Charge permutation reactions in tandem mass spectrometry



Central to the tandem mass spectrometry experiment is the process that gives rise to product ions, i.e. the reaction intermediate to stages of mass analysis. Changes in mass or charge of the parent ion (or both) are generally readily detected by all forms of tandem mass spectrometry. Charge changing, or charge permutation, reactions have a long history in mass spectrometry. However, with the advent of new ionization methods, such as electrospray ionization, and the expansion of tandem mass spectrometry instrumentation to include ion trapping instruments, the past decade has seen a major increase in the types of charge permutation reactions that can be studied. Most charge permutation reactions involve electrons or protons as the charge mediating agents. This report, therefore, provides an overview of charge permutation reactions involving protons or electrons. Particular emphasis is placed on processes that involve interactions of precursor ions with gaseous neutral species, electrons or oppositely charged ions. Charge permutation reactions involving electron gain/loss are described first according to a rough order of the energy required for the reaction beginning with the most endoergic reactions and ending with the most exoergic reactions. An analogous approach is then taken with charge permutation reactions involving proton gain/loss. Important charge permutation reactions discussed herein, among others, include those referred to as charge inversion, charge stripping, electron capture dissociation, collision-induced ionization and charge separation. These reaction types, and others described herein, are the subjects of active research and are also finding use in many current areas of application. Copyright © 2004 John Wiley & Sons, Ltd.


Tandem mass spectrometry (MS/MS)1, 2 is based upon a detectable (and informative) change in a mass-selected precursor ion prior to the second stage of analysis. As common mass spectrometers analyze on the basis of mass-to-charge ratio (e.g. mass filters, quadrupole ion traps, ion cyclotron resonance), kinetic energy to charge ratio (e.g. electric sectors), momentum-to-charge ratio (e.g. magnetic sectors) or velocity (e.g. time-of-flight instruments), processes involving changes in mass, charge, velocity or combinations thereof can be detected by one or more of the common types of tandem mass spectrometers. While at least small changes in velocity accompany many reactions that occur in tandem mass spectrometry, most instruments are designed to be insensitive to small changes in velocity. Instruments used to perform translational spectroscopy,3 in which high-resolution measurements of ion velocity or kinetic energy are made, are notable exceptions. The most common reaction in MS/MS is that of dissociation following some form of ion activation. In the simplest case, singly charged parent ions fragment to yield singly charged and neutral products. Reactions involving a change in charge, often referred to collectively as charge permutation reactions,4–6 are also fairly common and have been known since the beginning of mass spectrometry. For example, evidence for many charge permutation reactions was apparent in J. J. Thomson's pioneering work with his famous parabola mass spectrograph.7 As the origins of modern analytical MS/MS can be traced to multi-sector instruments,1, 2 the study of charge permutation reactions in tandem mass spectrometry was originally restricted largely to processes induced by kiloelectronvolt kinetic energy collisions between an ion and a neutral collision gas. The fate of the thermal energy neutral collision partner, even if it were ionized by the collision, could only be inferred from measurement of the fast products because the sector instrumentation was not suitable for collecting low translational energy products. However, as the array of instrumental configurations used for tandem mass spectrometry has grown to include instruments in which ions are often of low translational energies, including ion trapping instruments, charge permutation reactions arising from low relative collision energies are now commonly observed. Furthermore, neutral species that are ionized as a result of the reaction can often be readily analyzed. With the introduction of ionization methods capable of forming multiply charged ions, a significantly increased array of charge permutation reactions has become possible.

With advances both in ionization methods and in instrumentation for MS/MS, the scope of charge permutation processes that can occur within the context of an MS/MS experiment has expanded dramatically over the past two decades. The purpose of this report is to provide an overview of charge permutation reactions in contemporary MS/MS and to place them into context with respect to reaction conditions and exoergicities. While species such as metal cations or halide anions can serve as charge-carrying agents in charge permutation reactions, the most MS/MS studies reported to date that have centered on charge permutation reactions have involved either electrons or protons as charge-mediating agents. As a first step in classifying charge permutation reactions, it is useful to distinguish processes that involve the gain/loss of electrons from those involving the gain/loss of protons. While there are many analogies to be drawn between electron gain/loss and proton gain/loss, there are sufficient differences between many of these processes to warrant separate classifications. Most charge permutation reactions have involved the interactions of ions with neutral gaseous collision partners. However, interactions of ions with electrons, surfaces and ions of opposite polarity are expanding the scope of charge permutation processes in MS/MS and some of these processes are also discussed.


In the early development of analytical MS/MS emphasis was placed on reactions of metastable ions5 and reactions induced by collisions at relatively high laboratory frame collision energies (i.e. >5 keV).4, 8 At such high collision energies and short reaction times, charge permutation reactions are dominated by electronic processes such as electron transfer and collisional ionization (often referred to as ‘charge stripping’ in the original literature). Pioneering groups in the study of such processes included those of the groups of Beynon and Cooks,4, 5, 9, 10 McLafferty11 and Bowie,12, 13 among others.14 Many of the reaction types observed with sector tandem mass spectrometers are included in Table 1, which summarizes key charge permutation reactions involving electron gain/loss as a result of gaseous collisions between ions and neutral gases, electrons and oppositely charged ions. Table 1 includes a name commonly used to refer to each charge permutation process and a generalized reaction. Various processes are organized according to a relative ranking of the enthalpy or, in some cases, free energy associated with the reaction. The reactions are listed in rough ascending order of exoergicity such that the reactions at the top of the table are most endoergic and those at the bottom are most exoergic. Note that several of the listed reactions types share a common name, such as electron transfer. In these cases, the individual processes are distinguished either by different reaction energetics (i.e. exoergic versus endoergic) or they involve different types of reactants and products. Each reaction type is discussed briefly in turn below in order of increasing exoergicity with emphasis on typical conditions used to observe the charge permutation process and the context(s) within which the reaction has been studied.

Table 1. Collisional Δz reactions involving e
  Reaction No.Reaction nameReaction equationaAssociated energya
  • a

    M = intact parent species; N = neutral collision partner; m = fragment derived from M; IE(Y) = ionization energy of Y → Y+•; IEII(Y) = ionization energy of Y → Y2+; EA(Y) = electron affinity of Y; RE(Yn+) = recombination energy of Yn+ = −ΔHrxn(Yn+ + e → Y(n−1)+); εcrit(Y) = critical energy for dissociation of species Y; Y = M or N, as appropriate.

 +1.1Charge strippingM+• + N → M2+ + N + eIEII(M) − IEI(M)
 1.2Electron transferMn+ + N → M(n+1)+• + N−•IE(Mn+) − EA(N)
 1.3Charge inversionM−• + N → M+• + N + 2eIE(M) + EA(M)
 1.4Electron ionizationM+• + fast e → M2+ + 2eIE(M+•)
 1.5Collisional ionizationMfast + N → M+• + N + eIE(M)
 1.6Charge inversionM+• + N → M−• + N2+IEII(N) − RE(M+•) − EA(M)
   M+• + 2N → M−• + 2N+•2(IE(N)) − RE(M+•) − EA(M)
 1.7Charge separationM2+ + N → m1+ + m2+ + Nεcrit(M2+)
 1.8Electron transferM+/−• + N → M + N+/−•IE(N) − IE(M)/EA(M) − EA(N)
  Electron transferM+/−• + N → M + N+/−•IE(N) − IE(M)/EA(M) − EA(N)
 1.9Electron capture-induced dissociation (ECID)M2+ + N → m1+• + m2 + N+•IE(N) − (IEII(M) − IE(M)) + εcrit(M+•)
 1.10Electron transferM2+ + N → M+• + N+•IE(N) − (IEII(M) − IE(M))
 1.11Electron transferMn + N+• → M(n−1)−• + NEA(M(n−1)−•) − RE(N+)
   Mn+ + N−• → M(n−1)+• + NEA(N) − RE(Mnn+)
 1.12Electron capture dissociation (ECD)MHnn+ + e → fragmentsεcrit(MHn(n−1)+•) − RE(MHnn+)
 1.13Electron captureMHnn+ + e → MHn(n−1)+•−RE(MHnn+)

Reaction (1.1). Charge stripping: M+• + N → M2+ + N + e

This reaction involves the collisional ionization of a positive ion. The reaction shown indicates the collisional ionization of a singly charged radical cation, although collisional ionization of singly charged even-electron ions or ions of higher charge states also falls within the category of charge stripping. The energy required for this reaction is the ionization energy (IE) of the ground-state radical cation less any internal energy already present in the ion. In the case of ground-state reactants and products, the energy requirement is the adiabatic IE of the cation. However, these reactions typically require the conversion of at least 10 eV of translational energy to liberate an electron from a positive ion. High collision energies and short interaction times are required such that vertical transitions apply. Hence, under conditions of high collision energy and large projectile-to-target mass ratio, the minimum loss of relative translational energy associated with the reaction generally reflects the vertical IE of the cation. Although fragmentation of the multiply charged ion is not indicated in Table 1, the possibility for subsequent fragmentation of the doubly charged product ion is implied.

