Anal. Chem. 2009, 81, 1833–1840
Mass Selection of Ions from Beams Using Waveform Isolation in Radiofrequency Quadrupoles Qingyu Song,† Scott A. Smith,† Liang Gao,† Wei Xu,§ Michael Volny´,† Zheng Ouyang,*,‡,§ and R. Graham Cooks*,† Department of Chemistry, Weldon School of Biomedical Engineering, and School of Electrical and Computer Engineering, Purdue University, West Lafayette, Indiana 47907 A waveform isolation method is described for the massselective transmission of ions through quadrupole mass filters, and it is implemented on a new tandem mass analyzer instrument. The method features the application of broad-band waveforms comprising appropriate frequencies to cause mass-selective instability in ions of particular mass-to-charge (m/z) and to transmit all others. The experiment is implemented in a tandem quadrupole system in which the first mass filter is a rectilinear ion trap (RIT) operated in a continuous mass-selective mode to transmit ions of ions of one or more arbitrarily selected m/z value(s). The second analyzer was used to verify the quality of the mass selection achieved using the first analyzer via conventional quadrupole ion trap massselective instability scanning. A new subtype of product ion tandem mass spectrometry (MS/MS) scan, termed the summed product ion scan, is demonstrated with a mixture of biological compounds. It is used to characterize product ions arising after simultaneous isolation and collisional activation of multiple precursor species, in this case ions of the same analyte generated in different charge states. The summed product ion scan can be useful for enhancing sensitivity for the analyte of interest or for providing more comprehensive information about an analyte than is available by monitoring a single ionized form of the analyte. The analytical performance of the waveform isolation method is tested using simple drug mixtures, and its potential for increasing overall yields in preparative mass spectrometry is explored briefly. It is shown that efficiencies of ca. 70% of ion transfer to a surface for ion soft landing surface can be achieved. The upper mass range is limited by axial acceleration arising from the stretched geometry, and one solution to this problem is provided. Various combinations of mass analyzer geometries and operational methodologies have been explored since the advent of tandem mass spectrometry (MS/MS).1-8 They include those * To whom correspondence should be addressed. Phone: 765-494-5262(R.G.C.), 765-494-2214(Z.O.). Fax: 765-494-9421 (R.G.C.), 765-496-1912 (Z.O.). E-mail:
[email protected] (R.G.C.),
[email protected] (Z.O.). † Department of Chemistry. ‡ Weldon School of Biomedical Engineering. § School of Electrical and Computer Engineering. 10.1021/ac802213p CCC: $40.75 2009 American Chemical Society Published on Web 01/29/2009
which employ multiple mass analyzers (e.g., the tandem-in-space triple quadrupole) and those which operate with a single mass analyzer (e.g., the tandem-in-time quadrupole ion trap).9 Hybrid instruments have been particularly useful, including those which use quadrupoles as first-stage mass filters in tandem mass spectrometers.10-12 One notably successful combination is that of a quadrupole mass filter (Q) with a time-of-flight (TOF) analyzer. The Q-TOF configuration benefits from both the high efficiency of precursor ion selection using a quadrupole analyzer and the high sensitivity of a TOF mass analyzer.13 By adjusting the rf and dc components of a quadrupole filter, ions with a selected mass-to-charge ratio (m/z) can be placed at the apex of the Mathieu stability diagram, as shown in Figure 1a.14 In this experiment, known as mass-selective stability filtering, ions of a single or a small contiguous range of m/z values can be isolated from all others in an ion beam and be made to pass through the mass filter. The experiment provides high transmission efficiency and is well-suited to subsequent ion activation using collision-induced dissociation (CID).12 Despite the advantages associated with quadrupole filters as a means of primary ion selection, the method has disadvantages too. In the process of resolving a band of selected ions of particular m/z values, it is (1) Glish, G. L.; Burinsky, D. J. J. Am. Soc. Mass Spectrom. 