Anal. Chem. 2010, 82, 746–750
Technical Notes Inverse Ion Mobility Spectrometry Mahmoud Tabrizchi* and Elham Jazan Department of Chemistry, Isfahan University of Technology, Isfahan 84156-83111, Iran A novel method is proposed for enhancing the separation power of ion mobility spectrometry (IMS) and other similar pulsed techniques, such as time-of-flight mass spectrometry. In this technique, rather than generating an ion packet, a dip is created in the ion beam. This is achieved by an inverse pulse applied to the shutter grid. The dip moves with the same velocity as the ion packet, and the detector reads an inverse peak at the same drift time as that of the normal operation. Using this technique, we achieved 30-60% higher resolution compared to the normal method. In addition, two close peaks that were not resolved via normal IMS were well resolved to the baseline using this technique. The main reason for the increased resolution is likely the absence of space charge in the dip. Traditionally, pulsed techniques play a very important role in chemistry. They are commonly used in vital analytical instruments, such as time-of-flight (TOF) mass spectrometers and ion mobility spectrometers. In ion mobility spectrometry (IMS), an ion packet is moved in a gas by constantly applying an electric field to separate ions of different size. IMS is an analytical method that is used to investigate the composition of gas mixtures.1 It is a powerful tool for the study of molecular conformations, the separation of mass isomers, and the analysis of complex mixtures.2 Similar to TOF mass spectrometry, ions in IMS are dispersed in a time dimension but under atmospheric pressure. More specifically, the sample is ionized using various atmospheric pressure ionization techniques. The ions are then moved in an inert gas under a constant electric field. Although ions are continuously generated, the ion beam is chopped by a shutter grid to create an ion packet. To generate an ion pulse, the grid potential is removed for a short period of time by a pulse generator. The penetrating ion pulse moves in the drift region, where ions are separated according to their individual velocities, which depend on their size. A full description of the method is given in several books and review articles.1,3-5 Its main advantages include a low detection limit, fast response, simplicity, portability, and relatively * To whom correspondence should be addressed. E-mail: m-tabriz@ cc.iut.ac.ir. (1) Eiceman, G. A.; Karpas, Z. Ion Mobility Spectrometry, 2nd ed.; CRC Press: Boca Raton, FL, 2005. (2) Laiko, V. V. J. Am. Soc. Mass Spectrom. 2006, 17, 500–507. (3) Carr, T. W. Plasma Chromatography, Plenum Press: New York, 1984. (4) Eiceman, G. A. Crit. Rev. Anal. Chem. 1991, 22, 17–36. (5) Louis, R. H. St.; Hill, H. H. Crit. Rev. Anal. Chem. 1990, 21, 321–343.
746
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
low cost. A broad range of compounds, such as explosives,6 narcotics,7 herbicides,8 pesticides,9 commonly abused drugs,10 and biochemicals,11,12 have been detected by IMS. In addition, IMS has been used in many fundamental studies in the field of polymerization, including the determination of the initiation mechanism and structures of the early oligomers,13 as well as determination of the branching in oligomers.14 High-resolution ion mobility measurements can also be used for structural determinations, including the structural determination of metallofullerenes.15 Moreover, IMS has been successfully used to understand the oligomerization process for oligomer-specific therapeutic agents, such as amyloid β-protein, the key pathogenetic agent in Alzheimer’s disease.16 The unfolding, refolding, and hydration of proteins in the gas phase have also been studied using ion mobility measurements.17 The recent introduction of ion mobility into native mass spectrometry analyses (IM-MS) has added a new and exciting dimension to the field of structural biology. IM-MS allows for the assessment of gas phase ion collision cross sections of protein complex ions, which can be related to overall shapes/volumes of protein assemblies, and thus, IM-MS can be used to monitor changes in structure.18-20 Recently, the mass-selected ion mobility (MSIM) technique was used for (6) Khayamian, T.; Tabrizchi, M.; Jafari, M. T. Talanta 2003, 59, 327–333. (7) Khyamian, T.; Tabrizchi, M.; Jafari, M. T. Talanta 2006, 69, 795–799. (8) Clowers, B. H.; Steiner, W. E.; Dion, H. M.; Matz, L. M.; Tam, M.; Tarver, E. E.; Hill, H. H. Field Anal. Chem. Technol. 