Anal. Chem. 2000, 72, 2463-2467
A Multimicrospray Nebulizer for Microwave-Induced Plasma Mass Spectrometry Min Huang,*,† Atsumu Hirabayashi,† Toshihiro Shirasaki,‡ and Hideaki Koizumi†
Central Research Laboratory, Hitachi, Ltd., 1-280 Higashi-Koiga Kubo, Kokubunji, Tokyo 185-8601, Japan, and Hitachi Science Systems, Ltd., 882 Ichige, Hitachinaka, Ibaraki 312-8504, Japan
We have developed a nebulizer, called a multimicrospray nebulizer (MMSN), that efficiently introduces analytes for plasma mass spectrometry and plasma emission spectrometry. In this nebulizer, both the sample solution and the nebulizer gas are divided into several streams to produce a multispray. That is, the MMSN is a nebulizer that contains several micronebulization units, each unit including an orifice for passing the nebulizer gas and a capillary for introducing a sample solution. The microspray from each micronebulization unit can be operated at a microliter per minute sample uptake rate to achieve high nebulization efficiency. The multimicrospray nebulizer is capable of introducing more analyte to the plasma compared with a single-orifice micronebulizer, which has a very low sample uptake rate. In this work, an MMSN with three orifices was found to be suitable for microwave-induced plasma mass spectrometry (MIP-MS). The sample uptake rate can be varied within a range of 5-250 µL/min. Therefore, the nebulizer is unique in its ability to deal with various sample volumes and provide high nebulization efficiency. The sensitivity for all elements obtained with the MMSN was higher than that obtained with a conventional concentric nebulizer (CCN), which is difficult to achieve with other types of microintroduction nebulizers. For most elements, the MIP-MS sensitivity was improved about 2-fold at a sample uptake rate of 150 µL/min, a much lower rate than that for the CCN (usually 0.5-1.5 mL/min). The sensitivity for arsenic was improved by a factor of 5. The relative standard deviation was found to be less than 2.0%. The conventional concentric nebulizer (CCN) is widely used in plasma emission spectrometry and mass spectrometry because of its simplicity. However, the CCN usually requires an uptake rate of 0.5-1.5 mL/min and has poor nebulization efficiency. Recently, various micronebulizers were reported, such as the direct-injection nebulizer (DIN),1,2 the microconcentric nebulizer (MCN),3 the high-efficiency nebulizer (HEN),4-6 the oscillating capillary nebulizer (OCN),7 the direct-injection high-efficiency †Hitachi,
Ltd. Hitachi Science Systems, Ltd. (1) Lawrence, K. E.; Rice, G. W.; Fassel, V. A. Anal. Chem. 1984, 56, 289292. (2) Wiederin, D. R.; Smith, F. G.; Houk, R. S. Anal. Chem. 1991, 63, 219-225. (3) Vanhaecke, F.; Van Holderbeke, M.; Moens, L.; Dams, R. J. Anal. At. Spectrom. 1996, 11, 543-548. ‡
10.1021/ac991379e CCC: $19.00 Published on Web 04/21/2000
© 2000 American Chemical Society
nebulizer (DIHEN),8 and the sonic spray nebulizer (SSN).9 Micronebulizers are capable of introducing sample solutions at the microliter per minute level and offer higher analyte-transport efficiency than that of a conventional concentric nebulizer. Our previous work9 showed that the detection limit (DL) with the SSN is comparable to that of the CCN, yet with a sample uptake rate of only one-seventh the CCN rate. For the elements that have a high ionization potential, such as arsenic, selenium, and gold, the DLs were improved by a factor of about 7. We also demonstrated that this improvement is due to higher ionization efficiency caused by higher plasma temperature when the sample solution is introduced at microliter per minute uptake rates. However, with such a low sample uptake rate, no significant change in DL was found for easily ionized elements, indicating no significant increase in the amount of analyte transported to the plasma. This is a common problem for various types of micronebulizers. Unfortunately, we cannot improve the DL of the SSN simply by raising the sample uptake rate: as the sample uptake rate rises, the nebulization behavior will shift to that of the CCN and the nebulization efficiency will decrease. Thus, we have developed a new nebulizer called the multimicrospray nebulizer (MMSN). In this nebulizer, a sample solution is divided into several streams that are nebulized separately. The