Resolution Enhancement of Ion Mobility Spectrometry by Improving

Dec 23, 2011 - Unfortunately, in reality, this is impossible to ... function, which is retained by the leading edge of an ion packet. Also, the traili...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/ac

Resolution Enhancement of Ion Mobility Spectrometry by Improving the Three-Zone Properties of the Bradbury-Nielsen Gate Yongzhai Du,†,‡ Weiguo Wang,† and Haiyang Li*,† †

Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, People’s Republic of China Graduate School of Chinese Academy of Sciences, Beijing 100049, People’s Republic of China

‡

S Supporting Information *

ABSTRACT: A simple space compression-dispersion model for ion transport at ambient pressure was mathematically established. On the basis of this model and aided by SIMION simulation, a three-zone theory was proposed to characterize the Bradbury-Nielsen gating electric field features as three zones: the depletion zone, the dispersion zone, and the compression zone. Then, the influences of gating voltage difference increases on the full width at half-maximum of the Cl− peak were investigated in detail to verify the theory. For example, at a gating voltage difference of 350 V and a gate pulse width of 0.34 ms, the ion packets injected were reduced to as low as 60% of their original widths, with the peak height increased from 756 to 808 pA and the resolution from 18 to 33, enhanced by 7% and ∼80%, respectively. The ion mobility spectrometry (IMS) efficiency ratios, Rm/Rc and Rm/Rp, were also raised above theoretical values and reached about 182% and 175%, respectively. The experimental results were explained using the proposed theory with good consistency. Finally, a compression coefficient was extracted by fitting the experimental data to the applied gate pulse width, presenting a good linearity. All this shows a potential application in improving the performances of ion mobility spectrometry.

I

on mobility spectrometry (IMS),1,2 basically an ion separation technique at ambient pressure, is now widely employed for field screening chemicals like explosives and related compounds,3−6 chemical warfare agents,7−9 illicit drugs,3,4 etc. Over 65 000 standalone IMS instruments have been deployed at airports all over the world and in the militaries.1,10 This occurs due to the satisfactory features of IMS: compactness of instrumentation, good portability, fast response, low cost, and especially the high sensitivity working at ambient pressure. However, accompanying these merits comes the biggest disadvantage of this technology: the low resolving power and the consequent low selectivity and high false alarming rate. They have always been the concerns of IMS researchers trying to improve the separation capability while keeping its sensitivity. The measured resolution Rm of a conventional BradburyNielsen gate11 IMS drift tube is defined for a single peak as the drift time td of an ion species divided by the full width at halfmaximum (fwhm) of it.12

the following form with the initial pulse width.14

Rc = =

td (1) fwhm Considering the diffusion effect tdiff and neglecting the space charge effect and other factors,13 fwhm can be expressed as14 (2)

td 2

tg + (16kT ln 2/Vdez) ·td 2

(3)

Received: November 15, 2011 Accepted: December 23, 2011 Published: December 23, 2011

where tg is the gate pulse width to admit ions into the drift region. Then, the calculated resolution Rc can be turned into © 2011 American Chemical Society

tg 2 + (16kT ln 2/Vdez) ·(Ld 2 /KVd)2

where Ld, Vd, and T are the drift region length, drift voltage drop, and temperature, respectively; k, e, and z are Boltzmann’s constant, unit charge, and charge number the ions carry, respectively. Many factors influence the peak width,13,15 including the initial pulse width and shape, diffusion, 16 Coulombic repulsion,17−19 ion−molecule interactions,20 inhomogeneity of the drift field,21,22 temperature,23 pressure, etc.24 Besides, lengthening the drift region along with increasing the drift voltage and analyzing large and multiple charged ions will create very high resolutions.14,20,25,26 Actually, this kind of design is reducing the relative contribution of gate pulse width (tg) in eq 3 by lengthening the drift time, thus obtaining resolutions approaching the diffusion limit resolution.14,25 They have extended the application scope of ion mobility

