Axial Ion Focusing in a Miniature Linear Ion Trap - ACS Publications

A novel miniature linear ion trap with a total length of 19 mm and a quadrupole rod length of 15 mm has been fabricated to enable ion focusing in the ...
0 downloads 0 Views 243KB Size
Download PDF
Anal. Chem. 2007, 79, 3779-3785

Axial Ion Focusing in a Miniature Linear Ion Trap Gareth S. Dobson* and Christie G. Enke

Department of Chemistry, Clark Hall, University of New Mexico, MSC03 2060, Albuquerque, New Mexico 87131-0001

A novel miniature linear ion trap with a total length of 19 mm and a quadrupole rod length of 15 mm has been fabricated to enable ion focusing in the axial plane (between the end caps). Each end cap includes an inwardly projecting tubular section, which prevents dc fringe fields from penetrating to the center of the miniature linear ion trap and aids in ion extraction. Axial focusing of ion packets to dimensions of less than 1 mm through collisional cooling is predicted and demonstrated in the miniature linear ion trap. Due to this demonstrated collisional cooling, narrow kinetic energy distributions are also illustrated on batch ion extraction as might be useful for ion transfer to enable subsequent mass analysis. Both linear and three-dimensional (3D) ion traps are commonly used for mass/charge measurement as stand-alone mass spectrometers or as non-mass-selective devices in which they provide ion focusing for another mass measurement device, such as a time of flight.1-10 Time-of-flight mass analyzers have a higher mass/ charge measurement precision than ion traps, despite the continuing development and study of 3D ion trap mass measurement techniques.11-14 Ion traps have also been used in conjunction with orthogonal time-of-flight (o-TOF) analyzers, allowing improvement in transfer efficiency and duty cycle.10 Therefore, the development and characterization of ion traps as ion focusing devices is * Corresponding author. E-mail: [email protected]. Tel: (505) 2778684. Fax: (505) 277-2609. (1) Michael, S. M.; Chien, B. M.; Lubman, D. M. Rev. Sci. Instrum. 1992, 63 (10), 4277-84. (2) Michael, S. M.; Chien, B. M.; Lubman, D. M. Anal. Chem. 1993, 65 (19), 2614-20. (3) Chien, B. M.; Michael, S. M.; Lubman, D. M. Anal. Chem. 1993, 65 (14), 1916-24. (4) Marinach, C.; Brunot, A.; Beaugrand, C.; Bolbach, G.; Tabet, J.-C. Int. J. Mass Spectrom. 2002, 213, 45-62. (5) Eizoh, K.; Tanaka, K.; Ding, L.; Smith, A.; Kumashiro, S. 47th ASMS Conference on Mass Spectrometry and Allied Topics, Dallas, TX, 1999. (6) Hashimoto, Y.; Waki, I.; Yoshinari, K.; Shishika, T.; Tend, Y. Rapid Commun. Mass Spectrom. 2005, 19, 221-6. (7) Grebner, T. L.; Neusser, H. J. Int. J. Mass Spectrom. Ion Processes 1994, 137, L1-6. (8) Ji, Q.; Vlasak, P. R.; Davenport, M. R.; Enke, C. G.; Holland, J. F. J. Am. Soc. Mass Spectrom. 1996, 7 (10), 1009-17. (9) Chernushevich, I. V. Eur. Mas. J. Spectrom. 2000, 6 (6), 471-9. (10) Hashimoto, Y.; Hasegawa, H.; Satake, H.; Baba, T.; Waki, I. J. Am. Soc. Mass Spectrom. 2006, 17 (12), 1669-74. (11) Dobson, G.; Murrell, J.; Despeyroux, D.; Wind, F.; Tabet, J.-C. J. Mass Spectrom. 2004, 39 (11), 1295-304. (12) Dobson, G.; Murrell, J.; Despeyroux, D.; Wind, F.; Tabet, J.-C. J. Mass Spectrom. 2005, 40 (6), 714-21. (13) Dobson, G.; Murrell, J.; Despeyroux, D.; Wind, F.; Tabet, J.-C. Rapid Commun. Mass Spectrom. 2003, 17 (14), 1657-64. (14) Murrell, J.; Konn, D. O.; Underwood, N. J.; Despeyroux, D. Int. Mass J. Spectrom. Ion Processes 2003, 227 (2), 223-34. 10.1021/ac0620462 CCC: $37.00 Published on Web 04/13/2007

