Development of Multi-ESI-Sprayer, Multi-Atmospheric-Pressure-inlet

Nov 19, 1999 - The atmospheric pressure sampling nozzle (orifice, heated capillary, or inlet) of a high mass accuracy time-of-flight mass spectrometer...
0 downloads 0 Views 103KB Size
Anal. Chem. 2000, 72, 20-24

Articles

Development of Multi-ESI-Sprayer, Multi-Atmospheric-Pressure-inlet Mass Spectrometry and Its Application to Accurate Mass Measurement Using Time-of-Flight Mass Spectrometry Longfei Jiang and Mehdi Moini*

Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, Texas 78712

The atmospheric pressure sampling nozzle (orifice, heated capillary, or inlet) of a high mass accuracy time-of-flight mass spectrometer (TOF-MS) was modified by replacing its single nozzle with multiple atmospheric pressure nozzles. This allowed multiple streams of liquids to be introduced into the MS in parallel (an electrosprayer for each nozzle), with minimum analyte interactions between the streams. The chemical contents of all liquid streams were analyzed concurrently using a single mass spectrometer. To obtain a higher mass accuracy by providing internal reference on each scan (acquisition) and to evaluate the suitability of TOF-MS for molecular-formula confirmation, a dual-ESI-sprayer, dual-nozzle version of this design was used. The accurate masses of tens of organic compounds in the mass range of 200-3000 Da were measured, and the results were compared with those obtained using dual-sprayer, single-nozzle TOF-MS. A significant improvement in mass accuracy was observed when the former technique was used. Comparison between the mass accuracy using dual-ESI-sprayer, dualnozzle TOF-MS and that obtained using a double-focusing mass spectrometer operating under chemical ionization (CI) and fast atom bombardment (FAB) shows the suitability of the technique for elemental-composition confirmation. Approximately 85% of samples analyzed had mass errors of less than 5 ppm, and the other 15% had mass errors less than 8 ppm. Using a high-performance liquid chromatography (HPLC) as a device for introduction of one liquid stream (sample) and a syringe pump as a device for introduction of the second liquid stream (reference standard), the accurate mass of a tryptic digest of cytochrome c was measured. The range of mass errors was from -6.1 ppm to +3.6 ppm, a significant improvement over our previously reported mass accuracy for this digest using single-nozzle TOF-MS. The interactions between analytes in the liquid streams also were investigated using a variety of sample-introduction and nozzle-design 20 Analytical Chemistry, Vol. 72, No. 1, January 1, 2000

combinations, including single-ESI-sprayer, single-nozzle; dual-ESI-sprayer, single-nozzle; dual-ESI-sprayer, Yshaped inlet; and dual-ESI-sprayer, dual-inlet. The results demonstrated that the dual-ESI-sprayer, dual-inlet design provides reference peaks on every acquisition with minimum analyte-reference interaction and, therefore, higher consistent mass accuracy. With the advent of electrospray ionization (ESI),1,2 atmosphericpressure ionization techniques have become the ionization methods of choice for chemical analysis using mass spectrometry. Charged droplets formed in the ionization source are desolvated and transported into the mass-spectrometer sampling orifice through a heated-capillary inlet or a sampling cone/skimmer combination, hereafter called a “nozzle”. To date, most commercial atmospheric-pressure mass spectrometers use a single electrospray needle (hereafter called ESI sprayer) in conjunction with a single nozzle. However, multiple ESI sprayers have been used to enhance nebulization.3,4 In addition, a dual ESI sprayer in combination with a Y-shaped capillary interface/flow reactor has been used to investigate the gas-phase structure of electrosprayed proteins using ion-ion or ion-molecule reactions.5 Recent advances in TOF-MS have made it possible to obtain medium resolution (5000-7000 fwhm) and high mass accuracy (∼10 ppm) over a wide m/z range (100-3000) on a short time scale (one second or less).6 This, and the low cost of TOF-MS compared with other high-mass accuracy, high-resolution mass * Corresponding author. Tel: (512) 471-7344. Fax: (512) 471-1420. E-mail: [email protected]. (1) Dole, M.; Mach, L. L.; Hines, R. L.; Mobley, R. C.; Ferguson, L. P.; Alice, M. B. J. Chem. Phys. 1968, 49, 2240. (2) Whitehouse, C. M.; Dreyer, R. N.; Yamashita, M.; Fenn, J. B. Anal. Chem. 1985, 57, 675. (3) Kostiainen, R.; Bruins, A. Rapid Comm. Mass Spectrom. 1994, 8, 549-558. (4) Shia, J.; Wang, C. H. J. Mass Spectrom. 1997, 32, 247-250. (5) Ogorzalek Loo, R. R.; Smith, R. D. J. Am. Soc. Mass Spectrom. 1994, 5, 207-220, and references therein. (6) Cole, R. B., Ed. Electrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation & Applications; John Wiley & Sons: New York, 1997; Chapter 6. 10.1021/ac990777e CCC: $19.00

