Ion Proton Transfer

In this work, the mass/charge ratios of the various charge states of the parent compound were known and could be used to determine a correction for th...
0 downloads 0 Views 217KB Size
Download PDF
+

+

Anal. Chem. 1996, 68, 257-262

Product Ion Charge State Determination via Ion/Ion Proton Transfer Reactions William J. Herron, Douglas E. Goeringer, and Scott A. McLuckey*

Chemical and Analytical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6365

Proton transfer from protonated pyridine to product anions derived from quadrupole ion trap collisional activation of the triply charged anion of the oligonucleotide 5′-d(AAAA)-3′ and the 6- charge state of oxidized bovine insulin A-chain is shown to be a rapid and effective way to determine product ion charge states. The reactions are carried out in a quadrupole ion trap as part of a procedure involving three stages of mass analysis. It is demonstrated that the reactions can be driven at rates sufficiently high to convert 30-80% of the initial product anion population to second generation products in 50-200 ms. The use of ion/ion reactions enjoys significant advantages over the use of ion/molecule proton transfer chemistry. For example, ion/ion reactions are more universal than ion/molecule reactions due to their greater exothermicity, and ion/ion reactions allow for precise control over the timing of introduction and ejection of each reactant. Despite the high exothermicity of the reactions, no significant fragmentation of product ions derived from highmass multiply charged anions is observed. Multiply charged high-mass ions have become commonplace in biological mass spectrometry with the advent of electrospray,1-7 matrix-assisted laser desorption,8,9 and, most recently, massive cluster ion impact.10,11 The phenomenon of multiple charging has many implications in biological mass spectrometry. For example, it simplifies mass measurement from the point of view of the mass/charge range of the analyzer. However, it can complicate mass measurement as well, since some means for ascertaining ion charge must be available to enable ion mass determination, as most forms of mass spectrometry inherently yield information that is related to the mass-to-charge ratio rather than to mass directly. A set of simultaneous equations can be used to determine (1) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64-71. (2) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Mass Spectrom. Rev. 1990, 9, 37-70. (3) Smith, R. D.; Loo, J. A.; Edmonds, C. G.; Barinaga, C. J.; Udseth, H. R. Anal. Chem. 1990, 62, 882-899. (4) Smith, R. D.; Loo, J. A.; Ogorzalek Loo, R. R.; Busman, M.; Udseth, H. R. Mass Spectrom. Rev. 1991, 10, 359-451. (5) Huang, E. C.; Wachs, T.; Conboy, J. J.; Henion, J. D. Anal. Chem. 1990, 62, 713A-725A. (6) Mann, M.; Meng, C. K.; Fenn, J. B. Anal. Chem. 1989, 61, 1702-1708. (7) Covey, T. R.; Bonner, R. F.; Shushan, B. I.; Henion, J. D. Rapid Commun. Mass Spectrom. 1988, 2, 249-256. (8) Karas, M.; Bahr, U.; Giessmann, U. Mass Spectrom. Rev. 1991, 10, 335357. (9) Karas, M.; Hillenkamp, F. Anal. Chem. 1988, 60, 2299-2301. (10) Mahoney, J. F.; Perel, J.; Ruatta, S. A.; Martino, P. A.; Husain, S.; Lee, T. D. Rapid Commun. Mass Spectrom. 1991, 5, 441-445. (11) Mahoney, J. F.; Perel, J.; Lee, T. D.; Martino, P. A.; Williams, P. J. Am. Soc. Mass Spectrom. 1992, 3, 311-317.