Rm =

fwhm 2 = tg 2 + tdiff 2

Ld 2 /KVd

1725

dx.doi.org/10.1021/ac203013u | Anal. Chem. 2012, 84, 1725−1731

Analytical Chemistry

Article

Figure 1. BN gating field features. (a) SIMION model for the simulation and data extraction. GVD = 200 V. (b) Topographic of the three zones of Ey in the green-box area in (a). E0 = 50 V/mm, GVD = 200 V. (c) Ey curves along x1 = 20 mm and x2 = 21 mm at GVD = 0, 50, 100, 150, and 200 V, respectively.

is suddenly shut, slope less abruptly than a step function.14 This gating behavior has been proved by Tadjimukhamedov et al. and Puton et al. in their respective ion shutter simulations.10,32 From their results, it appears to be certain that ion peaks in BN gate drift tubes are always wider than the applied gate pulse width. On the basis of eq 3, Hill’s group introduced three parameters α, β, and γ to fit the measured resolutions as

spectrometry; however, they are not suitable for portable field screening deployment due to the size and weight increases as well as higher operating costs. Usually, narrowing the gate pulse width to inject thinner ion packets into the drift region improves the resolution, yet at the cost of sensitivity. Spangler et al.16 calculated and suggested 200 μs as a best compromise between resolution and sensitivity, and now in most situations, pulse widths of 100−300 μs are adopted.25 For larger ions, wider pulse durations are usually applied to increase sensitivity.27 Very little improvement in the resolution of standalone IMS instruments has been made since the first commercial ion mobility analyzer appearing in 1970, which gave a resolution of 30 for dimethyl sulfoxide.28 In practice, most field screening IMS analyzers have drift lengths of 6−10 cm and the resolving power is 20−60 at gate pulse widths of 100−300 μs.24 Yet, only about 80−90% of the theoretical resolution by eq 3 was achieved in most standalone designs,6,29,30 except in some cases as high as about 100% was reported.30 Tadjimukhamedov et al. pointed out that, when ion densities are low and the purity of the drift gas is high, performance in most well-designed mobility spectrometers is governed by the ion gate waveform.24 Ions are pulsedly injected into the drift region using a Bradbury-Nielsen (BN) gate,11 which consists of two sets of interdigitated metal wires. Theoretically, ideal ion packets are required to be chopped into the drift region to produce high resolution. Unfortunately, in reality, this is impossible to accomplish using a BN gate. Aronson’s mathematic model31 has shown that the ion density across a closed gate drops to zero considerably less rapidly than a step function, which is retained by the leading edge of an ion packet. Also, the trailing edges of the ion packet, created when the gate

Rm =

td 2

γ + βtg + αTtd 2/Vd

(4)

They also found that their fitting results were all larger than the theoretical values.14,25 In addition, as is well-known, the BN gate features a depletion effect,33 which occurs as an intrinsic characteristic of its structure and mechanism. It is created when the blocking field between the two sets of wires penetrates into the drift region, with its boundary becoming deeper and approaching a limit when increasing the gating voltage.27,34 This penetrated field will block ions, not exactly at the center of but a short distance after the gate wire surface, from entering the drift region. Thus, the depletion effect results in narrowed ion packets and decreased ion quantities by “cutting back” some of the ions that have passed the gate wire surface. However, to our knowledge, no studies have been reported on the electric features around the depletion boundary until now. Actually, the BN gating behaviors are still not understood clearly enough. In this work, aided by the SIMION simulation,35−38 a three-zone theory was proposed to characterize the electric features after and at the vicinity of the BN gate as the depletion zone, the dispersion zone, and the compression zone. It was found that ion packets will get dispersed or 1726