© 2007 American Chemical Society

important not only for ion trap mass spectrometry but also to enable them to be coupled with other mass analyzing techniques, where they can enhance the overall performance of the mass spectrometer. With an optimized ion trapping voltage applied to the ring electrode of a three-dimensional ion trap, focusing of the ion cloud sizes to ∼1 mm in diameter has been shown.15-17 Ion ejection from 3D ion traps requires ejection of ions through a rapidly changing rf field or stopping this rf field during the ejection of the ions. It is also possible to scan ions out of the 3D ion trap, but monochromatic kinetic energy pulses or beams of ions as required, for example, with a quadrupole ion trap coupled to an o-TOF, are difficult to produce. Such an example was illustrated by Marinach et al.,4 who demonstrated how the qz values influenced the kinetic energy of ions ejected from a 3D ion trap, through the use of a retarding potential applied to a grid situated between a 3D ion trap and a detector. Also, ions introduced into a 3D ion trap must penetrate through an rf field and will be trapped only when appropriate voltages and phase of the rf voltage are applied.18-20 Thus, 3D ion traps are difficult to employ as transfer and focusing devices between ion sources and mass analyzers (called intermediate devices in this study). In contrast with 3D ion traps, linear ion traps allow ion introduction and ejection along the central axis where the rf field is minimal. As the axial potential well is dc in linear ion traps, they are also easier to use as intermediate devices. The use of linear ion traps can also provide mass discrimination in a mass spectrometer. Ions are generally confined in the axial direction in linear ion traps by dc voltages applied to end cap lenses between quadrupole rods.21,22 With collisional cooling, radial and axial packet dimensions are determined by the shapes of the respective potential wells. The shape of the axial well confining these ions is intentionally broad to reduce charge density. In order to achieve 3D focusing in a linear ion trap, a narrower axial well is required while maintaining radial rf confinement. (15) Doroshenko, W. M.; Cotter, R. J. J. Mass Spectrom. 1998, 33, 305-18. (16) Alili, A.; Andre, J.; Vedel, F. Phys. Scr. 1988, T 22, 325-8. (17) Schubert, M.; Siemers, I.; Blatt, R. Appl. Phys. B 1990, 51 (6), 414-7. (18) Quarmby, S. T.; Yost, R. A. Int. Mass. J. Spectrom. 1999, 190/191, 81102. (19) Schwartz, J. C.; Senko, M. W.; Syka, J. E. P. J. Am Soc. Mass Spectrom. 2002, 13, 659-69. (20) Dolnikowski, G. C.; Kristo, M. J.; Enke, C. G.; Watson, J. T. Int. J. Mass Spectrom. Ion Processes 1988, 82, 1-15. (21) Hager, J. W. Rapid Commun. in Mass Spectrom. 2003, 6 (16), 512-26. (22) Ouyang Z.; Wu G.; Song Y.; Li H.; Plass, W. R.; Cooks, R. G. Anal. Chem. 2004, 76, 4595-605.

Analytical Chemistry, Vol. 79, No. 10, May 15, 2007 3779

Figure 1. (a) Illustration of a modified QqTOF instrument containing a miniature linear ion trap preceding a time of flight. (b) Illustration of section removed from the MLIT entrance lens allowing He gas flow into the MLIT. Helium gas flow into the MLIT also occurs through the hole of 1.5-mm diameter.