© 2000 American Chemical Society Published on Web 11/19/1999

spectrometers such as double-focusing and Fourier transform ion cyclotron resonance (FTICR), make TOF-MS ideal for analysis of complex biological mixtures.7-13 However, to use TOF-MS for reporting elemental composition, it is desirable to reduce the mass error of the TOF-MS to less than 5 ppm.14 Higher mass accuracy can usually be achieved when (internal) reference peaks are present on each scan (hereafter called “acquisition”, since TOFMS was used). Under electrospray ionization, this is usually achieved by mixing the sample and reference standard prior to mass analysis.13,15 Analyte mixing, however, results in suppression, discrimination, and/or adduct formation of one analyte by another analyte in the solution and/or during the ionization process. To prevent mixing of the analyte and reference compound prior to spray ionization, they are sprayed through separate (dual) ESI sprayers.16,17 However, since the solutions of the reference compound and the analyte are both sprayed at one nozzle, the sprayers’ mists (charged droplets) are mixed before they enter the first stage of the pumping and, therefore, result in undesirable interactions. To avoid the interaction of analytes under multi-ESIsprayer conditions, recently a multi-ESI-sprayer with a rotating sampling orifice was introduced.18 In this design, eight liquid streams are electrosprayed simultaneously at a rotating sampling orifice, but only one sprayer at a time has access to the inlet, and so analytes enter the mass spectrometer sequentially. The disadvantage of this design is that the reference and analyte are not recorded on the same or even adjacent acquisitions. To analyze multiple streams of liquid from multiple sources with minimum interaction between them, and to provide internal reference peaks on every acquisition, we have designed and constructed a mass spectrometer vacuum interface with multiple atmospheric pressure nozzles (one nozzle for each ESI sprayer). The focus of this article is on the dual-ESI-sprayer, dual-nozzle version of this design, in conjunction with TOF-MS, and its application to accurate mass measurement. Only a brief discussion regarding the construction of a quadro-ESI-sprayer, quadro-nozzle design has been provided. EXPERIMENTAL SECTION All electrospray ionization experiments were carried out using the Mariner time-of-flight mass spectrometer (PerSeptive Biosys(7) Lazar, I. M.; Xin, B.; Lee, M. L.; Lee, E. D.; Rockwood, A. L.; Fabbi, J. C.; Lee, H. G. Anal. Chem. 1997, 69, 3205-3211. (8) Takach, E. J.; Peltier, J.; Gabeler, S.; Verentchikov, A.; Vestal, M. L.; Martin, S. A. Rapid, High Resolution, Accurate Mass Characterization of Small Molecules and Biological Polymers by ESI-TOF Mass Spectrometry. Presented at the Association of Biomolecular Resource Facilities International Symposium, February 1997; ABRF, 1997. (9) Banks, J. F.; Dresch, T. Anal. Chem. 1996, 68, 1480. (10) Wu, J. T.; Qian, M. M. G.; Li, M. M. X.; Liu, L.; Lubman, D. M. Anal. Chem. 1996, 68, 3388. (11) Cao, P.; Moini, M. J. Am. Soc. Mass Spectrom. 1998, 9, 1081-1088. (12) Moini, M.; Cao, P.; Bard, A. J. Anal. Chem. 1999, 71, 1658-1661. (13) Cao, P.; Moini, M. Rapid Commun. Mass Spectrom. 1998, 12, 864-870. (14) Gross, M. L. J. Am. Soc. Mass Spectrom. 1994, 5, 57. (15) Palmer, M. E.; Clench, M. R.; Tetler, L. W.; Little, D. R. Rapid Commun. Mass Spectrom. 1999, 13, 256-263. (16) Andrien, B. A.; Whitehouse, C.; Sansone, M. A. Proceedings of the 46th ASMS Conference on Mass Spectrometry and Allied Topics, May 31-June 4, 1998, Orlando, FL; pp 889-890. (17) Dresch, T.; Keefe, T.; Park, M. Proceedings of the 47th ASMS Conference on Mass Spectrometry and Allied Topics, June 13-18, 1999, Dallas, TX; pp 18651866. (18) Bateman, R.; Jarvis, S.; Giles, K.; Organ, A.; de Biasi, V.; Haskins, N. Proceedings of the 47th ASMS Conference on Mass Spectrometry and Allied Topics, June 13-18, 1999, Dallas, TX; pp 2216-2217.