Most studies of charge stripping reactions of polyatomic ions have been conducted using sector mass spectrometers in which keV energy collisions are induced in a field-free region preceding an electric sector analyzer.9 The use of an electric sector analyzer facilitates the recognition of charge stripping products because, in the case of M+• → M2+, the product ions are nominally transmitted at half the electric sector field strength required for transmission of the reactant ion. The energy required to drive the reaction is derived from the relative translational energy of the collision partners, changes in which are largely reflected in the kinetic energy of the fast collision partners. Hence a small deviation in the kinetic energy-to-charge ratio of the product ion from half that of the parent ion reflects the reaction endothermicity (Q). The measurement of translational energy losses associated with charge stripping, as illustrated in Fig. 1 for a series of atomic ions, is an approach that has been used to estimate IE of the ion from the minimum energy loss, Qmin.15 In cases in which electronically excited ions are present, smaller values of Q can be observed, as are particularly apparent in Fig. 1(a) and (b).

Figure 1.

Charge stripping spectra representing the processes (a) P+ → P2+•, (b) S+ → S2+•, (c) Mn+ → Mn2+• and (d) Co+ → Co2+•. Reprinted from Porter CJ, Proctor CJ, Ast T, Beynon JH. ‘Charge-stripping spectra of monatomic ions’. Int. J. Mass Spectrom. 1982; 41: 265. Copyright 1982, with permission from Elsevier Science.

In addition to the measurement of the ionization energies of positive ions, charge stripping has been used as a means for differentiating isomeric ion structures.16 It is argued that the charge stripping process might be able to distinguish isomeric ions when collision-induced dissociation reactions do not. The two processes, charge stripping (∼10−15 s) versus collisional induced dissociation (>10−13 s), occur on different time-scales with charge stripping being much faster. Isomerization reactions, therefore, are less likely to play a role in charge stripping than in a unimolecular dissociation process.

Reaction (1.2). Electron transfer Mn+ + N → M(n+1)+• + N−•

The collision between a positive ion and a neutral gas atom or molecule at high translational energy may result in the transfer of one electron from the cation to the neutral. This highly endothermic reaction leads to the formation of a cation of one charge state higher and a radical anion. The energy required for this process is determined by the difference between the ionization energy of Mn+ and the electron affinity of N. It is therefore possible to affect the reaction energetics, as well as other factors that affect the reaction, by varying the neutral reactant. Noble gases, oxygen and other species have been employed.17–19 A detailed review has been published recently.20

Studies involving Reaction (1.2) have been conducted with an accelerator mass spectrometer21 (50 kV acceleration voltage) coupled with electrospray ionization (ESI), as shown schematically in Fig. 2. Ions formed by ESI are guided into the acceleration tube of the accelerator where they acquire a kinetic energy of 50q keV, where q is the charge state of the ion. In one particular case, the accelerated ions of interest are isolated by a 2 m radius, 72° bending magnet and are injected into a collision cell. The ions exiting the collision cell are separated based on their energy-to-charge ratios by an electrostatic analyzer that, when scanned, generates ion kinetic energy spectra.

Figure 2.

Schematic diagram of an accelerator-based mass spectrometer used to study charge permutation reactions involving ions derived from electrospray ionization. Reprinted from Jorgensen TJD, Andersen JU, Hvelplund P, Sorensen M. ‘High-energy collisions of multiply charged lysozyme ions in gases’. Int. J. Mass Spectrom. 2001; 207: 31. Copyright 2001, with permission from Elsevier Science.

High-energy collisions of protonated peptides have been used to prepare peptide ions in a charge state one charge higher than that obtainable directly by ESI. Like the charge stripping process, described above, this reaction can give rise to cations that are otherwise difficult to form. They may be of interest from a structural standpoint if dissociation of the multiply charged radical ions formed by electron transfer gives rise to fragmentation patterns different from even-electron species. Furthermore, the reactions involving collisional ionization processes (Reactions (1.1) and (1.2)) provide the possibility to form multiply charged ions with charge states higher than those formed via ESI.19

Reaction (1.3). Charge inversion:

equation image

This reaction involves the removal of two electrons from a negatively charged ion M−• to form a singly charged positive ion M+• as a result of a kiloelectronvolt (keV) energy collision of an anion with a neutral target gas atom or molecule.12 The capture of one of the electrons by the target to form an anion, thereby releasing only one electron, may also occur.2 The relative contributions of these two processes depend on the collision conditions and the identities of the collision partners. The overall process can also take place via two collisions4 (see also the discussion of Reaction (1.5)). Charge inversion of a negative ion is highly endothermic and the minimum energy involved is the sum of the ionization energy (IE) and electron affinity (EA) of M, assuming that the reactants are in their respective ground states.22, 23 In the case of anion formation, the minimum energy is the sum of the IE of M and EA of M, less the EA of N. If the minimum energy required for the charge inversion is exceeded, subsequent fragmentation of product cations may be observed. Since the removal of two electrons is a vertical process, the newly formed positive ion initially retains the geometry of the negatively charged precursor ion.12 Therefore, provided fragmentation is fast, the major fragments can reflect the original structure of the precursor anion.24

Most charge inversion experiments have been performed on sector instruments.25 Albeit early work involved the use of forward geometry sector instruments, most charge inversion experiments have been conducted with reverse geometry sector instruments. Typically, the magnetic sector is set to select the negatively charged precursor ion, which undergoes charge inversion in a collision region between the magnetic and electric sectors to form the oppositely charged product ion. A reversal of the electric sector field is necessary to pass the positive product ion to reach the final detector. Mass-analyzed kinetic energy spectrometry (MIKES) can be used to measure the m/z values of the positively charged product ions formed via this approach.26 The spectra acquired are referred to as ‘+ E’ spectra.12 The cross-section for charge inversion is generally very low at low collision energies. Therefore, triple-quadrupole or other instruments that use low collision energy conditions are not well suited for this reaction.2 Nevertheless, positively charged product ions from negative ion charge inversion have been observed using triple-quadrupole instruments at a collision energy ranging from 70 to 170 eV,27, 28 although these results were not likely to result from a single gaseous collision, as implied by Reaction (1.3). One study27 employed multiple collisions and the other28 employed collision with a surface.

Bowie and Blumenthal pioneered the charge inversion of negative ions as a method for investigating the structures of negative ions that do not yield negative product ions via conventional collisional activation.12 Since then, a variety of work has been conducted.24 A fundamental application is the measurement of a molecule's electron affinity via determination of the energy loss associated with the inversion reaction, assuming the ionization energy is known.22, 29 Dissociative charge inversion of negative precursor ions has played an important role in ion structure elucidation, particularly when the tandem mass spectrum of the negative precursor ion shows little, if any, fragmentation.12, 30 As demonstrated in Fig. 3, the negative ion tandem mass spectrum of the polyaromatic hydrocarbon (PAH) fluoranthene (m/z 202) from a coal liquid was compared in Fig. 3(a) with the charge inversion tandem mass spectrum in Fig. 3(b) of the same ion.31 Obviously, the charge inversion tandem mass spectrum yielded much more information than the negative ion tandem mass spectrum. Peaks present at every carbon number from C2 through C15 clearly matched in detail the data obtained for a fluoranthene standard. The absence of any extraneous peaks in the tandem mass spectrum of the coal liquid (Fig. 3(c)), despite the complexity of the sample, is attributable both to the high selectivity of the negative chemical ionization method and to the reduction of chemical noise associated with MS/MS.

Figure 3.

Comparison of fluoranthene tandem mass spectra: (a) negative ion MS/MS (− → −); (b) negative ion charge inversion MS/MS (− → +); (c) negative ion charge inversion MS/MS (− → +) spectrum of solvent-refined coal mass 202. Reprinted from Zakett D, Ciupek JD, Cooks RG. ‘Determination of polyacrylic aromatic hydrocarbons in solvent-refined coal by negative chemical ionization–charge inversion mass spectrometry/mass spectrometry’. Anal. Chem. 1981; 53: 723. Copyright 1981, with permission from the American Chemical Society.

Reaction (1.4). Electron ionization: M+• + fast e→M2+ + 2e

A range of phenomena can be associated with the interaction of electrons with positive ions. The relative likelihoods for these phenomena are highly sensitive to the kinetic energy of the electron. With relatively low energies (<10 eV), the likelihood for electron capture dissociation (ECD) and hot electron capture dissociation (HECD), which are discussed further below, is maximized. When the electron energy increases beyond about 10 eV, electron ionization of the cation can occur, according to Reaction (1.4), in competition with other processes. In analogy with the well known electron ionization of molecules, the energy required for this reaction is equal to the ionization energy of M+•. Electron ionization of cations has been applied to ions formed initially by electron ionization.32 Such studies have typically concentrated on atomic ions. However, electron ionization of protonated species within the context of a true MS/MS experiment has been demonstrated recently.33 In this case, the ionization energy of MH+ must be supplied. When the precursor ion is multiply charged, as with multiply protonated species, electron ionization results in the formation of ions of one charge state higher, viz.

equation image

In such cases, the energy involved is determined by the ionization energy of MHnn+, IE(MHnn+). As the charge state increases, the average ionization energy tends to increase as a result of the increase in energy cost associated with electron removal from an increasingly positive charge.33

It has been noted that secondary reactions can occur in ion trapping instruments, such as an ion cyclotron resonance mass spectrometer. For example, exothermic capture of a low-energy electron present in the ion trap may lead to the formation of an excited precursor ion MHnn+•*:

equation image

This combination of electron ionization followed by electron capture is equivalent to the electronic excitation of the precursor ion MHnn+. If fragmentation of MHnn+* occurs, the overall process is referred to as electronic excitation dissociation (EED), which is similar to that resulting from UV photodissociation. However, fewer secondary fragments are generated via EED.34, 35 A detailed discussion can be found in a recent review.35

The principle instrument used in studying the electron ionization of cations is the Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer.36 Typically, ions generated by electrospray ionization (ESI), matrix-assisted laser desorption/ionization (MALDI) or by other means are trapped in the ICR cell. The precursor ions are isolated and irradiated with electrons for 10–100 s. Afterwards, the excess kinetic and internal energies of the trapped ions are removed via collisional cooling and the ions are detected or subjected to collisional activation or electron capture dissociation. This process is referred as tandem ionization mass spectrometry (TIMS) and a full sequence of events corresponding to II–MSI–III–MSII–MSIII (two-stage ionization, three-stage mass spectrometry) is reproduced in Fig. 4.