In this study, a new miniature linear ion trap (MLIT) has been developed. The aim of this paper is to demonstrate and characterize the focusing capabilities of this novel miniature linear ion trap in the axial plane as well as the resulting kinetic energy dispersion that can be produced by batch axial ion ejection (as would be required when the ion trap is used as an ion focusing device before a mass analyzer). End caps of unusual shape are used in the novel MLIT, and the potential well formed by application of dc voltages is characterized. EXPERIMENTAL SECTION Materials. Angiotensin II (Asp-Arg-Val-Tyr-Ile-His-Pro-Phe) was purchased from Sigma (St. Louis, MO) and was prepared as a 1 µM solution of CH3OH/H2O (50:50) with 0.2% acetic acid. Mass Spectrometry. Experiments were carried out using an extensively modified QqTOF (Qstar Instrument, MDS Sciex, ON, Canada) with a novel miniature linear ion trap inserted between a shortened LINAC (MDS Sciex)23 and a grid (Figure 1). The quadrupole preceding the LINAC was used as an ion guide, and the einzel voltage was optimized for maximum ion transmission. Description of the Shortened LINAC and Voltages Applied. The LINAC is a quadrupole with four T rods inserted between the multipole rods.23 The T rods allow the ions to be pushed through the quadrupole with constant acceleration. Two electrodes have been placed between the MLIT and the LINAC, forming an exit lens to the LINAC and an entrance lens to the MLIT. The combination of the exit lens on the LINAC and the T-rods allows a wide dc potential well to be formed and, as a result, ion accumulation. As the LINAC is used as an ion source for the MLIT and is not the focus point of this publication it is not examined further here. A potential of 51 V was applied to the (23) Covey, T.; Stott, B.; Joliffe, C.; Thomson, B. 46th ASMS Conference on Mass Spectrometry and Allied Topics, Orlando, FL, 1998.

3780

Analytical Chemistry, Vol. 79, No. 10, May 15, 2007

T-rods with a 13.7-V dc offset relative to the first element of the einzel lens. The LINAC used was shortened by 3 cm (from ∼20 cm) so that the MLIT could be fitted into the space. The exit lens for the LINAC has a diameter of 34 mm and a hole of 1-mm diameter in the center. MLIT Assembly. The rods forming the quadrupole in the miniature linear ion trap have an interelectrode spacing between opposite rods (2r0) of 8.4 mm, and each rod has a diameter of 9.6 mm. Round cylindrical rods (not hyperbolic) constitute the quadrupole due to their easier fabrication. The entrance and exit tube lenses are 5 mm long with a hole in the center of 1.5-mm diameter. The entrance lens to the MLIT has 2 mm cut from the outside edge of the lens as illustrated in Figure 1b. The removal of part of this outside edge allows a faster decrease in the helium pressure when a helium gas pulse is stopped. The rf voltage applied to the quadrupole rods in the MLIT is the same as that applied to the quadrupole rods on the LINAC. Also, the rf voltage was applied to the quadrupole rods in each experiment so that the desired mass/charge range could be observed in the resulting mass spectra. Application of Buffer Gases: Allowing Ion Cooling. The gas pulse has been optimized with respect to the pressures and the pulse time by ion intensities (and therefore by ion loss). Ion loss occurs typically during the transfer of the ions from the LIT to the MLIT. The helium gas comes from an ultrahigh-purity helium bottle (Trigas, Irving, TX) with a fine regulation valve (BOC Edwards, Crawley, West Sussex, U.K.). Following the fine regulation valve, an Edwards two-stage 1.5 pump (BOC Edwards, Crawley, West Sussex, U.K.) was introduced to decrease the gas pressure. Between this reduced pressure region and the MLIT, a valve (Iota one fabricated by General Valve Corp., Parker Hannifin Corp., Fairfield, NJ) was introduced. This valve was controlled by a dc trigger pulse sent by the home-written Labview program

Figure 2. Timing diagram depicting the application of the dc voltages: to the tube lens entrance and exit lens; to the LINAC exit lens; and the helium pulse during an analytical scan for batch ion extraction.