charge states when ions of multiple charge states are derived from the same molecule,6 as is frequently the case in electrospray mass spectrometry. However, for cases in which multiple charge states derived from the same molecule are either not present or not apparent, as might be the case with complex mixtures, some other approach for determining charge state may be either desirable or necessary. Tandem mass spectrometry of multiply charged ions constitutes a particularly important case in which ion charge state determination can be problematic. For cases in which both product ions from a single dissociation reaction appear in the product ion spectrum, the charges of the products can be assigned on the basis of the ratio of m/z spacings between the product ion and the parent ion. However, when this is not the case, due either to instrumental discrimination or to further fragmentation of one or both of the initially formed product ions, some other means must be used to determine ion charge. A number of approaches to ion charge state determination have been used in cases in which multiple charge states are not apparent or when complementary ions are absent in tandem mass spectra. For example, as first pointed out and demonstrated by McLafferty and Henry,12 measurement of the spacings in massto-charge between adjacent peaks in the isotopic distribution can lead to a direct indication of ion charge state. The resolving power of the analyzer determines the upper limit to mass and charge of the ions amenable to this approach. Thus, Fourier transform ion cyclotron resonance (FTICR), due to its unparalleled resolving power, is particularly well-suited in this regard. Much lower resolving powers are needed when adduct ions are present, such as species with alkali ions in addition to or instead of protons, due to the much greater m/z spacings. Neubauer and Anderegg, for example, described the intentional addition of sodium salts to aid in charge state determination of peptides in liquid chromatography/electrospray ionization mass spectrometry.13 We described the use of ion/molecule reactions involving either clustering or proton transfer for charge state determination of product ions formed from collisional activation in the quadrupole ion trap.14 Hunter et al. demonstrated the use of proton transfer reactions in a triple quadrupole tandem mass spectrometer for the same purpose.15 Senko et al. described the use of the spacing of copper ion adducts present in both the parent and product ions for the facilitation of mass measurement in tandem mass spectrometry in an FTICR spectrometer.16 Bruce et al. described the charge state determination of an individual high-mass multiply (12) Henry, K. D.; McLafferty, F. W. Org. Mass Spectrom. 1990, 25, 490-492. (13) Neubauer, G.; Anderegg, R. J. Anal. Chem. 1994, 66, 1056-1061. (14) McLuckey, S. A.; Glish, G. L.; Van Berkel, G. J. Anal. Chem. 1991, 63, 1971-1978. (15) Hunter, A. P.; Severs, J. C.; Harris, F. M.; Games, D. E. Rapid Commun. Mass Spectrom. 1994, 8, 417-422.

+

charged ion on the basis of proton transfer reactions with background neutrals in an FTICR spectrometer.17 All of the approaches just mentioned rely on the measurement of the massto-charge spacings between two or more ions related in some fashion to the species of interest. That is, multiple charges of known spacing (usually assumed to be (1) and/or multiple masses of known difference must be present to assign charge. An approach recently described by Chen et al.18 is unique is this regard in that the response of a well-characterized detection circuit in FTICR has been shown to be effective in determining directly the charge of a single, multiply charged ion. This approach does not require the measurement of at least two mass-to-charge ratios, as do all of the others mentioned here. It has been shown that the ion/molecule reaction approach involving either clustering or proton transfer is conveniently employed in ion trapping instruments and requires significantly lower analyzer resolving power than is necessary for resolving 13C isotopes. However, we noted several important drawbacks to the use of ion/molecule chemistry for product ion charge state determination in our previous work.14 First, the reactions are not universal. We found that some product ions from multiply protonated parents underwent proton transfer with a given base, whereas others did not. In some cases, the base tended to cluster with the ion rather than to deprotonate it. As has been observed by a number of workers,19-22 for a given biopolymer, proton transfer reaction rates in the dilute gas phase decrease with charge state, and clustering tends to be observed for charge states that react either slowly or not at all by proton transfer. Although this appears to be a general trend, the current understanding of the relationship between charge state, ion composition and structure, and base strength is insufficiently detailed to make an a priori prediction as to which product ion/neutral combinations can be expected to proton transfer, cluster, or not react. Ion/molecule reactions for product ion charge state determination cannot, therefore, be regarded as a universal approach. A second significant practical limitation to the use of ion/molecule chemistry for product ion charge state determination is that the neutral reactant is usually always present in the vacuum system and can therefore react at inconvenient times, such as during ion accumulation. If the reactant is strong enough to neutralize product ions, it will likely be strong enough to neutralize the higher parent ion charge states. Therefore, the reactant can deplete parent ions as they are being accumulated, which can compromise significantly the ability to determine product ion charge states from highly charged parent ions. A pulsed valve might be considered as a means for restricting the presence of the neutral reactant to the defined reaction period. However, the time frames involved (16) Senko, M. W.; Beu, S. C.; McLafferty, F. W. J. Am. Soc. Mass Spectrom. 1993, 4, 828-830. (17) Bruce, J. E.; Cheng, X.; Bakhtiar, R.; Wu, Q.; Hofstadler, S. A.; Anderson, G. A.; Smith, R. D. J. Am. Chem. Soc. 1994, 116, 7839-7847. (18) Chen, R.; Wu, Q.; Mitchell, D. W.; Hofstadler, S. A.; Rockwood, A. L.; Smith, R. D. Anal. Chem. 1994, 66, 3964-3969. (19) McLuckey, S. A.; Van Berkel, G. J.; Glish, G. L. J. Am. Chem. Soc. 1990, 112, 5668-5670. (20) Cassady, C. J.; Wronka, J.; Kruppa, G. H.; Laukien, F. H. Rapid Commun. Mass Spectrom. 1994, 8, 394-400. (21) Schnier, P. D.; Gross, D. S.; Williams, E. R. J. Am. Chem. Soc. 1995, 117, 6747-6757. (22) McLuckey, S. A.; Van Berkel, G. J.; Glish, G. L.; Schwartz, J. C. In Modern Mass Spectrometry: Practical Aspects of Ion Trap Mass Spectrometry, Vol. 2: Fundamentals and Instrumentation; March, R. E., Todd, J. F. J., Eds.; CRC Press: Boca Raton, FL, 1995; Chapter 3, pp 89-141.