2002, 6, 302–312. ¨ anninen, O. Anal. Chim. Acta 2000, 404, (9) Tuovinen, K.; Paakkanen, H.; H 7–17. (10) Keller, T.; Miki, A.; Regenscheit, P.; Dirnhofer, R.; Schneider, A.; Tsuchihashi, H. als> Forensic Sci. Int. 1998, 94, 55–63. (11) O’Donnell, R. M.; Sun, X.; Harrington, P. B. Trends Anal. Chem. 2008, 27, 44–53. (12) Jafari, M. T.; Rezaei, B.; Zaker, B. Anal. Chem. 2009, 81, 3585–3591. (13) Alsharaeh, E. H.; Ibrahim, Y. M.; El-Shall, M. S. J. Am. Chem. Soc. 2005, 127, 6164–6165. (14) Dietiker, R.; di Lena, F.; Chen, P. J. Am. Chem. Soc. 2007, 129, 2796– 2802. (15) Sugai, T.; Inakuma, M.; Hudgins, R.; Dugourd, Ph.; Fye, J. L.; Jarrold, M. F.; Shinohara, H. J. Am. Chem. Soc. 2001, 123, 6427–6428. (16) Teplow, D. B.; Lazo, N. D.; Bitan, G.; Bernstein, S.; Wyttenbach, T.; Bowers, M. T.; Baumketner, A.; Shea, J.; Urbanc, B.; Cruz, L.; Borreguero, J.; Stanley, H. E. Acc. Chem. Res. 2006, 39 (9), 635–645. (17) Jarrold, M. F. Acc. Chem. Res. 1999, 32, 360–367. (18) Van Duijn, E.; Brendregt, A.; Synowsky, S.; Versluis, C.; J. R. Heck, A. J. Am. Chem. Soc. 2009, 131, 1452–1459. (19) Baker, E. S.; Hong, J. W.; Gidden, J.; Bartholomew, G. P.; Bazan, G. C.; Bowers, M. T. J. Am. Chem. Soc. 2004, 126, 6255–6257. (20) Baker, E. S.; Bushnell, J. E.; Wecksler, S. R.; Lim, M. D.; Manard, M. J.; Dupuis, N. F.; Ford, P. C.; Bowers, M. T. J. Am. Chem. Soc. 2005, 127 (51), 18222–18228. 10.1021/ac902009c 2010 American Chemical Society Published on Web 12/28/2009
the study of polymerization in the gas phase, in clusters, and on nanoparticle surfaces.21 As for other separation techniques, the ability of the IMS instrument to resolve two closely spaced peaks is of considerable interest. Generally, IMS is not a high-resolution technique. Many attempts, including increasing the length of the drift tube,1,22 reducing the pressure,23 increasing the electric field,24 and using different drift gases,25,26 have been made to enhance the resolution of IMS. The resolution of IMS and its dependency on various experimental parameters, such as temperature, the applied voltage, and the pulse duration, have been the subject of a number of studies. The effects of temperature27 and pressure28 on the resolution of IMS have previously been investigated. In this work, we describe a completely different approach and a novel solution to enhance the resolution of IMS. The method simply uses an inverse pulse applied to the shutter grid to create a dip in the ion beam rather than creating an ion packet. This method was practically evaluated using test compounds, and considerable improvement of resolution was achieved. EXPERIMENTAL SECTION The ion mobility spectrometer used in this study was constructed in our laboratory at the Isfahan University of Technology. The ionization region consisted of five 9.5 mm thick aluminum rings, with 20 mm ID and 55 mm OD. The drift tube consisted of 11 aluminum rings with the same OD size and a 36 mm ID. Each ring was connected to the adjacent one via a 5 MΩ resistor to create a potential gradient. A continuous corona discharge ionization source was used with a point-to-plane geometry, as described elsewhere.29 A Bradbury-Nielsen type gate30 was mounted between the ionization region and the drift tube. A 100-800 µs gate pulse was applied, and a Faraday cup type collector plate with an aperture grid was used to register the ion current. The ion current received on the collector plate was amplified by an electrometer with a gain of 109 V A-1 and then relayed to a computer via an A/D converter (Picoscope, UK). The digitized signal was averaged over a number of scans, and the resulting ion mobility spectrum was then displayed on the monitor. The IMS cell was housed in a thermostatic oven where the temperature could be adjusted from room temperature to 473 K within ±1 K. Nitrogen gas, after passing through a 13X molecular sieve (Fluka), was passed through the cell at 300 and 700 mL min-1 for the carrier and drift gases, respectively. The spectrometer was operated in the positive mode with various drift fields ranging from 375 to 625 V cm-1. The compounds selected for this experiment were purchased from (21) El-Shall, M. S. Acc. Chem. Res. 2008, 41 (7), 783–792. (22) Brokenshire, J. L. FACSS Meeting, Anaheim, CA, October, 1991. (23) Dugourd, P. H.; Hudgins, R. R.; Clemmer, D. E.; Jarrold, M. F. Rev. Sci. Instrum. 1997, 68, 1122–1129. (24) Leonhardt, J. W.; Rohrbeck, W.; Bensch, H. Forth International IMS Workshop, Cambridge, UK, 1995. (25) Asbury, G. R.; Hill, H. H., Jr. Anal. Chem. 2000, 72, 580–584. (26) Tabrizchi, M.; Khayamian, T. Int. J. Ion Mobility Spectrom. 2001, 4, 52– 56. (27) Tabrizchi, M. Talanta 2004, 62, 65–70. (28) Tabrizchi, M.; Rouholahnejad, F. Talanta 2006, 69, 87–90. (29) Tabrizchi, M.; Khayamian, T.; Taj, N. Rev. Sci. Instrum. 2000, 71, 2321– 2328. (30) Bradbury, N. E.; Nielsen, R. A. Phys. Rev. 1936, 49, 388–393.