allowing ∼100 ms between the gas pulses, which is used for ion accumulation in the LINAC. The helium gas pulse was applied 0.5 ms before the focus of the ions in the MLIT for a duration of 3 ms (Figure 2). This allowed the ions to be transferred to the MLIT through the tube lenses while there was an increase in the helium pressure in the tube lenses and the MLIT. The helium pressure decreased after the 3-ms pulse throughout the rest of the focus so that it was minimum during the extraction of the ions from the MLIT (different numbers of collisions between ions and helium atoms would cause wide kinetic energy distributions). The helium gas can escape through the holes in the plastic surrounding the MLIT illustrated in Figure 1, which have been partially covered using Teflon leaving only a hole of ∼1 cm2. Nitrogen flow in the LINAC was kept constant throughout the experiments and was ∼1.4 mTorr. This nitrogen pressure allowed collisional ion cooling as well as ion accumulation in the LINAC. Application and Timing of Applied dc Voltages. The timing of the QqTOF is controlled using a house-written Labview (National Instruments, Austin, TX) program requiring a National Instruments PXI chassis with DAQ and timer modules. In particular, this program allows us to control the dc voltages applied to the double lens following the LINAC as well as the MLIT exit lens. Three amplifiers purchased and assembled from Apex Microtechnology components (Tucson, AZ) allow fast response amplification of the time-controlled Labview signals up to 190 V, thereby allowing pulsed extraction scans to be written. The timing of the valve controlling the helium pulse is also controlled by the home-written Labview program. All other voltages were accessible through the Analyst software (MDS Sciex) available on the QqTOF. The timing of the applied voltages and helium pulse is illustrated in Figure 2. As previously described, the ions are accumulated in the LINAC and then transferred to the MLIT, which is undertaken by a decrease in the LINAC exit lens voltage from 20 to 12 V. Following the decrease in the LINAC exit lens voltage, 20 V is applied for the rest of the analytical scan. The entrance tube lens is 1.5 V less (10.5 V) than the LINAC exit lens during the transfer of the ions and 0.5 V higher than the dc offset applied to the quadrupole rods in the MLIT. During the transfer of the ions to the MLIT, the exit tube lens voltage is kept at 17 V

to prevent ions from passing through the MLIT. The transfer to the MLIT requires more than 0.035 ms in order to allow constant intensity mass spectra to be obtained. The intensity of ions in the resulting mass spectra was observed to be time dependent for transfer times less than 0.035 ms. Following a long transfer time of 0.14 ms, a linear increase in the confining (focus) voltage occurs. In order to limit the increase in the ions velocities, which can occur with a differing ramp in the entrance and exit lens voltages, the exit MLIT lens voltage was dropped for 0.002 ms from 17 to 15 V. The focus consists of a first linear ramp in the entrance and exit tube lens voltages from 15 to 30 V over 0.04 ms followed by a second slower ramp from 30 to 50 V over 2 ms. The focus voltages applied to the entrance and exit lens are then immediately increased to the value used for the rest of the focus time (with a maximum of 190 V). The extraction of the ions from the MLIT and their passage through the grid occurs over times much shorter than the times used in this part of the timing diagram (1 ms). The TOF push pulse (13 µs) consists of a nonmodified voltage applied to the accelerator in the TOF section and is scanned in time following the beginning of the extract from the MLIT. In other words, the time of application of this dc o-TOF push voltage required to accelerate the ions through the TOF section is changed linearly in time with the beginning of the extract from the MLIT (corresponding to a TOF push pulse time of 0 ms). This allows us to measure ion packet arrival times at this TOF push pulse region as well as the time of flight in the TOF section in the vertical plane (conventional TOF MS). Simulations. Simion 7.024 was used to compare the MLIT with and without novel tubular end caps as well as spatial ion focusing following ion transfer from the LINAC. User programs were written to allow change in the dc voltages for each time segment (transfer, focus ejection), as well as hard sphere collisional cooling with helium buffer gas. RESULTS AND DISCUSSION Ion Trapping in the MLIT. In a MLIT for a given radial interelectrode spacing of 2r0 between quadrupole rods (placed 180° from each other), the stability of an ion with a given mass/ charge value depends on the amplitude and frequency of the rf potential applied to the quadrupole rods. The application of the rf potential allows formation of a pseudopotential well and ions with a stable trajectory to be trapped in the radial plane. Direct current voltages applied to entrance and exit lenses form the axial potential well. The axial potential well depth is influenced by the distance between the entrance and exit lensessthe shorter the distance between these two lenses and the greater the applied dc voltage to the lenses, the deeper the potential well and the voltage difference between the lenses and the rods. Direct current voltages of equal value would produce an axial potential well centered half the distance between the entrance and exit lenses. Unlike the radial pseudopotential well, the axial potential well depth is independent of the m/z of the trapped ions. Influence of the Application of a dc Field in the Vicinity of an End Cap Hole. In order to introduce ions into the MLIT and to eject them, a hole must be present in both the entrance (24) Dahl, D. A. Proceedings of the 43rd ASMS Conference on Mass Spectrometry and Allied Topics, Atlanta, GA, 1995; p 717.