2008, 19, 161– 172. (2) McLafferty, F. W. Tandem Mass Spectrometry; Wiley, John & Sons, Incorporated, 1983. (3) Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry/Mass Spectrometry: Techniques and Applications of Tandem Mass Spectrometry, 1st ed.; Wiley-VCH, 1989. (4) Schwartz, J. C.; Wade, A. P.; Enke, C. G.; Cooks, R. G. Anal. Chem. 1990, 62, 1809–1818. (5) Wright, L. G.; Schwartz, J. C.; Cooks, R. G. TrAC, Trends Anal. Chem. 1986, 5, 236–240. (6) Vincenti, M.; Schwartz, J. C.; Cooks, R. G.; Wade, A. P.; Enke, C. G. Org. Mass Spectrom. 1988, 23, 579–584. (7) Cooks, R. G.; Amy, J. W.; Bier, M.; Schwartz, J. C.; Schey, K. Adv. Mass Spectrom. 1989, 11A, 33–52. (8) Schwartz, J. C.; Cooks, R. G. Spectroscopy (Amsterdam) 1998, 5 (1-6), 49–63. (9) Johnson, J. V.; Yost, R. A.; Kelley, P. E.; Bradford, D. C. Anal. Chem. 1990, 62, 2162–2172. (10) Dawson, P. H. Mass Spectrom. Rev. 1986, 5, 1–37. (11) Reinsfelder, R. E.; Denton, M. B. Int. J. Mass Spectrom. Ion Processes 1981, 37, 241–250. (12) Yost, R. A.; Enke, C. G. J. Am. Chem. Soc. 1978, 100, 2274–2275. (13) Glish, G. L.; Goeringer, D. E. Anal. Chem. 1984, 56, 2291–2295. (14) March, R. E., Todd, J. F. J., Eds. Practical Aspects of Ion Trap Mass Spectrometry; CRC Press: Boca Raton, FL, 1995; Vol. I: Fundamentals of Ion Trap Mass Spectrometry.
Analytical Chemistry, Vol. 81, No. 5, March 1, 2009
1833
Figure 1. (a) Concept of continuous mass-selective stability (rf/dc) filtering. (b) Concept of continuous mass-selective instability (waveform) filtering.
Figure 2. (a) Schematic representation of the product ion scan. (b) Schematic representation of the summed product ion scan.
not uncommon for the selected ions nearest the edges of the stability boundary to be discriminated against or completely ejected due to field imperfections.12 This effect can be avoided but only by widening the isolation window which reduces the resolution of ion isolation. The magnitude of the available rf/dc voltage, which is proportional to the m/z of the selected parent ions, can also limit the upper mass range in this type of ion selection experiment. Selection of ions of particular m/z values can also be achieved using quadrupole ion traps. In this case, the experiments are normally done using the mass-selective instability scan technique, but they can also be done using broad-band waveform excitation to cause all ions except those of a selected m/z ratio or narrow range of ratios to be excited and ejected. The usual method of implementing this concept is via stored-waveform inverse Fourier transform (SWIFT) excitation, an experiment first implemented by Guan and Marshall using an ion cyclotron resonance mass spectrometer (FTICR).15-17 In the SWIFT isolation experiment, a low-energy waveform is used to resonantly and simultaneously excite all trapped ions at their secular frequencies except for those corresponding to the ionic species of interest. Although SWIFT is typically used to isolate ions of a single m/z, multiple ions with different m/z values can be isolated at the same time by opening multiple notches in the frequency domain.18 In either case, ion isolation within traditional three-dimensional (3D) ion traps requires multiple sequential steps. These steps include the injection and trapping of ions in the trapping field, waveform excitation and ejection of all unwanted m/z, and then either mass analysis or further processing of the remaining mass-selected ions. Such pulsed isolation experiments have the advantage of allowing ion accumulation over time in order to increase sensitivity. On the other hand, the disadvantages of integration of ion signal intensity over time in an ion trap include the limited linear dynamic range due to limitations in ion trapping capacity and the length of time taken for the multistep isolation procedure. For continuous ionization sources, pulsed operation of the mass analyzer also reduces the efficiency with which the original ion beam is sampled due to the inherently low duty cycles for such experiments.
