Letter pubs.acs.org/ac
Dipolar Direct Current Driven Collision-Induced Dissociation in a Digital Ceramic-Based Rectilinear Ion Trap Mass Spectrometer Liang Wang, Fuxing Xu, and Chuan-Fan Ding* Department of Chemistry, Fudan University, Shanghai, 200433, China S Supporting Information *
ABSTRACT: A digital ion trap (DIT) and rectilinear ion trap (RIT) have been proven to be very useful technology in the past years. In this work, the digital ion trap technology was combined with the ceramicbased rectilinear ion trap (cRIT) system. The rectangular waveform was used for ion trapping. A dipolar excitation waveform which was formed by dividing down the trapping rectangular waveform was used for the ion ejection. We found that the high efficient collision-induced dissociation (CID) procedure could be obtained by simply manipulating the duty cycle of the dipole excitation waveform, and it could significantly simplify the tandem mass spectrometry analysis method and procedure with an ion trap, since the dipolar direct current (dc) voltage could be easily produced and applied to one of the pair of electrodes, which was fully controlled by the computer software and does not need any hardware modification. fixed qz, high precision control over the mass scan, lower power requirement than the sinusoidal waveforms, and so on. Although the ion trapping and ejection have been realized by a rectangular waveform, the CID experiment still cannot be completed digitally. In the previous study of DIT,19 the CID was realized by applying a sinusoidal dipole excitation waveform with a fixed frequency. This excitation frequency was lower than the secular frequency of the ion for limiting the amplitude of ion vibration. In this letter, we present a new method for CID experiments in a digital ceramic-based rectilinear ion trap mass spectrometer. 24 A well-defined dipolar dc potential (including amplitudes, duration time, phase) was applied to one pair of x-electrodes for ion excitation and collision induced dissociation by changing the duty cycle of the excitation waveform. The dipolar dc was simply produced by computer software without an additional power supply and any other experimental setup modification.
ince its first demonstration in a quadrupole ion trap mass spectrometer in 1987, collision-induced dissociation (CID) by resonance excitation has become the most popular method for tandem mass analysis.1−6 The method was usually accomplished by applying a small supplementary ac potential with the same frequency as the secular frequency of the parent ion to the end-cap electrodes of the ion trap. The singlefrequency resonance excitation will give rise to the first generation product ions that could be actively cooled-down by intermolecular collision with the bath gas inside the trap. Several other methods, such as the dc potential using ion excitation to improve the performance of the collisionalinduced dissociation, were studied and employed during the past decade.7−13 Dipolar dc (DDC) collision-induced dissociation technology in a quadrupole ion trap (QIT) was recently introduced by the McLuckey group. Their new technology provided a nonresonant means for ion acceleration by displacing the trapped ions from the center of the ion trap,14−16 so the precursor ions and all product ions were maintained at the same excitation condition with the probability of consecutive dissociation. Digital ion trap (DIT) technology was first introduced by Ding et al. in 2002 and has shown some special advantage compared to the traditional sinusoidal trapping waveform.17−23 For DIT, the rectangular waveforms were used to drive the ion trap. Instead of ramping the voltage of the sinusoidal trapping waveform, the digital ion trap performs ion trapping and mass selected ejection by scanning the frequency or changing the period of the rectangular waveform. The trapping and excitation waveform are all from the same digital circuitry. The digital waveform parameters, such as period, duty cycle, and relative phase, can be easily and precisely controlled. Many advantages can arise from the digital ion trap such as the capability of high mass analysis with a low trapping voltage and
S
© 2013 American Chemical Society
■
EXPERIMENTAL SECTION Instrumentation. All experiments were performed on a homemade three-stage differential pumping vacuum system as previously described (Figure 1).25 Briefly, ions from the electrospray (ESI) source were first introduced into a rf-only quadrupole ion guide and then into the ceramic-based rectangular ion trap (cRIT) through the aperture on the front end-cap electrode from the axial direction. In this work, a cRIT with x0 × y0 = 6 mm × 5 mm was used, and for details of its configuration, one can resort to ref 25. Helium gas was used Received: October 26, 2012 Accepted: January 8, 2013 Published: January 8, 2013 1271
dx.doi.org/10.1021/ac3031256 | Anal. Chem. 2013, 85, 1271−1275
Analytical Chemistry
Letter
Precision Instruments, Inc.) at a flow rate of 1 μL/min and a high voltage of ESI (typically 4 kV) was applied. The voltage of the orifice and skimmer was 150 V and 35 V, respectively. The electrospray ions directly went into the orifice which connected to the first differentially pumped region without a curtain gas.