Analytical Chemistry, Vol. 79, No. 10, May 15, 2007

3781

Figure 3. Diagram illustrating miniature linear ion trap geometry with and without tube end cap lenses and the resulting potential well. (a) Illustration of the dc voltages applied to the exit lens to the LINAC and to flat entrance and exit end cap lenses used with the MLIT. (b) The use of tubular end cap lenses with the MLIT and the voltages applied in (c). c) Illustration of the deformation of the axial potential well when a potential is applied in the vicinity of an end cap hole when the tube lenses are replaced by flat lenses of similar radius and when tube lenses are used.

and exit end cap lenses. On application of a dc voltage to flat end cap lenses (as illustrated in Figure 3a) relative to the grounded first lens of the einzel, in the vicinity of the hole, an altered dc field is observed compared to if the hole was not present. Such an example is illustrated in Figure 3c with flat end cap lenses. When 20 V was applied to the LINAC exit lens and 15 V applied to the flat end cap lenses, the dc potential well is asymmetric around the center of the MLIT (Figure 3c). The entrance and exit lenses used in linear ion traps are separated from the multipole rods by a distance sufficient to prevent a short circuit. Ideally, infinite length quadrupole rods would be used, but in the case of a MLIT with a quadrupole rod length of 15 mm, the distance between both the lenses and the quadrupole rods makes up ∼20% of the total axial distance. An rf voltage applied to quadrupole rods will appear to be greater between the rods than at a distance following the rods. A flat lens with a tube extending from the center into the quadrupole volume is called here a tube lens (illustrated in Figure 3b). The tube lenses enter into the quadrupole volume by 3 mm, implying that ions are transferred into the tube lens on introduction and ejection, without passing through the rf field outside the quadrupole volume. The tube lenses have been designed so that both helium gas and the ions that have accumulated in the shortened LINAC are transferred together into the MLIT. The aim of the tube lens in this role is to obtain a higher mean collision frequency between the gas and the ions, allowing for low kinetic energy ion introduction into the MLIT. The low kinetic energies 3782

Analytical Chemistry, Vol. 79, No. 10, May 15, 2007

of the ions obtained through collisional cooling, as well as the depth of both the axial potential well and the radial pseudopotential well, will define the spatial packet width of the ions for different mass/charge values in the ion trap. It is possible to create the tube lenses using a dielectric material such that it remains conductive to dc fields but nonconductive to ac fields. In the case of very low kinetic energy ejection from the MLIT, this could be useful for improving the radial focusing should the time spent by the ion be several orders less than the time of oscillation of the rf voltage applied to the quadrupole rods. In this paper, ions are typically ejected with kinetic energies of ∼l0 eV (so that they can be detected on a detector following a time of flight), rendering it advantageous to use stainless steel tube lenses, which are conductive to both ac and dc fields. When the flat lenses are changed to tube lenses, the dc potential well becomes symmetrical around the MLIT center for the same applied dc voltages (Figure 3c). The amplitude of a voltage applied at a given distance from the hole and having penetrated through the hole is a function of the: hole size, the amplitude of the applied voltage, and the distance from the hole at which it is applied. Tube lenses allow ions to be trapped in a symmetrical potential well uninfluenced by electric fields applied in the vicinity of the MLIT entrance and exit holes. However, when potentials are applied in the vicinity of the tube lenses, they can penetrate the end of the tube slightly, as illustrated by the curves of the voltage in the axial direction on the extremities of the 15-V lines in Figure 3c (at 77 and 96 mm). While the shape of the axial potential well formed by the entrance and exit lenses is symmetrical with respect to the ion trap center, it is not perfectly parabolic in form over the entire distance between the end cap lenses. However, the bottom of the potential well is a parabola (to 5 decimal places) over ∼2.5 mm and to a well depth of 9 V when focus voltages of 190 V are applied to the entrance and exit tube lenses (result obtained by simulation with Simion 7.0) illustrated by Figure 4. It should be noted that in Figure 4 the center of the miniature linear ion trap is now found at zero on the x axis of the graph, in order to illustrate the relationship between the voltage at each point in the axial direction and the calculated equation. The importance of a parabolic potential well is that it allows the frequency of ion oscillation to be independent of its kinetic energy and is observed here only when the amplitude of oscillation is much smaller than 2.5 mm. Influence of the End Cap dc Focusing Voltages on the Temporal Packet Widths on Batch Ion Extraction. In order to trap ions axially in a potential well in the MLIT, dc voltages are applied to the entrance and exit tube lenses while a radial pseudopotential well is formed by application of an rf voltage on quadrupole rods. The trapped ions will be cooled by collision with the buffer gas (helium) to the bottom of the potential wells. When the ions are cooled in the MLIT to near-thermal kinetic energy, the resulting spatial distribution will depend on the potential well width for that energy. An example would be if a wide axial potential well is formed using low dc voltages on the entrance and exit lenses (the application of the dc voltage to a tube lens to focus the ion packet is called here the focus voltage), the ions would occupy a larger spatial distribution than in a narrower potential well (using higher dc focus voltages)