It is possible to combine the two concepts just described, continuous beam sampling in a quadrupole mass filter and multispecies selection in an ion trap, by applying an appropriate waveform to the rods of a quadrupole mass filter. In such a case the mass filter is operated somewhat analogously to a 2D (linear) ion trap19,20 except that the ions are not trapped but continuously transmitted. Provided the operating conditions are selected so that the transmitted ions are allowed to undergo several secular cycles during their motion through the mass filter, it is possible to combine the arbitrary mass-selection capability and efficiency of broad-band excitation with the high duty cycle of ion isolation from an ion beam in a traditional linear quadrupole. With the use of a singly or multiply notched broad-band waveform created by SWIFT or a related method,21,22 beams of ions can be filtered so as to select ions of multiple m/z values according to any arbitrary criterion, as shown in Figure 1b. By isolating ions at low q values, beam-based waveform isolation also provides the ability to operate at lower rf and dc voltages when compared to mass-selective stability filtering, although lower voltages correspond to reduced resolution in both techniques. Although the concepts just outlined are straightforwardsand perhaps even foreseen in the early ion trap literature (several related experiments were reported earlier by different groups23-26)sthey have not entered the practice of ion trap mass spectrometry and for this reason are described here, and their use is illustrated. With appropriate selection of parent ions, a new subtype of the MS/MS product ion scan,4 termed a “summed product ion scan” as shown in Figure 2, can be performed. It could be useful in several types of applications (summarized in Table 1) in which the analyte of interest occurs in several different ionic forms, each with a characteristic m/z value. By simultaneously isolating the different forms of the ionized molecule and then subjecting them to activation and/or analysis, the detection sensitivity and the
(15) Guan, S. H.; Marshall, A. G. Int. J. Mass Spectrom. Ion Processes 1996, 158, 5–37. (16) Marshall, A. G. Int. J. Mass Spectrom. 2000, 200, 331–356. (17) Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Mass Spectrom. Rev. 1998, 17, 1–35. (18) Vachet, R. W.; McElvany, S. W. J. Am. Soc. Mass Spectrom. 1999, 10, 355– 359.
1834
Analytical Chemistry, Vol. 81, No. 5, March 1, 2009
(19) Schwartz, J. C.; Senko, M. W.; Syka, J. E. P. J. Am. Soc. Mass Spectrom. 2002, 13, 659–669. (20) Hager, J. W. Rapid Commun. Mass Spectrom. 2002, 16, 512–526. (21) McLuckey, S. A.; Goeringer, D. E.; Glish, G. L. Anal. Chem. 1992, 64, 1455–1460. (22) Julian, R. K.; Cooks, R. G. Anal. Chem. 1993, 65, 1827–1833. (23) Ross, C. W.; Guan, S. H.; Grosshans, P. B.; Ricca, T. L.; Marshall, A. G. J. Am. Chem. Soc. 1993, 115, 7854–7861. (24) Williams, E. R.; Loh, S. Y.; McLafferty, F. W.; Cody, R. B. Anal. Chem. 1990, 62, 698–703. (25) Maruyama, S.; Anderson, L. R.; Smalley, R. E. J. Chem. Phys. 1990, 93, 5349–5351. (26) Haebel, S.; Gaumann, T. Int. J. Mass Spectrom. Ion Processes 1995, 144, 139–166.
Table 1. Applications of Summed Product Ion Scan application
example type
typical system
analysis of multiply charged states of an analyte analysis of multiple oligomers analysis of multiple ionic forms of analyte
[M + nH]n+; n g 1 [nM + H]+; n g 1 [M + C]+ C ) cationizing species
protein ions generated by ESI oligomers of amino acids, e.g., serine clusters C ) H, Na, K,...