■
RESULTS AND DISCUSSION For conventional sinusoidal waveform-driven ion traps, the dc voltage should be superimposed on the main trapping field. So an additional dc power supply is needed. In the digital ion trap, the rectangular waveform is generated by the rapid switching between a high voltage level V1 and a low voltage level V2. The U and V (stand for dc and rf voltages) values can be expressed with V1, V2, and duty cycle values, as shown in eq 1:
Figure 1. Schematic diagram of the ion optics and mass analyzer configuration, indicating the pressure regimes in each vacuum stage.
U = dV1 + (1 − d)V2
as the buffer gas for cooling ions and performing collisioninduced dissociation (CID) experiments. The pressure of buffer gas was 1−2 × 10−5 Torr for a normal mass scan and 5−8 × 10−5 Torr for tandem mass spectrometry (MS/MS). A channeltron electron multiplier (CEM 4879, Burle/Photonis) was used as the ion detector. The rectangular waveform power for the digital ion trap mode system is mainly composed of two parts: the switches and the digital waveform generator. The switches were used for creating the high voltage digital trapping waveform constructed with power MOSFETs. The upper and lower dc levels were provided by two high-accuracy dc power supplies (DCS6001.7E, Sorensen power supplies, Elgar Electronics Corporation). The amplitude of digital rf can be adjusted freely between ±500 V. The digital waveform generator was specially designed based on direct digital synthesizing (DDS) technology. It can provide precise control of the waveform period at a resolution of 50 ps both for trapping waveforms and resonance excitation waveforms. A balanced rectangular waveform with the same amplitudes but opposite in phase was applied to the x and y electrode pairs, respectively. The excitation waveform was coupled to the digital trapping waveform similar to the conventional rf mode. Since the voltage of the digital trapping waveform is fixed, the mass scan in the digital ion trap mode is a period scan process as follows: starting at an initial waveform period, Tstart, the scan proceeds by incrementing the period by a fixed value Tstep after generating a certain number of waveforms. In addition, the frequency of the resonance excitation was derived digitally by dividing down the frequency of the trapping waveform. There is a specific relation between trapping and excitation waveform frequency values which can be expressed in the beta value. Lastly a data acquisition system was used to get a mass spectrum in this mode. The content of the way for the digital waveform generating and scanning was depicted clearly in Ding’s paper.18−20 In our work, the amplitude of trapping waveform was kept at 500 V0‑p. Subjected to the circuit design, the duty cycle of the trapping waveform has a limited range (from 31.25% to 68.75% for our equipment). However, the amplitude of excitation waveform is much lower, and its duty cycle can be set from near 0% to close to 100%. Materials. The reserpine was purchased from AladdinReagent Ltd. (Shanghai, China). Solutions of 5 × 10−5 M for reserpine were prepared in 50:50 methanol/water solutions with 0.5% acetic acid. The sample solution was pumped into an ESI capillary with an i.d. of 100 μm (TSP100200, Polymicro Technologies, L.L.C) by a syringe pump (SP100i, World
V = 2(V1 − V2)(1 − d)d
(1)
The duty cycle was defined as the ratio between t and the total period T of the waveform (shown in Figure S1, Supporting Information). For digital trapping waveform, the amplitude of V1 was equal to V2 but the phase was opposite (V1 = −V2), so there is a relation between the U and V values of the dc and ac components and the duty cycle d of the waveform, as indicated in eq 2 below: U 2d − 1 = V 4d(1 − d)
(2)