tems, Framingham, MA). The mass spectrometer was operated in the mass-to-charge ratio (m/z) range of 200-3000 (unless otherwise mentioned) at a rate of 10 000 acquisitions/second. In this study, 10 000 single-shot spectra were averaged to generate one averaged spectrum per second. The original pumping system of the Mariner includes a mechanical pump (Varian Vacuum Products, Lexington, MA) and a turbomolecular-drag pump (Pfeiffer Vacuum Technology, Inc., Nashua, NH) with respective pumping capacities of 7.5 L/s and 210 L/s. The mechanical pump evacuates the first stage of the vacuum housing (between the nozzle and the first skimmer) and also pumps the output of the turbo pump. The single nozzle (0.4-mm i.d., 19.8-mm-long) of the MS was replaced by an in-house-fabricated dual nozzle (each 0.33mm i.d., 28.5-mm-long). The two nozzles point to the instrument’s first skimmer. Using the dual-nozzle design, the analyzer’s pressure was increased from 1.4 × 10-6 to 2.44 × 10-6 Torr. However, the increased pressure had no adverse effect on the resolution of the instrument within the mass range studied. Therefore, no additional pump was used. However, when the quadro-nozzle (four sampling inlets) was installed, to maintain the analyzer pressure below 1 × 10-5 Torr, an additional mechanical pump was added to the original pumping system. The vacuum hose connecting the first stage of pumping to the original mechanical pump was disconnected and connected to the second mechanical pump (Edwards High Vacuum International, Crawley, Sussex, England) with a pumping capacity of 4.7 L/s. Under these conditions, the analyzer pressure was 6 × 10-6 Torr. Figure 1 depicts the quadro-ESI-sprayer, quadro-nozzle design. The dualsprayer, dual-nozzle version is identical with the exception that two nozzles are plugged and two ESI sprayers are absent. For all the dual-nozzle experiments, the ESI sprayers were at the same voltage (4 kV) using the instrument’s power supply. The nozzle temperature was maintained at 150 °C. For all experiments, the instrument was tuned to a resolution of 5000-6000 (fwhm) using the original nozzle. The instrument syringe pumps or HPLC injectors (Rheodyne, Rohnert Park, CA) were used for sample infusion or loop injections, respectively. The reference compound and sample were loop-injected into the stream of methanol + water solution (50: 50, v/v) using a loop volume of 20 µL and 2 µL, respectively. The large loop volume of the reference was used to ensure the presence of reference standard peaks when the sample ions reached the MS. The flow rate of the mobile phase (water + methanol, 50:50, v/v) for both the reference compound and samples was 2 µL/min. Nitrogen, with a flow rate of 0.5 L/min, was used as the nebulizing gas. A 5 ppm solution of sodium trifluoroacetate (NaTFA) in a water + methanol solution (50:50, v/v) was prepared according to our previously reported procedure19 and used as the reference compound. Under positive-ion electrospray ionization conditions using the TOF-MS, the solution of NaTFA provided reference peaks [(NaTFA)nNa+, n ) 1-21], covering the entire mass range studied. To examine the effect of selecting different pairs of reference peaks on the mass accuracy of the unknown, several different pairs were used to bracket the peak of the unknown. It was found that the closer the bracketing reference peaks were to (19) Moini, M.; Jones, B. L.; Rogers, R. M.; Jiang L. J. Am. Soc. Mass Spectrom. 1998, 9, 977-980.