information content associated with the analyte may be enhanced. For example, the summed product ion spectrum which includes all fragments arising from all charge states of an electrospray ionization (ESI)-generated protein present in a complex mixture should provide more complete and higher-quality information than the product ion spectra from any individual charge state. It is noted that it is difficult to dissociate all selected ionic species of different m/z values with comparable efficiencies and therefore that the expected increase in sensitivity may not be fully realized. Another application in which the continuous waveform isolation methodology would be useful is in preparative mass spectrometry, usually referred to as ion soft landing (SL).27 For a typical SL experiment, samples are first ionized in the gas phase followed by m/z selection and subsequent deposition of the intact purified species onto a surface.28-33 Quadrupole mass filters are the mass analyzers of choice for SL due to their high duty cycle; however, ions of only one m/z value can be selected at a time. Such a procedure will result in a reduction of the total soft-landing yield when multiple forms of a given analyte exist. Conventional linear ion traps have also been used for SL experiments28,29 but suffer the disadvantages of discontinuous operation, since they have been operated in the ion trapping mode. The use of continuous operation with 2D ion traps and continuous waveform selection should be advantageous over both previous methods, and this application was the driver of the developments described in this paper. We apply continuous mass-selective ion filtering with a multiply notched broad-band excitation waveform using a rectilinear ion trap of square cross section. A tandem analyzer mass spectrometer, capable of continuous isolation of ions from an ion beam, was developed in this work. A series of experiments was carried out to investigate the efficiency and resolution of continuous isolation. In addition, the potential of this method in preparative mass spectrometry is also explored. Several related experiments have been conducted in the past. In the first such case, Paul et al. reported applying a sinusoidal signal dipolarly on a quadrupole mass filter to create a notch filter for an isotopic enrichment study.34 Langmuir et al. extended Paul’s (27) Franchetti, V.; Solka, B. H.; Baitinger, W. E.; Amy, J. W.; Cooks, R. G. Int. J. Mass Spectrom. Ion Processes 1977, 23, 29–35. (28) Ouyang, Z.; Takats, Z.; Blake, T. A.; Gologan, B.; Guymon, A. J.; Wiseman, J. M.; Oliver, J. C.; Davisson, V. J.; Cooks, R. G. Science 2003, 301, 1351– 1354. (29) Blake, T. A.; Zheng, O. Y.; Wiseman, J. M.; Takats, Z.; Guymon, A. J.; Kothari, S.; Cooks, R. G. Anal. Chem. 2004, 76, 6293–6305. (30) Gologan, B.; Takats, Z.; Alvarez, J.; Wiseman, J. M.; Talaty, N.; Ouyang, Z.; Cooks, R. G. J. Am. Soc. Mass Spectrom. 2004, 15, 1874–1884. (31) Hadjar, O.; Futrell, J. H.; Laskin, J. J. Phys. Chem. C 2007, 111, 18220– 18225. (32) Volny, M.; Elam, W. T.; Ratner, B. D.; Turecek, F. J. Biomed. Mater. Res., Part B 2007, 80B, 505–510. (33) Volny, M.; Sengupta, A.; Wilson, C. B.; Swanson, B. D.; Davis, E. J.; Turecek, F. Anal. Chem. 2007, 79, 4543–4551. (34) Paul, W.; Reinhard, H. P.; von Zahn, U. Z. Phys. A: Hadrons Nucl. 1958, 152, 143–182.
