Improved Ion Mobility Resolving Power with Increased Buffer Gas

Pavel V. Shliaha, Rebekah Jukes-Jones, Andy Christoforou, Jonathan Fox, Chris ... Frank Lucus Steinkamp, Lauryn E. DeGreeff, Greg E. Collins, Susan L...
0 downloads 0 Views 400KB Size
Download PDF
Article pubs.acs.org/ac

Improved Ion Mobility Resolving Power with Increased Buffer Gas Pressure Eric J. Davis, Kristopher F. Grows, William F. Siems, and Herbert H. Hill, Jr.* Washington State University, Department of Chemistry, P.O. Box 644630, Pullman, Washington 99164, United States ABSTRACT: Security and military applications of analytical techniques demand a small, rugged, reliable instrument that has traditionally been served well by atmospheric pressure ion mobility spectrometry (IMS) systems. Modern threats stipulate these instruments must reliably operate in increasingly complex environments. Previous work has demonstrated that increasing the pressure of an IMS drift tube has the potential to increase the resolving power of IMS, but operation at low temperatures resulted in a leveling of the measured resolving power as a function of pressure. By creating a novel aperture grid/Faraday plate design, a high-pressure IMS (HPIMS) system has been created that maintains a resolving power efficiency of 80% regardless of the pressure applied to the cell. This allows previously unattainable resolving powers to be achieved utilizing a small (10.7 cm) IMS cell. Using high pressure, a stand-alone IMS cell of 10.7 cm length has demonstrated a resolving power of 102 when operated at 2.5 atm. An increase in peak-to-peak resolution was also noted as pressure increased. Finally, the slope of the resulting inverse mobility/ pressure curve for a single analyte has been shown to be proportional to the collision-cross-section of the analyte of interest, providing a novel method for the calculation of collision-cross-section of target ions from the HPIMS data.

T

increasing the length of the IMS cell. This has been previously demonstrated in both low pressure15 and atmospheric pressure systems.4,17 A longer drift tube allows for more ion/drift gas collisions and separates even closely related isomers by increasing the resolving power of the system. Raising the temperature of the drift cell tends to decrease resolving power but decreases clustering reactions within the drift region.13,18 Voltage has a complex relationship with resolving power due to its effect on both the IMS gating and simple diffusion as described by the Einstein relation.19 This is described theoretically by eq 5, and experimental results have shown an optimal voltage exists for every possible compound.1,19,20 Pressure in the drift cell has a dramatic effect on the maximum possible resolving power of the instrument. While low-pressure IMS cells have an advantage in overall signal and sensitivity (especially with low pressure interfacing to mass spectrometers and the ability to use ion funnels to focus ions into the detector1,15,21), atmospheric pressure IMS has consistently shown higher resolving powers for the same analytes.1,5,10,20−23 Pressures above atmospheric have also been investigated for improved IMS resolving power.21,23 However, these studies concluded that although high pressure has the potential to improve the resolving power of an IMS cell, ion clusters at elevated pressures limit the maximum resolving power obtained with high pressure. Thus the primary objective of this investigation was to use buffer gas temperature as a means to

he demand for analytical instrumentation with a small instrumental footprint has rapidly grown in recent years. In security and military applications, instrumentation must be small, rugged, and extremely reliable. Atmospheric pressure ion mobility spectrometry (APIMS) has traditionally performed this role exceptionally well due to its small size, ease of use, and sensitivity toward analytes such as explosives,1−6 narcotics,1 and chemical warfare agents (CWAs).7−9 However, the complicated matrixes of real-world samples makes the identification of a true threat difficult and create a propensity toward false positive results. Higher performance standards for IMS instruments are required to reduce the frequency of these false positive responses. Modern commercial IMS units used in security applications often have a resolving power no more than 40;10,11 while large, research-grade instruments coupled to mass spectrometers routinely achieve resolving powers greater than 100.12−16 Improved resolving powers for small stand-alone IMS instruments would lead to fewer false positive alarms. IMS systems coupled to mass spectrometers are often referred to as ion mobility mass spectrometry (IM-MS) systems due to the complementary nature of the combined techniques. These systems are typically large (IMS cells of 20 cm or longer in atmospheric pressure systems) and produce high resolving power through a combination of their length and the pinhole inlet leak into the mass spectrometer. This leak ensures only ions which have encountered a very smooth electric field enter the mass spectrometer, thus detecting only those ions which have been optimally separated in the IMS.12,13 Several previous studies in IMS have focused on improving resolving power. The most common method is through © 2012 American Chemical Society