nebulizer has several orifices, each containing a capillary for delivering the sample solution. The nebulizer gas is ejected from the area between the orifice and the capillary. A fine aerosol is produced due to collision between the high-speed gas and the sample solution. Each orifice and capillary set, we call a nebulization unit, performs like a micronebulizer. In this way, we have increased the sample uptake rate while keeping the nebulization efficiency high. We tested the MMSN by using it to introduce sample solutions for microwave-induced plasma mass spectrometry (MIP-MS). (4) Tan, H.; Meinhard, B. A.; Meinhard, J. E. Recent Investigations of Meinhard Concentric Nebulizers. Presented at the 19th Annual Meeting of the Federation of Analytical Chemistry and Spectroscopy Societies, Philadelphia, PA, Sept 1992. (5) Olesik, J. W.; Kinzer, J. A.; Harkleroad, B. Anal. Chem. 1994, 66, 20222030. (6) Liu, H.; Montaser, A. Anal. Chem. 1994, 66, 3233-3242. (7) Browner, R. F. Presented at the 21st Annual Meeting of the Federation of Analytical Chemistry and Spectroscopy Societies, St. Louis, MO, Oct 1995. (8) McLean, J. A.; Zhang, H.; Montaser, A. Anal. Chem. 1998, 70, 1012-1020. (9) Huang, M.; Shirasaki, T.; Hirabayashi, A.; Koizumi, H. Anal. Chem. 1999, 71, 427-432.
Analytical Chemistry, Vol. 72, No. 11, June 1, 2000 2463
Table 1. Instrumental Conditions and Measurement Parameters for MIP-MS microwave frequency microwave power plasma gas flow rate carrier gas flow rate sample uptake rate spray chamber spray chamber temperature mass analyzer sampling cone skimmer cone data acquisition mode mass species measured a
2.45 GHz 1.3 kW 13 L/min 1.3 L/mina 150 µL/mina double-pass 2 °Ca quadrupole analyzer copper, 0.8-mm diameter, 0.5 mm long copper, 0.4-mm diameter, 0.3 mm long peak hopping 51V+, 52Cr+, 55Mn+,59Co+, 63Cu+, 75As+
Unless otherwise noted.
EXPERIMENTAL SECTION Instrumentation. A high-power nitrogen microwave induced plasma mass spectrometer (P-6000, Hitachi, Ltd.) was used in this work. A nitrogen plasma (2.45 GHz, up to 1.5 kW) with a donutshaped structure was produced by using an Okamoto cavity.10 The mass spectrometer was a quadrupole type, with a three-stage vacuum system. The MIP torch consisted of two concentric tubes, rather than three tubes as in an ICP torch. The position of the MIP torch relative to the orifice of the sampling cone could be finely adjusted in three dimensions to allow ions being extracted to pass efficiently from the plasma into the mass analyzer. The ion focus lenses and other features were optimized to achieve a maximum signal-to-background ratio before the measurements. The operating conditions listed in Table 1 were used, unless otherwise specified. As used in most plasma mass spectrometers, a system for cooling the spray chamber to about 2 °C was employed. Besides reducing interference from oxide ions, solvent removal was also necessary to prevent excess water vapor from entering the MIP. The Multimicrospray nebulizer. The MMSN nebulizer we used contained three nebulization units. Each unit was composed of an orifice (170 µm) and a fused-silica capillary (127-µm o.d.; 50-µm i.d.) and worked like a micronebulizer. However, these nebulization units shared the same sources of nebulizer gas and sample solution, and so affected each other. To achieve highly efficient and stable nebulization, it was important to equally divide the nebulizer gas among the orifices, as well as the sample solution among the fused-silica capillaries. The MMSN is shown schematically in Figure 1. The three orifices form the points of an equilateral triangle and are 2 mm apart. Each of the three capillaries is 80 mm long. The end of each capillary is 0.1 mm from the orifice. The MMSN was designed to be directly inserted into the endcap of the spray chamber and sealed with two internal O-rings in our system. Therefore, the MMSN is easy to attach or remove. The sample solution can be precisely fed, at an uptake rate as low as 5 µL/min, by using a syringe pump (JP-S1 microfeeder, Furue Science, Inc.) and a 1-mL syringe (Hamilton 1000 gastight syringe, Hamilton, Ltd.). To avoid any contamination from the metal needle of the syringe, the needle was replaced with a homemade plastic connector attached with a Teflon tube. The flow (10) Okamoto, Y. Anal. Sci. 1991, 7, 283-288.