work and patented a method using a filtered noise field (FNF) applied to the quadrupole structure to deflect undesirable ions into unstable trajectories.35 Belov et al. reported dynamic range enhancement by using a data-dependent rf-only dipolar excitation to eject the most abundant ion species in the course of an LC/ FTICR experiment.36 A relatively recent method of mass scanning with quadrupoles without using a dc voltage, called mass-selective axial ejection, was developed by Hager37 and later optimized by Moradian and Douglas.38 This method uses the fringing field at the exit of the quadrupole to couple the ion’s x-y motion into the z direction, enabling mass-selective axial ejection when ions overcome the exit lens potential barrier to reach the detector. The same concept was subsequently used in a linear quadrupole ion trap for axial ejection.20 A broad-band waveform excitation technique for mass-selective ion injection into quadrupole ion traps was demonstrated by Goeringer et al.39 and independently in work from this laboratory;40 by applying a notched waveform signal to the endcap electrodes during ion injection, the accumulation of matrix ions was effectively suppressed. Several related experiments on quadrupole ion traps have been reported earlier, including a method of single-ion or single-reaction monitoring which avoids high rf voltage scanning by using rf/dc isolation or waveform isolation in combination with the axial dc pulses to eject and detect ions of particular m/z.41 EXPERIMENTAL SECTION A mixture of compounds consisting of 20 ng/µL methamphetamine, 10 ng/µL caffeine, 2 ng/µL cocaine, and 2 ng/µL heroin in 50:50 methanol/water with 0.5% acetic acid was used to evaluate the continuous waveform filtering method. All of the compounds were purchased from Sigma-Aldrich (St. Louis, MO). A mixture of biological compounds was used in the MS/MS experiment, including the following: VVR (Genscript Corp., Piscataway, NJ), MRFA (Research Plus Inc., Manasquan, NJ), ALILTLVS (Bachem Bioscience Inc., King of Prussia, PA), and reserpine (SigmaAldrich, St. Louis, MO); the compounds were diluted to 50 µM in 50:50 methanol/water acidified with 0.5% acetic acid. Instrumentation. A schematic diagram showing the key components of the new tandem mass spectrometer built for this study is shown in Figure 3a, and details of the ion optics and vacuum system appear in Figure 3b. A two-stage differential pumping system is used with the first chamber being evacuated (35) Langmuir, R. V. United States, 1964; p 4. (36) Belov, M. E.; Anderson, G. A.; Angell, N. H.; Shen, Y. F.; Tolic, N.; Udseth, H. R.; Smith, R. D. Anal. Chem. 2001, 73, 5052–5060. (37) Hager, J. W. Rapid Commun. Mass Spectrom. 1999, 13, 740–748. (38) Moradian, A.; Douglas, D. J. J. Am. Soc. Mass Spectrom. 2008, 19, 270– 280. (39) Goeringer, D. E.; Asano, K. G.; McLuckey, S. A.; Hoekman, D.; Stiller, S. W. Anal. Chem. 1994, 66, 313–318. (40) Soni, M. H.; Cooks, R. G. Anal. Chem. 1994, 66, 2488–2496. (41) Zhang, C.; Chen, H. W.; Guymon, A. J.; Wu, G. X.; Cooks, R. G.; Ouyang, Z. Int. J. Mass Spectrom. 2006, 255, 1–10.
Analytical Chemistry, Vol. 81, No. 5, March 1, 2009
1835
Figure 3. (a) Schematic diagram of new tandem quadrupole mass filter instrument built for this study. (b) Details of ion optics, vacuum system, and operating conditions.
by two Edwards E2M30 rotary vane mechanical pumps (BOC Edwards, Wilmington, MA). With the use of a stainless steel heated capillary interface of dimensions 500 µm i.d. × 100 mm (Thermo Fisher Scientific Inc., San Jose, CA), an ultimate pressure of ∼1 Torr was achieved in the first stage of the vacuum chamber. An ion funnel based on the Pacific Northwest National Laboratory design42 was used to contain and guide the ions in the first vacuum chamber. The second chamber was pumped using the first stage of a SplitFlow turbomolecular pump (Pfeiffer Vacuum Inc., Nashua, NH). The ultimate pressure of the second chamber was maintained at ∼5 mTorr by adjusting the rate at which helium buffer gas was introduced. Inside the second vacuum chamber, a square quadrupole was used to transport the ion beam to