Figure 4.

Typical sequence of events in tandem ionization mass spectrometry. Reprinted from Zubarev RA, Nielsen ML, Budnik BA. ‘Tandem ionization mass spectrometry of biomolecules’. Eur. J. Mass Spectrom. 2000; 6: 235. Copyright 2000, with permission from IM Publications.

Since the first application of this approach to form doubly charged radical cations from singly charged precursor ions, electron ionization of cations has been demonstrated for polypeptides,36, 37 proteins, fullerenes and porphyrins.38 Ionization energies of cations with single or multiple charges can be determined with this method via threshold techniques.33, 36 Activation of radical ions formed via tandem ionization may provide complementary fragmentation pathways to even-electron ions and also to the distinct radical species formed via ECD.37 Negative ions have also been subjected to similar investigation. The anions derived from acidic peptides, [M − 2H]2−, yielded novel radical [M − 2H]−• species after irradiation.38 Fragmentation that results from this process has been referred to as electron detachment dissociation (EDD).39 An example is demonstrated in Fig. 5, where EDD cleaved all inter-residue bonds including several N[BOND]Cα bonds of the dianion of a sulfated peptide caerulein. On the other hand, collisionally activated dissociation (CAD) of the dianion produced mainly neutral group losses with little backbone cleavage.

Figure 5.

EDD spectrum of dianions of the sulfated peptide caerulein. The inset shows the result of the capture of hydrogen ions by the dianions. Reprinted from Budnik BA, Haselmann KF, Zubarev RA. ‘Electron detachment dissociation of peptide dianions: an electron–hole recombination phenomenon’. Chem. Phys. Lett. 2001; 342: 299. Copyright 2001, with permission from Elsevier Science.

Reaction (1.5). Collisional ionization: Mfast + N → M+• + N + e

This reaction represents the ionization of a fast neutral species, Mfast, upon collision with a target neutral species, N.40 The minimum energy required is equal to the adiabatic ionization energy of M. This reaction has been studied most extensively in MS/MS within the context of the neutralization–reionization mass spectrometry (NRMS) experiment, first reported by McLafferty and co-workers.41, 42 In NRMS, a precursor ion of multi-kiloelectronvolt kinetic energy is mass selected to produce the desired neutral by electron transfer or detachment, depending upon precursor ion charge, or by decomposition of metastable or collisionally activated ions. Residual ions are deflected electrostatically to yield a beam of fast-moving neutrals, which is subjected to high-energy collision with a neutral target gas. Electron removal from a fast neutral upon collision can occur by electron detachment, as indicated in Reaction (1.5), or by electron transfer to the target. If the product ions are formed with sufficient energy, further dissociation may occur.42 This has been referred to as collision-induced dissociative ionization (CIDI) and has been used to probe transient as well as relatively stable species formed via neutralization43, 44 (see also other references in this section).

In the neutralization step of NRMS experiments, the internal energy of the neutral formed and the extent of fragmentation can be controlled by the choice of targets. For example, vaporized metals are found to be the most efficient targets for charge exchange neutralization and they yield much less fragmentation than is observed with helium as the target.45 In the reionization step, oxygen is found to be the most efficient collision gas, and it also minimizes the extent of fragmentation. Therefore, when structural information is desirable, increased target pressure (to maximize collisional excitation) and/or helium gas should be used. Maximum neutralization and reionization efficiencies vary substantially between species but generally are found at pressures corresponding to ∼30% transmittance of the precursor beam. As a rule, the highest precursor ion kinetic energies possible are recommended to minimize losses due to scattering.45, 46

Many NRMS experiments have been performed in the field-free region of a double-focusing mass spectrometer, where neutralization and reionization occur in the same collision cell. This arrangement may lead to ambiguity with respect to the origin of fragments, as they might arise from either neutralized or ionized parent species.41–43 For this reason, a sector mass spectrometer was equipped with multiple collision cells in a single field-free region to allow for spatial separation of the neutralization, collisional activation and reionization processes.11, 45–47

One major application of NRMS has been the generation and characterization of elusive and/or hitherto unknown radicals and molecules.48 Several cases have been reported that demonstrate the distinction of isomeric compounds via NRMS,42, 49, 50 while corresponding CID experiments only gave rise to identical spectra for the isomers. It has been reported that NRMS can also provide valuable information about the structures and chemistries of neutrals and ions derived from small organic and inorganic molecules,11, 42, 43, 51, 52 organometallic compounds53 and relatively large molecules including fullerenes,54 peptides,55, 56 and nucleotides and related species.57, 58 The studies alluded to here constitute a small subset of the applications of NRMS. An example of NRMS applied to a relatively large species is given in Fig. 6, where both the neutralization–reionization mass spectrum of FcCH2+ (Fig. 6(a)) and the CIDI spectrum of FcCH2CH2Fc+• (Fig. 6(b)) confirmed the formation of stable neutral FcCH2(Fc = CpFeC5H4) having a formal 19-electron configuration.59

Figure 6.

(a) Neutralization–reionization mass spectrum of FcCH2+ ions and (b) collision-induced dissociative ionization mass spectrum of FcCH2CH2Fc+•. Reprinted from Zagorevskii DV, Holmes JL. ‘Generation and identification of neutral CpFeC5H4[DOUBLE BOND]X (X = O, CH2, CO) complexes in the gas phase by tandem mass spectrometry’. Organometallics 1997; 16: 1969. Copyright 1997, with permission from the American Chemical Society.

Reaction (1.6). Charge inversion:

equation image

Charge inversion has also been observed for positive ions. The collision between the positively charged ion, M+•, and the target neutral gas molecule, N, may give rise to the formation of negatively charged M−•, in addition to a positively charged ion, N2+, in a single encounter. Alternatively, a net charge inversion can occur via sequential electron transfer reactions that yield two singly charged cations, N+•. Analogous reactions can be written for even-electron precursor ions. Although N is usually in the form of rare gas, alkali metal targets have also been used as the collision gas.60, 61

The energetic requirement for the single collision mechanism is greater than that of the two-collision mechanism for a given neutral target.62 In the former case, the minimum energy of the process requires the double ionization energy of the neutral gas molecule, IEII(N), minus the total of electron affinity (EA) of neutral molecule, M, and the recombination energy (RE) of singly charged cation, M+: Emin = IEII(N) − RE(M+•) − EA(M) (Note that the recombination energy is usually reported as a positive number63 and this convention is used here.) When this reaction occurs in two steps, it can be represented as

equation image

The minimum energy required for this process is the total of twice the ionization energy IE(N) minus the total of EA(M) and RE(M+•):Emin = 2(IE(N)) − RE(M+•) − EA(M). Since twice the first ionization energy of the neutral gas molecule is smaller than the energy required to remove two electrons from a single molecule of M, the minimum energy required for the single-step reaction will be higher than that for the two-step process. These two mechanisms can be distinguished by their collision gas pressure dependences. The single-step reaction has a linear relationship with the gas pressure whereas the two-step process shows a pressure-squared dependence.64, 65 In some studies, both mechanisms have been implicated.66 The double ionization energies of the target molecules,5, 65, 67 or the single ionization energies of M68, 69 can be determined when the positive ion charge inversion proceeds by a one-step mechanism or a two-step mechanism, respectively.

Cooks' group initiated this type of experiment using a double-focusing mass spectrometer of forward geometry.70 When the charge inversion occurs in the first field-free region preceding the electric sector, the direction of the electric field must be reversed from the normal positive ion transmission mode to transmit the negative product ions. A detector situated between the two sectors may be used to obtain an ion kinetic energy spectrum without performing mass analysis. Alternatively, the direction of the magnetic field can also be reversed to pass ions to the final detector for mass analysis, thereby allowing for the acquisition of the so-called ‘−E’ mass spectrum.4, 70 If the charge inversion process occurs between the two sectors, the electric field must pass positive ions while the direction of magnetic field must be reversed to pass negative ions to acquire charge inversion mass spectra.4 With the widespread adoption of reverse-geometry sector instruments, charge inversion reactions have subsequently been induced between the magnetic and electric sectors with the electric sector set to pass negative ions.68, 71 Positive ion charge inversion has also been explored using a tandem quadrupole28 and multiple-sector mass spectrometers.72, 73

Although not investigated as extensively as negative ion charge inversion owing to the low yields resulting from the small cross-section of this process,2 positive ion charge inversion has been demonstrated to be a useful tool in the structural study of some ions of interest. It has been reasoned that the nature of the rapid conversion process, at least for the one-step mechanism, leads to the product negative ions maintaining essentially the same structure as the precursor positive ions. Extensive fragmentation of the anion has been observed to provide structural information that could be complementary to that obtained from direct dissociation of the positive precursor ions.5, 70, 74 Charge inversion mass spectrometry was found to be useful in differentiating isomeric ions, sometimes more clearly than CID of the positive ions.60, 72, 75

Reaction (1.7). Charge separation: M2+(+N)→m1+ + m2+(+N)

The charge separation process, represented here by the simplest case, which involves the decomposition of a doubly charged ion into two singly charged fragments, may be either unimolecular or collision induced.4, 5 In the former case, M2+ formed as a metastable ion can undergo fragmentation due to excess internal energy associated with the ionization process. In the latter case, the reaction involves the activation of a doubly charged ion with neutral gas followed by decomposition to give rise to two singly charged fragments, m1+ and m2+. The energy required for this reaction is equal to the lowest energy threshold for dissociation of the precursor ion. Products are not restricted to two singly charged ions as the formation of a neutral product, in conjunction with the two product ions, has been noted.76 Charge separation mass spectrometry has also been applied to multiply charged ions to give rise to multiple singly charged products.77, 78 Furthermore, with the advent of ionization methods that can provide a wide array of multiply charged ions, MS/MS of such species gives rise to a wide variety precursor/product charge state combinations, many of which involve charge separation.