+

in pulsing in and pumping out the relatively polar molecules required for reaction compromise the utility of this option. The disadvantages just mentioned can be overcome by using a universal reactant that can be exposed to the ion of interest at precisely defined times. We suggested in the original work14 that an oppositely charged ion might serve as a more universal reactant and certainly as one that can be more readily manipulated in the quadrupole ion trap than a neutral reactant. We have recently reported on the study of ion/ion proton transfer23 and electron transfer24 reactions involving oppositely charged ions in the quadrupole ion trap. We report here on the application of proton transfer for the determination of product ion charge states using ion/ion reactions. We illustrate the approach with multiply charged parent anions and protonated pyridine as reactants. The process is first illustrated with a well-understood parent ion, in which product ion charge states are already known, and further illustrated with a parent ion in which the product ion charge states cannot be determined either by the presence of complementary ions or by the isotopic spacings using a resolving power of roughly 1000. EXPERIMENTAL SECTION Samples and Apparatus. The oligonucleotide 5′-d(AAAA)3′ was obtained from Pharmacia (Milwaukee, WI) as the sodium salt. The oxidized A-chain of bovine insulin was obtained from Sigma (St. Louis, MO). Solutions were prepared by dissolving the sample in a drop of HPLC-grade water and diluting with HPLCgrade methanol to give a concentration of 10-20 µM in at least 9:1 methanol/water (vol/vol). All solutions were infused at a rate of 1.0 µL/min through a 120 µm i.d. needle held at a potential of -3000 to -3500 V. Multiply charged parent anions were formed by electrospray using a home-made source coupled with a Finnigan-MAT (San Jose, CA) ion trap mass spectrometer modified for injection of ions formed external to the ion trap. Details of the electrospray/ ion trap interface have been described.25 Electron ionization served as the primary source of cations and was effected by drilling a 3 mm diameter hole through the ring electrode and by mounting a heated filament near the ring electrode such that electrons could be gated into the trapping volume. The electron gate electrode was controlled by a variable length pulse generator (Stanford Research Systems, Model 535), actuated by a trigger pulse from the ion trap mass spectrometer electronics. This arrangement allowed for independent control of anion accumulation and cation formation. In all cases, the multiply charged parent anions were accumulated and isolated prior to cation formation. For all ion/ion proton transfer experiments, pyridine vapor was leaked into the vacuum chamber to a pressure of (1-3) × 10-6 Torr. At this pressure, the radical cation of pyridine is converted quantitatively to protonated pyridine via self-chemical ionization within 10 ms. This time frame is short relative to the reaction times of 50-200 ms required for significant formation of ion/ion reaction products. Ion Manipulation and Mass/Charge Analysis. Anions were injected axially into the ion trap for periods ranging from 0.3 to (23) Herron, W. J.; Goeringer, D. E.; McLuckey, S. A. J. Am. Soc. Mass Spectrom. 1995, 6 529-532. (24) Herron, W. J.; Goeringer, D. E.; McLuckey, S. A. J. Am. Chem. Soc., submitted for publication. (25) Van Berkel, G. J.; Glish, G. L.; McLuckey, S. A. Anal. Chem. 1990, 62, 1284-1295.