Figure 4. Illustration of the axial dc potential well in the miniature linear ion trap with 200 V applied to the entrance and exit tube lenses. The R2 indicates a parabola to at least five decimal places. It should therefore be noted that the potential well is more parabolic the closer the ion oscillation is to the center of the MLIT.

The spatial distribution of ions in the MLIT prior to batch ion extraction will result in a spatial and temporal packet width at the o-TOF push pulse for each mass/charge ion packet. The ion intensity can be plotted as a function of the time of the o-TOF push pulse following the batch ion extraction from the MLIT. Different mass/charge values can be identified by their time of arrival at the TOF push pulse region due to their mass/charge dependent velocities. However, as the o-TOF push pulse is only ∼5 cm from the MLIT exit lens, considerable overlap between the different mass/charge ion packets can occur (resolution between 10 and 40 no units). The individual mass/charge ion intensity data are therefore obtained by selecting it from the total ion current (TIC) (called XIC in Analyst, MDS Sciex). When the intensity is plotted against the o-TOF push pulse time (µs) following the start of the batch ion extraction, the full width at half-maximum (fwhm) in microseconds is calculated from the resulting peak, for each mass/charge value. The focus voltage applied prior to batch ion ejection is shown in this study to influence this temporal packet widthsfwhm (µs), at the o-TOF push pulse. An increase in the focus voltage applied to the entrance and exit lens decreases globally the temporal spread at the o-TOF push pulse (and therefore the spatial distribution) for ions of different mass/charge values (350 Th [M + 3H]3+, 466 Th [{Arg-Val-TyrIle-His-Pro-Phe} + 2H]2+, and 524 Th [M + 2H]2+), as illustrated by Figure 5. Increasing the dc focus voltage applied to the tube lenses from 50 to 190 V was observed to decrease the fwhm from 4.5 to 2.5 µs at the TOF push pulse for the 350 Th ion packet. The same increase in the dc voltage to the tube lenses allowed a decrease in the temporal distribution of the 466 Th ion packet at the TOF push pulse from 4.7 to 3.1 µs and for the 524 Th ion packet from 11.1 to 4.8 µs. While the temporal fwhms of the different mass/charge ion packets are decreased to half their initial times by increasing the focus voltage applied to the tube lenses from 50 to 190 V, the temporal fwhm of 4.8 µs for the 524 Th ion packet remains much larger after 14-ms focusing than that of the 350 Th ion packet (2.5 µs).

Figure 5. Influence of the focus voltages applied to the entrance lens and the exit lens on the fwhm for different mass/charge ions in the axial plane in microseconds. The fwhm was obtained from a plot of intensity as a function of o-TOF push time following ion extraction from the MLIT.