dx.doi.org/10.1021/ac203013u | Anal. Chem. 2012, 84, 1725−1731

Analytical Chemistry

Article

of 1 mm, is placed at y = 6 mm and separates the gap into two regions: the reaction region and the drift region. The potential for one set of the gate wires was set constantly at 500 V, and the other set is applied with an extra gating voltage difference (GVD). Within this coordination system configuration, positive ions in the drift region will fly in the −y direction where the ycomponent of electric field Ey > 0 and toward +y direction where Ey < 0 (Ey = dV/dy, as is expressed in the SIMION software). Figure 1a displays the iso-potential diagram for GVD = 200 V. Clearly, at the vicinity of the 500 V wire set in the drift region, there are potential valleys formed, while at the 700 V wire set, there are potential ridges. To further understand how the potential valleys and ridges are connected, which should be important to understand the BN gating behaviors, the Ey within the green-boxed area (19.5 mm ≤ x ≤ 22.5 mm and 2 mm ≤ y ≤ 6 mm) for GVD = 200 V is extracted and plotted topographically in Figure 1b. According to our previous compression-dispersion model, the postgate region for a closed BN gate can be divided into three zones by two curves, shown in Figure 1b. The first curve has Ey = 0 on it, and we name it the depletion-dispersion curve, within which is the known depletion zone, where ions will fly back toward the gate wire plane because Ey < 0. The second curve has Ey = E0 = 50 V/mm and is named the dispersioncompression curve. We name the zone outside it the compression zone, where with Ey > E0 = 50 V/mm, ion packets will get compressed. The zone between the two curves meets 0 < Ey < E0 = 50 V/mm, and ion packets inside it will get dispersed and become wider; so, we name it the dispersion zone. These three zones combine to form the complex electric features after the BN gate. To further understand the evolution of the BN gate electric field with the GVD so as to explain its gating behaviors, Ey is extracted with the GVD being varied from 0 to 200 V at a step of 50 V. Data along two typical lines marked in Figure 1a, one at x1 = 20 mm and through the center of a 700 V gate wire and the other at x2 = 21 mm and through the center of a 500 V gate wire, are plotted as a function of y-coordinate in Figure 1c. By Figure 1b,c, the electric features after a closed BN gate are concluded as follows. As shown in Figure 1b, after the high-voltage wires (x1 = 20 mm), there is only a compression zone. However, after the lowvoltage wires (x2 = 21 mm), there is a depletion zone, a dispersion zone, and a compression zone appearing in turn into the drift region. (1) Interestingly, at different GVDs, all the Ey lines through the low-voltage gate wire at x2 = 21 mm (500 V) intercept the line Ey = E0 = 50 V/mm at the same point M, as shown in the inset of Figure 1c, which belongs to the dispersion-compression curve. It is the same case with all other points on the dispersion-compression curve, implying that the position of the dispersion-compression curve does not change with the GVD when E0 is fixed for a given BN gate in the present gate operation mode (See Figure S-1, Supporting Information, for more information proving this). (2) By Figure 1c, it is found that, with the increase in the GVD, the electric field in the compression zone is enhanced after all wires. The depletion zone is enlarged both in depth and width. At GVD = 50, 100, 150, and 200 V, the Ey lines at different GVDs intercept the line Ey = 0 at different points, marked as A, B, C, and D, with DC < CB < BA. As these points belong to the depletion-dispersion curve (by its definition), so the depletiondispersion curve approaches a limit with increase in the GVD.

compressed in certain zones after the gate. Increasing the gating voltage difference can enlarge the depletion zone, shorten the dispersion zone, and enhance the compression zone in our instrumental setup, thus reducing the gate pulse width contribution to the fwhm and enhancing the resolution and sensitivity. Experiments by varying the gate pulse width and gating voltage difference illustrated the improvements and verified the three-zone theory.

