Selective Extraction of a Monoisotopic Ion While Keeping the Other Ions in Flight on a Multi-Turn Time-of-Flight Mass Spectrometer

Using a multi-turn time-of-flight (TOF) mass spectrometer, we have extracted a single xenon isotope ion, 129Xe+, from its orbit at given a lap number without disturbing the rest of isotopes. After detecting the 129Xe+ at 20 laps, the rest of the xenon isotope spectrum was obtained at 30 laps, which generated a TOF spectrum where the TOF difference between 129Xe+ and 130Xe+ was 87.4 μs while 130Xe+ and 131Xe+ were 1.03 μs. The time distance between 129Xe+ and other isotopes can be set by any lap difference that is a factor of 8.7 μs, which depends on the acceleration voltage and the mass of the ion. Method accuracy was verified by comparing the isotopic abundance ratio of the xenon sample after withdrawing one of the ions from the isotope cluster to the abundance ratio obtained from the conventional method. The TOF stability was also evaluated at various lap numbers between 10 to 230.

INTRODUCTION e mass accuracy of the miniaturized multi-turn timeof-ight (TOF) mass spectrometer using a unique mass assignment algorithm has been reported. 1) e design of the multi-turn TOF mass analyzer provides an in nite ight path by keeping ions in an in nite orbit using perfect-ion focusing. 2) Ions can be ejected out from orbit and introduced into the detector at a given timing. Unlike re ectron TOF mass spectrometers, mass assignment for the multiturn instrument is highly predictable due to the linear "TOF equation," 3) which is advantageous for automatically ejecting a speci c ion out of orbit at a desired timing by a simple analyzer control system. Using the determined "TOF equation," the location of a given ion in the analyzer at a known time can be calculated quickly and precisely. erefore, it is possible to eject and detect an ion at a given lap number and then introduce the remaining ions into the detector some laps later. Such a control protocol can generate a series of TOF di erence between given ions, which is a powerful tool to investigate the microchannel plate gain drop issue a er intense ion ux detection 4) by using ion. Furthermore, we can hold an ion cluster keep ying and extract one by one at a given timing that may be a good tool for the rst stage analyzer on the tandem TOF mass spectrometry (TOF/TOF).
Xenon is a good sample for a spectral pro le testing since it has a unique isotope distribution. As shown in Table 1 EXPERIMENTAL Instrumentation e miniaturized multi-turn time-of-ight (TOF) mass spectrometer 5) in TOF-UHV (MSI.TOKYO, Inc., Tokyo, Japan) was used with in-house modi cations reported previously. 1,6) Ions were detected by MIGHTION (Hamamatsu Photonics K.K., Hamamatsu, Japan), 7) which is a microchannel plate (MCP) combined with an avalanche diode. e detector signal was passed through a model C11184 preampli er (Hamamatsu Photonics K.K., Hamamatsu, Japan), followed by waveform acquisition using an Acqiris U5303A 1 GS s −1 high-speed digitizer (Acqiris, Geneva, Switzerland). e detector operation conditions are 350 V for avalanche diode voltage, −4.96 kV for MCP-In potential, and 560 V for between MCP-In and MCP-Out. Xenon gas (Takachiho Chemical Industrial Co., Ltd.) was introduced into the electron ionization (EI) chamber using a 1 m length of 0.1 mm inner diameter PEEK (Polyether Ether Ketone) capillary tubing. e vacuum condition in the ionization chamber during sample measurement was 3.4×10 −3 Pa ( e pressure when the sample introduction valve was closed was 2.2×10 −4 Pa). Ionization energy was set to 30 eV and lament current to 3200 mA for xenon pro le spectrum monitoring, or 2880 mA for xenon ion counting experiments. Figure 1 illustrates the relationship between analyzer timings and the detector signal. e conventional timing for acquiring a xenon spectrum is illustrated in the bottom of Fig. 1, where the in TOF is operating in multi-turn mode, and the ejection sector at the halfway point of the orbital path is closed for the sample orbital period until the xenon ion cluster arrives a er 30 laps.

Analyzer timings
One additional functionality was added to the control system, which allows additional, precise ejection sector timing. Figure 1 top illustrates selecting and ejecting 129 Xe + at 20 laps, while the rest of the control timings remain the same. In this case, we have a xenon TOF spectrum at 30 laps with the absence of 129 Xe + . e ejection sector has consisted of the high-voltage MOSFET push-pull mode switch, 8) which both rise and fall time is about 100 ns.

