Helium Plasma Source Time-of-Flight Mass Spectrometry: Off-Cone

Apr 21, 2000 - studied can be achieved with a digital oscilloscope. These detection limits can be easily improved with an advanced detection system, w...
0 downloads 0 Views 119KB Size
Anal. Chem. 2000, 72, 2455-2462

Helium Plasma Source Time-of-Flight Mass Spectrometry: Off-Cone Sampling for Elemental Analysis Yongxuan Su, Yixiang Duan,* and Zhe Jin

Chemical Science and Technology Division, CST-9, MS K484, Los Alamos National Laboratory, Los Alamos, New Mexico 87545

In this paper, an atmospheric pressure, helium microwaveinduced plasma (MIP) ion source coupled with an orthogonal acceleration time-of-flight mass spectrometer (TOFMS) is explored for elemental analysis. Studies of the relationship between ion signals and sampling distance of the MS reveal that background signals can be suppressed dramatically without sacrificing the signal intensities of analytes when the microwave plasma plume is off the tip of the sampler orifice. This “off-cone” ion sampling mode provides a technique to obtain nearly “clean” background spectra and, thus, eliminates the spectral interference from entrainment air and the working-gas species, making it possible to sensitively determine isotopes that suffered from spectral interference in ICPMS and MIPMS (such as 40Ca, 52Cr, 55Mn, and 56Fe). On the other hand, since the high-temperature plasma is kept away from the sampler aperture, off-cone sampling places little demand on the cooling device and the lifetime of the sampler plate can be extended. The instrumental system can provide a fairly good mass resolution of 1100 (fwhm). The detection limits (3σ) in the tens of picograms per milliliter level for the elements studied can be achieved with a digital oscilloscope. These detection limits can be easily improved with an advanced detection system, which is currently available in commercial markets. Over the past decade, plasma source mass spectrometry (PSMS) has emerged as a powerful technique for trace elemental analysis. PSMS provides a method to analyze samples for various elements at ultratrace levels by combining the ability of a plasma source to efficiently ionize elements with the sensitivity of quadrupole MS detection. The technique provides the capability for multielement detection and isotope ratio information. In addition, PSMS can be coupled with other analytical approaches, such as chromatographic techniques, to provide more detailed information about samples of interest. Inductively coupled plasmas (ICP) operated with argon gases have been widely used as ion sources for elemental mass spectrometry. ICPMS provides the analytical advantages of spectral simplicity, multielement analysis, low detection limits, and * To whom correspondence should be addressed. Email: [email protected]. 10.1021/ac991374h CCC: $19.00 Published on Web 04/21/2000