■

BN GATE FEATURES: THE THREE-ZONE THEORY Compression-Dispersion Model for Ambient Pressure Ion Transport. As shown in the TOC figure (left), at ambient pressure, a positive ion packet of the same species, which has a mobility coefficient K and an original width δ, is traveling along the x-axis direction in a homogeneous positive low electric field E0 ( 0 is still satisfied), which is similar to the gate shutting in an ion mobility spectrometer. Taking into consideration no other factors that influence the ion transport except the electric field, the drift time difference for the leading edge and trailing edge to reach the Faraday plate at x = d (d > δ, similar to that in IMS) can be expressed as Δti = ttrailing − tleading =

∫0

=

∫0

d

d dx dx − KE(x) δ KE(x)

δ

dx KE(x)

∫

(5)

A compression coefficient η can be defined and deducted as

η=

Δti E 1 1 = 0 = = tg E(ξ) 1 + Ei(ξ)/E0 1 + Ec /E0

(6)

ξ is a value on the x-axis which meets 0 < ξ < δ and the Lagrange’s Mean Value Theorem. Here, we define the item Ei(ξ) as the equivalent compression electric field (Ec). Obviously, increasing Ec reduces the value of η and thus enhances the compression. (1) If Ei = 0, then η = 1, indicating that the packet will keep its thickness and shape no matter how long a distance it travels; that is, δ′ = δ. The time span Δti = δ/ KE0= tg. (2) If Ei ≠ 0, then η ≠ 1; we have the following two situations: (a) When Ei > 0, the electric field becomes stronger, η < 1; Δti < tg, and the ion packet is compressed; the smaller is the value of η, the narrower the ion packet becomes. (b) When −E0 < Ei < 0, the electric field becomes weaker, η > 1; Δti > tg, and the ion packet is dispersed and becomes wider. Three-Zone Theory: SIMION-Aided Characterization of the BN Gating Behaviors. On the basis of the above theory, the electric features of a closed Bradbury-Nielsen gate were examined, aided by SIMION simulation with the model shown in Figure 1a. Two 1 mm-thick circular electric poles are positioned parallel at y = 0 mm and y = 11 mm and applied with a potential of 250 and 750 V, respectively. Thus, an electric field of E0 = 50 V/mm is established between them. A BN gate, the same as used in the following experiments,39 with a wire diameter of 0.1 mm and an adjacent wire-center distance 1727

dx.doi.org/10.1021/ac203013u | Anal. Chem. 2012, 84, 1725−1731

Analytical Chemistry

■

Article

RESULTS AND DISCUSSION To verify the above three-zone theory, experiments were carried out on the drift tube shown in Figure S-2, Supporting Information. The drift tube settings for the experiments are summarized in Table 1.

The depletion zone enlarges and the dispersion zone shrinks, with their total area unchanged. Predictably, these features will affect the peak width and signal intensity in the following situations. (1) With the increase in the GVD, the enlarged depletion zone, the reduced dispersion zone, and the electrically enhanced compression zone consistently help to produce narrower peaks. Obviously, the enhanced depletion zone will deplete more ions, resulting in reduced peak area. (2) The peak height (or sensitivity) reflects the ion density and should be more complex, depending on the initial gate pulse width (GPW, denoted as tg in equations) or, rather, ion packet width (δ = KE0tg). (a) Peak height increases: For wide enough GPWs, a small portion of the initial thickness lies in the depletion region and compression and reduced dispersion dominate, resulting in larger peak heights by their compensation for the depletion losses of ion quantities. (b) Peak height decreases: For narrower GPWs, a large portion of the initial thickness is located in the depletion zone so depletion dominates and a large portion of the ions are depleted, causing lowered ion quantities and peak heights, which cannot be compensated for by the other two effects. (c) Peak height keeps constant: For intermediate GPWs, the peak height remains almost unchanged with the GVD, as a result of evenly matching contributions of them. Suppose the average depth of the depletion zone is D and the injection electric field is E0, for an ion species of mobility K, the three-zone effects can be combined into eq 7 (see the Supporting Information for detailed deduction)

Table 1. Drift Tube Settings for the Experiments parameter

setting

drift region length drift voltage drop GVD GPW drift gas carrier gas moisture level drift gas temperature drift gas flow carrier gas flow chemical chemical concentration

6.25 cm 1440 V 50−350 V 0.26−1.07 ms air air Rm/Rp because Rc < Rp, so it is even harder to surpass. The two ratios were calculated for GPW = 0.34 ms and plotted as a function of GVD in Figure 6. From Figure 6, we