Procedure
A er setting analyzer conditions and experimental system equilibrations were set, mass assignment needs to be veri ed using data from 20 and 30 laps of 132 Xe + . 1) e data acquisition so ware determines the timings for the ejection sector for a given chemical formula and lap number by using the "TOF equation" determined from the mass assignment process. To obtain the TOF spectrum of xenon at 20 laps, two parameters were set in the so ware: the formula for 131 Xe + and the number of laps to 20, which produced the spectrum shown in Fig. 2 bottom. At this point, the so ware adjusts the timing for 131 Xe + to be the center of the region of  interest for the xenon isotope cluster. A er this spectrum is visible on the real-time monitor, the duration parameter can be narrowed down for the ejection sector timing until single 129 Xe + peak is obtained without any other ions visible on the spectral monitor screen as shown in Fig. 2 top. e timing for the ejection sector was determined to be open at 176.03 µs for 0.7 µs, which means that 129 Xe + is always passing through the ejection sector during this time frame as long as "TOF equation" does not change. A second ejection sector timing was added to the so ware, which can either be enabled or disabled during each acquisition protocol. Using this modi cation, 129 Xe + at 20 laps can be monitored while acquiring the remaining xenon spectrum at a minimum of 21 laps or higher by setting the above timing values into the second ejection sector parameter and enabled. Table 2 shows the abundance ratio of xenon isotopes acquired from 30 laps of xenon. e count rate for 132 Xe + was adjusted to 36.3%, where 36.3 132 Xe + ions were counted of 100 ion push triggers. "Protocol 0" (P0) is acquiring a xenon spectrum at 30 laps with our normal monitoring setup; "protocol 1" (P1) is also acquiring a xenon spectrum at 30 laps, but 129 Xe + was ejected and monitored at 20 laps. ese two protocols were alternated for every ion push trigger event. As shown in the Table 2, the isotope abundance ratio for 130 Xe + and 131 Xe + against 132 Xe + showed excellent agreement between P0 and P1. e obtained counts for 20 laps of 129 Xe + is lower than 30 laps on Table 2, which is instrument tuning dependent, however, the obtained count ratios within the same lap numbers were stable. We can conclude that the 129 Xe + was successfully ejected from a cluster of xenon isotopes without a ecting the abundance ratio, i.e., no ion loss was observed by removing it from a sample cluster. Figure 3 shows the corresponding spectra for P0 and P1. e spectra presented in this paper uses ejectionsector open duration time of 700 ns, however, spectral pro le change did not observe until it narrows down to 640 ns (data were not shown). erefore, any ion peak, where the TOF away from at least 640 ns from adjacent ion peaks, can be withdrawing without disturbing other ions by using this method. Figure 4 shows the mass errors for 132 Xe + at various numbers of laps. e standard deviation of mass error was 3.32 mDa, and the mass resolving power at 20 and 200 laps were 8,700 and 40,700, respectively. e standard deviation of time errors compared to the TOF computed using the "TOF equation" for 132 Xe + in the range of 94.0 to 2016.7 µs was approximately 8.4 ns, which is two orders of magnitude lower than the duration of the ejection sector was open. Both m/z and TOF for 132 Xe + stayed the same between protocols, whether the 129 Xe + was withdrawn at 20 laps. is demonstrates how closely the ion ight path follows the "TOF equation," and that ion ight is not disturbed by the withdrawal of a monoisotopic ion during TOF separation in the analyzer.

Capability of the method for the TOF/TOF application
e results shown here also suggests that this analyzer is an excellent candidate as a rst mass analyzer in a TOF/TOF instrument 9) with the capability of selecting multiple precursor ions from the single ion push cycle. In the case of ultra high-performance liquid chromatography (UHPLC) analysis of complex sample matrices, such as protein and drug metabolites identi cation, many single chromatographic peaks consist of multiple compounds. e peak width on a UHPLC chromatogram, in general, is approximately one second or less, and about 10 sample points per peak are required to determine chromatographic peak area e ectively. So it is challenging to apply datadependent TOF/TOF for multiple components in a single peak, which has led to investigation of several non-tandem TOF approaches 10) . However, if the analytes are more than one dalton m/z di erence in an ion trigger cycle, it is possible to select one precursor ion to introduce to the second mass analyzer (MS2) and keep the other ions in the multiturn analyzer (MS1). e orbital period is mass-dependent, corresponding to the TOF for the length of the gure-eight orbit (0.662 13 m), e.g., 13.23 µs and 13.45 µs for m/z 300 and 310 ions, respectively. Assuming ions of m/z 300 and 310 co-exist in a spectrum, m/z 300 could be ejected into MS2 at an early lap number, and then wait for an additional 134.5 µs (10 more laps) to introduce m/z 310 into MS2.

CONCLUSION
New, unique ejection sector control capabilities were added to the multi-turn TOF mass spectrometer to extract 129 Xe + from a xenon isotope cluster. Setting a 700 ns duration for the ejection-sector shows excellent results for monitoring the 129 Xe + peak at 20 laps while the rest of the isotopes remain in orbit for subsequent measurement at 30 laps. e peak abundance ratio obtained from xenon isotopes at 30 laps, both with and without 129 Xe + shows less than 5% error on the pro le spectrum under rich ion ux conditions. e accuracy of the abundance ratio determined by ion counting shows less than 1.55% error for 130 Xe + and less than 0.59% error for 131 and 132 xenon.
We also investigated the m/z errors at various laps from 10 to 230. Since the mass assignment method is a linear equation, mass error is re ecting the TOF uctuations directly. e standard deviation of m/z error for 10 to 230 laps was 3.32 mDa, which corresponds to 8.4 ns on average, which is accurate enough to select an ion by controlling the ejection sector. e m/z accuracy was not a ected by the use of the ejection sector for ion selection. Use of the ejection sector did not a ect either peak intensity or m/z errors, where the m/z error is the TOF conformity to the "TOF equation" on this instrument.