Detection of inorganic ions from water by electrospray ionization-ion mobility spectrometry

Talanta 57 (2002) 1161– 1171 www.elsevier.com/locate/talanta

Detection of inorganic ions from water by electrospray ionization-ion mobility spectrometry Heather M. Dion a,b,c,*, Luke K. Ackerman b,c, Herbert H. Hill Jr. b,c a

Sa6annah Ri6er Ecology Laboratory, Uni6ersity of Georgia, Sa6annah Ri6er Site, Drawer E, Aiken, SC 29802, USA b Department of Chemistry, P.O. Box 644630, Washington State Uni6ersity, Pullman, WA 99164 -4630, USA c Center for Multiphase En6ironmental Research, P.O. Box 642710, Washington State Uni6ersity, Pullman, WA 99164 -2710, USA Received 26 November 2001; received in revised form 12 April 2002; accepted 12 April 2002

Abstract The results from this study illustrate the first time electrospray ionization-ion mobility spectrometry (ESI-IMS) has been used to separate inorganic cations in aqueous solutions. Using ESI-IMS nine inorganic cation solutions were analyzed. Counter ions affected both the sensitivity and the identity of the response ions. Aluminum sulfate, lanthanum chloride, strontium chloride, uranyl acetate, uranyl nitrate, and zinc sulfate produced spectra containing a single response ion. Aluminum nitrate and zinc acetate solutions produced multiple ion peaks, which increased the detection limits and the difficulty of identification. Cation detection limits ranged from 0.16 to 13 ng ml − 1 depending on the solution studied. The identities of the ion species detected were unconfirmed, but mass spectrometry literature suggested the detection of positively charged cation-solvent or cation-solvent-anion complexes. Finally, cations from strontium and lanthanum chloride solutions were separated with a resolution of 2.2. The results from this study suggest that ESI-IMS has potential as a field technique for the detection of metal cations and their complexes in the environment. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Ion mobility spectrometry; Electrospray ionization; Inorganic cation detection

1. Introduction Electrospray ionization (ESI) was initially employed by Chapman in 1937 to study the gasphase mobility of several charge carrying liquids in an Erikson mobility tube [1]. It was not until 1972 that ESI was attempted as an ionization source for ion mobility spectrometry (IMS) using * Corresponding author. Tel.: +1-803-725-2907; fax: + 1803-725-3309. E-mail address: [email protected] (H.M. Dion).

organic compounds [2] and successful electrospray ionization-ion mobility spectrometry (ESI-IMS) spectra were not obtained for organic compounds until the mid to late 1980’s [3–5]. During the 1990’s, ESI-IMS has been developed almost exclusively as an organic [6] or bio-organic [7–9] species separation and detection method. Resolving power and sensitivity have increased greatly since then, increasing the value of ESI-IMS as an analytical tool for organic compounds [10]. Inorganic compounds such as nitrate, nitrite, chloride, and acetate ions have been evaluated using negative

0039-9140/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 3 9 - 9 1 4 0 ( 0 2 ) 0 0 1 9 7 - 2

1162

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

mode ESI-IMS-mass spectrometry (MS) from surface water [11]. Additionally, laser desorption-IMS and matrix-assisted laser desorption/ ionization have been used to elucidate the structure and behavior of silicon, metallic clusters, and alkali metal– crown ether complexes [12 – 17]; however, ESI-IMS remains unexplored as a method for the separation and detection of aqueous inorganic cationic species. The ions produced by ESI are dependent on the spray conditions. Over 20 studies of aqueous cation solutions have been conducted using ESI with MS ([18–27] and references therein). Although, the exact mechanism by which gasphase ions of inorganic species are generated by ESI is not fully understood, operating conditions affect the identity of ions produced [19].

The ESI-MS literature [20–27] for the species studied in this work provide examples of the types of ions produced including bare and complexed metal ions. Complexation occurred with the cation and the solvent (e.g. Sr(CH3OH)3 + 2) and the original salt anion (e.g. LaCl2 + ) [11,13]. Of the literature cited, five papers observed the bare cations of La + 3, Sr + 2, and UO2 + salts; these cations were thought to arise from high fields within collision-induced dissociation (CID) regions rather than from direct ESI [20–24]. The instrumental operating conditions reported for the detection of metal ions and ion complexes vary widely throughout the ESI-MS literature. One example was the widely varying operating conditions particularly within CID regions. In general, the electric field strength and pressure were inversely proportional to the extent of complexation and stability of ESI generated gaseous ions. Higher field regions produced ions with less clusters and complexes, and in some cases, bare ions were observed [19]. Additionally, the sheath flow rate surrounding the ESI needle also showed significant effect on the species observed [19]; however, these effects were not consistent between studies. Lastly, complexation appeared to be proportional to the curtain gas flow rate. Lower flow rates generally produced less complexation, potentially due to incomplete desolvation of the cation. In this study, ESI-IMS was investigated for the detection of inorganic cations in solution. Do different anions have an effect on the detection of those cations? Can simple separations of inorganic cations be accomplished? The effectiveness of ESI-IMS to produce simple reproducible spectra was evaluated. Metal salts with different anionic composition were used to determine anion effects and a mixed salt solution was used to determine if ESI-IMS was capable of simple cation separations.