■

CONCLUSION A three-zone theory was proposed to describe the BradburyNielsen gate behaviors as three effects: the depletion, dispersion, and compression effects. Experimental results verify that increasing the GVD helps to enhance the depletion and the compression while reduce the dispersion, which is consistently good for improving the resolution of ion mobility spectrometry. The peak area is always reduced with increased GVDs while the peak height takes on complexity as a result of the competitive and contrary contributions between the depletion zone and the other two zones. In the experimented GPW range and for a fixed GVD, the contribution of GPW to fwhm was found to increase with the GPW by a constant compression ratio η smaller than 1. The ratio η decreased with increased GVDs, and the value of 1/η was a linear function of GVD. Two IMS efficiency ratios were calculated and found to reach as high as about 180% of the theoretical values. The three-zone theory points out potential methods for improving the resolution of ion mobility spectrometry while keeping acceptable sensitivities, which will help to increase the performances of miniaturized drift tubes, too. If the depletion and the dispersion zones can be diminished and the compression zone further enhanced, then even narrower peaks and higher heights may be obtained. More work will be done on this in our future work.

■

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■ ■

AUTHOR INFORMATION

Corresponding Author

*Fax: +86-411-84379517. E-mail: [email protected].

ACKNOWLEDGMENTS This work is supported by High-Tech Research and Development Plan (No. 2011AA060602), National Natural Science

Figure 6. Two IMS efficiency ratios Rm/Rc and Rm/Rp as a function of GVD, GPW = 0.34 ms. 1730

dx.doi.org/10.1021/ac203013u | Anal. Chem. 2012, 84, 1725−1731

Analytical Chemistry

Article

(34) Shamlouei, H. R.; Tabrizchi, M. Int. J. Mass Spectrom. 2008, 273, 78−83. (35) Lai, H.; McJunkin, T. R.; Miller, C. J.; Scott, J. R.; Almirall, J. R. Int. J. Mass Spectrom. 2008, 276, 1−8. (36) Prasad, S.; Tang, K.; Manura, D.; Papanastasiou, D.; Smith, R. D. Anal. Chem. 2009, 81, 8749−8757. (37) Dahl, D. A. Int. J. Mass Spectrom. 2000, 200, 3−25. (38) Appelhans, A. D.; Dahl, D. A. Int. J. Mass Spectrom. 2005, 244, 1−14. (39) Du, Y.; Cang, H.; Wang, W.; Han, F.; Chen, C.; Li, L.; Hou, K.; Li, H. Rev. Sci. Instrum. 2011, 82, 086103.

Foundation of China (Grant Nos. 11004190 and 20877074), Natural Science Foundation of Liaoning Province (No. 201102220), and Science and Technology Plan Project of Dalian City (No. 2010J21DW028).