Received: February 15, 2012 Accepted: May 15, 2012 Published: May 15, 2012 4858

dx.doi.org/10.1021/ac300452j | Anal. Chem. 2012, 84, 4858−4865

Analytical Chemistry

Article

decrease ion-neutral clustering and increase resolving power under high pressure conditions.

eff =

■

THEORETICAL CONSIDERATIONS Ion mobility spectrometry measures ion current as a function of drift time. Mobility (K) is a function of the drift field applied, the length of the tube, and the drift time that is measured. Mobility is often adjusted for standard temperature and pressure in a form known as reduced mobility (K0): K0 =

L2 ⎛⎜ P ⎞⎟⎛⎜ 273 ⎞⎟ Vtd ⎝ 760 ⎠⎝ T ⎠

Rp Rc

× 100

(6)

Figure 1 shows the results of eq 5 when high pressure is applied. Figure 1A shows the conditional resolving power

(1)

where L is the length of the drift tube in centimeters, V is the voltage applied to the drift tube, td is the drift time of the ion in seconds, P is the pressure of the drift tube in Torr, and T is the temperature of the drift tube in Kelvin. A common method for quantifying the field strength of an IMS is through the use of E/ N (electric field divided by the number density of the drift gas), where N is defined as

N=

P k bT

(2)

where N is measured in atoms/molecules per unit volume. Drift-time IMS measurements are typically made under E/N values below 10 × 10−17 or 10 Td (Townsends). In order to compare instrument performance between laboratories, resolving power (Rp) is often used and is defined as

Rp =

td FWHM

(3)

where FWHM is the full-width-half-maximum of the peak of interest in the IMS spectrum. Resolving power provides a convenient method for comparing the relative ability of two IMS cells to separate closely spaced peaks but is calculated using a single peak in the spectrum. If separation performance is being tested, resolution (Rs) provides a direct measurement of peak-to-peak separation between two IMS peaks and is defined as Rs =

0.589(td1 − td2) (FWHM1 + FWHM 2)/2

Figure 1. Conditional resolving power as calculated using eq 5 for a drift length of 10.7 cm and a buffer gas temperature of 50 °C. (A) Conditional resolving power versus voltage at increasing pressure. As pressure increased, the maximum conditional resolving power shifted to a higher value, at a higher voltage. (B) Conditional resolving power versus pressure, plotted using the optimal voltage for each pressure. Increased pressure has the most dramatic effect at lower pressures but increases steadily with pressure up to a value greater than 300 at 100 atm.

(4)

where td1 and FWHM1 correspond to the longer drifting peak. This equation provides a method for comparing the relative ability of an IMS to separate closely related peaks. The maximal resolving power an IMS cell is capable of achieving can be calculated for any set of instrumental parameters:10,23,24 Rc =

plotted versus voltage at various pressures. A higher theoretical resolving power is clearly possible under increased pressure conditions. These plots show a maximum resolving power at each specific pressure. The decreased resolving power in the lower voltage regime of these plots is referred to as the diffusion-limited regime. The ion velocity is low enough under these conditions that diffusion dominates the peak-broadening observed in the experimental peak. The decreased resolving power in the higher voltage portion of these curves is referred as the gate-pulse-width limited regime. Under these conditions, the high electric field causes the width of the ion packet initially inlet into the IMS to dominate the peak width that is observed experimentally. Figure 1B shows eq 4 plotted versus pressure, with voltage calculated as the optimal voltage for each pressure.24 For the IMS cell used in this study (10.7 cm length), pressure tends to directly increase resolving power and can result in a resolving power of over 300 at 100 atm of pressure.