2. Materials and methods

2.1. Instrumentation Fig. 1. Electrospray ionization-ion mobility spectrometer built at Washington State University.

The ESI-ion mobility spectrometer was con-

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

structed at Washington State University (Fig. 1) and has been described extensively elsewhere [28,29]. Two major modifications were made to the IMS, which are not described in the works cited. One, the two 63Ni foils, previous ionization source, were removed from the desolvation region and three additional metal rings were added to lengthen the desolvation region, which resulted in a total length of 8.5 cm. The extra length added to the desolvation region aids in the complete desolvation of the electrosprayed ions. Two, ESI was used as the ionization source with the needle shielded in a polyethylene jacket and cooled by nitrogen flowing along the axis of the needle, which has been found useful in preventing Corona discharge [28]. Nitrogen flowing through the system at 1 l min − 1 provided a flow counter to the ion motion, which aided in desolvation of ions and evacuation of uncharged species. The ESIIMS was operated in the positive mode with the electrospray needle held at a positive 10 000 V, the target screen held at a positive 7000 V, and a constant temperature of 250 °C. This difference created a spray voltage of 3000 V and an electric field of about 305 V cm − 1 from the ion gate to the detector, total length of 13.2 cm. A dual piston syringe pump (Brownlee Labs, Santa Clara, CA) provided continuous flow of solvent to the ESI. A six-port injector (Valco C6W, Valco Industries, Houston, TX) with an external injection loop was used to introduce samples with a sample loop volume of 50 ml. The mobile-phase solvent consisted of 47.5:47.5:5 methanol:water:acetic acid (HPLC grade chemicals, Fisher Scientific, Fair Lawn, NJ) and was held constant at 10 ml min − 1, conditions similar to other studies [11]. The current signal was collected with a Keithley Model 427 amplifier (Keithly Instruments, Cleveland, OH), amplified (109 gain), and then sent to a Labview™ v. 5.1 (National Instruments, Austin, TX) based data acquisition system designed at Washington State University. All spectra shown were an average of 500 individual spectrum taken at 25 ms intervals with a gate pulse of 0.300 ms. Due to overhead from the Labview data processing, each spectrum took 22 s to obtain.

1163

2.2. Chemicals Stock solutions of 1000 ng ml − 1 of Al + 3, La + 3, Sr + 2, UO2 + 2, and Zn + 2 were prepared from reagent-grade salts of aluminum nitrate, aluminum sulfate, lanthanum chloride, strontium chloride, strontium nitrate, uranyl acetate, uranyl nitrate, zinc acetate, and zinc sulfate (JT Baker Co. Phillipsburg, NJ). Dilutions of the stock solutions were made in 50:50 methanol:water (Fischer Scientific) at the 10, 50, and/or 100 mg l − 1 level. Nitrogen used for coolant, nebulization, and drift gas was purified prior to use by passing the gas through a Supelcarb™ HC filter (Supelco, Bellefonte, PA).

3. Results and discussion Several metal salts including: aluminum nitrate and sulfate, uranyl acetate and nitrate, zinc acetate and sulfate, lanthanum chloride, and strontium chloride were investigated for response differences due to anionic composition using ESIIMS. Ion mobility response was accomplished with varying results. Three solvent peaks were prevalent in all spectra obtained in this study. Reduced mobility values (K0) were determined for the solvent peaks and compared to previous results using an ESIIMS-MS system. The K0 values determined in this study indicate that two of the solvent peaks in the spectra were protonated water (H3O+) and protonated methyl acetate (MeAcH+) [11]. The calculated K0 values for each peak were compared to literature values for H3O+ (2.67 cm2 V − 1 s − 1) and MeAcH+ (3.50 cm2 V − 1 s − 1) [11] and were within 4% agreement with the values obtained in this work, which were 2.719 1.47% (H3O+) and 3.529 0.57% cm2 V − 1 s − 1 (MeAcH+). The third peak could not be positively identified through comparison with any literature value. A complete list of the metal salts studied, drift times, calculated peak resolution, and reduced mobility values (K0) are reported in Table 1. The resolving power for each peak was calculated using Eq. (1).