© 2000 American Chemical Society

wide dynamic range.1-6 Despite the success of the technique, there are still some elements that cannot be easily determined due to spectral interference. In a background spectrum of an argon ICPMS, polyatomic ions such as 38ArH, 40Ar, 40ArH, 38Ar14N, 36Ar16O, 40Ar15N,40Ar16O, 40Ar etc. are found to be significant, and 2 these ions interfere the isotopes of 39K, 40Ca, 41K, 52Cr, 55Mn, 56Fe, and 80Se.2,6 On the other hand, the use of argon as a plasma gas for ICPMS has limitations for effective ionization of elements such as Se, Hg, P, S, and halogens.7 For these reasons, interest in alternative ion sources for MS has been increased over the past decade,8-12 one of which is microwave induced plasma (MIP) ion source. The MIP sources have significant advantages, including reduced gas flow rate and power consumption and an ability to sustain plasmas with alternative gases. Several working gases have been explored in MIPMS research with different levels of success.13-16 Research work on moderate-/ high-power nitrogen MIPMS has been investigated and shown to have some promising features for elemental analysis.17 However, strong background signals in nitrogen MIPMS make it ineffective at determining elements in the low mass range.17,18 A helium(1) Houk, R. S.; Fassel, V. A.; Flesch, G. D.; Svec, H. J.; Gray, A. L.; Taylor, C. E. Anal. Chem. 1980, 52, 2283-2289. (2) Vela, N. P.; Olson, L. K.; Caruso. J. A. Anal. Chem. 1993, 65, 585A-597A. (3) Douglas, D. J.; Quan, E. S. K.; Smith, R. G. Spectrochim. Acta, Part B 1983, 38, 39-48. (4) Aggarwal, J. K.; Shabani, M, B.; Palmer, R. A.; Gnarsdottir, K. V. Anal. Chem. 1996, 68, 4418-4423. (5) Horlick, G.; Shao, Y. In Inductively Coupled Plasma in Analytical Atomic Spectrometry, 2nd ed.; Montaser, A., Golightly, D. W., Eds.; VCH: New York, NY, 1992. (6) Tan, S. H.; Horlick, G. Appl. Spectrosc. 1986, 40, 445-460. (7) Houk, R. S. Anal. Chem. 1986, 58, 97A-105A. (8) Wilson, D. A.; Vickers, G. H.; Hieftje, G. M. Anal. Chem. 1987, 59, 16641670. (9) Houk, R. S.; Montaser, A.; Fassel, V. A. Appl. Spectrosc. 1983, 37, 425428. (10) Satzger, R. D.; Fricke, F. L.; Caruso, J. A. J. Anal. At. Spectrom. 1988, 3, 319-323. (11) Chan, S.; Montaser, A. Spectrochim. Acta, Part B 1985, 40, 1467-1472. (12) Zhang, H.; Nam, S. H.; Cai, M. X.; Montaser, A. Appl. Spectrosc. 1996, 50, 427-435. (13) Okamoto, Y. J. Anal. At. Spectrom. 1994, 9, 745-749. (14) Shen, W.; Davidson, T. M.; Creed, J. T.; Caruso, J. A. Appl. Spectrosc. 1990, 44, 1003-1010. (15) Ohata, M.; Furuta, N. J. Anal. At. Spectrom. 1998, 13, 447-453. (16) Shirasaki, T.; Hiraki, K. Bunseki Kagaku 1994, 43(1), 25-29. (17) Oishi, K.; Okumoto, T.; Iinot, T., Koga, M.; Shirasaki, T.; Furuta, N. Spectrochim. Acta, Part B 1994, 49, 901-914. (18) Shen, W.; Davidson, T. M.; Creed, J. T.; Caruso, J. T. Appl. Spectrosc. 1990, 44, 1011-1014.

Analytical Chemistry, Vol. 72, No. 11, June 1, 2000 2455

supported ion source offers some promising characteristics that make it even more attractive when compared with the argon. First, helium is a monoisotopic gas with low mass and has a significantly higher ionization potential (24.5 ev) than argon does (15.8 ev). In addition, helium provides greater thermal conductivity than argon (He 1.520 mW/cm/K versus Ar 0.1772 mW/cm/K). As a result, more efficient ionization could be expected for metals and nonmetals. Since the ionization potentials of most nonmetal elements are close to or even higher than that of argon (15.8 eV), it is difficult to use argon plasma to ionize these elements.19 Second, helium metastable states are believed to play a major role in energy transfer within the plasma. The lowest metastable state of atomic He is 19.73 eV, whereas the lowest metastable state of atomic Ar is only 11.50 eV. The higher ionization potential and metastable states of helium should provide better ionization efficiency. Furthermore, the spectral interference induced by the argon isotopes could be reduced or eliminated if pure helium is used. Because of these characteristics, the helium plasma source becomes an attractive alternative in elemental mass spectrometry. The helium microwave plasma coupled with a mass spectrometer for elemental analysis was first described by Satzger et al. in 1987.20 Although detection limits obtained were promising, the analyses were impaired by the complex mass-spectral background, presumably from air entrainment. Fecher and Nagengast explored using helium MIPMS for the analysis of biological materials.21 Compared with argon ICPMS, the rate of polyatomic species is high and the sensitivity for As, Se, Br, and I is low despite the high ionization potential of the helium.21 Chambers and Hieftje conducted fundamental studies on the sampling process of a helium MIPMS.22 They believed that two obstacles needed to be overcome if helium PSMS could become a viable alternative to argon ICPMS.22 The first obstacle would be to eliminate the air entrainment and the second to maintain a high ion flux through the MS interface. Wu et al. coupled a helium microwave plasma torch (MPT) ion source to a quadrupole mass spectrometer for elemental determination.23 Using a pneumatic nebulization and a desolvated sample introduction system, detection limits for seven nonmetals range from 12 ng/mL to 1.0 µg/mL.23 Most research work on MIPMS has been performed with a quadrupole-based ion filter. Since a quadrupole filter is usually operated with only unit-mass resolution, isobaric interference, such as 40Ar16O+ with 56Fe+, is especially problematic. In addition, because a quadrupole mass analyzer is operated under scanning mode, it is difficult to handle transient signals, such as pulsed ion source and continuous ion source recruited electrothermal vaporization or flow injection techniques. These difficulties in quadrupole mass filter have stimulated scientists to explore alternative mass analyzers, among which, a time-of-flight (TOF) mass analyzer has been found to be one of the most promising alternatives.24-30 (19) Koppenaal, D. W.; Quinton, L. F. J. Anal. At. Spectrom. 1988, 3, 667-672. (20) Satzger, R. D.; Fricke, F. L.; Brown, P. G.; Caruso, J. A. Spectrochim. Acta, Part B 1987, 42, 705-712. (21) Fecher, P. A.; Nagengast, N. J. Anal. At. Spectrom. 1994, 9, 1021-1027. (22) Chambers, D. M.; Carnahan, J. W.; Jin, Q.; Hieftje, G. M. Spectrochim. Acta, Part B 1991, 46, 1745-1765. (23) Wu, M.; Duan, Y.; Jin, Q.; Hieftje, G. M. Spectrochim. Acta, Part B 1994, 49, 137-148. (24) Hieftje, G. M.; Myers, D. P.; Li, G.; Mahoney, P. P.; Burgoyne, T. W.; Ray, S. J.; Guzowski, J. P. J. Anal. At. Spectrom. 1997, 12, 287-292.