the mass analyzer. A rectilinear ion trap (RIT) with rods of square cross section and dimensions of 5 × 4 × 40 mm3 (x0, y0, and z)43 was used as the quadrupole mass filter, hereafter termed Q1; the design, fabrication, and characterization of this particular RIT is described elsewhere.44 Although the resolution of the RIT is compromised due to the nonlinear field created by the flat electrodes, the RIT demonstrates adequate analytical performance for low-resolution protein separation.45 In comparison to hyperbolic electrodes, the wider pseudopotential well associated with the higher-order field components provided by the rectilinear geometry makes it favorable for operation at higher pressure. Q1 was operated simultaneously as an rf ion guide and as a single- or multiple-notch mass filter by applying a supplemental broad-band waveform comprising sinusoidal ac signals with predefined frequencies. The entire control signal for Q1 was generated from modified LCQ electronics (Thermo Fisher Scientific Inc.). For the purpose of qualitative verification of the waveform-selected ion beam methodology, a second RIT, (42) Tang, K. Q.; Tolmachev, A. V.; Nikolaev, E.; Zhang, R.; Belov, M. E.; Udseth, H. R.; Smith, R. D. Anal. Chem. 2002, 74, 5431–5437. (43) Ouyang, Z.; Wu, G. X.; Song, Y. S.; Li, H. Y.; Plass, W. R.; Cooks, R. G. Anal. Chem. 2004, 76, 4595–4605. (44) Song, Q.; Xu, W.; Gao, L.; Ouyang, Z.; Cooks, R. G. Manuscript in preparation, 2008. (45) Song, Q.; Kothari, S.; Senko, M. A.; Schwartz, J. C.; Amy, J. W.; Stafford, G. C.; Cooks, R. G.; Ouyang, Z. Anal. Chem. 2006, 78, 718–725.
1836
Analytical Chemistry, Vol. 81, No. 5, March 1, 2009
Q2, was installed in tandem with Q1. Associated with Q2 are independent lens controls, rf generation and amplification circuits including a coil tank, an ion detector, data acquisition and a data processing system (Griffin Analytical Technologies, West Lafayette, IN). Q2 was normally operated using the massselective instability mode. However, it was operated as a simple rf ion guide in experiments in which ions selected in the mass filter Q1 were passed to the SL surface via Q2. The ion deposition current was measured using a Keithley 6485 picoammeter with signal averaging turned on (Keithley Instruments, Inc., Cleveland, OH); the picoammeter was interfaced to a computer to allow data transcription into an Excel spreadsheet using ExceLINX software (Keithley Instruments, Inc.). The ESI ion source used for these experiments was homemade, and its construction is described elsewhere.46 RESULTS AND DISCUSSION Ions with different m/z values will have different secular frequencies (ω) in a quadrupolar rf field. Theoretically, the fundamental secular frequencies can be expressed as14
ω)
βx Ω 2
(1)
where Ω is the drive rf frequency and βx is a function of the Mathieu parameters a and q and must have a value between 0 and 1. For an rf-only quadrupole, the Mathieu parameter a is zero. The secular frequency is related only to the Mathieu parameter q; the q value at a constant rf voltage (Vrf), rf frequency (Ω), and for a given geometry (x0) is in turn inversely proportional to m/z according to the Mathieu equation for q:
q)
4zeVrf mx02Ω2
(2)
(46) Takats, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Anal. Chem. 2004, 76, 4050–4058.
The approximately harmonic motion (also know as pseudomotion) of an ion in a quadrupole field can be modeled by using the ideas of the pseudopotential well approximation.47,48 This approximation assumes that the ion motion is composed of a small ripple at the rf frequency Ω, superimposed on a simple harmonic motion at much lower secular frequency ω. In the presence of an ac dipolar excitation, ion/neutral collisions can be represented as a damping term c,49 and the time-dependent differential equation which describes the ion pseudomotion is then ∂2x ∂x + c + ω2x ) fejωst ∂t ∂t2
(3)
where ωs is the applied ac signal frequency and f ) (Uac/2u0)(e/m) where Uac represents the applied ac signal voltage. Following the treatment of Major and Dehmelt50 and Goeringer et al.,49 a fast harmonic motion at the applied ac frequency (ωs) together with a slowly varying envelope motion A(t) of the fast x(t) oscillation describes the ion motion.51 x(t) ) A(t)ejωst