+

Figure 1. Product ion spectrum derived from ion trap collisional activation of the triply charged anion of 5′-d(AAAA)-3′. An asterisk and dashed line indicate the m/z location of the parent ion.

1.0 s. The radio frequency (rf) sine wave amplitude applied to the ring electrode during ion injection ranged from 700 to 1000 V0-p. In all cases, helium was admitted into the vacuum system to a total pressure of 1 mTorr, with a background pressure in the instrument of 2 × 10-5 Torr without the addition of helium. A single resonance ejection scan was used for isolation of parent ions. Low m/z ions were ejected by passing the ions through a qz value of 0.908 by scanning the amplitude of the ring electrode rf sine wave. High m/z ions were ejected by dipolar resonance ejection scan using a 12 Vp-p sine wave signal applied to the endcaps at a frequency selected to eject ions at an m/z value slightly greater than that of the parent ion. Parent ions were isolated prior to the 0.05-0.2 s ion/ion reaction period at less than unit resolution to avoid parent ion loss, due either to dissociation or to ejection from off-resonance power absorption. Mass/charge analysis was effected after the completion of all ion isolation and reaction periods using resonance ejection26 to yield a mass/charge range of 100-1300. The mass/charge scales for the ion/ion reaction spectra were calibrated using the electrospray mass spectra of the parent compounds. In this work, the mass/charge ratios of the various charge states of the parent compound were known and could be used to determine a correction for the mass scale provided by the ion trap data system. The mass accuracy associated with mass/charge assignments in the MS/MS and MSn experiments is on the order of 0.2% or better. The spectra shown here were the result of an average of 50-100 individual scans. RESULTS AND DISCUSSION The spectrum resulting from ion trap collisional activation of the mass-selected triply charged anion derived from the small oligonucleotide 5′-d(AAAA)-3′ using electrospray is shown in Figure 1. This ion has been studied extensively in our laboratory, and its decomposition products have been identified.27 Interpretation of tandem mass spectra derived from ion trap collisional activation of multiply charged oligonucleotide anions is facilitated (26) Kaiser, R. E., Jr.; Louris, J. N.; Amy, J. W.; Cooks, R. G. Rapid Commun. Mass Spectrom. 1989, 3, 225-229. (27) McLuckey, S. A.; Van Berkel, G. J.; Glish, G. L. J. Am. Soc. Mass Spectrom. 1992, 3, 60-70.

+

Figure 2. Product ion spectrum arising from the sequence (M 3H)3- f w32- f ??, in which the w32- product ion resulting from collision-induced dissociation is stored in the presence of a population of protonated pyridine ions for 200 ms.