The rectangular waveform can be manipulated by the control of software, and the switch timing of the rectangular waveforms can be easily and precisely controlled. The well-defined dc component applied to the electrodes can be obtained by adjusting the duty cycle (d) of the waveform. It has been demonstrated that the ion isolation and stability diagram mapping were achieved by variation of the duty cycle of the trapping waveform in DIT.26−28 In both cases, the dc offset was applied to the main trapping field to change the a-working point (value of “a” in the stability diagram of the quadrupole ion trap). According to eq 2, the value of U (dc value) was the same when d was x% and 100% − x% (i.e., d = 10% and d = 90%) for dipolar dc. For DIT, the dipolar excitation waveform is derived digitally from the trapping waveform by dividing down the frequency of the trapping waveform. For resonance ejection, the division rate n = 3 were employed at fixed β values of 2/3 which may be a nonlinear resonance point for our cRIT. The oscilloscope (DPO 3034, Tektronics, Inc.) traces of the trapping and excitation waveforms (see Figure S2, Supporting Information). Similarly, a well-defined dc component can be obtained by changing the duty cycle of the excitation waveform without any additional dc power supply. In our experiment, two opposite phases of rectangular waveforms were symmetrically applied to the x- and y-pairs of electrodes. Ions in the trap will eject from the x-direction, and a dipolar excitation voltage was coupled to the trapping waveform of the x-electrodes for ion ejection. When the duty cycle of the dipolar excitation waveform deviates from 50%, an offset voltage of ±n V (a pair of dc voltages with the same amplitude and opposite phase, namely, dipolar dc) will be set on the x-pair of the electrodes, respectively, as shown in Figure 2. There is no hardware change for the instrument system, and the circuit is also much simpler. Since the amplitude of the dipolar excitation waveform was fixed, different duty cycle 1272
dx.doi.org/10.1021/ac3031256 | Anal. Chem. 2013, 85, 1271−1275
Analytical Chemistry
Letter
isolation of the precursor ion, the trapping waveform period was fixed at 1.2 μs and the β value was set at 0.3478 which was deviated from the resonance excitation point. The digital dipolar excitation waveform with different duty cycle values was applied, and the duration of the CID course was controlled at 36 ms. The amplitude of the dipolar excitation waveform was 3.3 V. During the experiment, the duty cycle of trapping waveform were kept at 50% all the while. Figure 3b−f shows the different CID results when the duty cycle of the dipolar excitation waveform was 82.82%, 73.06%, 67.36%, 66.00%, 64.65% separately, and the dipolar dc amplitude can be calculated based on eq 2 as follows in turn: 3.75 V, 1.90 V, 1.28 V, 1.15 V, and 1.04 V. During these values, the duty cycle of 64.65% gave the highest CID efficiency. No fragmentation was observed even varying other parameters when the duty cycle of the dipolar excitation waveform was 50%. It is interesting that the CID efficiency become lower with increasing the dc amplitude probably because for the lower duty cycle dc voltage was too small to fragment the precursor ion; but for a much higher duty cycle, the precursor ion could be almost ejected from the trap before fragmentation. There is an optimal duty cycle for dc driven CID. For reserpine the β value was 0.3478 in Figure 3, 64.65% was the optimal value. Here CID efficiency is defined as the ratio of the total fragment intensity to the precursor ion intensity prior to the fragmentation. Furthermore, several other β values of the dipolar excitation waveform were chosen for the CID experiment, as shown in Figure 4. When β was 0.4324, 0.4, 0.3721, and 0.3333, respectively, comparative high efficiency CID mass spectra were obtained by varying the duty cycle values. Compared to the previous work about DDC collision-induced dissociation in a 3D ion trap, a lower amplitude and a longer duration were used. In Figure 4a−d, different fragmentation patterns were observed under different conditions. We are not sure whether the dc potential is the crucial factor for the fragmentation patterns. Other parameters such as trapping waveform
Figure 2. Schematic of circuit for cRIT in the digital cRIT mass spectrometer.
values will give different amplitudes of the offset dc. A series of CID mass spectra were shown in Figure 3 under several given duty cycle values of the dipolar excitation waveform. The amplitude of the trapping waveform is 500 V0‑p, while the amplitude of excitation waveform was much lower (