(4)
Substituting eq 4 back into eq 3, and because A(t) is a slowly varying function with respect to time, the second-order derivative of A(t) is ignored, then we have (c + 2jωs) ∂ A/ ∂ t + (jcωs + ω2 - ωs2)A ) f
(5)
Solving eq 5 near resonance (ωs - ω,ω), the ion motion envelope (A(t)) is obtained as A(t) ) f/m + K0e-nt
(6)
where m ) jcωs + ω2 - ωs2, n ) m/(c + 2jωs), and K0 is a function of the initial conditions for a ion with initial conditions of x(0) ) 0 and x′(0) ) 0, K0 ) -f/m. An example of an ion motion envelope in a quadrupole field subjected to dipolar excitation is shown in Figure 4a. At resonance, where the dipole excitation frequency equals the ion secular frequency, the amplitude of motion increases rapidly. Given sufficient kinetic energy from the resonance excitation process, the amplitude of the ion’s oscillation will grow larger than the physical size of the ion trap electrodes and it will therefore be ejected. When the ion secular frequency is far from the excitation frequency, collisional cooling is dominant and ions will eventually be thermalized. The ion motion envelop is related to the excitation time and excitation voltage Uac. In order to eject ions by dipolar excitation in a short time frame, a high excitation voltage is required. Figure 4b illustrates the effect of the mass filter length on the excitation bandwidth. Provided ions enter the filter with the same z velocity, the longer ions stay in the rf field, the better excitation resolution will be achieved. To selectively pass ions of (47) Wuerker, R. F.; Shelton, H.; Langmuir, R. V. J. Appl. Phys. 1959, 30, 342. (48) Makarov, A. A. Anal. Chem. 1996, 68, 4257–4263. (49) Goeringer, D. E.; Whitten, W. B.; Ramsey, J. M.; McLuckey, S. A.; Glish, G. L. Anal. Chem. 1992, 64, 1434–1439. (50) Major, F. G.; Dehmelt, H. G. Phys. Rev. 1968, 170, 91–107. (51) Plass, W. R.; Li, H. Y.; Cooks, R. G. Int. J. Mass Spectrom. 2003, 228, 237– 267.
Figure 4. (a) Ion motion envelope showing responses under resonance and off-resonance conditions. (b) Quadrupole mass filter length effect on the resonant excitation bandwidth.
particular m/z values through this filter, frequency notches corresponding to particular secular frequencies (all but those of the selected ions) must be present in the broad-band ac waveform. Continuous mass filtering was implemented using the RIT, Q1. While a constant differential rf voltage was applied to the x and y electrodes to create a quadrupolar field to guide the ion beam, a broad-band ac waveform (5-500 kHz) with one or more frequency notches in the frequency domain was applied simultaneously to the x electrodes as shown in Figure 1b. The lower-frequency components in the waveform were given increasingly greater amplitudes in order to provide more excitation energy to ions with greater m/z values. The optimized amplitude for the excitation waveform was chosen so as to use the minimal voltage required to resonantly ejection the ions when they fell in any of these isolation windows. There is no confinement along the z-axis, so the time that ions spend in Q1 depends only on their initial velocity, the pressure, and the physical length of the trap. Given an ion with m/z of 150 and an empirically determined kinetic energy of ∼2 eV, the travel time through Q1 is about 30 µs, corresponding to about 10 cycles of secular motion under typical experiment conditions where βx ) 0.67 and Ω ) 1 MHz. To achieve continuous filtering in such a short time and using such a small number of cycles requires higher ac amplitudes to excite ions out of their stable trajectories than would be necessary in a quadrupole ion trap with similar dimensions and Analytical Chemistry, Vol. 81, No. 5, March 1, 2009
1837
Figure 5. (a) Continuous waveform single-peak isolation (back to front): full-scan spectrum of drug mixture (scan 1), waveform-isolated ions of m/z 304 (scan 2) and m/z 150 (scan 3); (b) secular frequency distribution for m/z 150 at q ) 0.7; (c) secular frequency distribution for m/z 304 at q ) 0.4.