Table 1. Comparison of the Resolving Power for Normal and Inverse Modes for T ) 303 K and a 375 Vcm-1 Drift Field resolution compound
td (ms)
normal
inverse
improvement
RI acetophenone butanone pentanone methyl isobutyl ketone (MIBK) 2,4-dimethyl pyridine (DMP)
12.78 19.28 16.20 16.96 18.13
49.2 58.4 50.6 62.8 53.3
79.9 80.3 81.0 80.8 75.5
62% 37% 60% 29% 42%
18.25
60.8
82.9
36%
Fluka and used without further purification (See Table 1). Pure morphine and codeine were prepared from Temad Co (Iran). RESULTS AND DISCUSSION The gate consisted of two sets of parallel wires mounted on a ceramic frame in a plane perpendicular to the moving direction of ions. When an adequate potential was applied to the wires, the electric field created between the wires could stop penetration of ions; hence, the gate was closed. Conversely, removing the potential opened the gate. In the normal operation, the gate was mostly closed, and it was open only for a short period of time. Figure 1a shows the shape of the applied pulse to the gate for normal operation, as well as the peak shape resulting from such a pulse. In this case, the shutter grid allows a swarm of ions to diffuse into the drift region, forming an ion packet. However, if the applied electric pulse is inversed, as shown in Figure 1b, the shutter grid is always open except for a short time. Unlike the case of normal operation, this situation creates a dip in the ion beam. As seen in Figure 1, the depth and width of the dip are reduced in comparison with the peak. The resolution, based on a singlepeak definition (t/∆t), is increased from 44 to 72, i.e., or by 63%. The improvement of resolution could be partially due to the lower intensity of the inverse peak. In order to more accurately investigate the change in resolution, the lower intensity of the inverse peak was compensated for by widening the pulse width to obtain the same intensity as the normal one. The dip was then
Figure 1. Ion mobility signal for acetophenone recorded in (a) normal operation and (b) inverse operation. The applied pulse to the shutter grid is shown for each case. Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
747
Figure 2. Increase in resolution due to application of an inversed pulse to the shutter grid for acetophenone.
inverted and plotted, along with the peak, in Figure 2. Both signals were normalized to unity to adjust for the small differences in their intensity. Clearly, the inverse peak is narrower than the normal peak. In this case, the resolving power shows an increase of about 37%. The experiment was repeated for several compounds. In all cases, the resolving power was enhanced considerably, as tabulated in Table 1. For the cases of the reactant ion (RI) and butanone, the increase was as high as 60%. A typical way to increase resolution in the normal operation is to apply a narrower pulse to the shutter grid. However, reducing the pulse width results in a weaker signal, i.e., resolution is gained at the expense of less intensity. Figure 2 and Table 1 demonstrate that by inverting the shutter grid pulse; it is possible to gain resolution without decreasing the intensity. The effect of pulse width on resolution and peak height is compared for both normal and inverse modes of operation in Figure 3. In both the inverse and normal cases, the resolution can be further increased by reducing the pulse width. However, the resolution is always better for the inverse mode at the same pulse width as that in the normal mode. In addition, for the inverse case, there is an optimum pulse width for gaining maximum resolution. The intensity is mostly better for the case of normal operation. Figure 4 is the plot of resolution versus the peak height recoded at various pulse widths for both normal and inverse mode.
Figure 4. Resolution versus the peak height for acetophenone peak recorded at various pulse widths in normal and inverse operation.
Figure 4 shows that, at the same intensity, resolution is better for the inverse mode and, for the same resolution, the peak intensity is higher for the inverse mode. Interestingly, the height of the best-resolved peak in the inverse mode was 10 times higher than the narrowest peak in the normal mode as demonstrated in Figure 5. The reason for the narrower peak in the inverse mode is not known. However, it may be explained by considering the space charge repulsion between the ions. The importance of space charge on resolving power in IMS was recently described by Mariano et al.31 In the case of normal operation, a swarm of ions is traveling, and the ions diffuse out of the swarm during their travel, which results in broadening. In addition, the ions inside the swarm push each other away, causing further broadening. In fact, diffusion and repulsion both contribute to the broadening. However, in the case of inverse operation, the width of the traveling dip tends to decrease because of the repulsion between the ions outside the dip. This is demonstrated in Figure S-1 (in the Supporting Information). In reality, the dip between two long ion pulses, which tend to expand, is thus compressed, as opposed to the tendency of broadening due to diffusion. As a result, higher resolution is achieved in the inverse mode. This assumption was
Figure 3. Plots of resolution and peak height versus pulse width for acetophenone in inverse and normal operation modes. 748
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
Figure 5. Comparison of the best resolved peaks for acetophenone in the normal and inverse mode.