Application of Extraction Voltages. As we wish to perform low kinetic energy batch ion extraction from the MLIT, low dc voltage gradients are applied to the tubular end caps during ion ejection. If 20 V is applied to the entrance lens and 5 V to the exit lens with a 10-V dc offset to the MLIT quadrupole rods, a voltage gradient of 1 V/mm is obtained across the center of the MLIT (Figure 6). Assuming cooling to near-thermal kinetic energy, the kinetic energy distribution of the ions on ejection from the MLIT will therefore be principally due to the spatial distribution of the ion packet prior to ejection. Also, the cooling of the ions in the MLIT will influence the kinetic energy distributions on batch ion extraction by the size of the initial ion packet in the axial direction. Difference between Kinetic Energy Distributions of Ion Packets with Little and Significant Cooling. When a retarding dc voltage applied to the grid (following the MLIT exit lens) is varied, ions with a kinetic energy less than the grid voltage cannot be observed. The derivative of the variation in intensity with the grid voltage illustrates the kinetic energy distribution of the ions. In Figure 7a, a 1-ms focus time with 110 V applied to the entrance Analytical Chemistry, Vol. 79, No. 10, May 15, 2007

3783

Figure 6. Illustration of the extraction field applied for ion ejection from the miniature linear ion trap calculated using Simion 7.0. The center of the miniature linear ion trap is at 86.3 mm in the simulation, and in this region, the linear extraction voltage is ∼1 V/mm.

and exit lenses was used with helium being pulsed into the MLIT for the first 0.75 ms. A kinetic energy distribution of ∼5 V resulted from the 1-ms focus. However, when a typical 14-ms focus was used with helium being pulsed into the MLIT for 2.5 ms, a narrow kinetic energy distribution was observed (∼1 V). The 1-V kinetic energy distribution can be attributed to the spatial distribution of the ions prior to ejection, thus demonstrating focusing to ∼1-mm total axial dispersion. Ion Loss with Increasing Focus Voltages Applied to Tube Lenses in the MLIT. An increase in the dc voltages applied during the focus of the ions in the MLIT will narrow the potential well in the axial direction. However, unlike in the 3D ion trap where the frequency of oscillation in the axial plane defines whether it has a stable or unstable trajectory, the dc voltages focusing the ions in linear ion traps do not. In the MLIT, high dc voltages are applied to the tubular end caps (only a few millimeters from the ion packet) form very steep potential wells as illustrated by Figure 4. In order to elucidate whether ion loss occurs due to the narrowing of these steep dc potential wells, a plot of intensity as a function of the focus voltage (applied for 14 ms) was undertaken for different mass/charge ions (Figure 8). Figure 8 also illustrates the intensity of the TIC. With the increase in focus voltage, the intensities of the multiply charged ions (350 , 466 , and 524 Th) and of the 720 Th singly charged ion (present in the angiotensin II mass spectrum) decrease as well as the TIC. Focus voltages higher than 150 V cause no ions of mass/charge 720 Th to be observed. However, it remains possible to observe the lower mass/charge values between 150 and 190 V despite modest qx values of ∼0.4 for the 720 Th ion packet and qx of ∼0.8 for the 350 Th ion packet, which should allow the ions to have stable