2464 Analytical Chemistry, Vol. 72, No. 11, June 1, 2000
Figure 1. Cross-sectional view of the multimicrospray nebulizer (MMSN).
rate of the nebulizer gas could be adjusted from 0.2 to 1.45 L/min by a thermal mass flow controller (model 3650, Kofloc, Ltd.), with an error of 2.0% of the flow scale used. Evaluation of the MMSN. We evaluated the analytical capabilities of the nebulizer for MIP-MS by measuring prepared standard solutions. We compared the results obtained with the MMSN to those obtained with a CCN system. The amounts of analyte transported to the plasma under various conditions for the MMSN and the SSN were measured and compared by collecting and weighing the amounts of the drained solution after spraying a certain amount of sample solution. The drained solution was collected in a beaker in a system that included a double-pass spray chamber connected to a plasma torch, but without lighting the plasma. The beaker used to collect the drain was isolated from the outside atmosphere to prevent vaporization of the drain. Although this method introduced a relatively large error and the precise amount of analyte could not be obtained due to the very low uptake rates, the results were useful for a relative comparison between the MMSN and the SSN. To reduce the error, we sprayed 5-15 mL quantities of solution for these measurements, depending on the sample uptake rate used. Reagents. Standard solutions of As, Cd, Cr, V, Co, and Mn in a 0.1% HNO3 medium were prepared by diluting stock solutions (1000 µg/mL, SPEX Plasma Standard, SPEX Industries, Edison, NJ) to the concentrations required for determinations with deionized water (18 MΩ) and high-purity nitric acid. Blank solutions of deionized water with the 0.1% HNO3 medium were also prepared. RESULTS AND DISCUSSION Optimization of the Nebulizer. As shown in Figure 1, the MMSN has a spacious gas pathway in which the gas pressure from the gas inlet to the orifice is not greatly reduced. Therefore, a very high gas pressure, as used in many other types of nebulizers, was not necessary. A gas pressure of 5 atm could be used efficiently for nebulization. Although the nebulizer is capable of introducing a sample solution at a low microliter per minute rate, we mainly evaluated the nebulizer at a sample uptake rate from 100 to 200 µL/min to achieve high sensitivity. Unlike that of a CCN nebulizer, the sample uptake rate of the MMSN is independent of the flow rate of the nebulizer gas. However, the nebulizer gas flow rate does affect the aerosol produced. In our previous SSN experiments, we found that a good aerosol could only be produced with a rate higher than 0.4 L/min when the diameter of the orifice and the outer diameter of the
Table 2. Nebulization Stabilities for the MMSN (RSD, %) elements, amt (ng/mL) uptake rate (µL/min)
Cr, 5
Mn, 5
Co, 5
Cu, 5
As, 100
7 20 30 60 80 100 150 200 250
1.43 1.84 0.20 2.20 1.97 1.03 0.43 1.77 0.72
1.13 1.53 1.00 0.56 0.52 1.55 0.19 2.03 0.88
1.74 1.96 0.87 1.40 0.96 0.83 2.09 1.09 1.15
1.35 1.28 0.42 0.69 0.38 0.54 1.43 0.16 0.72
1.90 2.52 0.44 0.97 1.04 1.55 0.40 0.98 1.24