1 2 tg 2K 0 2T 2V 2

760 ( 273.15 )

L4P 2

+

16kBT ln 2 qV

(5)

where Rc is the conditional resolving power, tg is the gate pulse width (s), T is the temperature (K), V is the voltage, L is the length (cm), P is the pressure (Torr), kB is Boltzmann’s constant, and q is the charge on an electron. The conditional resolving power provides a convenient benchmark where the operating parameters of the instrument can be optimized for a given compound. When an IM-MS instrument is used, most instruments produce measured Rp values approximately equal with the conditional resolving power. However, in stand-alone IMS cells, a lower value is often observed.10 To quantify this value, an IMS efficiency term is defined as10 4859

dx.doi.org/10.1021/ac300452j | Anal. Chem. 2012, 84, 4858−4865

Analytical Chemistry

Article

Figure 2. (A) Resolving power versus voltage of DtBP for several pressures using a 0.2 ms gate pulse width and traditional aperture grid design. An increase in resolving power was expected as pressure increased, but the data shown does not indicate a significant increase. (B) Efficiency versus voltage for several pressures. Efficiency was shown to decrease with voltage regardless of the pressure applied to the drift tube. (C) Efficiency versus E/N (Td) for several pressures. As pressure increased, the efficiency stepped downward, contrary to the slope found in part B. Efficiency must remain at a constant value for both voltage and Td in order to produce an increase in resolving power as pressure increases. All plots represent averaged triplicate measurements; error bars are neglected for readability.

■

EXPERIMENTAL SECTION Gases. Ultrahigh purity compressed air was used as the drift gas throughout the experiments described, with a flow rate of 1.0 L/min. Reagents. Samples tested were di-tert-butylpyridine (DtBP) (neat liquid, Aldrich, St. Louis, MO), dimethyl-methylphosphonate (DMMP) (neat liquid, Aldrich, St. Louis, MO), trinitrotoluene (TNT) (1000 mg/mL, Accustandard, New Haven, CT), and 2,4-dinitrotoluene (2,4 DNT) (1000 mg/mL, Supelco, St. Louis, MO). IMS Cell. The IMS used for this study was a traditional, stacked-ring design of 22 electrodes and a 10.7 cm drift tube. Ions were injected into the drift region of the IMS cell using a Bradbury-Nielsen style ion gate, with an aperture grid of identical design with the same voltage on both sets of wires. The IMS cell has been described in detail previously.24 Ionization was achieved through a 63Ni foil positioned on a screen on the first ring of the IMS tube. The Faraday plate was modified for this study by using an adapter to insert a 1 cm diameter plate in the precise center of the tube. This design avoids collecting ions whose paths included the “bumpy” electric field near the inside edges of the IMS tube and could be moved closer or farther from the aperture grid as needed. Inhouse software, programmed in the LabVIEW (National Instruments, Austin, TX) programming language was used for IMS instrumental control and data analysis and has been described previously.25 Pressure Chamber. The pressure chamber used in this study has been described previously,23 with some modifications. The chamber was extended by 12 in. to insert a larger IMS cell,

and HV feedthroughs with a voltage rating of 20 kV (Solidsealing, Watervliet, NY) were installed to allow higher voltages to be applied to the IMS cell. Incoming drift gas pressurized the chamber and was regulated through a Swagelok (Solon, OH) back-pressure spring regulator attached to the pressure chamber. The pressure was monitored through a Swagelok (Solon, OH) pressure transducer rated for 300 psi. Using this method, pressures were maintained at ±0.01 atm for the duration of the experiments described herein. Sample Introduction. All samples were continuously injected into a custom-built heated injection port where they were allowed to volatilize and the vapors were swept into the IMS through a 50 μm i.d. capillary at a constant flow rate regardless of pressure. This capillary entered the IMS through a specially designed drift ring positioned immediately in front of the IMS gate. The vaporous samples were swept by the countercurrent drift gas toward a 63Ni ionization source where they reacted with reactant ions to produce the observed product ions. A syringe pump (Harvard Apparatus, Holliston, MA) was used to allow continuous and precise sample flow into the heated injection port. Flow rates were modified depending on the compound tested and the pressure differential to maintain even peak heights between the reactant ion peak (RIP) and product ion peaks (PIP). Compounds were diluted to 100 ppm in methanol and injected at various flow rates to ensure even peak heights for the reactant ion peak (RIP) and product ion peak (PIP). 4860

