A Miniaturized Linear Wire Ion Trap with Electron Ionization and Single Photon Ionization Sources
- 650 Downloads
A linear wire ion trap (LWIT) with both electron ionization (EI) and single photon ionization (SPI) sources was built. The SPI was provided by a vacuum ultraviolet (VUV) lamp with the ability to softly ionize organic compounds. The VUV lamp was driven by a pulse amplifier, which was controlled by a pulse generator, to avoid the detection of photons during ion detection. Sample gas was introduced through a leak valve, and the pressure in the system is shown to affect the signal-to-noise ratio and resolving power. Under optimized conditions, the limit of detection (LOD) for benzene was 80 ppbv using SPI, better than the LOD using EI (137 ppbv). System performance was demonstrated by distinguishing compounds in different classes from gasoline.
KeywordsLinear ion trap Wire electrodes Vacuum UV ionization Miniaturized ion trap
Portable mass spectrometry has wide applications, including public safety, environmental protection, and industrial process monitoring [1, 2, 3, 4, 5, 6, 7, 8]. Miniaturization of a conventional mass spectrometer includes reducing the size of the mass analyzer and coupling it to one or more appropriate ion sources. Nearly all types of mass analyzers have been miniaturized , including a magnetic sector mass analyzer , quadrupole mass filters [10, 11], quadrupole ion traps [12, 13, 14, 15, 16], a Fourier-transform ion cyclotron resonance mass analyzer , and time-of-flight (TOF) mass analyzers [18, 19]. Among miniature mass analyzers, quadrupole ion traps are distinguished because of their compact size, relatively high working pressure and the unique capability to perform multistage tandem mass analysis (MSn). The linear ion trap (LIT) has attracted recent attention because of higher ion trapping efficiency from an external ion source, increased ion storage volume, and good resolving power . In our laboratory we have built a miniaturized linear wire ion trap (LWIT) and found it has additional advantages for portable mass analysis because of high ion transmission rates, low capacitance, and low weight .
Ion sources coupled to miniature mass analyzers include electron ionization (EI) [17, 22], discharge ionization , electrospray ionization (ESI) [24, 25, 26], single photon ionization (SPI) [27, 28, 29, 30, 31, 32, 33], and desorption electrospray ionization (DESI) [34, 35]. Because samples are usually mixtures in a real test environment, the ability to distinguish single compounds from a mixture is very important when no separation devices, such as a gas or liquid chromatograph, are coupled to the ion trap. Single photon ionization (SPI), as a soft ionization technique, has the ability to detect organic compounds in mixtures with good sensitivity [28, 29]. In addition, a vacuum ultraviolet (VUV) lamp is also a suitable photon source for portable mass spectrometry because of its small size, low power consumption, variety of choices for photon energy, and strong photon flux.
VUV lamp ion sources have been coupled to miniature TOF mass spectrometers and proved to be a good combination [27, 28, 29, 36, 37, 38, 39]. However, the high vacuum and resolution requirements in a TOF analyzer make the further size reductions very difficult in such systems. In a recent study, the combination of a VUV lamp and a commercial ion trap showed good sensitivity . In addition, the MS/MS ability in ion traps can provide structure information and, thus, is very helpful in identifying compounds. However, challenges were encountered when coupling the VUV lamp to the miniature ion trap. First, the photons emitted from the VUV lamp can generate a strong signal on the detector through the photoelectric effect, which increases background noise and even causes the detector to saturate. Therefore, it significantly reduces the lifetime of the detector. Second, to avoid space charge effects inside the ion trap, a control system is needed to control the amount of ions inside the ion trap. A design that simply radiates VUV light into the ion trap is unable to meet this requirement.
Building on our previous work with the LWIT , here we demonstrate both SPI and EI ion sources for the miniature LWIT. In the SPI ion source, a VUV lamp is pulsed ON and OFF to control the ionization and avoid spurious photon signals during m/z analysis. The SPI ion source is set outside the ion trap and the transmission rate can be controlled by a potential on the end electrode of the trap. The limit of detection (LOD) for benzene is shown to be 80 ppbv under routine working conditions. Instrument performance is further demonstrated by analysis of compounds present in gasoline.
Dual Ion Source Design
A gated electron gun was set on the right of the LWIT. The electrons generated by the electron gun were pushed into the LWIT under a potential of –60 V. The electron gate alternated between 22 V (allowing electrons through) and –100 V (closed). Neutral sample molecules were ionized inside the LWIT and were confined by two end electrodes, each at 12 V. A magnetic ring was set around the electron gun and used to confine the trajectory of electrons. Previous work shows that the magnetic ring can improve the electron ionization efficiency by up to two orders of magnitude in the photo ionization source . In our setup, simulations show that the magnetic ring can increase the transmission efficiency from 6.8% to 10.6%. In addition, the helix path of the electrons during weak rf field significantly increases their path and, thus, increases the ionization efficiency. The ions are less affected by the magnetic field due to their relatively heavy mass. A study shows that a magnetic field lower than 0.1 T does not significantly perturb the ion’s frequency of motion . The magnetic field in our setup (0.07 T close to the ring and 0.015 T in the position 5 mm away) is far below 0.1 T, so the impact of magnetic field on the ions is negligible.