Most of the early work involving charge separation reactions was carried out with sector mass spectrometers for reactions slow enough to give rise to metastable ions.79 If the charge separation takes place in the first field-free region, an electric sector voltage scan is usually applied to record the products.80 In these cases, the metastable dications have at least microsecond lifetimes. Such studies involving either metastable or collisionally activated ions have also been performed with reverse-geometry sector instruments with reactions that occur in the second field-free region. In recent years, coincidence techniques, such as photoelectron–photoion coincidence, have made fast charge-separation reactions of multiply charged ions with nanosecond or shorter lifetimes amenable to study by time-of-flight mass spectrometers.81, 82 A key feature of all of these approaches is the ability to measure the recoil energies associated with the fragmentation, as reflected by the kinetic energy distributions of the products. Charge separation reactions have also been studied with other forms of instrumentation, such as triple-quadrupole83 and ion trapping mass spectrometers.84 These tools, however, are insensitive to product ion kinetic energies.

Charge separation mass spectrometry has been used to obtain structural information about target molecules. A practical advantage of the charge separation process is that one of the fragment ions has a mass-to-charge ratio larger than that of the precursor ion. Therefore, this product can be clearly distinguished from the often more abundant product ions formed via neutral loss.85 Typically, charge separation is accompanied by a large amount of kinetic energy release (KER), thereby yielding very broad peaks when the products are analyzed on the basis of kinetic energy-to-charge ratio. KER carries valuable information concerning both the structures of the species involved and the energetics and dynamics of the reaction.80, 86 The amount of kinetic energy released can be calculated from the widths of peaks in ion kinetic energy spectra and the distance between the two charges at the moment of decomposition, estimated by5, 87

equation image

can be used to infer information regarding the structures and geometries of precursor ions.88, 89 For example, the MIKE spectrum and B/E spectrum of the M2+ ion of indole90 are shown in Fig. 7, where broad metastable peaks were observed due to the charge separation processes. KER as well as the intercharge distance were obtained from the top spectrum in characterizing the precursors ion and determining mechanistic aspects associated with the fragmentation.

Figure 7.

(a) MIKE spectrum of the M2+ of indole and (b) B/E spectrum of the M2+ of indole. Reprinted from Cardoso AM, Ferrer Correia AJ. ‘Fragmentation reactions of molecular ions and dications of indoleamines’. Eur. J. Mass Spectrom. 1999; 5: 11. Copyright 1999, with permission from IM Publications.

Although the intercharge distance obtained from a KER measurement must be interpreted cautiously, it has been used in discriminating isomers.79, 82, 88, 91 Recently, charge separation mass spectrometry has been applied to relatively large systems, including multiply charged fullerenes92 and porphyrins.93 Furthermore, doubly charged cluster ions have also been subjected to study by charge separation mass spectrometry and possible structures of such clusters have been proposed.84, 94, 95

Reaction (1.8). Electron transfer: M+/−• + N→M + N+/−•

Electron transfer reactions (charge transfer and charge exchange are other commonly used terms for this class of reactions) are among the most common in all of chemistry and there is a long history of the study of such reactions in mass spectrometry. No attempt is made here to review this enormous body of work. It is included here for completeness in that electron transfer is inherently a charge permutation process. The energy associated with electron transfer involving a positive ion and a neutral species is determined by the sum of the recombination energy of the positive ion and the ionization energy of the neutral. When the ion–neutral interaction occurs at thermal energies, endoergic reactions are unlikely. However, such reactions can be driven by collision energy in systems that allow for acceleration of the ion.2 In favorable cases, it is possible to select cationic reagents for electron transfer that allow for a narrowly defined and known internal energy to be present in the ion produced via the reaction.2, 63, 96 A breakdown curve can be constructed directly in this way.63 With this information, charge exchange reactions have been applied to distinguish isomers and obtain the critical energies of isomerizations.97–99 It has been argued that, in many cases, isomer distinction is more facile using charge exchange ionization than using electron ionization.99

In the interaction of an anion with a neutral species, the sum of the electron detachment threshold of the anion and the electron affinity of the neutral determines the energetic requirement for the reaction. A major difference between a reaction occurring in negative polarity and that in positive polarity lies in the fact that electron detachment from M−• may compete with the electron transfer process:100

equation image

In some cases, this reaction may also occur at low collision energies by forming a short-lived MN−• complex followed by ejection of the electron.101

A variety of tandem mass spectrometers have been employed in the study of this type of reaction. Double-focusing instruments were first employed.2, 63, 97, 102 Later, quadrupole mass spectrometers,103 triple-sector mass spectrometers,98, 104 FT-ICR instruments105 and trap mass spectrometers106 have been used to study electron transfer.

Reaction (1.9). Electron capture-induced dissociation (ECID): M2+ + N → m1+• + m2 + N+•

Electron transfer is energetically favorable for many dication/neutral target combinations (i.e., M2+/N) to yield two singly charged products, M+• and N+. If either product is sufficiently excited by this exoergic process, it may undergo unimolecular dissociation.2, 107–110 The process for M2+, therefore, involves two steps:

equation image

The energy involved in this reaction is the total of the ionization energy of N and the critical energy for dissociation of M+• less the first recombination energy of M2+. Therefore, the internal energy available for deposition into M+• varies with the ionization energy of N.107, 109, 111, 112 This overall process has been titled electron capture-induced dissociation (ECID). Unfortunately, this is a somewhat misleading description as it would be more accurately termed electron transfer-induced dissociation.

ECID in MS/MS was first performed with a reverse-geometry sector mass spectrometer where collisions between the isolated doubly charged precursor ion, M2+, with the target gas occurred in the second field-free region.107, 112, 113 ECID products formed in the second field-free region were analyzed by scanning the electrostatic analyzer (ESA) voltages between E and 2E, a range of potentials within which the singly charged ions formed from doubly charged precursor ions in the second field-free region are transmitted. Products in this kinetic energy-to-charge region can also be formed via the charge separation process of M2+ to generate m1+ and m2+. The latter products are easily distinguished from ECID products because they appear as dished peaks. As an example, the MIKE spectrum for the CID of CS22+ to form S+ ions in collision with Ar is shown in Fig. 8. Three distinguishable processes, two collision-induced charge separation dissociation (CICSD) reactions and one ECID reaction, are observed. Methods to better separate peaks from the two sources have been developed, such as floating the collision cell, varying the accelerating voltage and changing the collision gas.85 A forward geometry double-focusing mass spectrometer114 and time-of-flight mass spectrometers115 have also been used to study this process.

Figure 8.

MIKE spectrum for the CID of CS22+ to S+ ions in collision with Ar. Reprinted from Zhou XD, Shukla AK, Tosh RE, Futrell JH. ‘Unimolecular and collision-induced dissociation study of CS22+ with Ar at high collision energy’. Int. J. Mass Spectrom. 1997; 160: 49. Copyright 1997, with permission from Elsevier Science.

ECID is useful in providing information regarding threshold energies for various dissociation processes, the electronic states of the reactants, total and differential cross-sections, the internal energy distribution in the product ions and energy transfer in the collision.113 In the last two decades, ECID has been applied for the characterization of the structures of doubly charged precursor ions,107, 116, 117 differentiation of isomers112, 118 and the determination of fragmentation pathways.114, 119 It is possible to use ECID as an alternative to charge stripping for the determination of the ionization energies of singly charged ions. It has an advantage over charge stripping in that with a low-IE target, the electron-capture cross-section is higher.107, 113 ECID has also been effected between doubly protonated peptide ions, [P + 2H]2+, at an energy of 100 keV with metal vapor, such as Na or Mg, on an accelerator mass spectrometer to produce mainly [P + 2H]+•.120 Fragments corresponding to backbone cleavages were shown to give rise to both c′ and z ions, which are similar to the fragmentations arising from electron capture dissociation (see below).