■

REFERENCES

(1) Eiceman, G. A.; Karpas, Z. Ion Mobility Spectrometry, 2nd ed.; CRC press: Boca Raton, FL, 2005. (2) Borsdorf, H.; Mayer, T.; Zarejousheghani, M.; Eiceman, G. A. Appl. Spectrosc. Rev 2011, 46, 472−521. (3) Kanu, A. B.; Hill, H. H. Talanta 2007, 73, 692−699. (4) Guerra-Diaz, P.; Gura, S.; Almirall, J. R. Anal. Chem. 2010, 82, 2826−2835. (5) Hilton, C. K.; Krueger, C. A.; Midey, A. J.; Osgood, M.; Wu, J. L.; Wu, C. Int. J. Mass Spectrom. 2010, 298, 64−71. (6) Roscioli, K. M.; Davis, E.; Siems, W. F.; Mariano, A.; Su, W.; Guharay, S. K.; Hill, H. H. Anal. Chem. 2011, 83, 5965−5971. (7) Zimmermann, S.; Barth, S.; Baether, W. K. M.; Ringer, J. Anal. Chem. 2008, 80, 6671−6676. (8) Mäkinen, M. A.; Anttalainen, O. A.; Sillanpäa,̈ M. E. T. Anal. Chem. 2010, 82, 9594−9600. (9) Yamaguchi, S.; Asada, R.; Kishi, S.; Sekioka, R.; Kitagawa, N.; Tokita, K.; Yamamoto, S.; Seto, Y. Forensic Toxicol. 2010, 28, 84−95. (10) Tadjimukhamedov, F. K.; Puton, J.; Stone, J. A.; Eiceman, G. A. Rev. Sci. Instrum. 2009, 80, 7. (11) Bradbury, N. E.; Nielsen, R. A. Phys. Rev. 1936, 49, 388. (12) Rokushika, S.; Hatano, H.; Baim, M. A.; Hill, H. H. Anal. Chem. 1985, 57, 1902−1907. (13) Stlouis, R. H.; Hill, H. H. Crit. Rev. Anal. Chem. 1990, 21, 321− 355. (14) Siems, W. F.; Wu, C.; Tarver, E. E.; Hill, H. H.; Larsen, P. R.; McMinn, D. G. Anal. Chem. 1994, 66, 4195−4201. (15) Revercomb, H. E.; Mason, E. A. Anal. Chem. 1975, 47, 970− 983. (16) Spangler, G. E.; Collins, C. I. Anal. Chem. 1975, 47, 403−407. (17) Xu, J.; Whitten, W. B.; Ramsey, J. M. Anal. Chem. 2000, 72, 5787−5791. (18) Mariano, A. V.; Su, W. S.; Guharay, S. K. Anal. Chem. 2009, 81, 3385−3391. (19) Tolmachev, A. V.; Clowers, B. H.; Belov, M. E.; Smith, R. D. Anal. Chem. 2009, 81, 4778−4787. (20) Watts, P.; Wilders, A. Int. J. Mass Spectrom. Ion Processes 1992, 112, 179−190. (21) Kwasnik, M.; Fernandez, F. M. Rapid Commun. Mass Spectrom. 2010, 24, 1911−1918. (22) Eiceman, G. A.; Nazarov, E. G.; Rodriguez, J. E.; Stone, J. A. Rev. Sci. Instrum. 2001, 72, 3610−3621. (23) Tabrizchi, M. Talanta 2004, 62, 65−70. (24) Snuggs, R. M.; Volz, D. J.; Gatland, I. R.; Schummer, J. H.; Martin, D. W.; McDaniel, E. W. Phys. Rev. A 1971, 3, 487−493. (25) Wu, C.; Siems, W. F.; Asbury, G. R.; Hill, H. H. Anal. Chem. 1998, 70, 4929−4938. (26) Dugourd, P.; Hudgins, R. R.; Clemmer, D. E.; Jarrold, M. F. Rev. Sci. Instrum. 1997, 68, 1122−1129. (27) Tabrizchi, M.; Shamlouei, H. R. Int. J. Mass Spectrom. 2010, 291, 67−72. (28) Cohen, M. J.; Karasek, F. W. J. Chromatogr. Sci. 1970, 8, 330− 337. (29) Kanu, A. B.; Gribb, M. M.; Hill, H. H. Jr. Anal. Chem. 2008, 80, 6610−9. (30) Adamov, A.; Mauriala, T.; Teplov, V.; Laakia, J.; Pedersen, C. S.; Kotiaho, T.; Sysoev, A. A. Int. J. Mass Spectrom. 2010, 298, 24−29. (31) Aronson, E. A. An analysis of ion spectrometry; No. SAND870072; Sandia Cop.: Albuquerque, NM, 1987. (32) Puton, J.; Knap, A.; Siodłowski, B. Sens. Actuators, B: Chem. 2008, 135, 116−121. (33) Eiceman, G. A.; Vandiver, V. J.; Chen, T.; Ricomartinez, G. Anal. Instrum. 1989, 18, 227−242. 1731

dx.doi.org/10.1021/ac203013u | Anal. Chem. 2012, 84, 1725−1731