2456

Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

A TOF mass analyzer is perhaps the simplest type of mass spectrometer. Ions are formed in a source and accelerated to a constant kinetic energy as they enter a drift region. These ions arrive at a detector following flight time that is proportional to the square root of their masses. Therefore, the time-dependent output of the detector located at the exit of the field-free region represents a complete mass spectrum each time. Because ions of all masses within each packet are detected, the TOF mass analyzer possesses high ion-utilization efficiency in comparison with scanning mass spectrometers, where only a single mass per charge can be monitored at any given time. The compromise between mass coverage and sensitivity inherent in scanning systems, such as quadrupole and sector instruments, does not apply to the TOF mass analyzer. Since the field-free region of a TOF mass analyzer is only a hollow stainless steel tube with a few grids, the transmission efficiency of TOF is the highest among all mass analyzers, and thus superior sensitivity can be expected. Furthermore, the repetition rate of a TOF instrument is governed only by the flight time of the heaviest ions through the field-free region. Since the flight time of atomic species (amu e 238) is generally less than 100 µs, more than 10 000 spectra covering the full atomic mass range can be generated in each second. This high repetition rate and simultaneous detection capabilities of the TOF mass spectrometer not only reduce the analysis time, but also make the TOF instrument an ideal choice for transient signal analysis. On the basis of the above considerations, it is desirable to couple a helium microwave plasma ion source with a TOFMS for elemental analysis. In this work, instrumentation of an atmospheric pressure helium microwave plasma source coupled with a TOF mass analyzer is described. A new ion sampling method, off-cone sampling, was first explored and investigated for elemental analysis. This off-cone sampling technique provides a new method of determining elements with very low background spectral interference. The influences of the operating parameters of this instrumental system were studied, and the analytical figures of merit are also discussed and evaluated.

EXPERIMENTAL SECTION Microwave Plasma Ion Source. The microwave plasma ion source used in this study consists of a commercial 2450 MHz microwave generator (Model GMP-03K/SM, Sairem) and an inhouse fabricated MPT source. This microwave plasma source is a modified version of the one originally introduced by Jin et al.,31 and has been demonstrated elsewhere for analytical atomic (25) Myers, D. P.; Li, G.; Yang, P.; Hieftje, G. M. J. Am. Soc. Mass Spectrom. 1994, 5, 1008-1016. (26) Myers, D. P.; Li, G.; Mahoney, P. P.; Hieftje, G. M. J. Am. Soc. Mass Spectrom. 1995, 6, 400-410. (27) Myers, D. P.; Li, G.; Mahoney, P. P.; Hieftje, G. M. J. Am. Soc. Mass Spectrom. 1995, 6, 411-420. (28) Myers, D. P.; Mahoney, P. P.; Li, G.; Hieftje, G. M. J. Am. Soc. Mass Spectrom. 1995, 6, 920-927. (29) Mahoney, P. P.; Li, G.; Hieftje, G. M. J. Anal. At. Spectrom. 1996, 11, 401405. (30) Mahoney, P. P.; Ray, S. J.; Li, G.; Hieftje, G. M. Anal. Chem. 1999, 71, 1378-1383. (31) Jin, Q.; Zhu, C.; Borer, W.; Hieftje, G. M. Spectrochim. Acta, Part B 1991, 46, 417-430.