trajectories in the radial pseudopotential well. In fact, increasing the dc focus voltages causes the axial potential well to become narrower and radial excitation can occur resulting in loss of ions from the radial pseudopotential well. Ion Focusing in the MLIT Following Transfer from a Shortened LINAC. Using Simion 7.0, it was possible to simulate ion focusing in the MLIT with tube lenses following ion introduction from a shortened LINAC (Figure 9). A hard-sphere collision cooling model and developed user programs for Simion 7.0 were used in the simulation. Ions enter into the MLIT (0 ms) with a KE of ∼3 eV from the LINAC, which has allowed ion accumulation near to the LINAC exit lens. The entrance tube lens is at 76.8 mm, and the center of the ion trap is at 86.3 mm. During the transfer into the MLIT, a dc voltage of 15 V was applied to the exit lens (95.8 mm), preventing ions from simply passing through the MLIT. Following the transfer of ions into the MLIT (0.03 ms), they were confined in a dc potential well formed by application of 110 V to the tubular end caps and 9-V dc offset to the quadrupole rods. Focusing was carried out here during 10 ms for a helium pressure (constant throughout the 10 ms) between 30 and 40 mTorr. On entrance into the MLIT, the ions oscillate with a trajectory of ∼4 mm around the ion trap center. Collisions with the helium buffer gas gradually allow a reduction of the ions’ kinetic energy and allow them to oscillate with increasingly smaller trajectories around the ion trap center. However, despite the use of a 10-ms focusing time, little change is observed in the spatial focusing after 4 ms and the ion packet width remains less than 1 mm in the axial direction. After 10 ms, the different mass/charge ions oscillate with a trajectory with similar axial and radial dimensions and possess diameters of 0.5 mm in the axial plane (around the center of the MLIT at 86.3 mm) and 0.8 mm in the radial plane. Following the focus of the ions to the center of the MLIT, it was possible to undertake batch ion extraction with 20 V applied to the MLIT tubular entrance end cap lens, 5 V applied to the exit end cap lens, and a 10-V dc offset applied to the quadrupole rods. While this simulation demonstrates that good axial focusing is possible in a MLIT with tubular end cap lenses, it also suggests that much of the decrease in the spatial distribution can occur during a short helium pulse (3 ms) as used previously in this study. It is worth noting that the simulation illustrated by Figure 9 does not include Coulombic repulsion, and this is expected to

Figure 7. Illustration of the derivative of the intensity of a single mass/charge value as a function of the grid voltage (V) (a) with almost no ion cooling in the miniature linear ion trap and (b) with typical ion cooling in the miniature linear ion trap. 3784

Analytical Chemistry, Vol. 79, No. 10, May 15, 2007

Figure 8. Illustration of the loss in intensity with an increase in focus voltage due to instability of the ions in the radial pseudopotential well.

Figure 9. Simulation of the trajectories of five ions of 350 and 524 Th entering from a shortened LINAC, focused in a MLIT with tube lenses to less than a 1-mm ion packet in the axial plane using helium buffer gas at a pressure of 30-40 mTorr, before simultaneous ejection toward an o-TOF.

increase the global axial ion packet size to the concluded experimental value of ∼1-mm diameter. CONCLUSION A novel MLIT has been created, and the ion packet focusing provided by the formation of a dc potential well between unusual tubular end cap lenses and the introduction of helium collision gas has been characterized. The use of tube lenses decreases the penetration of dc fields applied near to the end cap holes into the MLIT volume, allows a symmetrical potential well to be formed, and enables a greater extraction field strength. In particular, the bottom of the dc potential well is a parabola over a distance of 2.5 mm. Tube lenses can therefore be used to change the demonstrated physical properties of a MLIT. Increasing the dc voltages applied to the tubular end cap lenses for ion focusing of collisionally cooled ions was observed to

decrease the temporal distribution of the ion packet at an o-TOF push pulse. However, the use of focus voltages greater than 150 V applied during collisional cooling of the ion packets gave rise to loss of high mass/charge ions. High dc voltages should therefore be used with a MLIT with tubular end cap lenses to allow ion focusing (by collisional cooling) in a narrow potential well, but they should be sufficiently low not to limit the overall mass/charge range. It is possible to obtain axial and radial spatial ion focusing in the MLIT of less than 1 mm in all directions, which is comparable to that described in a 3D ion trap.15-17 On batch ion extraction from the MLIT, this good spatial ion focusing influences the kinetic energy distribution of the ions. A kinetic energy distribution of 1 V was illustrated by use of a retarding potential applied to a grid after the MLIT and before a time of flight (following a 14-ms focus with a 1 V/mm gradient across the MLIT center on extraction). The ability of the MLIT with tubular end cap lenses to allow such good 3D ion focusing as well as introduction of ions from a prior focusing quadrupole, low kinetic energy batch ion extraction with a narrow kinetic energy distribution, and control of the ion packet through application of dc voltages suggests that this novel ion trap would be a good intermediate device in a mass spectrometer. ACKNOWLEDGMENT We thank MDS Sciex for funding this project and UNM for the facilities. We particularly thank Bruce Thomson for his help and technical advice.

Received for review November 1, 2006. Accepted March 14, 2007. AC0620462

Analytical Chemistry, Vol. 79, No. 10, May 15, 2007

3785