dx.doi.org/10.1021/ac300452j | Anal. Chem. 2012, 84, 4858−4865

Analytical Chemistry

Article

Figure 3. (A) Resolving power versus voltage of DtBP for several pressures after moving the Faraday plate to within 0.25 mm of the aperture grid. As pressure increased, resolving power increased as expected. (B) Efficiency versus voltage for several pressures after moving the Faraday plate to within 0.25 mm of the aperture grid. An average value of 75% was observed with a standard deviation of 3% regardless of the pressure applied to the tube. (C) Efficiency versus Td under the same conditions. With the changed aperture grid/Faraday plate arrangement, efficiency remained constant under all voltage and pressure regimes. All plots represent averaged triplicate measurements; error bars are neglected for readability.

■

RESULTS AND DISCUSSION High-Temperature, High Pressure Ion Mobility Spectrometry. The most common method for reducing the effects of clustering reactions in IMS is through increased temperature operation.26,27 Figure 2 shows the initial results of an increased (50 °C) buffer gas temperature in a high-pressure environment. Figure 2A shows the measured resolving power versus voltage under various pressures at 50 °C. Some improvement in resolving power with respect to pressure was noted, increasing from 50 at atmospheric pressure to 70 at 3.0 atm. However, the gains were much lower than that expected by theory. At 1.0 atm, the conditional resolving power of this instrument at 6000 V was 64, while at 3.0 atm the conditional resolving power was 105. These values produce an efficiency of 78% and 67%, respectively. A decrease in efficiency was noted as pressure increased and is shown in Figure 2B where the efficiency of the resolving power is plotted versus voltage at various pressures and 50 °C. Parts B and C of Figure 2 plots of resolving power efficiency versus voltage and E/N, respectively, under increasing pressures and 50 °C. As shown, the efficiency remained constant with respect to voltage but decreased with respect to an increase in pressure. Figure 2C clearly demonstrates this effect, as both voltage and pressure are incorporated into the E/N value. In this plot, higher voltages produce a higher E/N value, and an increase in pressure shifts the operating E/N to a lower value. As demonstrated with Figure 1, higher pressures require increased voltages, so if pressure did not affect the separation, the efficiency plots should overlay each other. As shown in Figure 2C, each increase in pressure coincided with a direct decrease in efficiency, which resulted in the observed resolving powers. For example, at 1.0 atm, the average efficiency was 80% for all Td values. At 2.0 atm, the average decreased to approximately 65%. This decrease in resolving power efficiency

as a function of increasing pressure was observed for room temperature, 50, 100, 150, and 200 °C, though this data is not displayed herein. Thus, it was concluded that temperature did not appreciably affect the observed decrease in efficiency with respect to pressure. During the initial temperature studies shown in Figure 2, the distance between the aperture grid and Faraday plate was set at half the distance between drift rings (approximately 1.4 mm) so as to ensure an even electric field. When the aperture grid was positioned closer to the Faraday plate, the resolving power efficiency was found to remain constant as a function of increasing pressure. Figure 3 shows the results of the identical study described in Figure 2 performed after the Faraday plate was moved as close as physically possible to the aperture grid (about 0.25 mm). Figure 3A shows the measured resolving power versus voltage at various pressures and 50 °C, and Figure 3B shows the efficiency of the IMS versus voltage at increasing pressure at 50 °C. Once the Faraday plate was moved, the efficiency stayed constant as voltage and pressure increased, remaining at approximately 75% for a gate pulse width of 0.2 ms. Compared to the results indicated in Figure 2, the movement of the Faraday plate removed the deleterious effect of pressure on efficiency and allowed the instrument to operate according to theory as pressure increased. Figure 3C plots the efficiency of the instrument versus E/N and indicates that a constant efficiency was obtained regardless of the voltage or applied pressure. As expected, neither pressure nor voltage had an effect on resolving power efficiency for this experiment. While voltage limitations prevented pressures above 3.0 atm for the instrument used in this study, it may be envisioned (Figure 1) that increased pressures would greatly improve the resolving power above that shown in Figure 3. The effect of mirror currents on the resolving power of an IMS separation has been shown previously.28 However, the 4861