capillary were 250 and 150 µm, respectively.9 In this study (with diameters of 127 and 50 µm, respectively), a minimum gas flow rate of 0.25 L/min for each orifice was required to achieve a stable spray. The variation in the signal intensity due to the gas flow rate represents a compromise between a high nebulization efficiency and a long plasma retention time for analytes in the plasma. A high gas flow rate helps to produce a fine aerosol but shortens the retention time and lowers the ionization efficiency. Results obtained with the CCN showed that when the gas flow rate was higher than 1.35 L/min, the signal intensity of MIP-MS began to decrease because of a short retention time. In the case of the MMSN used in this work, the maximum gas flow rate for the nebulizer was 1.30 L/min under 5 atm; i.e., the nebulizer was operated under the highest possible nebulizer gas speed and the flow rate was matched to the MIP. We also installed and evaluated an MMSN with a maximum gas flow rate of 1.45 L/min; the highest sensitivity obtainable was 30% lower. Therefore, the nebulizer with the maximum gas flow rate of 1.30 L/min was used and the flow rate of the nebulizer gas was fixed at 1.30 L/min in this study, unless otherwise specified. Nebulization Stability. It is important to divide the nebulization gas and the sample solution equally to ensure a stable multispray. We evaluated the nebulization stability for the MMSN through 10 successive measurements of a standard solution containing 5 ng/mL Cr, Mn, and Co and 100 ng/mL As. The results are given in Table 2. In most cases, the relative standard deviation was found to be less than 2% when the sample uptake rate was varied from 7 to 250 µL/min. This stability is sufficient for trace analysis. Transportation Efficiency and Sensitivity. Transportation efficiency is defined as the amount of analyte transported to the plasma divided by the total amount of analyte nebulized. The efficiencies for the MMSN and the SSN were measured and compared by collecting and weighing the drained sample solutions, as described in the Experimental Section. The efficiencies for both the MMSN and the SSN appeared to be considerably higher than that for a CCN (usually less than 3%). With an increasing sample uptake rate, the transportation efficiency decreased in both the MMSN and the SSN (Figure 2), but the transportation efficiency for the MMSN was always higher than that for the SSN, regardless of the sample uptake rate. Clearly, the MMSN was more efficient than the SSN. This conclusion can also be drawn from examining Figure 3. The amounts of analyte transported to the plasma in the figure were calculated by taking the sample uptake rate into consideration. No significant increase
Figure 2. Comparison of the transportation efficiencies of the MMSN and the SSN.
Figure 3. Comparison of the amounts of analyte transported to the plasma by the MMSN and the SSN.
Figure 4. Effect of the nebulizer gas pressure on the analyte transportation efficiency for the MMSN.
in the amount of analyte transported to the plasma was found for the SSN when the sample uptake rate was above 90 µL/min, while an increase was seen for the MMSN up to 200 µL/min. This confirms the advantage of the MMSN. We also investigated the effect of the nebulizer gas pressure on the MMSN’s transportation efficiency. The efficiencies obtained with gas pressures of 3.3 and 5 atm are compared in Figures 4 and 5. Although the MMSN provided high efficiency at a high uptake rate even at 3.3 atm, higher gas pressure increased the efficiency. Usually, it is not difficult to use a 5 atm gas supply Analytical Chemistry, Vol. 72, No. 11, June 1, 2000
2465
Figure 5. Effect of the nebulizer gas pressure on the amount of analyte transported to the plasma for the MMSN.
Figure 7. Sensitivity improvement factors at various sample uptake rates for the MMSN compared with the conventional concentric nebulizer (CCN). Here, the factor for the CCN is defined as 1.