The VUV lamp used to provide soft ionization (PKR 106; Heraeous Noblelight GmbH, Hanau, Germany) was set on the left of the LWIT. The energy of the photons is 10.6 eV (117 nm) with FWHM of about 0.1 eV. A sample inlet and a helium inlet were drilled in the middle electrode on which a DC 20 V potential was applied. A repel electrode at 23 V pushed ions into the ion trap. A tune electrode was used to focus ions into the hole on the wire support plate. The optimized potential on this electrode is around 15 V. The VUV lamp was powered by a 13.5 MHz rf power supply (C210; Heraeous Noblelight GmbH), which needed a 25 V DC potential as input. To avoid the strong noise generated by the VUV light on the detector, a pulse amplifier (Model: GRX-1.5 K-L; Directed Energy Inc., Fort Collins, Colorado, USA), controlled by a pulse generator, served as the power supply for the VUV lamp. VUV lamp ignition takes about 0.02 s, setting a lower bound for the repetition rate on a scan. A copper foil was used to shield the lamp power supply to avoid coupling to the detector signal (not shown).
A model in SIMION 8.1 (Scientific Instrument Services, Inc., Ringoes, NJ, USA) was built for verifying the configuration, as shown in Figure 1. In the study of the electron transmission rate, the time of birth for electrons is set as a unified distribution from 0 to 650 μs (1000 periods of rf power). The rf is 1.53 MHz and the amplitude is 600 Vp-p. The birth position of electrons is set as the same position as the hot filament. The magnetic ring is also built into the model and the magnetic field is set according to a measured value from the magnetic ring (0.07 T close to the ring and 0.015 T in the position 5 mm away). Under such conditions, the simulated electron transmission rate through the electron gate is 6.8% without the magnetic ring and 10.6% with the magnetic ring.
The trapping efficiency was also evaluated under the same rf power supply. The birth position of the ions is a cylinder under the VUV lamp with diameter of 8 mm and length of 10 mm with a random distribution. The initial ions are set as m/z = 100 and 10,000 ions are used to get precise results. The hard sphere collision model is introduced in the simulation. The buffer gas is set as helium and the pressure is 0.35 Pa, the same as the experimental condition. The electrodes potential is based on the optimized potential in the experiment. Under such conditions, the simulation shows that the trapping rate for ions from the ion source into the trap is estimated to be 6%.
The vacuum was supported by a turbo pumping station (HiCube; Pfeiffer, Nashua, NH, USA) with a pumping speed of 35 L/s for N2. The sample gas was leaked into the vacuum through a leak valve (Granville-Phillips Co., Boulder, CO, USA) to reach an indicated pressure of 0.002 Pa. The gas flow rate was estimated to be 0.042 mL/min at atmosphere pressure. After the pressure reached a stable status, helium was introduced as the buffer gas and the pressure of the chamber kept at 0.35 Pa, as measured using a calibrated Bayert-Alpert-type ionization gauge (FRG-720; Agilent, Santa Clara, CA, USA).
Electronics and Time Sequence
A 1.53 MHz rf signal from a power supply (Ardara Quadrupole Power Supply; ArdaraTech, Ardara, PA, USA) was applied to the wires of two opposite sides of the LWIT (wires in the Y direction). A positive DC voltage of up to 12 V was applied to the axial confinement electrodes to provide a DC trapping potential well for positive ions along the Z axis. The static potential applied on the filament was –60 V. Trapped ions were mass selectively ejected by scanning the rf amplitude at a rate of 3830 Th/s. An AC signal at a frequency of 520 kHz, generated by a waveform generator (DS345; Stanford Research Systems, Sunnyvale, CA, USA) was applied on six X wires on one side to provide the quasi-dipolar field that induced ion ejection at q = 0.78 during the rf scan. The ions ejected from the LWIT were detected using an electron multiplier (DeTech XP-2200; Palmer, MA, USA) with estimated gain of ~106. The signal was first amplified using a preamplifier at a gain of 107 (Keithley 427 current amplifier; Beaverton, OR, USA), and then acquired using a digital oscilloscope (Teledyne LeCroy Wavesurfer 3024; Chestnut Ridge, NY, USA) at a sampling rate of 500 k samples/s.