dx.doi.org/10.1021/ac300452j | Anal. Chem. 2012, 84, 4858−4865

Analytical Chemistry

Article

Figure 4. HPIMS spectrum of DtBP in the positive mode with a gate pulse width of 0.1 ms, a pressure of 2.5 atm, and a voltage of 5 750 V. The DtBP peak has a measured resolving power of 102 and a conditional resolving power of 126, providing an efficiency of 81% for this spectrum.

By decreasing the gate pulse width, the conditional resolving power shifts to a higher value, though increased voltages are required to obtain the optimal voltage.23,24 Figure 5 plots the experimental resolving power of the described HPIMS system with a gate pulse with of 0.1 ms versus voltage at increasing pressures. A maximum resolving power of 102 was obtained at 2.5 atm (this point is plotted as Figure 4), and increased pressures produced significant increases in resolving power from 1.5 atm through 2.5 atm The pressure 3.0 atm was investigated at a 0.1 ms, but the maximum voltage limit of 10 kV provided signals below the limit of detection for the instrument, so this data was neglected. However, 2.5 atm is not a fundamental limit on the pressure that may be applied to an IMS. An IMS could be envisioned wherein pressures as high as 100 atm could be applied. Since the breakdown voltage of nitrogen is nearly 300 000 V/cm at this pressure,29,30 the calculated 83 000 V required to achieve the optimal voltage is attainable. An IMS of 10.7 cm length would have a theoretical resolving power greater than 350 under these conditions (see Figure 1). Pressure Effects on Resolution. As predicted through IMS theory, resolving power has been shown to increase with respect to pressure. This coincided with a simultaneous increase in peak-to-peak resolution for mixtures within the IMS cell. Figure 6 shows a two-dimensional plot of drift time versus pressure versus intensity. The pressure within the IMS cell was allowed to increase from 1.0 atm up to 1.5 atm as spectra were obtained. The upper peak plotted in this spectrum is [DtBP]H+ ion, while the lower trace is the positive mode reactant ion peak (RIP). The increase in pressure caused these

effect of pressure on aperture grid efficacy and the resulting observation of mirror currents in the signal has not been previously demonstrated. In Figure 2, the increased pressure increased the time the ions spent between the aperture grid and Faraday plate, resulting in the observed decrease in resolving power due to mirror currents. As the number density of the buffer gas increased, the slower movement of the ions under the same electric field conditions as at atmospheric pressure caused the time in which mirror currents could occur to increase before ion impact. By reducing the physical distance between the aperture grid and Faraday plate, this effect was mitigated and efficiency increased as noted in Figure 3. Though the effect was not observed under the pressures studied, it is also anticipated that decreased aperture grid wire diameters and wire spacing will produce a significant effect on the efficiency of the IMS cell as greater pressures are applied due to this “lengthening” effect of pressure on the IMS cell components, though careful ion modeling is required for further investigations. High-Resolution HPIMS. The modification of the Faraday plate/aperture grid arrangement allowed a stand-alone IMS cell to reach resolving powers previously attainable only with IMMS systems. Figure 4 shows the first IMS spectrum taken with a small IMS cell at high pressure and Faraday plate detection with a resolving power greater than 100. This is a spectrum of DtBP taken at 2.5 atm, 50 °C, and a gate pulse width of 0.1 ms. This spectrum has an experimentally measured resolving power of 102. At this pressure, the conditional resolving power was calculated to be 126, producing an efficiency of 81%. 4862

dx.doi.org/10.1021/ac300452j | Anal. Chem. 2012, 84, 4858−4865

Analytical Chemistry

Article

Figure 5. Resolving power versus voltage for several pressures with a 0.1 ms gate pulse width. As pressure increased, the measured resolving power increased according to theory, producing consistent resolving powers of near 102 for 2.5 atm. This plot represents averaged triplicate measurements; error bars indicate 1 standard deviation.