Figure 6. Relationship between the signal intensity and the sample uptake rate for the MMSN.
system in a laboratory. Therefore, nebulizer gas at 5 atm was used for this work. Although the MMSN can introduce sample solutions at the microliter per minute rate, we focused on achieving high nebulization efficiency at high sample uptake rates in this work. The relationship between the signal intensity and the sample uptake rate at the optimized gas flow rate (1.3 L/min) is shown in Figure 6. Except for the case of As, the intensity increased continuously for V, Mn, Co, and Cu, with a rising sample uptake rate up to 200 µL/min. In the single-orifice SSN, the signal intensity only increased when the sample uptake rate was less than 90 µL/min.9 It is clear that dividing the sample solution into several streams for the multimicrospray improved the nebulization efficiency at a high sample uptake rate. One of the main goals in developing the MMSN was to obtain a higher sensitivity than that of a CCN. The sensitivities of the MMSN and the CCN are compared in Figure 7. For all the elements we considered, higher sensitivity was obtained by using the MMSN. Most elements had an improvement factor of about 2. This confirmed the advantage of the multimicrospray. It is worth noting here that the improvement factor for arsenic was as high as 5. This result suggests that the MMSN retains the advantages of microliter sample introduction and helps to maintain a high plasma temperature, which favors arsenic ionization, by producing a very fine aerosol even at a high sample uptake rate. The signal intensity of arsenic, however, decreased as the sample uptake rate increased when the uptake rate was higher 2466 Analytical Chemistry, Vol. 72, No. 11, June 1, 2000
Figure 8. Comparison of signal intensities of the MMSN and the SSN at various sample uptake rates for an easily ionized element (Co) and a difficultly ionized element (As).
than 80 µL/min, (Figures 6 and 7). The behavior of arsenic can be attributed to the cooling effect on the MIP plasma source from the sample aerosol, as discussed in our previous work.9 Since the amounts of solvent introduced into the plasma by using the MMSN and the SSN could be different, we compared the two nebulizers at various sample rates. Arsenic and cobalt were used for this experiment to represent elements that are sensitive and insensitive, respectively, to a change in the plasma temperature. Arsenic has a high ionization potential and requires a high plasma temperature for efficient ionization. Cobalt is easily ionized and has nearly 100% ionization efficiency. Therefore, any change in the signal intensity of cobalt reflects only a change in nebulization efficiency. As shown in Figure 8, the total amount of the analyte transported to the plasma by the MMSN is about double that transported by the SSN. Also, a continuous increase in the signal intensity of cobalt was obtained. However, in the case of arsenic, the contribution from the ionization efficiency, rather than the nebulization efficiency, dominates the signal change. Therefore, the signal is sensitive to a change in the plasma temperature. When the solvent is overloaded, the plasma temperature will drop, leading to a decrease in the signal intensity. The signal intensity for the SSN did not decrease until the sample uptake rate rose above 120 µL/min, while that for the MMSN began to decrease even at 80 µL/min (Figure 8). This result suggests that more of
the aerosol was brought to the plasma by the MMSN. In other words, the MMSN is more efficient than the SSN. Generally, the sample uptake rate can be set at 200 µL/min to achieve maximum sensitivity for most elements when the MMSN is used. However, to determine As, Se, or other elements with a very high ionization potential, a lower sample uptake rate can be used. CONCLUSION On the basis of results showing that a low sample uptake rate allows high nebulization, we developed a multimicrospray nebulizer (MMSN) to increase analyte nebulization efficiency. In this work, an MMSN with nebulization from three orifices was investigated. For all elements, the sensitivity of MIP-MS using the MMSN was higher than that of MIP-MS using the conventional concentric nebulizer (CCN) or the sonic spray nebulizer (SSN) with a single orifice. The sensitivities for most elements
were improved by a factor of 2 compared with those of a conventional nebulizer; for arsenic, the improvement factor was 5. The MMSN is flexible enough to allow sample introduction at microliter per minute rates while providing increased sensitivity at higher sample uptake rates. ACKNOWLEDGMENT This work was partly supported by CREST (Core Research for Evolutional Science and Technology) of the Japan Science and Technology Corp. We wish to thank Mr. Hisao Kojima (Central Research Laboratory, Hitachi, Ltd.) and Ms. Kazuko Yamamoto (Hitachi Science Systems, Ltd.) for their experimental help.
Received for review November 30, 1999. Accepted March 3, 2000. AC991379E
Analytical Chemistry, Vol. 72, No. 11, June 1, 2000
2467