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

1164

Table 1 List of metal salts studied, drift times, calculated resolving power of IMS for each peak, and reduced mobility values in cm2 V−1 s−1 Salt

Chemical form

Drift time (ms)

Resolving power

Observed K0

Aluminum nitrate Aluminum sulfate Lanthanum chloride Strontium chloride Uranyl acetate

Al(NO3)3 Al2(SO4)3·18H2O LaCl3·H2O

7.5, 9.4, 11.4 11.1 10.8

22.1, 23.4, 30.7 41.6 31.3

2.27, 1.82, 1.50 1.54 1.58

SrCl2·6H2O

9.8

38.0

1.74

UO2(CH3COO)2

12.3, 16.3

47.6

1.39, 1.05

Uranyl nitrate Zinc acetate

·2H2O UO2(NO3)2·6H2O 10.7 Zn(CH3COO)2 12.0, 14.4, 18.8, 20.1, 24.5

31.4 28.8, 34.6, 36.6, 34.5, 36.8

1.60 1.42, 1.19, 0.91, 0.85, 0.70

Zinc sulfate

·2H2O ZnSO4·7H2O

30.2

1.67

10.2

Table 2 Calculated s/n ratios and predicted detection limits for the salt solutions which yielded only one major peak in the IMS spectra Salt

Concentration (ppm)

s/n ratio

Predicted detection limit (ppm)

Aluminum sulfate Lanthanum chloride Strontium chloride Uranyl acetate Uranyl nitrate Zinc sulfate

10 100 100 50 100 100

45 45 73 41 6.0 28

0.16 1.7 4.5 0.90 13 2.7

r=

  td w1/2

(1)

where td, drift time of an ion in milliseconds; w1/2, width of ion peak at half height in milliseconds. Reduced mobility values were calculated using Eq. (2). K0 =

    6 E

273 T

P 760

(2)

where K0, cm2 V − 1 s − 1; 6, velocity of selected ion (cm s − 1); E, strength of electric field (V cm − 1); T, experimental temperature (K); P, experimental pressure (Torr). The electric field strength was determined to be 330.4 V cm − 1 from the difference in the voltage between the gate and aperture grid (length of drift tube), the temperature was held constant at 548 K, and the pressure was 695 Torr. The reduced mobility values were validated using lutidine as an external standard, which has a literature K0 value

of 1.91 cm2 V − 1 s − 1 [30,31]. Lutidine spectra were taken on several days and had an experimental K0 value of 1.999 4.02% cm2 V − 1 s − 1. Signal-to-noise (s/n) ratios were calculated in order to estimate the detection limit at three times the standard deviation of the noise (Table 2). The detection limits were determined only on spectra that had one major ion peak: aluminum sulfate, lanthanum chloride, strontium chloride, uranyl acetate, uranyl nitrate, and zinc sulfate (Figs. 2 and 3). It appeared that clusters of ions, as seen in the case of aluminum nitrate and zinc acetate decreased the overall sensitivity and increased detection limit (Fig. 4). The presence of ion clusters has been seen throughout ESI-MS spectra presented in the literature (Table 3). Additionally, the ESIMS studies have shown multiply charged ions for several of the ions studied in this work; however, in this study, fewer multiply charged species were

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

expected due to softer ionization in ESI-IMS than in ESI-MS. In order to examine the possible presence of multiply charged ions, the resolving power of the ions studied was plotted as a function of drift time (Fig. 5). Resolving power is a function of the drift time and the peak width at

1165

half height (Eq. (1)), with peak width a function of the gate pulse, inhomogeneity in the electric field, Coulombic repulsion, and the charge state of the ion [32]. Given that all factors controlling peak width, except charge state of the ions, were held constant in this study, differences in resolv-

Fig. 2. ESI-IMS spectra for (a) aluminum sulfate (10 ng ml − 1 Al), (b) strontium chloride (100 ng ml − 1 Sr), and (c) zinc sulfate (100 ng ml − 1 Zn). Spectra collected at 0.300 ms pulse width and averaged 500 times.