Previous work by Tabrizchi and Rouholahnejad concluded that while pressure and temperature both affect the number density of the drift gas, only temperature changes the identity of the ions through ion-neutral clustering while pressure did not appreciably affect clustering reactions.31 Figure 7A plots the inverse mobility of [DtBP]H+ versus pressure. In this plot the points indicate experimental data, and the lines indicate theoretical data as calculated using eq 1. If clustering were present as a function of pressure, high pressures would result in a lower-than-expected mobility as the peak drifts longer than anticipated, resulting in a nonlinear curve. Since linear curves were observed, clustering was not present with respect to pressure, regardless of the applied temperature. Figure 7A also shows the temperature dependence of mobility for a single ion. The temperature for each line is reflected in the slope of the line, with higher temperatures producing lower slopes. Since inverse mobility (1/K) is directly proportional to drift time and pressure and inversely proportional to temperature, a plot of inverse mobility versus pressure indicates the ability of pressure to separate similar compounds. Figure 7B provides inverse mobility versus pressure plots of multiple compounds, with theory plotted as solid lines. This plot is a mixture of both positive and negative mode ions run individually, all at 200 °C. DtBP, DMMP, and positive RIP were all positive mode ions, while TNT, 2,4 DNT, negative RIP, and Cl − were negative mode ions. If plotted together, several of these ions would be difficult to distinguish from one another at atmospheric pressure, as indicated by the proximity of their respective 1/ K values at 1.0 atm. However, each ion produced a unique slope. These differences in slope allowed for the ions to be

Figure 6. Successive IMS scans over time as pressure increases for DtBP and RIP. Pressure increased from 1.0 atm up to 1.5 atm in the data displayed. All IMS spectra were obtained at a voltage of 5 400 V and a gate pulse width of 0.2 ms. As the pressure increased, the resolution of the peaks increased with pressure, causing the peaks to spread apart.

two ions to separate, producing linear curves with different slopes for each analyte. In this plot, the peak-to-peak resolution (eq 3) is initially 9.09 and increases to 11.86, a 30% increase with a pressure increase of 0.5 atm. The screenshot presented in Figure 6 is analogous to a plot of inverse mobility versus pressure, as mobility and drift time are inversely proportional (eq 1). 4863

dx.doi.org/10.1021/ac300452j | Anal. Chem. 2012, 84, 4858−4865

Analytical Chemistry

Article

is the reduced mass of the ion and drift gas, kb is Boltzmann’s constant, and T is the temperature. If the inverse is taken of eq 7and combined with eq 2, the resulting equation is easily applied to the plots presented in Figure 7: 1/2 1 16 P Ω ⎛ μk bT ⎞ ⎜ ⎟ = K 3 k bTq ⎝ 2π ⎠

(8)

Thus, if we take the slope of the plots presented in Figure 7 to be the change in 1/K with pressure, the slope (m; d(1/K)/ dP) may be defined as ⎛d 1 ⎜ K m⎜ dP ⎝

( ) ⎞⎟ = 16 ⎟ ⎠

Ω ⎛ μk bT ⎞ ⎜ ⎟ 3 k bTq ⎝ 2π ⎠

1/2

(9)

This may be rearranged and simplified to show that the CCS is a function of the slope, temperature, and reduced mass for any given IMS: Ω = mT1/2μ−1/2 C

where C is a constant which combines the Boltzmann’s constant, elementary charge, 16/3, and 2π terms found in eq 9. Equation 10 provides a novel method for the determination of CCS values from data in an IMS cell with variable pressure capabilities similar to plotting K versus the applied voltage. A plot of 1/K versus voltage is a simpler experiment to perform, but the ability to use pressure data for the calculation of collision cross section values has not been previously demonstrated. The results of this theory as applied to the data found in Figure 7 is found in Table 1, wherein the CCS has been calculated using this method for each compound in Figure 7b.