Analytical Chemistry, Vol. 72, No. 1, January 1, 2000

21

Figure 1. Schematic of multi-ESI-sprayer, multinozzle time-of-flight mass spectrometry. Table 1. Comparison of Mass Accuracies Obtained Using Different Sprayer/Nozzle Combinations measurement error (ppm) molecular formula

calculated mass

method 1

method 2

method 3

method 4

method 5b

(1) C24H24O8S + Na (2) C13H52N2O6SiS + Na (3) C34H41PO10 + Na (4) C53H53NO10 + Na

495.1090 631.3213 663.2335 886.3567

7.5 6.3 (6.6)a 1.4 2.7

6.5 7.1 (7.8) 4.5 3.4

9.9 -4.7 (2.2) 7.7 5.4

-6.7 -1.2 (6.4) 3.3 -1.4

-4.4 0.9 (2.0) 3.8 2.9

a The values in parentheses are the standard deviations of measured mass error for three replicate measurements. b Sodium adducts were the base peaks in methods 1-4, while the protonated molecule of each compound was the base peak in method 5. The base peaks were used for accurate mass measurements.

the unknown, the lower was the measured mass error. For example, the standard deviation in mass error for seven samples in the mass range of 400-1000 Da was 3.3 ppm for the closest pair of bracketing peaks and 7.0 ppm for the second closest set of bracketing peaks. Therefore, NaTFA was especially advantageous, since its peaks were separated by only 136 Da. As such, for all experiments explained above, the closest bracketing reference peaks were used to obtain the accurate masses. Cytochrome c (24 pmol/µL) was digested according to the previously reported procedure.20 For HPLC/MS, 5 µL of the digest solution was analyzed. The HPLC (Magic 2002, Michrom BioResources, Inc., Auburn, CA) mobile-phase flow rate was 50 µL /min, and the gradient was from 95% A (0.1% TFA-1% acetonitrile) to 65% B (0.1% TFA-90% acetonitrile) in 15 min. One ESI sprayer was connected to the HPLC column’s outlet, while the other ESI sprayer was connected to a syringe pump, delivering the 5 ppm solution of NaTFA at a flow rate of 2 µL/min. All chemicals were purchased from SIGMA (P.O. Box 14508, St. Louis, MO), except those provided by various professors in the Department of Chemistry and Biochemistry of the University of Texas at Austin, and were used as supplied. RESULTS AND DISCUSSION Interaction of Analytes and Reference Compound Under Dual-ESI-Sprayer Conditions. To study the interactions of (20) Takada, Y.; Nakayama, K.; Yoshida, M.; Sakairi, M. Rapid Commun. in Mass Spectrom. 1994, 8, 695-697.

22 Analytical Chemistry, Vol. 72, No. 1, January 1, 2000

analytes in two separate liquid streams under different dual-ESIsprayer configurations, three experimental designs were investigated: dual-sprayer, mono-nozzle; dual-sprayer, Y-channel; and dual-sprayer, dual-nozzle. For this experiment, the sample and the reference compound were each loop-injected into a separate liquid stream leading to a separate ESI sprayer. The idea was to study the degree of sodiation of a sample with the reference compound using the three experimental setups. Since the degree of sodium adduct formation is compound-dependent, we chose a compound with chemical composition of C32H43NO10 (m/z 601) which could easily be sodiated. The results are summarized in Figure 2. Under the dual-ESI-sprayer, mono-nozzle condition, the molecules were completely sodiated (Figure 2a). Under dual-ESI-sprayer, Ychannel condition, almost 75% of the molecules were sodiated (Figure 2b). Under dual-ESI-sprayer, dual-nozzle, however, only 10% of the molecules were sodiated (Figure 2c). The results clearly indicate that, compared with the first two designs, the dual-ESIsprayer, dual-nozzle design dramatically reduces the interaction between the sample and reference compounds. The main reasons for reduced interaction under the latter technique were: (1) analytes of the two streams were separated until the first stage of pumping (see Figure 1), where they were already desolvated; and, (2) because of the low pressure of this region, the probabilities of ion/ion or ion/molecule interactions were low. Mass Accuracy Comparison of Different Sprayer/Nozzle Combinations. Table 1 shows the comparison between mass

Table 2. Comparison of Mass Accuracies Obtained Using Method 1 (Single-ESI-Sprayer, single nozzle) and Method 5a measurement error (ppm) sample no.

molecular formula

calculated mass

singleb

dualc

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

C14H17N3O2 + H C14H17O2Br + H C15H24O7S + H C19H28O5Si + H C21H28O7 + H C26H26NO7P + H C32H30N2O6 + H C28H35N4I + H C31H50N2O7 + H C31H26N4O6S + H C31H50N2O6SiS + H C32H52N2O6SiS + H C31H52N2O4116Sn + H C18H23N6O14P3 + H C44H51N3O8 + H C63H92N18O10 C92H91N13O31 C115H109N17O39 C136H133N19O45