by the fact that complementary ions are usually present.28 Product ion charge state determinations can therefore often be made without recourse to measurement of 13C spacings, ion molecule chemistry, or any of the other methods mentioned above for charge state assignment. For example, the adenine anion and the ion labeled as (-A)2-, the product ion resulting from the loss of the adenine anion, are implicated as complementary ions due to the 2:1 ratio in m/z distance of the respective ions from the m/z ratio of the triply charged parent ion. If it is recognized that the remaining product ion signal arises from further decomposition of the ion arising from loss of the adenine anion, the charge states of all but one of the remaining major product ions in the spectrum can be assigned on the basis of there being complementary pairs of ions present in the spectrum. For example, the equidistance of the ions labeled as w1- and (a3 - A3)- from the anion labeled as (-A)2- suggests that these ions constitute a complementary pair of singly charged anions. (The nomenclature and structures associated with these ions have been discussed.24,27,28) Likewise, the ions labeled as (a2 - A2)- and w2are implicated as a complementary pair of singly charged ions on this basis. There is no complementary ion apparent in the spectrum for the ion labeled as w32- that might arise from either the parent ion or the ion formed by loss of A-. The absence of a complementary ion alone might suggest that this ion is formed by the loss of a neutral from the doubly charged (-A)2- anion, but the evidence is certainly not conclusive in the absence of more information, such as the 13C spacings or an MS3 experiment, etc. Figure 2 shows the results of an MS3 experiment in which the parent ion (M - 3H)3- derived from 5′-d(AAAA)-3′ is subjected to collisional activation and the ion labeled as w32-, which is formed upon collisional activation, is isolated and stored simultaneously with a population of protonated pyridine cations for 200 ms. The appearance of the high-mass ion labeled as (w3 + H)- clearly shows that the ion is multiply charged and, assuming that a change in mass of 1 Da is associated with a change in charge of +1, indicates that the reactant anion is doubly charged, with a mass corresponding to the expected w3 fragment. (Signal at m/z 80 is also observed whenever a cation formation pulse is activated and arises from protonated pyridine, despite the fact that the (28) McLuckey, S. A.; Habibi-Goudarzi, S. J. Am. Chem. Soc. 1993, 115, 1208512095.

+

+

conversion dynode voltage is optimized for anion detection.) Two important points are illustrated in Figure 2. First, a very simple set of reaction products arises from the reaction, as indicated below:

w32- + C5H5NH+ f (w3 + H)- + C5H5N

As was observed in our original report on ion/ion reactions,14 ion/ ion proton transfer does not appear to give rise to observable fragmentation of the (w3 + H)- product ion, despite the large exothermicity expected for an ion/ion reaction. This is very important in the present context, because it must be assumed that changes in mass arise solely from transfer of a proton. If the product ion were to fragment extensively, serious ambiguity would be introduced into the measurement. Thus far, we have noted a minor degree of fragmentation associated with ion/ion proton transfer for several small doubly charged anions.29 However, we have not yet observed significant fragmentation from anions arising from biopolymers when they are subjected to ion/ ion proton transfer. Second, despite the relatively low ion number densities generally present in ion trapping instruments, the ion/ ion proton transfer reaction can proceed to an easily measurable extent in tens to a few hundred milliseconds, depending upon the number of cations present. The rapidity with which these reactions can be driven in the ion trap, in addition to the charge state information that they provide, indicates that they can be analytically useful. That is, the product of the cation number density and the reaction rate constant is sufficiently high (2-20 s-1) and the spatial overlap of the anion and cation populations sufficiently great to deplete 30-80% of the reactant anion population in 50-200 ms. This reaction efficiency is comparable to the highest efficiencies used for product ion charge state determination using ion/molecule reactions.14 The example just given illustrates the essential features of charge state determination using ion/ion proton transfer reactions. Product ion charge state determination following ion trap collisional activation, however, is not a particularly pressing problem for small multiply charged oligonucleotides, due to the fact that complementary product ions are generally observed. For multiply charged peptides and proteins, on the other hand, it has often been noted that complementary ion pairs are frequently not observed,3 possibly due to further fragmentation of one of the initially formed products. The use of ion/ion proton transfer reactions is therefore illustrated further for a case in which product ion charge states have not already been established and their values are not readily determined by identifying likely complementary pairs. Figure 3 shows the spectrum resulting from the ion trap collisional activation of the 6- charge state derived from electrospray of bovine insulin A-chain under negative ion conditions. (The MS/MS spectra of the 3- and 4- charge states of bovine insulin A-chain obtained using a triple quadrupole tandem mass spectrometer have been reported by Loo et al.,30 but the product ion assignments were not made with the benefit of unambiguous charge state determination.) The charge states of all of the ions giving rise to labeled peaks were determined by ion/ion proton transfer experiments. All of the ions have been (29) McLuckey, S. A. Unpublished results, Oak Ridge National Laboratory, 1995. (30) Loo, J. A.; Ogorzalek Loo, R. R.; Light, K. J.; Edmonds, C. G.; Smith, R. D. Anal. Chem. 1992, 64, 81-88.