Figure 7. (A) Inverse mobility versus pressure for DtBP at increasing temperatures. Points indicate experimental results, while lines indicate theoretical values. If clustering were present, as pressure increases, the value obtained for 1/K would decrease relative to the theoretical line and produce a nonlinear curve. As this is not present, it may be concluded that clustering is not a major contributor to decreased resolving power in HPIMS data. (B) Inverse mobility versus pressure for several compounds. At atmospheric pressure, many of these peaks overlap, but with increased pressure, they become well separated due to the CCS dependent slopes of these curves. All plots represent averaged triplicate measurements; error bars are neglected for readability.

■

CONCLUSIONS By reducing the effect of mirror currents on the Faraday plate detector and increasing the pressure, a stand-alone IMS system has been demonstrated for the first time with a resolving power greater than 100. This increase in resolving power is limited only by the maximum pressure and voltage that may be applied to an IMS cell and creates HPIMS as a technique which may be quickly adapted to any IMS system. Raising the pressure of the buffer gas also increased the peak-to-peak resolution between two ions, more easily separating closely related compounds as pressure increases. Finally, a plot of inverse mobility verses pressure provided a novel method for the calculation of collision cross sections of ions that is analogous to changing the voltage on the tube for an averaged CCS result across multiple conditions. The application of high pressure to a small tube allows resolving powers previously unattainable even in large, research-grade IMS instruments. In theory, a 10.7 cm drift tube such as the tube used in this study can achieve a resolving

easily distinguished from one another at high pressures with the corresponding increase in resolving power demonstrated in Figures 3 and 5. The slopes as calculated from this plot are tabulated in Table 1. As the curves produced are linear, it is a simple matter to apply a slope analysis. Mobility is inversely proportional to the collision cross section (Ω) of any particular ion:32 1/2 3 q ⎛ 2π ⎞ K= ⎜ ⎟ 16 N Ω ⎝ μk bT ⎠

(10)

(7)

where q is the elementary charge, N is the number density of the drift gas, Ω is the collision cross section of the ion (CCS), μ

Table 1. Experimental Reduced Mobility (K0 Average), Literature K0 (K0 Lit.), Molecular Weight (MW), Slope, and Collision Cross Section Values As Calculated Using Equation 10 for Compounds from Figure 7B

2

K0 (average) (cm /vs) K0 (lit.)(cm2/vs) MW slope CCS-ang2 CCS-nm2

[DtBP]H+

[DMMP]2H+

[DMMP]H+

[TNT]−

[2,4 DNT]−

+RIP

−RIP

[CI]−

1.43 1.42 191.3125 0.446 126.2 1.262

1.38 1.45 247.16 0.441 122.9 1.229

1.93 1.91 124.08 0.315 92.1 0.921

1.54 1.54 227.13 0.392 109.7 1.097

1.70 1.68 182.13 0.365 103.5 1.035

2.26

2.59

2.99

37 0.283 99.5 0.995

32 0.248 90.3 0.903

35 0.201 71.7 0.717

4864

dx.doi.org/10.1021/ac300452j | Anal. Chem. 2012, 84, 4858−4865

Analytical Chemistry

Article

(29) Burm, K. R. A. L. Contrib. Plasma Phys. 2007, 47 (3), 177−182. (30) Lieberman, M. A.; Lichtenberg, A. J., Principles of Plasma Discharges and Materials Processing; John Wiley and Sons: Hoboken, NJ, 2005; p 757. (31) Tabrizchi, M.; Rouholahnejad, F. J. Phys. D: Appl. Phys. 2005, 38, 857−862. (32) Revercomb, H. E.; Mason, E. A. Anal. Chem. 1975, 47 (7), 970− 983.