Three sample gases were used in this study. Toluene and deuterated toluene (toluene-D8) were used to demonstrate the differences between SPI and EI ion sources. The gas sample was prepared by ejecting 2.85 μL toluene and 2.85 μL toluene-D8 with a 5 μL precise syringe into a 6 L gas bag filled with pure nitrogen, which resulted in a mixture of toluene and toluene-D8 at 115 ppmv. The lower atmospheric pressure (about 86450 Pa) in Provo, UT was considered in the calculation. After 2 h to completely evaporate the liquid samples, a 600 mL gas sample was transferred into a new gas bag with 5.4 L pure nitrogen. In this way, a concentration of 12 ppmv of toluene and toluene-D8 was prepared. A 1.8 ppmv benzene gas sample was prepared in the same way with different values (2.38 μL liquid benzene to generate a concentration of 115 ppmv in a 6 L gas bag and 94 mL sample gas transferred to a 6 L bag filled with 5.9 L nitrogen) and used for evaluating the impact of sample pressure on signal-to-noise ratio and resolution. A gasoline sample gas was produced by evaporating 20 μL gasoline (85-octane regular; Chevron, Provo, UT, USA) into a 6 L gas bag filled with pure nitrogen. This gas sample was used to demonstrate the analytical ability of the SPI ion source.
Results and Discussion
The sample gas pressure has a significant impact on the resolution and signal-to-noise ratio. Under higher sample gas pressure, more molecules are introduced into the ion trap and, thus, more ions are generated, which increases the detected signal. However, the higher sample gas pressure has a negative impact on the resolution in three ways. First, the collisions between the introduced gas and ions reduce the resolution. Helium was usually used as the background gas because it can smoothly reduce the energy of ions and cool them to the center of the linear ion trap, whereas the introduced gas molecules are usually heavier than helium and warm up the cooled ions. Second, during the excitation process, the collisions between the high molecular weight molecules and ions significantly deviate the trajectories of ions and, therefore, reduce the resolution. In addition, the resolution is reduced due to space charge effects if the ion intensity inside the trap is too high.
Mass Spectrum of Gasoline
Two ion sources, SPI with soft ionization and EI with hard ionization, were built for the LWIT and tested. The soft ionization in SPI provides information about molecular ions whereas the hard ionization in EI can provide structural information about analytes. These features were demonstrated in the analysis of gasoline, in which compounds from different classes were easily identified. Under optimized conditions, the limit of detection for benzene using SPI was 80 ppbv, and under the routine working conditions, as extrapolated from the 1.8 ppmv data, better than that using EI (137 ppbv). The MSn ability was not available in our current setup, but it could significantly enhance the ability of identifying unknown compounds, especially in SPI mode. The LWIT with dual ion sources is a promising candidate for portable mass spectrometry because of its compact design, good sensitivity, and the ability to detect organic mixtures.
The authors acknowledge that this work was funded by the National Science Foundation (USA) Chemical Measurement and Imaging Program.
- 17.Vilkov, A.N., Gamage, C.M., Misharin, A.S., Doroshenko, V.M., Tolmachev, D.A., Tarasova, I.A., Kharybin, O.N., Novoselov, K.P., Gorshkov, M.V.: Atmospheric pressure ionization permanent magnet Fourier transform ion cyclotron resonance mass spectrometry. J. Am. Soc. Mass Spectrom. 18, 1552–1558 (2007)CrossRefGoogle Scholar
- 18.Bryden, W.A., Benson, R.C., Ecelberger, S.A., Phillips, T.E., Cotter, R.J., Fenselau, C.: The tiny-TOF mass-spectrometer for chemical and biological sensing. J. Hopkins APL Tech. Dig. 16, 296–310 (1995)Google Scholar
- 21.Wu, Q., Li, A., Tian, Y., Zare, R.N., Austin, D.E.: Miniaturized linear wire ion trap mass analyzer. Anal. Chem. 88, 7800–7806 (2016)Google Scholar
- 30.Huang, Y.G., Li, J.X., Tang, B., Zhu, L.P., Hou, K.Y., Li, H.Y.: Development of a portable single photon ionization-photoelectron ionization time-of-flight mass spectrometer. Int. J. Anal. Chem., article 581696 (2015). doi: 10.1155/2015/581696
- 32.Muhlberger, F., Saraji-Bozorgzad, M., Gonin, M., Fuhrer, K., Zimmermann, R.: Compact ultrafast orthogonal acceleration time-of-flight mass spectrometer for on-line gas analysis by electron impact ionization and soft single photon ionization using an electron beam pumped rare gas excimer lamp as VUV-light source. Anal. Chem. 79, 8118–8124 (2007)CrossRefGoogle Scholar
- 33.Muhlberger, F., Zimmermann, R., Kettrup, A.: A mobile mass spectrometer for comprehensive on-line analysis of trace and bulk components of complex gas mixtures: parallel application of the laser-based ionization methods VUV single-photon ionization, resonant multiphoton ionization, and laser-induced electron impact ionization. Anal. Chem. 73, 3590–3604 (2001)CrossRefGoogle Scholar
- 36.Wu, Q.H., Hua, L., Hou, K.Y., Cui, H.P., Chen, P., Zhao, W.D., Wang, W.G., Li, J.H., Li, H.Y.: Characterization and applications of combined single photon ionization and photoelectron ionization source. Chin. J. Anal. Chem. 39, 1465–1469 (2011)Google Scholar