Reaction (1.10). Electron transfer: M2+ + N → M+• + N+•

This reaction is the first step of the ECID process when the doubly charged positive ion, M2+, undergoes single electron transfer with a target gas, N, to form two singly charged positive radical ions, M+• and N+•. The energy associated with this process is the difference between the ionization energy of the neutral target less the first recombination energy of the doubly charged precursor ion, M2+. Typically, the singly charged M+• ion is formed in a vibrationally or electronically excited state. A distribution of such states often results in broad peaks in kinetic energy distribution of the product ions. Extrapolation to the high-energy side of the peaks can provide the translational energy change at threshold and the lowest ionization energy of M+• can be obtained from this value.85, 107, 113 The reaction energetics can be varied by the choice of neutral molecules with different ionization energies. Although this reaction has been studied primarily at high collision energies (keV),111, 117, 121–123 it is usually exoergic and can be studied at low translational energies (eV or sub-eV).124, 125 Based on Landau–Zener theory,125, 126 the electron transfer is pictured as occurring at the intersection of two diabatic potential energy curves with interpartical separations of 2–6 Å, corresponding to the so-called ‘reaction window’ resulting in a high probability for the net electron transfer for reactants at this range of distances.110, 115, 127

This reaction has been studied via tandem mass spectrometry with both reverse- and forward-geometry sector mass spectrometers by Beynon and co-workers86, 123 and Moran and co-workers121, 128 to obtain fundamental information regarding doubly charged ions derived from polyatomic molecules. In these experiments, M+• formed via charge transfer acquires twice the kinetic energy of M+• ions formed in the ion source. Therefore, by setting the electric sector plate voltages at twice their normal voltages, 2E, M+• can be separated from all other ions and analyzed free of interference to give rise to a ‘2E mass spectrum’. The abundances of these M+• ions are determined by a number of factors, including the relative abundances of the M2+ precursor ions, the relative cross-sections for reaction, and the relative efficiencies for charge transfer in competition with other processes, such as fragmentation, double electron transfer and scattering.2, 129 A cross-beam spectrometer composed of a quadrupole mass spectrometer and a time-of-flight mass spectrometer has been developed at University College London for electron transfer studies of dictations at low collision energy.125, 126, 130

It has been demonstrated that isomeric compounds can be readily distinguished by comparing their 2E mass spectra, even if their singly charged ion mass spectra are identical.131 Although most of the work has been carried out with small dictations, multiply charged macro-ions, including C60 and related fullerene compounds, have been subjected to studies of this kind.132, 133 Models of charge distribution including charge motion on the ion microsurface of C60n+ during the interaction trajectory have been proposed.133, 134 In other cases, C60 has been used as the neutral target and collided with a highly charged ions to undergo electron transfer.135

Reaction (1.11). Electron transfer (between ions):

equation image

Several combinations of cation and anion types can, in principle, react by electron transfer. For example, even-electron multiply charged anions might react by electron transfer with either even- or odd-electron cations. The latter reaction is shown above. This particular ion–ion reaction variation involves the transfer of one electron from a multiply charged even-electron, closed-shell negative ion, Mn, to a singly charged positive radical ion, N+•, to generate a negative radical ion, M(n−1)−•. The reaction enthalpies are determined by the difference between the electron affinity (EA) of M(n−1)−• and the recombination energy (RE) of N+•. In the case of even-electron cations, a radical is formed upon electron transfer. In most studies to date, rare-gas cations have been used as an electron acceptor and reaction exothermicities have been ≥10 eV. In one case, CCl3+ has been used and the reaction was expected to be exothermic by ∼5 eV(115 kcal mol−1(1 kcal = 4.184 k J)).136 As a result of the relatively high exothermicities of these reactions, fragmentation of the radical anions is often observed following electron transfer, particularly when the anions are relatively small. The extent to which the dissociation occurs depends on the kinetic stability and size of the initially formed ion–ion reaction product, as well as the ion–ion reaction exothermicity.136, 137 A detailed discussion can be found in the literature.138, 139

In the case of multiply charged cations reacting by electron transfer with anions, relatively few examples have been described.138 However, it has been reported recently that electron transfer to multiply protonated polypeptides has been observed and that fragmentation of the cations to yield products similar to those arising from electron capture dissociation (see below) is noted.140, 141 All electron transfer ion–ion reactions, regardless of ion polarity combination, have been studied with electrodynamic ion traps.136, 137, 140–145 ESI has been employed to produce multiply charged ions. Several species capable of forming multiply charged ions have been subjected to electron transfer reactions including polypeptides,140, 141 nucleic acids136, 137, 143–145 and various species with multiple acidic functional groups.142, 143 Fig. 9 shows the spectrum obtained by subjecting anions derived from a 5′-pd(A)40–60-3′ mixture to electron transfer with O2+•. Ions corresponding to each of the 40–60-mer components are clearly present. Relatively abundant signals corresponding to 20–39-mers and even smaller oligomers are also noted and arise from fragmentation of the larger oligomers as a result of electron transfer reactions with O2+•.

Figure 9.

Electron transfer product spectrum obtained from the reactions between the multiply charged nucleic acid anions derived from a 5′-pd(A)40–60 − 3′ mixture with O2+•. Reprinted from McLuckey SA, Wu J, Bundy JL, Stephenson JL Jr, Hurst GB. ‘Oligonucleotide mixture analysis via electrospray and ion/ion reactions in a quadrupole ion trap.’ Anal. Chem. 2002; 74: 976. Copyright 2002, with permission from the American Chemical Society.

Reaction (1.12). Electron capture dissociation (ECD): MHnn+ + e→fragments

Gas-phase electron capture by multiply protonated molecules followed by dissociation of multiply protonated ions was first noted in the late 1990s. Similar to the well-known dissociative recombination (DR) process,146 the energy involved in this reaction is the critical dissociation energy of the charge reduced radical, εcrit(MHn(n−1)+•) minus RE(MHnn+). Here, RE(MHnn+) can be calculated based on the proton affinity (PA) and hydrogen atom affinity (HA) of MHn−1(n−1)+ according to the equation

equation image

A detailed description of energy considerations can be found in the literature.146

The cross-section for electron capture has been demonstrated to be inversely related to the electron energy and directly related to the square of cation charge.147, 148 The velocity dependence leads to a maximum efficiency of conventional ECD at near zero electron energy. A broad local maximum of ECD cross-section, which is two orders of magnitude smaller than that at <0.2 eV, has also been found at roughly 10 eV.149 This maximum has been attributed to electronic excitation prior to electron capture. ECD under such conditions is referred as ‘hot-electron’ capture dissociation (HECD). Secondary fragmentation arising due to the higher electron energy has been noted and used to differentiate isomeric residues of amino acids.149, 150

Unlike other dissociation methods, such as CAD and IRMPD, evidence has been presented to suggest that ECD is a non-ergodic process. That is, dissociation is not slow relative to the rate of redistribution of internal energy such that a key assumption of statistical theories of unimolecular dissociation, which are based on dynamic equilibration of internal energy prior to dissociation of an activated molecule/ion, does not apply. Furthermore, the fragmenting species of ECD is the radical cation, [M + nH](n−1)+•, formed from electron capture of the precursor ion, while the other dissociation methods apply to even-electron precursor ions. While the details of the dynamics and mechanisms underlying ECD are still under examination, the dominant fragmentation pathways associated with ECD clearly differ from those of other approaches. For example, mainly c′ and z ions have been observed from fragmenting peptides or proteins via ECD whereas b and y′ ions are the major fragments formed by low-energy CAD. In the HECD experiments, w and d ions have also been reported.149 In addition to these major fragmentation pathways, small molecule losses also contribute to the product ion spectra.151 This topic has recently been reviewed by Zubarev et al.152

To date, ECD has been mainly effected in FT-ICR instruments.151 Cell modification for simultaneous trapping of cations and electrons has been useful in such investigations.148, 151, 153, 154 Modification by coupling the external accumulation (XA) technique to FT-ICR, which effects the continuous storage of ions generated from ESI in an r.f.-only multipole followed by pulsed injection into the ICR cell,155, 156 has provided a significant efficiency improvement, reducing reaction times from seconds to milliseconds. Furthermore, the replacement of a tungsten filament by an indirectly heated dispenser cathode with a much larger emitting area and a lower working temperature can greatly reduce the time required for electron irradiation.156, 157 More recently, efforts have been made to develop a quadrupole ion trap capable of performing ECD experiments.158, 159

Since the first application of ECD to the characterization of peptides and proteins,160 the combination of high resolving power FT-ICR with ECD and other activation methods has been successfully applied to de novo protein sequencing148, 154, 161–163 and the location of labile post-translational modifications in polypeptides and proteins.153, 164–166 Fig. 10 illustrates an example of an ECD spectrum of a 4+ enzymatic peptide ion from the bovine milk protein PP3.166 The ECD process gave rise to 25 c′ and 24 z fragment ions corresponding to 27 possible inter-residue cleavages. The resulting sequence tag combined with its intact mass uniquely identified this protein via a database search. This spectrum also confirmed five phosphorylation sites.

Figure 10.

ECD FT mass spectrum of the 4+ of peptide 25–53 enzymatically derived from the bovine milk protein PP3. Phosphorylation sites were determined at positions Ser29, Ser34, Ser38, Ser40 and Ser46. Reprinted from Kjeldsen F, Haselmann KF, Budnik BA, Sorensen ES, Zubarev RA. ‘Complete characterization of posttranslational modification sites in the bovine milk protein PP3 by tandem mass spectrometry with electron capture dissociation as the last stage’. Anal. Chem. 2003; 75: 2355. Copyright 2003, with permission from the American Chemical Society.

Positive ions other than those formed from proteins and peptides have also been subjected to study with ECD, including peptide nucleic acids (PNAs),167 DNA,168 oligosaccharides169 and synthetic polymers.170, 171 An ECD spectrum of polypropylene glycol (PPG) ions [H(C3H6O)82OH + nH]n+ is shown in Fig. 11. The major ECD products are A ions, fragments that contain a new terminal-OH, and the complementary odd-electron B species. Unlike the case for protein ions, ECD and CAD of polyglycol ions dissociated by similar mechanisms, except that B ions arising from CAD were even-electron species.