operating conditions. In addition, higher buffer gas pressures were also required to damp the trajectories of the ions of interest in this short time frame. When Q1 was operated in the filtering mode, Q2 was used as a linear ion trap to collect ions passing through Q1 and to generate a representative mass spectrum using the standard mass-selective instability mode. In other words Q2 was used to test, by recording a standard mass spectrum, the success with which Q1 operated as a filter for the ions of interest. With no ac excitation applied to Q1, the full spectrum of the sample mixture was acquired by Q2, as shown in Figure 5a scan 1. In the ion isolation experiments, a broad-band ac waveform was applied with a frequency notch corresponding to a q value of 0.7 and a width of 20 Th (m/z units, Thomson).52 An rf voltage of appropriate magnitude was applied to Q1 in order to associate protonated methamphetamine (m/z 150) with this isolation q value. The mass selection was validated by transmitting the trapped ions in Q2, then scanning Q2 to provide real-time feedback which was useful for the optimization of the ac waveform amplitude for both isolation efficiency and mass resolution. A mass spectrum of the isolated ions of m/z 150 is shown in Figure 5a scan 3. Protonated cocaine ions (m/z 304) were also continuously isolated with a notch width of 80 Th using a q value of 0.4 as shown in Figure 5a scan 2. The frequency distribution of the selected ions, viz. the isolation resolution, was studied by varying the Q1 rf amplitude in order to change the selected ion frequency range. This experiment can be mathematically described as a convolution of the frequency distribution function over a rectangular function (the frequency notch). The maximum isolation efficiency was (52) Cooks, R. G.; Rockwood, A. L. Rapid Commun. Mass Spectrom. 1991, 5, 93–93.
1838
Analytical Chemistry, Vol. 81, No. 5, March 1, 2009
obtained when the frequency notch covered the entire frequency distribution associated with the ion(s) of interest. The distribution of frequencies in the isolated ions can be interrogated by plotting the signal as a function of frequency (using a relatively narrow band of frequencies), and this data is shown in Figure 5, parts b and c, for m/z 150 and m/z 304. Ion frequencies plotted in units of Th which is inversely proportional to the q value at each point. With a frequency notch width of 20 Th for isolation of m/z 150, a plateau in the plotted intensity occurs at the maximum efficiency showing that the frequency distribution of the transmitted ions is less than this notch window width; hence, this width is sufficiently wide to capture the entire population of ions of m/z 150. The same experiment showed the frequency distribution for the population of m/z 304 is ∼80 Th at q ) 0.4. From the previous discussion, the ion excitation bandwidth is related to excitation time and the damping term c with an appropriate excitation voltage chosen to cause the displacement x of a resonant ion to reach the value x0. The optimum isolation efficiency was measured by comparing the isolated peak area to the same peak area in the full mass spectrum. Isolation efficiencies of 63.5% ± 1.5% and 41.8% ± 0.3% were achieved for single-ion filtering for 150 and 304 Th, respectively, as determined by triplicate measurements. The experimental data from the RIT are comparable to the analytical solution from a pure quadrupole excitation model as shown in Figure 4b. This suggests that at higher pressures the geometrical imperfections are increasingly overwhelmed by the gas collision effect. Note again that the relatively poor isolation resolution is now understood to be mainly a consequence of the short length of the RIT waveform isolation unit, Q1. In a test of performance of continuous mass-selective filtering, a probe-mounted surface was inserted through an isolation
Figure 6. Continuous waveform multipeak isolation from a drug mixture: (a) full-scan mass spectrum; (b) simultaneous isolation of ions of m/z 150 and m/z 304.
chamber without breaking vacuum and used as the substrate for ion SL. This allowed a mass-filtered ion current to be delivered to the surface after passage through Q2, which was operated as an rf-only ion guide. Due to buffer gas collisional cooling at 5 mTorr, the soft-landing kinetic energy was mainly controlled by the potential difference between Q2 and the surface. In the SL current measurement experiments, stainless steel was used as the SL surface and the potential difference was set at 5 V. Without an isolation signal applied to Q1, the total ion current from the mixture measured in front of Q2 at the intertrap lens was 170 pA and the current on the SL surface was 110 pA; for each measurement, the dc voltage applied to Q2 was adjusted to maximize signal. For all current measurements, the noise was