Figure 1. Schematic diagram of the MIP-TOFMS system. E, extraction cone; L1, L2, L3, cylindrical ionic lenses; S, slit; R, Repeller; G1, G2, TOF entry grids; X1, X2, x-direction steering plates; Y1, Y2, y-direction steering plates; G3, reflectron entry grid; R1, retarding grid; R2, reflecting grid; MCP, dual microchannel plates.

spectrometry.32-34 A X-Y-Z three-dimension translatable stage is used to precisely adjust the position of the plasma torch and to facilitate optimization of the ion sampling position. Time-of-Flight Mass Spectrometer. Figure 1 shows the schematic diagram of the MIPTOFMS system. The TOFMS is an angular-type instrument (model Angular reflectron D-850, R. M. Jordan Co., Grass Valley). The microwave plasma ion source is interfaced to the TOFMS via a self-designed, three-stage differentially pumping interface. A sampler plate with an aperture diameter of 0.5 mm samples the ions from the atmospheric plasma source into the first vacuum stage where they are pumped with a 18 L/s mechanical pump (model E1M 18, Edwards High Vacuum). Located 8 mm downstream is a 0.5-mm skimmer orifice. The second vacuum stage, between the skimmer plate and the extraction cone (E), is pumped by a 330 L/s Turbo molecular pump (model TPH 330, Pfeiffer Balzers). The extraction cone contains a 1.5-mm orifice and is positioned 12 mm behind the skimmer aperture. Behind the extraction cone is the third vacuum stage, which is also pumped with a 330 L/s Turbo molecular pump (model TPH 330, Pfeiffer Balzers). Four more ion optical elements (L1, L2, L3, S), as shown in Figure 1, were installed in this stage. The cylindrical lenses (L1, L2, L3) have a diameter of 12 mm and a length of 7 mm, respectively, and a rectangular slit (S) 2 mm in width and 12 mm in length. The third-stage ionic elements further focus ions into the region between repeller plate (R) and the first entrance grid (G1) of the TOFMS (Figure 1) and offer improved resolution power, increased sensitivity, and lower noise level. The distance from the tip of the sampler orifice to the center of the repelling region is designed to be as short as 12 cm. This short (32) Madrid, Y., Wu, M.; Jin, Q.; Hieftje, G. M. Anal. Chim. Acta 1993, 277, 1-8. (33) Prokisch, C.; Broekaert, J. A. C. Spectrochim. Acta, Part B 1998, 53, 11091119. (34) Pack, B. W.; Hieftje, G. M. Spectrochim. Acta, Part B 1997, 52, 21632168.

distance is desirable for high ion throughput from the atmospheric pressure ion source. The TOFMS used in this work is operated in a pulsed extraction mode, followed by an acceleration zone. The pulsed extraction is achieved by applying a positive pulse with a controllable repetition rate of several kilohertz to the repeller plate, which sends the ions from a continuous beam to the second step of acceleration (between G1 and G2 entrance grids). This second acceleration zone consists of a linear electrostatic field maintained with six stainless steel rings. The entrance grids G1 and G2 have a diameter of 12 mm, and all other rings have a diameter of 25 mm. The grid G2 is biased at -2000 V, and a linear potential gradient across the second acceleration region is achieved by using equal resistance of 2.2 MΩ between the ring elements. The drift tube is insulated from the vacuum chamber walls and receives the same potential as in G2. Currently, this second acceleration region is, in length, approximately three times the distance between the repeller and the first entrance grid (G1). Two steering plates, X1 and X2 (25 mm long and separated by 25 mm), are located after the second entrance grid (G2) to compensate for the initial perpendicular velocity of the supersonic beam. A pulse with a peak height of 200 V can be applied to a deflection plate (Y1) at an adjustable delay time after the repeller pulse. The time delay and pulse width of this pulse are adjusted to reduce the signal intensity of dominant background species. The deflection plates (Y1, Y2) are perpendicular to the steering plates (X1, X2). The ion packet enters a field-free drift region about 1.0 m in length. At the end of this region, ion velocity is slowed and then reversed by the ion reflector installed. The newly focused ion packet is detected with a dual 40-mm microchannel plate (MCP) detector (Galileo Electrooptic Corp.). The total ion flight path is about 1.5 m. Signals from MCP are then sent to a wide-band preamplifier (model SR445, Standford Research Systems) and collected with a digital oscilloscope (model 54520A, Hewlett Packard Company). Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