power of over 300 at 100 atm of pressure. The application of high-pressure IMS can maintain a small instrument footprint while producing higher resolving powers and improved resolution.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: (509) 335-5648. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Eiceman, G. A.; Karpas, Z. Ion Mobility Spectrometry; CRC Press, Inc.: Boca Raton, FL, 1994; p 217. (2) Eiceman, G. A.; Nazarov, E. G.; Stone, J. A. Anal. Chim. Acta 2003, 493, 185−194. (3) Daum, K. A.; Atkinson, D. A.; Ewing, R. G. Int. J. Mass Spectrom. 2002, 214, 257−267. (4) Wu, C.; Steiner, W. E.; Tornatore, P. S.; Matz, L. M.; Siems, W. F.; Atkinson, D. A.; Hill, H. H. Talanta 2002, 57, 123−134. (5) Kanu, A. B.; Hill, H. H. Talanta 2007, 73 (4), 692−699. (6) Babis, J. S.; Sperline, R. P.; Knigh, A. K.; Jones, D. A.; Gresham, C. A.; Denton, M. B. Anal. Bioanal. Chem. 2009, 395, 411−419. (7) Klepel, S.; Grafenhain, U.; Lippe, R.; Stach, J.; Starrock, V. Third International Workshop on Ion Mobility Spectrometry, Galveston, TX, October 16−19, 1994; pp 274−280. (8) Cao, L.; Harrington, P. d. B.; Liu, C. Anal. Chem. 2004, 76, 2859−2868. (9) Kanu, A. B.; Haigh, P. E.; Hill, H. H. Anal. Chim. Acta 2005, 553, 148−159. (10) Aanu, A. B.; Gribb, M. M.; Hill, H. H. J. Anal. Chem. 2008, 80, 6610−6619. (11) Tolmachev, A. V.; Clowers, B. H.; Belov, M. E.; Smith, R. D. Anal. Chem. 2009, 81, 4778−4787. (12) Wu, C.; Siems, W. F.; Asbury, G. R.; Hill, H. H. Anal. Chem. 1998, 7, 4929−4938. (13) Kanu, A. B.; Dwivedi, P.; Tam, M.; Matz, L. M.; Hill, H. H. J. Mass Spectrom. 2008, 46, 1−22. (14) Collins, D. C.; Lee, M. L. Anal. Bioanal. Chem. 2002, 372 (1), 66−73. (15) Baker, E. S.; Clowers, B. H.; Li, F.; Tang, K.; Tolmachev, A. V.; Prior, D. C.; Belov, M. E.; Smith, R. D. J. Am. Soc. Mass Spectrom. 2007, 18, 1176−1187. (16) Dwivedi, P.; Bendiak, B.; Clowers, B.; Hill, H. H. J. J. Am. Soc. Mass Spectrom. 2007, 18, 1163−1175. (17) Asbury, G. R.; Hill, H. H. J. J. Microcolumn Sep. 2000, 12 (3), 172−178. (18) Tabrizchi, M. Talanta 2004, 62 (1), 65−70. (19) Siems, W. F.; Wu, C.; Tarver, E. E.; Hill, H. H. J. Anal. Chem. 1994, 66, 4195−4201. (20) Rokushika, S.; Hatano, H.; Baim, M. A.; Hill, H. H. J. Anal. Chem. 1985, 57 (9), 1902−1907. (21) Tabrizchi, M.; Rouholahnejad, F. Talanta 2006, 69, 87−90. (22) Davies, J. H.; Jackson, R. A.; Kuja, F. J. Pressure stabilized ion mobility spectrometer. U.S. Patent 5,552,600, September 3, 1996. (23) Davis, E. J.; Dwivedi, P.; Tam, M.; Siems, W. F.; Hill, H. H. Anal. Chem. 2009, 81, 3270−3275. (24) Davis, E. J.; Williams, M. D.; Siems, W. F.; Hill, H. H. Anal. Chem. 2011, 83 (4), 1260−1267. (25) Davis, E. J.; Clowers, B. H.; Siems, W. F.; Hill, H. H. J. Int. J. Ion Mobility Spectrom. 2011, 14, 117−124. (26) Rokushika, S.; Hatano, H.; Hill, H. H. J. Anal. Chem. 1986, 58 (2), 361−365. (27) Berant, Z.; Karpas, Z.; Shahal, O. J. Phys. Chem. 1989, 93 (21), 7529−7532. (28) Hill, H. H., Jr.; Siems, W. F.; St. Louis, R. H. Anal. Chem. 1990, 62 (23), 1201A−1209A. 4865

dx.doi.org/10.1021/ac300452j | Anal. Chem. 2012, 84, 4858−4865