Figure 3. Product ion spectrum derived from ion trap collisional activation of the 6- charge state of bovine insulin A-chain. An asterisk and dashed line indicate the m/z location of the parent ion.

Figure 4. Product ion spectrum arising from the MS3 experiment whereby the product ions formed in the mass-to-charge range of 475505 from collisional activation of the 6- charge state of bovine insulin A-chain were allowed to react with a population of protonated pyridine ions for 200 ms. The peaks labeled as 4- and 3- correspond to (Z16 - 4H)4- and (Z16 - 3H)3-, respectively.

assigned according to the Roepstorff nomenclature,31 as indicated in the figure insert. Charge states were determined for 13 of the product ions using several MS3 experiments. In most cases, the charge states of several product ions could be determined in a single MS3 experiment. For example, Figure 4 shows the results of an MS3 experiment in which product ions over an m/z range of roughly 475-500 were allowed to react with protonated pyridine for 200 ms. The reactions were sufficiently fast that several charge states for one of the products (i.e., the C205- ion) were formed. This allowed for a straightforward charge state determination for both of the major product ions that fall within the selected mass window. In the case of the C205- product anion, the complementary Z1ion (m/z 116) fell below the low m/z cutoff of the ion trap during the collisional activation period, thereby preventing the identification of the complementary C20/Z1 ion pair as a means for charge state determination. In some cases, no complementary ions are (31) Roepstorff, P.; Fohlman, J. Biomed. Mass Spectrom. 1984, 11, 601.

+

Figure 5. Product ion spectrum arising from the MS3 experiment whereby the product ions formed in the mass-to-charge range of 400410 from collisional activation of the 6- charge state of bovine insulin A-chain were allowed to react with a population of protonated pyridine ions for 200 ms. The peaks labeled as 4- and 3- correspond to (C14 - 4H)4- and (C14 - 3H)3-, respectively.

Scheme 1

present due to the fact that the product ion is formed by loss of a neutral. This is the case for the C206- anion, the charge of which was identified from the spectrum in Figure 5. In this case, as many as three protons were transferred within the 200 ms reaction period, thereby yielding four distinct m/z ratio values for mass assignment. Note also that an ion of relatively low abundance at a slightly higher m/z ratio than that of the C206- anion remained after the ion isolation period and also reacted with protonated pyridine. The mass-to-charge values of the parent and product ion indicate that they carry four and three charges, respectively. This spectrum, and that of Figure 4 as well, shows how two parent ions of similar, or even identical, mass-to-charge ratio but different mass and charge can be dispersed by proton transfer reactions. Based on the mass assignments made possible by the ion/ ion reactions and the assumption that the oxidized cysteine residues were favored charge sites, Scheme 1 could be constructed, showing the likely major structurally informative reactions in the decomposition of the 6- parent ions. In some cases,