1166

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

Fig. 3. ESI-IMS spectra for (a) uranyl acetate (50 ng mL − 1 U) and (b) uranyl nitrate (100 ng ml − 1 U). Spectra collected at 0.300 ms pulse width and averaged 500 times.

ing power should be due to charge state differences on the ion. Recently Valentine and coworkers expanded on this relationship using triptic digest peptide fragments and mass spectrometric confirmation [33]. In our work, most points fall on the lower resolving power curve with three exceptions: uranium acetate, aluminum sulfate, and strontium chloride. It can be hypothesized that these three salts form multiply charged complexes in the ESI process, which is substantiated by the ESI-MS literature. Aluminum can hydrolyze and complex with small inorganic anions, e.g. Al(NO3)(OH)(H2O) + 2 or Al(NO3)(OH)(H2O) + 3, and strontium has been shown to form multiply charged complexes with methanol including Sr(CH3OH)3 + 2, Sr(CH3OH)4 + 2, Sr(CH3OH) + 2, and Sr(CH3OH)2 + 2 [21].

Aluminum sulfate was detected at 10 ng ml − 1 with a s/n ratio of 45 (Fig. 2) and had the lowest calculated detection limit at 0.16 ng ml − 1 (Table 2). Most of the salt solutions were measured at 100 ng ml − 1 and the s/n ratios were 73 for strontium chloride, 45 for lanthanum chloride, 28 for zinc sulfate, 6.0 for uranyl nitrate, and uranyl acetate measured at 50 ng ml − 1 had a s/n ratio of 41. The linear extrapolated detection limits for the salt solutions determined at 100 ng ml − 1 ranged from a low of 1.7 ng ml − 1 for lanthanum chloride to a high of 13 ng ml − 1 for uranyl nitrate; the detection limit for uranyl acetate, which was calculated from a 50 ng ml − 1 spectrum, was 0.90 ng ml − 1. The detection limits for each of the salt solutions studied may be much lower than predicted by the 100, 50, and 10 ng ml − 1 spectra,

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

1167

Fig. 4. ESI-IMS spectra for (a) aluminum nitrate (100 ng ml − 1 Zn) and (b) zinc acetate (100 ng ml − 1 Zn). Spectra collected at 0.300 ms pulse width and averaged 500 times.

Table 3 List of chemical species observed in electrospray and ion spray mass spectrometry Metal cation

Chemical form

Species detected

Matrix

Reference

Aluminum

Al(NO3)3

Al(NO3)(OH)(H2O)+2\Al(NO3)(OH)(H2O)+3 \Al(OH)2(C6O7H8)+\Al(OH)2(H2O)+3

Citric acid

[11]

Lanthanum

La(NO3)3

La(OH)(H2O)6+2\La(OH)(H2O)5+2\La(OH)(H2O)7+2 \La(OCH3)(H2O)5+2 = La(OCH3)(H2O)6+2 \La(OH)(H2O)4+2\La(NO3)(H2O)7+2\La(OH)(H2O)8+2 =La(NO3)(H2O)8+2 La(OH)(H2O)2+2\La(OH)(H2O)5+2\La(OH)(H2)7+2 =La(OH)2+\La(OH)2(H2O)+ LaO+\La+\LaH(OCH3)+\LaCl+\LaCl2+ =La+2 Sr(CH3OH)3+2\Sr(CH3OH)4+2\Sr(CH3OH)+2 \Sr(CH3OH)2+2 UO+\U+ = UO2(OH)+\UO3+ Zn+\ZnO+\ZnOH(H2O)+\ZnO(OH)+

Methanol

[13]

Methanol

[13]

Methanol Methanol

[12] [11]

Methanol

[11] [11]

Strontium

LaCl3 SrCl2

Uranium Zinc

UO2(NO3)2 Zn(NO3)2

1168

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

Fig. 5. Plot of resolving power as a function of drift time. Proposed singly charged ions ( ) and multiply charged ions ( ). Curve follows pattern of singly charged ions.

because the standards used were at concentrations far above the calculated detection limit. In order to obtain a more accurate measure of the detection limit, a calibration curve down to the detection limit would need to be obtained. Figs. 3 and 4, uranium and zinc salts, show distinctly different ion peaks depending on the anions present in solution. It should be noted that while every sample was carried by a mobile-phase with 5% acetic acid, zinc sulfate and uranium nitrate did not produce response peaks similar to their acetate counterparts. The ESI-MS literature did not give any insight into elucidating the specific ion species for either of these two metal ions, because the experimental conditions were very different [24]. The ability of ESI-IMS to detect and separate multiple cations in solution was tested using lanthanum and strontium chlorides (Fig. 6). Two important aspects of this separation include baseline resolution between the peaks in Fig. 6c, and similarity between the mixed solution and the individual solutions (Fig. 6a and b). Resolution was calculated to be 2.2 using Eq. (3), more than adequate for this type of separation.