2457

Table 1. Typical Operating Conditions of the Helium MIP-TOFMS

Kn )

Plasma Parameters microwave forward power (W) microwave reflected power (W) carrier gas flow rate (L/min) plasma gas flow rate (L/min) sample up-take rate (mL/min) sampling distance (mm)

100 16 mm), signals of both NO and the elements most studied decrease with Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

2459

Figure 3. Influence of working gas flow rates on signal intensities. Sampling distance, 12 mm; other parameters are the same as in Figure 2, part b.

an increase in sampling distance. It is interesting that the signal of Bi still shows a trend of increase in the third sampling region. A sampling distance around 12 mm was used in the experimental work. B. Working Gas Flow Rate. The influence of the working gas flow rate on ion signals was further investigated to determine (39) Niu, H.; Houk, R. S. Spectrochim. Acta, Part B 1996, 51, 779-815.

2460

Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

the optimal operating conditions of the microwave plasma ion source under off-cone sampling mode. The influence of the working gas flow rate on the selected elements is given in Figure 3. Both the flow rates of plasma and carrier gases were found to have significant influence on the ion signals. This finding is different from the previous observation reported by Wu et al. with a quadrupole MS where the plasma gas flow rate was found to

Figure 4. Influence of microwave forward power on signal intensities. Sampling distance, 12 mm; other parameters are the same as in Figure 2, part b.

have little influence on ion signals.23 As shown in Figure 3, either a lower plasma gas flow rate combined with a higher carrier gas flow rate or a higher plasma gas flow rate associated with a lower carrier gas flow rate is desirable for Cu, Ni, Cr, and Bi. However, for Zn, Mn, and Ba, a low carrier gas flow rate around 0.6 L/min gives a maximum of signal intensity when the plasma gas flow rate is tuned from 0.4 to 0.8 L/min. The carrier gas flow rate not only alters the nebulization efficiency, droplet size distribution, sample aerosol transport efficiency, and the analyte residence time in the plasma, but also changes the plasma characteristics such as plasma temperature and volume. Although more aerosols can be delivered into the plasma with the increase of carrier gas flow rate, the ionization efficiency of the microwave plasma may be lowered through the central “cooling” effect due to the aerosol introduction. In addition, a significant change in carrier gas flow rate results in a variation of plasma configuration and appearance and thus causes a shift of the optimum sampling position. Moreover, as shown in Figure 3, the maximum signal intensity varies with the gas flow rates and the elements, which makes the optimization of the experimental conditions for multielement analysis complicated. As a compromise among these effects, a carrier gas flow rate of 0.7 L/min and a plasma gas flow of 0.6 L/min were usually used in this work. C. Microwave Forward Power. Figure 4 depicts the influence of microwave forward power on the ion signal intensities with a sampling distance of 12 mm. The ion signals of the analyte seem to be power-dependent. Increments in microwave power between 80 and 140 W produce a rise in ion signals for Zn and Ba. In contrast, maximum signals for Ni and Cr appear when the microwave power is around 90 W, and a power about 120 W is desirable for the Mn ion signal. It is noticeable that the increase of microwave power results in a slight signal decrease for Cu and Bi. Generally, a microwave power of 100 W was employed in the work. Analytical Figures of Merit. Operating the helium microwave plasma TOFMS system in the off-cone sampling mode provides several analytical advantages. First, sampling ions with the offcone mode can effectively suppress the background signals and,

Figure 5. Mass spectra of 20 ng/mL Mn, Ni, and Cu solution. Sampling distance, 12 mm; carrier gas, 0.7 L/min; plasma gas, 0.5 L/min; microwave forward power, 110 W; Detector voltage, -1800 V; signal preamplified factor, 125.