260.1399 297.0490 349.1321 365.1784 393.1913 496.1525 539.2182 555.1985 563.3696 583.1651 607.3237 621.3393 633.3023 641.0563 750.3754 1260.7244 1873.5944 2351.7069 2751.8703

-8.8 +9.8 -8.0 +2.7 -6.9 +6.9 -8.5 +2.3 +5.3 -5.7 -7.2 +2.7 +2.4 -8.0 +4.8 6.3d -3.5d +2.7d +2.8d

-5.4 +0.3 +0.9 +2.7 +2.8 +7.7 0.0 0.0 +2.7 +3.8 +4.0 +0.2 -2.7 0.0 +1.9 -1.8e +1.3e +0.4e -0.6e

a The sample was infused into a 5 ppm NaTFA water-methanol buffer by injection. b b ) (M+Na)+. c c ) (M+H)+. d d ) (M+2Na)+2. e e ) (M+2H)+2.

Figure 2. Interaction of an analyte with a reference compound under three different experimental setups: dual-sprayer, mono-nozzle (A); dual-sprayer, Y-channel (B); and dual-sprayer, dual-nozzle (C). 602 (m/z) is the protonated analyte with the chemical composition of C32H43NO10 , 624 (m/z) is the sodium adduct of the analyte, and 567 and 703 (m/z) are the peaks of the reference compound in the mass range shown.

accuracies obtained using four commonly used sample introduction techniques and the dual-ESI-sprayer, dual-nozzle method: 1) samples were loop injected into the stream of the NaTFA solution; 2) samples were mixed with the NaTFA solution, and then loop injected into the stream of the water + methanol solution (50/ 50, v/v); 3) samples and 5 ppm NaTFA were sequentially loop injected into the stream of the water + methanol solution; for these three experiments, a single ESI sprayer was used; 4) samples and the NaTFA solution were loop injected into two separate liquid streams, going into two separate ESI sprayers, each pointed at the MS’s original (mono) nozzle; and 5) samples and the NaTFA solution were loop injected into two separate ESI sprayers pointing to two separate nozzles. In addition, the standard deviation of measured mass error for three replicates of a compound with chemical composition of C13H52N2O6SiS (m/z 608.3316) was calculated and tabulated (compound 2 of Table 1). As is shown (Table 1), the dual-ESI-sprayer, dual-nozzle combination (method 5), gave the lowest errors and the most consistent results. This was because the dual-ESI-sprayer, dualnozzle combination provides references peaks on each acquisition (true internal calibration), with minimum interactions between analyte and reference compound. The second best mass accuracy

was provided by those techniques for which the reference compound and sample were mixed before (methods 1, 2) or during (method 4) the ionization process. This was because, while the internal reference peaks were present on each acquisition, strong interactions between the sample and the reference compound were observed (for example, the sodiated molecules were usually the most abundant peaks of the samples). This interaction may result in formation of complexes that can interfere with the peak of interest. The resolution of TOF-MS (5000-6000 fwhm) is not usually high enough to resolve these interferences which can lead to lower mass accuracies. Method 3 gave the highest error, since the sample and the reference compounds were introduced sequentially, and therefore, to have the reference peaks present on the spectrum of the sample, the chromatographic apex of the analyte had to be averaged with that of the reference. To evaluate the suitability of dual-ESI-sprayer, dual-nozzle TOFMS for elemental composition confirmation, the mass accuracies of 64 samples were measured using this technique. The mass range of the samples studied spanned 200 - 3000 Da. The range of error for TOF-MS was -5.8 to +7.2 ppm, where 85% of the analyses gave mass error less than 5 ppm. The results demonstrate the suitability of the dual-ESI-sprayer, dual-nozzle TOF-MS for reporting chemical composition.14 Table 2 provides a pool of data to compare method 1 and method 5 for samples in the 200-3000 Da mass range. As is shown, a significant enhancement in mass accuracy was achieved by using the dual-ESI-sprayer, dual-nozzle design. Comparison between the Sensitivity of Detection of the Dual Nozzle vs the Single Nozzle. To compare the sensitivity of detection between the original single nozzle and our dual nozzle, a 5 ppm solution of neurotensin (MW ) 1671) in water + acetonitrile (50:50, v/v) was loop-injected into the stream of the mobile phase using the 2-µL-volume loop. In each experiment, the parameters were optimized for maximum sensitivity. The triply Analytical Chemistry, Vol. 72, No. 1, January 1, 2000