R=





2Dt wb1 + wb1

(3)

where Dt, drift time of peak 2 minus drift time of peak 1; wb1, width of peak 1 at base; wb2, width of peak 2 at base. Identification of the lanthanum species from the IMS spectra alone was not possible. ESI-MS investigations have demonstrated that a number of structures are possible (Table 3). Most of the selected samples yielded single peak spectra, which indicated that ESI-IMS was capable of producing uncomplicated spectra (Figs. 2 and 3). Aluminum nitrate, lanthanum chloride, strontium chloride, zinc sulfate, uranyl acetate, and uranyl nitrate were detected as single peak species. Aluminum nitrate appeared to be the most sensitive and was detected at 10 ng ml − 1 Al. The response of strontium chloride was strong at both 100 ng ml − 1 Sr in Fig. 3b and at 50 ng ml − 1 Sr in Fig. 6b. Uranyl acetate and uranyl nitrate were detected in solution and the resulting peaks had drift times of 12.3 and 10.7 ms, respectively. The instrumental conditions were identical for the two uranyl species, indicating that different complexes were formed depending on the anion in solution.

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

The ESI-IMS spectra for aluminum sulfate (10 ng ml − 1 Al) and aluminum nitrate (100 ng ml − 1 Al) are illustrated in Fig. 2a and Fig. 4a. ESI-IMS was more sensitive to aluminum sulfate than to aluminum nitrate. This may be explained by the fact that aluminum nitrate formed several gasphase ion species, and the available charge was divided among several peaks (7.5,9.4,11.4 ms)

1169

whereas the aluminum sulfate salt only displays one predominant peak (11.1 ms). The peak at 11.4 ms for the nitrate and the 11.1 ms peak for the sulfate were assumed to be the same response ion since the difference in drift times is only 2.6%. Thus, the 11.4 and 11.1 ms peaks for aluminum were either a charged hydrolysis or acetate complex while the additional peaks in the nitrate

Fig. 6. ESI-IMS spectra for (a) lanthanum chloride (100 ng ml − 1 La), (b) strontium chloride (100 ng ml − 1 Sr), and (c) lanthanum (50 ng ml − 1 La) and strontium (50 ng ml − 1 Sr) separation. Spectra collected at 0.300 ms pulse width and averaged 500 times.

1170

H.M. Dion et al. / Talanta 57 (2002) 1161–1171

spectrum may have been a mixture of nitrate complexes such as Al(NO3)(OH)(H2O) + 2 or Al(NO3)(OH)(H2O) + 3 as detected by ESI-MS [21].

4. Conclusions The results from this study show that aqueous phase metal cations can be detected by IMS, and that most cations produce single, easily quantifiable responses. Average detection limits for the cations studied were in the low ng ml − 1 range. Uranium acetate and uranium nitrate produced response ions whose identities were dependent on the nature of the anion in solution. Additionally, many ions could be separated and directly analyzed by ESI-IMS. In some cases, multiple response ions were observed. For these cations, simultaneous determination of ions is more difficult and sensitivity is decreased. The ESI-MS literature, while very specific in metal species identification, is not readily transferable to ESI-IMS due to the differences in experimental conditions. It should be noted that the ESI-MS literature rarely reports bare cations, and even when using high-field CID regions, cation-solvent, and cation-solvent-anion complexes were the norm (Table 3). Thus, it is very likely that the species being detected were cationsolvent or cation-solvent-anion complexes, whose exact identity is a function of concentration, charge-competition at the needle, and gas-phase stability. With further refinement it may be possible to control the complexation so that the majority of the current is carried by a single species, making it possible to develop a sensitive method for the direct field analysis of inorganic cations in water.

Acknowledgements Heather Dion was supported by the National Science Foundation’s Integrative Graduate Education and Research Training Grant to Washington State University under Grant c9972817. The National Science Foundation’s Integrative Gradu-

ate Education and Research Training Grant also provided support for Luke Ackerman through the form of an undergraduate summer research experience fellowship. The authors also wish to acknowledge the Center for Ion Mobility Spectrometry sponsored by the Idaho National Environmental and Engineering Laboratory for supplies and operating expenses. The Savannah River Ecology Laboratory operated by the University of Georgia and supported by Financial Assistance Award Number DE-FC09-96SR18546 from the United States Department of Energy to the University of Georgia Research Foundation supported manuscript preparation.