Figure 11.

ECD spectrum of (PPG82 + 6H)6+ ions. Reprinted from Cerda BA, Breuker K, Horn DM, McLafferty FW. ‘Charge/radical site initiation versus coulombic repulsion for cleavage of multiply charged ions. Charge solvation in poly(alkene glycol) ions’. J. Am. Soc. Mass Spectrom. 2001; 12: 565. Copyright 2001, with permission from Elsevier Science.

Reaction (1.13). Electron capture: MHnn+ + e → MHn(n−1)+•

A reaction between a polyatomic cation with an electron leading to the reduction of the cation charge without subsequent fragmentation can occur at very low electron energies (≤0.1 eV). Reaction exothermicity is given by the negative of the recombination energy of the precursor ion. As electron energy increases, other reaction types, including ECD and electron ionization of the precursor ions, may occur.35 Usually appearing as an accompanying reaction to ECD, the electron capture charge reduction reaction has been noted unambiguously only in FT-ICR instruments. Charge reduction products of multiply protonated protein/peptide ions149, 151 and polymers171 were observed to be 1 Da more massive than the MHn−1(n−1)+ ions formed by ESI, indicating that capture of electrons was the mechanism of formation (as opposed to proton transfer to background neutral species). Multiple electron capture has also been observed.148, 154, 156, 166, 172 For example, Fig. 12 shows the ECD spectrum of [M + 5H]5+ ions derived from insulin,154 which is dominated by electron capture products ranging from [M + 5H]4+• to [M + 5H]2+•.

Figure 12.

5+ ions from the electrospray spectrum of insulin subjected to ECD. Reprinted from Zubarev RA, Kruger NA, Fridriksson EK, Lewis MA, Horn DM, Carpenter BK, McLafferty FW. ‘Electron capture dissociation of gaseous multiply-charged proteins is favored at disulfide bonds and other sites of high hydrogen atom affinity’. J. Am. Chem. Soc. 1999; 121: 2857. Copyright 1999, with permission from the American Chemical Society.

Electron capture charge-reduced products have also been observed via interactions of charged precursor ions with high-energy electrons. For instance, the reaction between +16 ions of cytochrome c (Mr = 12 360) and 70 eV electrons was found to lead to the formation of [M + 16H]15+• and [M + 16H]14+ ions along with the electron ionization product [M + 16H]17+•. It was postulated that the charge-reduced products were originated via capture of a slow electron ejected from another molecular ion via the following two-step process:38

equation image


With the advent of ionization methods capable of generating singly or multiply protonated and deprotonated ions from macromolecules of interest,173 charge permutation reactions involving proton gain/loss have been playing an increasingly important role in MS/MS studies. A variety of charge permutation reactions involving proton gain/loss (see Table 2) are discussed in this section. As in Table 1, the reactions here are listed in rough ascending order of exoergicity. Reactions sharing the same name are distinguished from one another by their reaction energetics and reactant or product types.

Table 2. Collisional Δz reactions involving H+
  Reaction No.Reaction nameReaction equationaAssociated energya
  • a

    M = intact parent species; B = neutral base; X = species from which an anion is formed; m = fragment derived from M; εcrit(MHnn+) = critical energy for dissociation of MHnn+;GB(Y) = gas-phase basicity of species Y; ΔGacid(M) = free energy associated with the reaction M → (M − H) + H+. Note that much of the literature associated with the proton transfer reactions listed herein report values in terms of free energies. To be fully consistent with Table 1, which lists in terms of enthalpies, GB is replaced by proton affinity (PA) and ΔGacid is replaced by gas-phase acidity (GA).

 +2.1Charge separationMHnn+ → m1Hmm+ + m2H(nm)(nm)+εcrit(MHnn+)
 2.2Proton transferMH+ + B → M + BH+GB(M) − GB(B)
  Proton transferMH+ + B → M + BH+GB(M) − GB(B)
 2.3Proton transferMH22+ + B → MH+ + BH+GB(MH+) − GB(B)
 2.4Proton transferMH22+ + X → MH+ + XHGB(MH+) − ΔGacid(XH)
 2.5Multiple protonMH+ + X2− → (M − H) + XH2GB(M) + ΔGacid(M)−
 transfer 1st and 2nd ΔGacid(XH2)
    MH33+ + X2− → MH+ + XH21st and 2nd GB(MH+)−
  1st and 2nd ΔGacid(XH2)

Reaction (2.1). Charge separation: MHnn+ → m1Hmm+ + m2H(nm)(nm)+

Unimolecular dissociation historically has been the most important reaction in MS/MS. The dissociation of multiply protonated ions often involves charge separation process described in Reaction (2.1). Dissociation reactions that involve the loss of a neutral fragment may also compete. In general, losses of neutral species become less prominent as the charge of the parent ion increases. In analogy with charge separation processes involving electrons, the energy required is equal to the critical dissociation energy of the precursor ion MHnn+. For highly charged ions, the distribution of charges between the two first-generation products can range from an even distribution to one in which all but one of the charges are present in only one of the products. Sequential fragmentation can ensure if the products are excited,163, 174 to the extent that, in extreme cases, first-generation fragments are not observed. It has been demonstrated for macro-ions derived from proteins or peptides that the critical energy usually decreases as the charge state increases, as reflected in the dependence of fragmentation rates on the charge states of the ions.85, 175

A variety of dissociation methods, with activation times ranging from 10−15 to 103 s,152, 176 have been applied to multiply charged ions. Slow activation methods such as IRMPD, BIRD and ion trap collisional activation that involve stepwise excitation are generally restricted to ion trapping instruments. Fast activation methods, such as electron capture and electron transfer processes, which are expected to give rise to fragmentation upon a single ‘collision’, are also well suited for ion trapping instruments. Surface-induced dissociation (SID), a fast activation technique, of multiply charged ions has also been effected in ICR instrumentation.177 Beam-type instruments, such as the Q-TOF combination, triple-quadrupole mass spectrometers, accelerator mass spectrometers and multi-sector tandem mass spectrometers are restricted to faster activation methods and to the study of metastable ions. An example of the latter involving charge-separation processes of metastable multiply protonated ions investigated by MIKES has been described.178

Although the dissociation of multiply charged macro-ions is not reviewed here, within the context of charge permutation it is important to recognize that the magnitude and locations of charge play major roles in determining favored fragmentation channels. Fragmentation of multiply charged ions involving charge separation has provided valuable information about macro-ion sequence derived from a variety of molecules including proteins,148, 151, 154, 162, 172, 174, 179, 180 peptides,156, 163, 164, 175, 181, 182 DNAs,168 oligosaccharides169 and synthetic polymers.171, 183 Gas-phase conformational information can also sometimes be extracted from dissociation studies.171, 174, 178, 180, 182, 184 Furthermore, fragmentation data are useful in locating post-translational modification sites of peptides and proteins.153, 156, 164, 165, 172, 185 As an example, Fig. 13 reproduces the SID and sustained off-resonance irradiation CID (SORI-CID) spectra of the doubly charged peptide RLDIFSDFR ions.182 In both experiments, a charge separation process of the doubly charged peptide gave rise to the complementary fragments b7 and y2, which dominated the spectra.

Figure 13.

(A) ESI–SID spectrum of doubly protonated RLDIFSDFR at a collision energy of 55 eV on a 2-(perfluorooctyl)ethanethiolate (CF3(CF2)7CH2CH2SAu) monolayer surface. (B) Sustained off-resonance irradiation mass spectrum in a Fourier tandem mass spectrometer of doubly protonated RLDIFSDFR at an excitation energy of 13 eV. Reprinted from Tsaprailis G, Nair H, Somogyi A, Wysocki VH, Zhong WQ, Futrell JH, Summerfield SG, Gaskell SJ. ‘Influence of secondary structure on the fragmentation of protonated peptides’. J. Am. Chem. Soc. 1999; 121: 5142. Copyright 1999, with permission from the American Chemical Society.

Charge separation is also commonly observed in the dissociation of non-covalently bound systems composed of, for example, protein–peptide,186 protein–protein,187, 188 protein–oligosaccharides189 and also DNAs and RNAs.145, 190

Reaction (2.2). Ion–molecule proton transfer: MH+ + B → M + BH+

Proton transfer is one of the most commonly observed processes in ion–molecule chemistry.191–193 It plays a major role in most moderate-to-high pressure ionization techniques, such as conventional chemical ionization and atmospheric pressure chemical ionization. Within the context of MS/MS, proton transfer reactions have often been studied to determine relative gas-phase basicities (GB), proton affinities (PA) and gas-phase acidities (GA). The measurements of the gas-phase basicities or proton affinities of bases are usually performed by measuring the equilibrium constants of proton transfer,194 by the dissociation of proton-bound mixed dimers,195 or by the method of bracketing.191 In both the equilibrium and bracketing methods, a proton is transferred from a monocation, MH+, to a neutral base species, B, with known gas-phase basicity, according to the Reaction (2.2). The free-energy change for this reaction, ΔGrxn, is the difference in gas-phase basicities of M and B, ΔGrxn = GB(M) − GB(B), while the enthalpy change can be obtained from the difference in gas-phase proton affinities of M and B: ΔHrxn = PA(M) − PA(B). For the negative ion analog of Reaction 2.2,

equation image

the free energy change is determined by the difference in the ΔGacid values of BH and MH and the enthalpy change ΔHrxn is determined by the difference of the gas-phase acidities of BH and MH, ΔHrxn = GA(BH) − GA(MH).