23

Table 3. HPLC/MS of the Tryptic Digest of Cytochrome c Using the Dual-ESI-Sprayer, Dual-Nozzle Designa peak no.

a.a. sequence

calculated m/z

measured m/z

error (ppm)

1 2 3 4 5 6 7 8 9 10

CAQCHTVEK EETLMEYLENPK TGQAPGFTYTDANK TGPNLHGLFGR EDLIAYLK MIFAGIK YIPGTK IFVQK GITWK Acetyl-GDVEK

817.3213 (+2) 748.3535 (+2) 735.8472 (+2) 584.8153 (+2) 964.5355 779.4489 678.3827 634.3850 604.3458 589.2833

817.3216 748.3508 735.8465 584.8133 964.5368 779.4516 678.3838 634.3870 604.3466 589.2869

-0.4 +3.6 +1.0 +3.4 -1.4 -3.5 -1.5 -3.2 -1.3 -6.1

a A 5-µL tryptic digest of horse cytochrome c was injected into an HPLC column while 2 ppm NaTFA in 50:50 water + methanol (v/v) was continuously infused into the other ESI needle.

protonated peak of neurotensin (m/z 558) was used to compare the sensitivities. The average of three consecutive experiments showed that the peak intensity using the original (single) nozzle design was 386 with S/N of 179 compared with 314 with S/N of 150 for one nozzle of the dual-nozzle design. The higher peak intensity of the single-nozzle design is attributed to its larger opening, its shorter length, and the fact that the nozzle is coaxial to the skimmer compared with the off-axis design of the dual nozzle. Application of Dual-ESI-Sprayer, Dual-Nozzle TOF-MS to Accurate Mass Measurement of a Protein Digest. To examine the mass accuracy of the dual-ESI-sprayer, dual-nozzle design under HPLC/MS conditions, a tryptic digest of cytochrome c was analyzed. In this experiment the outlet of the HPLC column was fed into one ESI sprayer while a constant stream of 5 ppm NaTFA was infused into the second ESI sprayer. Each sprayer pointed toward a separate nozzle. The mass spectrometer was operated in the m/z range of 400-1500. In this manner, each spectral acquisition of the digest contained several peaks of the reference compound, providing a true noninterfering internal reference standard. By using the two closest, bracketing peaks of NaTFA, the accurate mass of each peptide was measured (Table 3). The results were superior to our previously reported data13 using a

24 Analytical Chemistry, Vol. 72, No. 1, January 1, 2000

single-ESI-sprayer, single-nozzle TOF-MS. In our previously reported experiment, to obtain the accurate mass of each peptide, its electrophoretic peak was averaged with peaks of two reference compounds that had been added to the protein digest prior to separation. That technique provided an average mass error of 8 ppm, and five out of 10 peaks had errors greater than 10 ppm. In the present technique, the average mass error was only 2.5 ppm, and only one out of 10 peaks had mass errors of greater than 5 ppm (-6.1 ppm). CONCLUSION The main advantage of the new design is parallel introduction of multiple liquid streams into a mass spectrometer through multiple atmospheric pressure nozzles (one for each ESI sprayer). A dual-ESI-sprayer, dual-nozzle version of this design is especially useful for accurate mass measurement of organic compounds using TOF-MS. It provides reference peaks on every acquisition without significant interaction between the reference standard and the analyte of interest. The mass accuracy obtained under dualESI-sprayer, dual-nozzle conditions is good enough to be used for chemical-composition confirmation. The four-ESI-sprayer, fournozzle version of the device is under investigation for simultaneous, or pulsed, introduction of two or three liquid samples, in addition to continuous introduction of the solution containing reference standard. Work is underway on a nine-ESI-sprayer, ninenozzle design to analyze eight streams of liquid and one reference. This design is especially useful for combinatorial chemistry and genome and proteomics projects where parallel sample analysis is important. Other advantages include: (1) decrease in sample analysis time due to parallel analysis of multiple samples, (2) significant reduction in cost of using a mass spectrometer for multiple analyses, and (3) present spectrometers can be upgraded with minimal modifications. ACKNOWLEDGMENT Presented in part at the 47th ASMS Conference on Mass Spectrometry and Allied Topics, Dallas, Texas, June 1999. Received for review July 15, 1999. Accepted October 4, 1999. AC990777E