References [1] S. Chapman, Phys. Rev. 52 (1937) 184. [2] J. Geiniec, H.L. Cox, D. Teer, M. Dole, Abstracts 20th Annual Conference of Mass Spectrometry, Allied Top, 1972, 276. [3] R.H. St. Louis, W.F. Siems, H.H. Hill Jr., J. Chromatogr. 479 (1989) 221. [4] R.W. Hallen, C.B. Shumate, W.F. Siems, T. Tsuda, H.H. Hill Jr., J. Chromatogr. 480 (1989) 233. [5] R.H. St. Louis, W.F. Siems, H.H. Hill Jr., J. Microcol. Sep. 2 (1990) 138. [6] G.R. Asbury, H.H. Hill Jr., Anal. Chem. 72 (2000) 580. [7] C. Wu, W.F. Siems, H.H. Hill Jr., Anal. Chem. 72 (2000) 396. [8] C.A. Srebalus, J. Li, W.S. Marshall, D.E. Clemmer, Anal. Chem. 71 (1999) 3918. [9] S.C. Henderson, S.J. Valentine, A.E. Counterman, D.E. Clemmer, Anal. Chem. 71 (1999) 291. [10] C. Wu, W.F. Siems, G.R. Asbury, H.H. Hill Jr., Anal. Chem. 70 (1998) 4929. [11] G.R. Asbury, C. Wu, W.F. Seims, H.H. Hill Jr., Anal. Chem. Acta 404 (2000) 273. [12] K.-M. Ho, A.A. Shvartsburg, B. Pan, Z.-Y. Lu, C.-Z. Wang, J.G. Wacker, J.L. Fye, M.F. Jarrold, Nature 392 (1998) 582. [13] J.L. Fye, M.F. Jarrold, Int. J. Mass Spec. 185/186/187 (1999) 507. [14] R.R. Hudgins, M. Imai, M.F. Jarrold, P. Dugourd, J. Chem. Phys. 111 (1999) 7865. [15] A.A. Shvartsburg, M.F. Jarrold, Chem. Phys. Lett. 317 (2000) 615. [16] P.R. Kemper, M.-T. Hsu, M.T. Bowers, J. Phys. Chem. 95 (1991) 10600. [17] S. Lee, T. Wyttenbach, G. von Helden, M.T. Bowers, J. Am. Chem. Soc. 117 (1995) 10159. [18] S. Mollah, A.D. Pris, S.K. Johnson, A.B. Gwizdala III, R.S. Houk, Anal. Chem. 72 (2000) 985.

H.M. Dion et al. / Talanta 57 (2002) 1161–1171 [19] I.I. Stewart, Spectrochimi. Acta B 54 (1999) 1649. [20] J.J. Corr, J.F. Anacleto, Anal. Chem. 68 (1996) 2155. [21] G.R. Agnes, G. Horlick, Appl. Spectrosc. 49 (1995) 324. [22] I.I. Stewart, G. Horlick, Anal. Chem. 66 (1994) 3983. [23] I.I. Stewart, G. Horlick, Anal. Chem. 15 (1996) 80. [24] G.R. Agnes, G. Horlick, Appl. Spectrosc. 46 (1992) 401. [25] G. Wang, R.B. Cole, Anal. Chem. 70 (1998) 873. [26] O. Schramel, B. Michalke, A. Kettrup, J. Chromatogr. A 819 (1998) 231. [27] G. Zoorob, F.B. Brown, J. Caruso, J. Anal. Atom. Spec. 12 (1997) 517.

1171

[28] P.S. Tornatore, Design Enhancements for High-Flow, High Resolution Stand Alone Ion Mobility Spectrometry, M.S. Thesis, Washington State University, May 2000. [29] L.M. Matz, P.S. Tornatore, H.H. Hill Jr., Talanta 54 (2001) 171. [30] C. Shumate, R.H. St. Louis, H.H. Hill Jr., J. Chromatogr. 373 (1986) 141. [31] Z. Karpas, Anal. Chem. 61 (1989) 684. [32] W.F. Siems, C. Wu, H.H. Hill Jr., P.R. Larsen, D.G. McMinn, Anal. Chem. 66 (1994) 4195. [33] S.J. Valentine, A.E. Counterman, C.S. Hoaglund-Hyzer, D.E. Clemmer, J. Phys. Chem. B 103 (1999) 1203.