therefore, simplify the background mass spectra. A nearly “clean” background mass spectrum can be obtained with only the 30NO peak when the sampling distance is greater than 9 mm (see Figure 2a). Since spectral interference caused by background peaks is minimized or eliminated, isotopes that suffer from spectral interference in ICPMS and MIPMS, such as 52Cr and 55Mn, can be easily determined with the off-cone sampling mode. Second, acquiring the clean spectral background does not sacrifice the analyte signal. On the contrary, even more intense analyte signals can be achieved in the sampling range of 10-16 mm (see Figure 2b). Such experimental results are the most desirable for elemental mass spectrometry because low background interference and high detection capability can be achieved simultaneously. In addition, off-cone sampling presents a minor requirement of the cooling device for the sampler plate because the hightemperature plasma is kept away from the tip of the sampler aperture. As a result, the lifetime of the sampler plate can be significantly extended. In this work, no obvious erosion of the sampler aperture was found after the instrument had been operating for six months. Figure 5 shows the mass spectrum of a 20 ng/mL solution of Mn, Cu, and Ni obtained without any spectral interference. Operational parameters of the instrumental system were optimized with respect to the Mn signal. As shown in Table 2, the present helium MIP-TOFMS can provide detection limits of about 30170 pg/mL for the elements studied. Since the helium MIP ion source was operated with the off-cone sampling mode in this work, the pressure in the drift tube region is about 9.0 × 10-6 Torr with a sampling distance of 12 mm. As a result, a potential of -1800 V is usually used in the MCP detector with the consideration of its safety. Though the first vacuum stage is pumped with an 18 L/s mechanical pump, the diameters of the two pumping ports are only 4 mm and they therefore restrain the actual pumping capacity of the mechanical pump. Modification of the pumping ports of the first vacuum stage can further lower the backing pressure, and therefore a higher working potential of the MCP detector up to -2200 V can be applied for a better detection power. On the other hand, the current spectral processing system is simply a Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

2461

spectral processing would significantly improve the detection power. It is noticeable in Figure 5 that a tailing ringing occurred at the anode output of the MCP detector. The dual channel MCP detector is susceptible to electric ringing if a large ion packet reaches the detector. Usually, a capacitor is needed between ground and each electrode of the detector to stabilize the potential of the detector assembly. Figure 6 shows mass spectra of Pb and Bi obtained with the helium off-cone sampling MIP-TOFMS. The current MIP-TOFMS can provide a mass resolution of about 1100, which is fairly good for elemental determination. Further modification of the ion optics system, including the recruitment of a quadrupole lens right after the third cylinder lens (L3), could possibly eliminate the tradeoff problem between ion transmission efficiency and mass resolution that frequently occurred in a three-cylindrical-lens system.25,26 Figure 6. Mass spectra of 200 ng/mL Pb and Bi solution. Operating parameters are the same as in Figure 2, part b. Table 2. Detection Limits (3σ) with Off-Cone Sampling Helium MIP-TOFMS element

first ionization potential (eV)

detection limit (pg/mL)

Cr Mn Ni Cu Zn Ba Pb Bi

6.77 7.44 7.64 7.73 9.39 5.21 7.42 7.29

76 52 47 32 167 45 83 148

digital oscilloscope with signal average function. Although the TOFMS can produce several thousands of mass spectra with full elemental mass range coverage in each second, the digital oscilloscope used in this work could only average about 105 mass spectra in a second. Therefore, most of the data produced by the TOFMS are wasted with the current detection system. The employment of a state-of-the-art detection device for signal and

2462

Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

CONCLUSION The combination of helium microwave plasma ion source and a TOF mass spectrometer was explored for elemental analysis. Although the pressure in the drift region is currently limited by the pumping capacity of the first stage, operating the helium MIPTOFMS system with an off-cone sampling mode demonstrates its abilities to eliminate background signals and increase the detection power of the analytes. This ion-sampling mode provides a new method to minimize the spectral interference resulting from background gas species in MIPMS and ICPMS. A mass resolution of 1100 and detection limits of tens of picograms per liter for most elements studied were achieved. With a modified ion optics system and a better data acquisition device, a significant improvement of these values can be expected. ACKNOWLEDGMENT Financial support from the Department of Energy and the Los Alamos National Laboratory program office is gratefully acknowledged by the authors. Received for review November 29, 1999. Accepted February 29, 2000. AC991374H