings used in the reagent compartment should be noted. Over the course of several months of experiments, we have not noticed any detrimental effects on instrument performance, nor any chemical erosion downstream of the leak valve, probably because very little reagent is used up in these experiments. We judge that the rate of reagent consumption is exceptionally low (

A modified travelling wave ion mobility mass spectrometer as a

Recommend Documents