+

ions arising from more than one fragmentation step could be identified. Many ions, including the parent ion, showed loss of water and/or ammonia as an important process. Ions indicated in brackets did not appear in the MS/MS spectrum but were implied by the appearance of the complementary fragment. The ions in brackets did not appear due either to the mass/charge being below the exclusion limit of the ion trap or, presumably, to further fragmentation. Unfortunately, only limited structural information is available from ion trap collisional activation of the 6- parent anion, reflecting, perhaps, a limitation of the activation method or an inherent characteristic of negatively charged peptides with oxidized cysteines. Interestingly, C- and Z-type product ions were formed from the A-chain bovine insulin polyanion, in contrast to the usual observation of B- and Y-type product ions from positively charged peptides. Given the extremely limited data set reported for collisional activation of peptide anions (this work and ref 30), little can be said about the generality of the tendency for C- and Z-type product ion formation from peptide anions. The triple quadrupole MS/MS data of the 4- and 3- anions of oxidized A-chain bovine insulin show some product ions in common with the ion trap MS/MS data of the 6parent ion, as well as some additional product ions. A detailed comparison of these data is beyond the scope of this paper, but it is likely that both the different charge states interrogated and the significantly different collisional activation conditions that prevail in the ion trap and triple quadrupole experiments must be considered in rationalizing the product ions not held in common. It is noteworthy that the charge states of the product ions could be determined without recourse to searching for complementary ion pairs. The likelihood that the latter approach can lead to erroneous conclusions increases with the complexity of the spectrum. That is, as the number of product ions of discrete massto-charge ratio increases, the likelihood that two unrelated ions might fall at mass-to-charge values corresponding to those of a possible complementary pair also increases. The proton transfer approach is therefore more reliable than the identification of possible complementary pairs in determining charge states. It is also noteworthy that, unlike ion/molecule proton transfer, every multiply charged product ion so far interrogated by ion/ion chemistry involving a protonated Brønsted base has reacted by proton transfer. All singly charged ions also apparently react, as reflected by a rapid rate of disappearance in the presence of cations. Presumably, the singly charged anions also react by proton transfer, but no firm conclusions can be drawn in the absence of a mass measurement for either of the neutral products. While much more experience with ion/ion reactions will be needed to characterize well all aspects of ion/ion chemistry involving multiply charged polyatomic ions, it seems clear that ion/ion proton transfer between ions derived from Brønsted acids and bases is a significantly more universal reaction than the analogous ion/molecule reaction. CONCLUSIONS The dual-polarity ion storage capability of the quadrupole ion trap allows for the use of ion/ion proton transfer reactions involving oppositely charged ions for ion charge state determination. Ion/ion proton transfer reactions involving high-mass multiply charged ions studied to date appear to result in essentially no fragmentation of the high-mass ion despite the high exothermicity of the reaction. They are, therefore, an effective means

+

for inducing known shifts in mass and charge for high-mass multiply charged ions. They can be considered as a candidate tool for charge state determination in cases where no clear relationship between ions in a spectrum can be drawn on an a priori basis, as is often the case in tandem mass spectra of highmass multiply charged ions. The use of proton transfer reactions, involving either ion/ion or ion/molecule chemistry, is an alternative that requires far lower analyzer resolving powers than measuring the spacings between 13C peaks. Ion/ion proton transfer enjoys significant advantages over ion/molecule proton transfer in this application. Ion/ion proton transfer chemistry appears to be significantly more universal than ion/molecule chemistry. The highly exothermic nature of mutual neutralization and the absence of Coulomb barriers, which are involved in ion/ molecule proton transfer reactions,32,33 make ion/ion proton transfer reactions more likely to be observed than ion/molecule reactions. A second major advantage is the much more precise control afforded over introduction and elimination of the reactants when they are both charged. The introduction of each reactant is controlled electronically so that the period over which reactions can proceed can be precisely controlled. Reactions occurring at (32) McLuckey, S. A.; Glish, G. L.; Van Berkel, G. J. Proceedings of the 39th ASMS Conference on Mass Spectrometry and Allied Topics; Nashville, TN, May 1924, 1991; pp 901-902. (33) Bursey, M. M.; Pedersen, L. G. Org. Mass Spectrom. 1992, 27, 974-975.

+

inconvenient times can therefore be avoided. Reactions occurring during ion accumulation, ion isolation, mass analysis, etc. can be problematic for the application of ion/molecule proton transfer reactions and limit the reactant neutral number density that can be introduced into the vacuum system. We have found that ion/ ion proton transfer reactions in the ion trap can be driven at rates at least as great as those observed for ion/molecule proton transfer reactions at the “best compromise” neutral reactant number densities. The advantages of ion/ion proton transfer reactions over ion/molecule proton transfer reactions can therefore be realized without a compromise in duty cycle due to longer reaction times. ACKNOWLEDGMENT Dr. Joe Loo is gratefully acknowledged for discussions regarding anions of bovine insulin. This work was supported by the National Institutes of Health under Grant GM45372. Oak Ridge National Laboratory is managed for the U.S. Department of Energy by Lockheed Martin Energy Systems, Inc., under Contract DEAC05-84OR21400. Received for review September 1, 1995. October 12, 1995.X

Accepted

AC950895B X

Abstract published in Advance ACS Abstracts, December 15, 1995.