Figure 6. Dip and peak recorded at a very large pulse width (830 µs) for acetophenone.
evaluated by recording the spectra at a very large pulse width (830 µs) for both inverse and normal modes. The resulting peak and dip are presented in Figure 6. The full width at half-maximum (fwhm) for the peak is 1460 µs, which is 1.76 times larger than that of the initial pulse. This is
quite natural due to the diffusion of ions out of the initial packet. Surprisingly, the width of the dip is less than that of the initial width. More specifically, it is reduced from 830 µs to about 400 µs. This confirms that repulsion is acting against diffusion in the inverse mode. For two overlapping peaks, we expect a better separation when IMS operates in the inverse mode. This hypothesis was examined for two test compounds that produce two close peaks. Morphine and codeine both show a double peak in their ion mobility spectra. Figure 7 compares their spectra for the cases of normal and inverse operation. The normal spectra were recorded at optimum conditions for resolving the two peaks as much as possible. The inversed spectra were recorded under conditions that resulted in almost the same intensity as that in the normal mode. Obviously, unlike in the normal mode, the overlapping peaks are resolved to the baseline in the inverse mode. In the normal mode, the space between the two close peaks is readily filled due to diffusion and repulsion inside the two adjacent ion packets. However, in the inverse mode, in which two neighboring dips are traveling, a layer of ions is trapped between the two dips. These ions are under a sandwich force from the ion clouds on both sides. As a result, the trapped swarm of ions gets narrower, thus increasing the baseline resolution for the two neighboring dips. Two additional experiments were performed to show the ability of the inverse mode of operation in resolving two adjacent peaks. In the first experiment, a spectrum of morphine was recorded under nonoptimal separation conditions in the normal mode such that the two peaks did not exhibit a clear valley. Then, the shutter grid pulse was inverted. Spectra corresponding to this experiment is shown in Figure S-2 (in the Supporting Information). Comparison of the two spectra indicates that the use of the inverted mode does in fact separate the two peaks. This is a clear indication of enhanced resolution regardless of differences in baselines, signal levels, etc. The second experiment was performed with a more complex sample (more than one component). The proton-bound dimer peak of 2,4-dimethyl pyridine (DMP) and methyl isobutyl ketone (MIBK) are very close to each other. A mixture of MIBK and DMP was recorded in two different modes of operation under optimal separation conditions in the normal mode such that the
Figure 7. Ion mobility spectra of morphine and codeine in normal and inverse mode. The overlapping peak is well resolved using the inverse mode. Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
749
two peaks exhibit the best valley (Figure S-3, Supporting Information). Then, the shutter grid pulse was inversed. As demonstrated in Figure S-3 (in the Supporting Information), the two peaks are better separated in the inversed mode. This illustrates that the enhanced resolution provides an improved resolving power, which has a great analytical utility. CONCLUSION IMS was first introduced in 1970 under the name of “plasma chromatography.”.3 Thus, linear-IMS can be conceptualized as an analog of the gas chromatography method in which molecules of different sizes under the influence of a carrier gas flow move through a column of defined porosity at different rates. The species of interest in IMS are separated while they travel in a gaseous medium. The major difference is that the species are charged and derived by an electric field. As with chromatography, the peak in IMS tends to broaden, largely due to diffusion. In addition, the charge on the species in IMS causes further broadening. In this work, we demonstrated that the charge of the species, which is the origin of further broadening, can be employed to counteract the broadening by applying an inverted (31) Mariano, A. V.; Su, W.; Guharay, S. K. Anal. Chem. 2009, 81 (9), 3385– 3391.
750
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
pulse to the shutter grid. This creates a gap that is traveling in an ion bath in which repulsion favors narrowing of the gap. This method is a simple but very effective way of increasing resolution. It is a promising technique for enhancing the utility of IMS-MS in structural analysis, where ions are first separated in a drift tube based on their size/shape and the ions then enter the mass spectrometer. In addition, this method may be applied to other similar pulsed techniques dealing with charged particles, including TOF mass spectrometry. ACKNOWLEDGMENT This work was financially supported by Isfahan University of Technology (IUT), Iran. The Center of Excellence for Sensor and Green Chemistry of IUT also deserves our gratitude for their support. Authors are grateful to Prof. G. Spangler for valuable discussions. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review December 4, 2009. AC902009C
September
6,
2009.
Accepted