Hundreds of measurements have been performed based on the proton transfer Reaction (2.2) or its negative ion analog, many of which involved MS/MS, using a variety of mass spectrometers capable of MS/MS experiments. Principle among them are ICR and FT-ICR mass spectrometers.196, 197 Other tools, such as triple-quadrupole instruments,198 quadrupole ion traps199, 200 and guided ion beam mass spectrometers,201 have also been employed. ‘Flow reactor’ or ‘flow tube’ techniques, including flowing afterglow (FA) and selected ion flow drift tube (SIFDT), have been used for the determination of thermodynamic values associated with proton transfer.193, 202 As one of the most widely studied types of ion–molecule reactions in the gas phase, studies involving this reaction have provided a substantial database of thermochemical information for organic compounds,197, 199, 203 monosaccharides,204 terpenes,198 multidentate ligands205 and biomolecules.206

Reaction (2.3). Ion–molecule proton transfer: MHnn+ + B → MHn−1(n−1)+ + BH+(n ≥2)

Proton transfer reactions between multiply protonated species and neutral bases have been studied using a variety of mass spectrometric techniques including ion trapping approaches based on the quadrupole ion trap207–209 and FT-ICR tandem mass spectrometry.210–214 Other instrumental approaches include those based on the selected-ion flow tube,215–218 triple-quadrupole mass spectrometers219–221 and an accelerator mass spectrometer.17

The proton transfer reaction between a multiply protonated species MHnn+ and a neutral base B leads to the formation of protonated B in addition to partially neutralized species, MHn−1(n−1)+. The free energy change associated with this reaction corresponds to the difference in gas-phase basicities of MHn−1(n−1)+ and B. The energy surface associated with Reaction (2.3) differs significantly from that associated with Reaction (2.2) in that the exit channel in Reaction (2.3) involves the separation of like charges. This leads to long-range repulsion as the proton is transferred to B. As a result, the reverse reaction is characterized by a long-range repulsive barrier. As a consequence of the lack of a dielectric medium, the establishment of a proton transfer equilibrium in the gas phase is impractical. Therefore, the equilibrium method cannot be applied to the measurement of the PA or GB values of multiply charged ions.215, 217, 222 Fig. 14 illustrates qualitative energy diagrams for proton transfer reactions of singly and doubly protonated M, a hypothetical symmetrical molecule with two identical protonation sites separated by a distance >10 Å, and neutral reference bases (A, B, C). Potential surface ‘a’ represents a reaction of MH+ with A where PA(M) = PA(A) and ‘d’ is the hypothetical reaction of MH22+ with A in the absence of Coulombic repulsion. The difference between these two energy surfaces corresponds to PA(M). When Columbic repulsion is considered, reaction of MH22+ with C where PA(MH+) = PA(C), represented as surface ‘b’, is slow owing to the activation barrier created by the Coulomb interaction. For this reason, the true PA of MH+ cannot be determined directly via the bracketing method for the determination of proton affinities. However, the barrier can be lowered by increasing the basicity of the neutral base, as demonstrated by ‘c’, where PA(B) > PA(A).

Figure 14.

Quantitative interaction potentials for proton transfer reaction of singly and doubly protonated M and neutral reference bases (A, B, C) where M is a symmetric molecule with two identical sites of protonation. (a) Reaction of MH+ with A, PA(M) = PA(A); (b) reaction of MH22+ with C, PA(MH+) = PA(C); (c) reaction of MH22+ with B, PAapp(MH+) = PA(B); (d) the hypothetical reaction of MH22+ with A in the absence of Coulombic repulsion. Reprinted from Schnier PD, Gross DS, Williams ER. ‘Electrostatic forces and dielectric polarizability of multiply protonated gas-phase cytochrome c ions probed by ion/molecule chemistry’. J. Am. Chem. Soc. 1995; 117: 6747. Copyright 1995, with permission from the American Chemical Society.

The bracketing method is based on negligible intrinsic barriers to proton transfer.194 This approach, therefore, cannot be directly used to determine gas-phase basicities of multiply charged ions. However, the occurrence or non-occurrence of a reaction can be used as a basis for the determination of relative ‘apparent’ or kinetic gas-phase basicity (GBapp) or acidity (GAapp) values.212, 213, 216, 221, 223, 224 Recently, equations based on electrostatic interaction have been developed for two-proton and multiproton systems to allow for the evaluation of GBapp of the basic sites on proteins.221 Ewing et al. have proposed an efficiency value of 0.1, approximately 10% of all collisions resulting in a deprotonation reaction, as the ‘break point’ for peptide ions.225 With these values, quantitative measurements of Coulomb energy and intrinsic dielectric polarizability in gas-phase ions with multiple protons have been demonstrated to provide information about ion conformation in the gas phase.192 For instance, multiple distinct conformations have been reported for several proteins based on proton transfer kinetics.210, 214, 224 It is noteworthy that the GB values of amino acid residues in a protein are significantly higher than that of the individual amino acid. This observation is attributed to extensive intramolecular interactions that stabilize the charge.213, 226 However, as the charge state of a polypeptide increases, the charge tends to be increasingly destabilized owing to electrostatic repulsion. Furthermore, some of the intramolecular interactions that stabilize charges can be disrupted as the molecule adjusts to minimize electrostatic repulsion. Therefore, the GB values of proteins decrease with increasing charge state.192 As an example, Table 3 lists the apparent gas-phase basicities of a range of charged states derived from cytochrome c.221

Table 3. Gas-phase basicities of cytochrome c varying with charge statea
Charge zGBjapp (kcal mol−1)GBjappGBmath image (kcal mol−1)Residueb
  • a

    Total of 29 basic sites considered.

  • b

    Residue that is easiest to be deprotonated in the presence of a base such as NH3.

  • c

    zobs = observed charge with highest intensity; however, zobs = 9 was also present.

  • d

    Charge provided by charge residue model (CRM) from precursor droplet of same radius as protein, R = 21 Å (based on x-ray data).

  • e

    Largest number of charges that available basic sites can hold.

zobsc 8208.512.5LYS53
zCRMd 9203.37.3LYS27
NSBe 10198.22.2LYS99

Ion–molecule proton transfer reactions have been used to determine the charges of product ions formed from CID208, 219 and also to improve the ‘effective’ mass resolution of mass spectrometers by manipulating the charge states formed via ESI.209 Ion–molecule proton transfer reactions of dications derived from fullerene adducts217, 221 and metallic compounds have also been explored.192, 220 For example, a study with gramicidin S suggested that [M + X]+ (X = Li, Na, K) ions have a higher apparent basicity than [M + H]+ owing to a larger separation between charges.192 The analogous reaction to Reaction (2.3) in the negative mode, [M − nH]n + BH → [M − (n − 1)H](n−1)− + B, has also been investigated with M being proteins/peptides211 or olignucleotides.209

Reaction (2.4). Ion–ion proton transfer: MHnn+ + X → MHn−1(n−1)+ + XH

Proton transfer between multiply charged positive ions and singly charged negative ions is one of the most commonly observed and studied ion–ion reactions. The free energy of this reaction is determined by the difference in the gas-phase basicity of MHn−1(n−1)+ and X, or its equivalent, the ΔGacid of XH, viz.

equation image

Accordingly, the enthalpy of the reaction is determined by the respective proton affinities, viz. ΔHrxn = PA(MHn−1(n−1)+) − PA(X) The negative analog of this reaction is:

equation image

and the free energy associated with it is determined by the relative gas-phase basicities of [M − nH]n and X. Deprotonation of cations and protonation of anions are expected to be highly exothermic for virtually every value of n and for virtually any combination of ions, provided that proton transfer is the preferred recombination mechanism. Details about the thermodynamics and kinetics associated with this reaction have been discussed.138, 139, 227–229

Ion–ion proton transfer reactions have been carried out mainly in high-pressure ion sources prior to sampling into a mass spectrometer and in electrodynamic ion traps. In the former case, the first studies involved ions of opposite polarity introduced into separate arms of a Y-tube where they merged and reacted in the flow at near atmospheric pressure prior to sampling into the atmosphere/vacuum interface of a quadrupole mass filter.230, 231 Proton transfer between multiply protonated ions and singly charged anions has also been investigated with time-of-flight mass spectrometers by Smith and co-workers using either a polonium decay ionization source or a corona discharge ionization source to generate singly charged ions.232 The quadrupole ion trap is particularly well suited to the study of ion–ion reactions owing to its ability to store ions of opposite polarity simultaneously and in overlapping regions of space. In cases where ion–ion proton transfer reactions have been studied in quadrupole ion traps, a variety of ion source configurations have been employed. In these studies, multiply charged ions formed by ESI have been injected through an aperture in the entrance end-cap electrode. Singly charged ions have been formed by electron ionization, ion trap chemical ionization, glow discharge ionization and laser desorption.138, 233 Recently, instrumentation has been developed to facilitate the study of reactions between ions of opposite polarities formed by ESI. An instrument with two electrospray sources234 and another with three independent electrospray sources and one glow discharge source (Fig. 15)235 have been described. Ions from each of the ESI sources can be injected into the ion trap via the entrance end cap in these instruments. These modifications have allowed the study of a wide variety of ion–ion reactions. Very recently, proton transfer reactions between multiply charged cations with singly charged anions have been effected in linear ion trap mass spectrometers140, 236

Figure 15.

Schematic diagram of multiple-source quadrupole ion trap mass spectrometer. Reprinted from Badman ER, Chrisman PA, McLuckey SA. ‘A quadrupole ion trap mass spectrometer with three independent ion sources for the study of gas-phase ion/ion reactions’. Anal. Chem. 2002; 74: 6237. Copyright 2002, with permission from the American Chemical Society.

Most proton transfer reaction studies have involved biopolymers, such as proteins, peptides and oligonucleotides,138, 229 and synthetic polymers, such as dendrimers and polyethylene glycols.183, 237 The highly exothermic nature of ion–ion reactions and the square of the charge (z2) dependence of the kinetics make them particularly useful for charge state manipulation within the context of mixture analysis. They are most useful when they can be effected between stages of mass selection/analysis. The development of ion–ion proton transfer reactions for polymer mixture analysis has been focused most heavily on strategies for ‘top-down’ proteomics using electrodynamic ion traps as both ion reactor and mass analyzer.237–241 Important uses include the reduction of whole protein charge states in electrospray mass spectra of mixtures to mitigate or eliminate the charge state overlap problem typically encountered in the ESI of protein mixtures.240, 241 Another application is the charge state reduction of product ion charge states to facilitate the interpretation of the whole protein tandem mass spectra.239, 240 A technique for concentration of ions initially dispersed over a range of charge states into a single charge state by use of the so-called ion parking technique has been developed.228 Fig. 16 demonstrates an example of the application of ion parking to concentrate cytochrome c ion signals from a range of charge states to mainly +10. The technique takes advantage of the fact that ions execute mass-to-charge dependent frequencies of motion in a quadrupole ion trap. The resonant acceleration of ions of a particular m/z ratio can greatly diminish the ion–ion reaction rate of the accelerated ions. The selective inhibition of ion–ion reaction rates allows for the accumulation of a large fraction of ions in an electrospray charge state distribution into a single charge state, provided that it is lower than the bulk of the ions formed via ESI.

Figure 16.

Mass spectra of bovine cytochrome c ions acquired in (a) pre-ion–ion, (b) post-ion–ion and (c) ion parking modes. The anion injection and mutual cation/anion storage periods used in (b) and (c) were 1 and 150 ms, respectively. Reprinted from McLuckey SA, Reid GE, Wells JM. ‘Ion parking during ion/ion reactions in electrodynamic ion traps’. Anal. Chem. 2002; 74: 336. Copyright 2002, with permission from the American Chemical Society.

Two sequential ion parking steps have been applied in deriving a charge state purified ion population from the ESI of whole protein mixtures.240 The subsequent collisional activation of the concentrated and purified parent ion population can then be followed by another ion–ion reaction period to reduce product ion charge states to facilitate interpretation of the resulting product ion spectrum. The overall process involves three distinct charge permutation steps and one collisional activation step. This sophisticated experiment is made possible by the robust nature of ion/ion proton transfer reactions and the facility with which ion trapping instruments can effect MSn experiments.

Reaction (2.5). Multiple proton transfer: MHnn+ + Xm → [M − (np)H](np)+ + XHp(mp)−

Similar to single ion–ion proton transfer discussed above, multiple proton transfer between two ionic species with opposite polarities is highly exothermic for virtually any combination of reactants. In fact, such reactions are more highly exothermic than single proton transfer reactions because they involve multiple neutralization steps. In the case of partial neutralization of a multiply protonated molecule with a doubly charged anion via the following reaction:

equation image(1)

the free energy change, ΔGrxn, of the reaction is given by

equation image

An important special case of multiple proton transfer is the reaction that leads to the charge inversion of the analyte species of interest:

equation image(2)

The free energy change, ΔGrxn, associated with this reaction is given by

equation image

In both cases, the proton transfer reagent Xn receives multiple protons in one step with the only distinction being the polarity of the product ion associated with the analyte molecule, M.

Charge inversion by proton transfer was first observed using a quadrupole mass spectrometer interfaced with two ESI sources via a Y-shaped capillary reactor. Unlike Reaction (2.5b), inversion of charge from negative to positive was noted. Specifically, the reaction of singly deprotonated adenosine 5′-monophosphate (AMP) and fluorescein ions with multiply protonated myoglobin ions were found to form the singly protonated forms of the nucleic acid and fluorescein.230 Limitations of the experimental apparatus made it difficult to draw firm conclusions regarding the reactions between multiply charged ions of opposite polarity in the Y-tube reactor. Recently, multiple proton transfer in a single step in an ion trap mass spectrometer from the reaction between multiply protonated ions with multiply deprotonated ions has been reported.188, 242–244 Both proton transfer and complex formation have been observed to be major reaction pathways. Several mechanisms have been examined to account for the observations and a model allowing for proton transfer and complex formation via bound orbits has best accounted for the experimental data available.188 Charge inversion by proton transfer has also been successfully applied to change the charge states and polarities of protonated peptides. For example, the use of sequential ion/ion charge inversion steps to form a doubly protonated peptide from a singly protonated peptide using a multi-source ion trap instrument has been reported.243 The individual steps are represented as

equation image

Here, reaction of the MH+ ion with a multiply charged anion can give rise to charge inversion of the peptide to produce the singly deprotonated species [M − H]. In the next step, the singly charged anion is reacted with a highly charged reagent cation. In the complex formed from these reactants, it is possible for charge to be redistributed between the two reactants giving rise to at least some products in which M carries more than one positive charge. The net effect is an increase in the charge of M from +1 to some higher charge state. The extent of charging is expected to depend both on the nature of M and on the nature and charge of the reagent ion. Various combinations of reactant charge states have been observed to give rise to charge inversion. Charge increase in the negative polarity by similar approach has also been carried out with macro-anions derived from a polypeptide and an oligonucleotide.245 Further work demonstrated the possibility to perform charge inversion of multiply charged precursor ions and form multiply charged product ions with opposite polarity by choosing suitable charge inversion agents.246


The scope of charge permutation processes occurring within the context of an MS/MS experiment has expanded dramatically during the last two decades, with advances in both ionization methods and instrumentation for MS/MS. In particular, the introduction of electrospray ionization and ion trapping instrumentation has given rise to new dimensions of charge permutation chemistry. Given the increasing range of species for which multiply charged ions can be formed and the continued development of tools that are capable of supporting multiple reactions in a single experiment, it is likely that charge permutation reactions in MS/MS will continue to see expanded scope and application. For example, the expanded availability of linear ion traps, either as stand-alone instruments or coupled with other mass analyzers, as well as expanded adoption of FT-ICR mass spectrometers, is likely to make many of the exoergic charge permutation reactions discussed here more widely accessible to the mass spectrometry community. This overview emphasizes the gain/loss of electrons and the gain/loss of protons in gas phase collisions, as most charge permutation reactions studied to date fall into one of these categories. However, charge permutation reactions resulting from interactions of ions with surfaces or photons are also well known. Furthermore, a wide range of reactions that involve changes in ion charge state that do not fall clearly into either the proton transfer or electron transfer categories are possible. These include, for example, metal ion transfer and halide ion transfer reactions. In addition, reactions that are more complex than simple charged particle transfer may also involve charge permutation. A specific example is the nucleophilic substitution reactions that lead to the attachment of trimethylsilyl moieties to phosphodiester linkages of oligonucleotide anions.247 This ion–molecule reaction provides an example of the covalent modification of a gas-phase ion. Ion–ion reactions can also give rise to new bond formation. In fact, the attachment of anions to cations in ion–ion reactions is commonly observed,242, 248 but remains a relatively little explored area in the chemistry of gaseous multiply charged ions. The reactions just mentioned add to the high dimensionality associated with charge permutation reactions in MS/MS. The continued expansion in the scope and application of charge permutation reactions can therefore be expected in the foreseeable future.


The authors acknowledge support by the Chemical Sciences, Geosciences and Biosciences Division, Office of Basic Energy Sciences, Office of Science, US Department of Energy under Award No. DE-FG02-00ER15105 and the National Institutes of Health under Grant GM 45372.


adenosine 5′-monophosphate


blackbird infrared radiative dissociation


collisionally activated dissociation


collision-induced charge separation dissociation


collision-induced dissociation


collision-induced dissociative ionization


charge residue model


deoxyribonucleic acid


dissociative recombination


critical energy for dissociation


electron affinity


electron capture-induced dissociation


electron capture dissociation


electron detachment dissociation


electronic excitation dissociation


electrostatic analyzer


electrospray ionization


flow afterglow


Fourier transform ion cyclotron resonance


gas-phase acidity


apparent gas-phase acidity


gas-phase basicity


apparent gas-phase basicity


hydrogen atom affinity


hot electron capture dissociation


ionization energy


infrared multiphoton dissociation


kinetic energy release




matrix-assisted laser desorption/ionization


mass-analyzed kinetic energy (spectrometry)


molecular mass


mass-to-charge ratio


neutralization–reionization mass spectrometry


proton affinity


polyaromatic hydrocarbon


peptide nucleic acid


polypropylene glycol


quadrupole time-of-flight


recombination energy


ribonucleic acids


surface-induced dissociation


selected-ion flow drift tube


sustained off-resonance irradiation collision-induced dissociation


tandem ionization mass spectrometry


external accumulation