ARTICLE pubs.acs.org/ac
Electron-Induced Dissociation of Singly Charged Organic Cations as a Tool for Structural Characterization of Pharmaceutical Type Molecules Jackie A. Mosely,†,* Michael J. P. Smith,† Aruna S. Prakash,† Martin Sims,‡ and Anthony W. T. Bristow‡ † ‡
Department of Chemistry, Durham University, South Road, Durham, DH1 3LE, United Kingdom Analytical Sciences, Pharmaceutical Development, AstraZeneca, Macclesfield, Cheshire, SK10 2NA, United Kingdom ABSTRACT: Collision-induced dissociation (CID) and electron-induced dissociation (EID) have been investigated for a selection of small, singly charged organic molecules of pharmaceutical interest. Comparison of these techniques has shown that EID carried out on an FTICR MS and CID performed on a linear ion trap MS produce complementary data. In a study of 33 molecule-cations, EID generated over 300 product ions compared to 190 product ions by CID with an average of only 3 product ions per precursor ion common to both tandem MS techniques. Even multiple stages of CID failed to generate many of the product ions observed following EID. The charge carrying species is also shown to have a very significant effect on the degree of fragmentation and types of product ion resulting from EID. Protonated species behave much like the ammonium adduct with suggestion of a hydrogen atom from the charge carrying species strongly affecting the fragmentation mechanism. Sodium and potassium are retained by nearly every product ion formed from [M þ Na]þ or [M þ K]þ and provide information to complement the EID of [M þ H]þ or [M þ NH4]þ. In summary, EID is proven to be a fitting partner to CID in the structural elucidation of small singly charged ions and by studying EID of a molecule-ion holding different charge carrying species, an even greater depth of detail can be obtained for functional groups commonly used in synthetic chemistry.
M
ass spectrometry is one of the most valuable analytical techniques employed by the pharmaceutical industry in the quest of compound characterization, to the extent that virtually every analytical or synthetic laboratory has access to this kind of technology. This is due, in part, to online compatibility with separation techniques such as gas or liquid chromatography, which allows even highly complex mixtures to be simplified ahead of the mass spectrometer. Sheer sensitivity toward compounds in very low abundance, particularly in the presence of more concentrated samples, as is typical of impurities arising from synthesis, makes up-front separation a necessity in many cases. Tandem mass spectrometry, the ability to further isolate desired precursor ions in the gas phase and subject them to conditions that induce fragmentation, undoubtedly adds that extra dimension of molecular information. For the pharmaceutical industry, the resulting fragmentation patterns from the tandem mass spectrometry of small organic molecules can provide a means to structural characterization, identification of impurities or degradation products and metabolites. Historically, characteristic fragmentation patterns from electron impact (EI) ionization have been established to the point that spectral libraries from EI ionized chemicals are routinely used to identify compounds “on-the-fly”.1 True tandem MS can be achieved in numerous ways including collision-induced dissociation (CID)2,3 and surface-induced dissociation (SID),4 infrared multiphoton dissociation (IRMPD),5,6 electron capture dissociation r 2011 American Chemical Society
(ECD)7 and electron transfer dissociation (ETD).8 Most common by far is CID, and recent studies into the reproducibility of CID product ion spectra have demonstrated that CID MS data is sufficiently reproducible across a range of different instrument types to allow instrument-independent CID libraries to be considered a viable approach to structural identification.9 ECD, a relatively new technique, has found its niche in the proteomics arena and routinely provides data, complementary to CID data, for the tandem MS studies of peptides10 and intact proteins.11 Whereas CID induces bond cleavage via the lowest energy pathways,12 crucially ECD has been shown to fragment some bonds leaving labile functional groups attached, thus able to locate and identify such modifications as phosphorylation,13 glycosylation,14 acetylation15 or sulphation.16 Subsequently, a great deal of research has been carried out into the fragmentation mechanisms occurring during ECD.7,17 The concept of ECD has been mainly considered for multiply charged species, whether for multiply protonated species18 or species containing a metal atom with a valence of n (where n > 1).1922 These studies have led to the proposal of two different mechanisms. The first mechanism, the hot hydrogen atom mechanism, was described by Zubarev and co-workers and suggests an incoming electron neutralizing Received: January 7, 2011 Accepted: April 7, 2011 Published: April 07, 2011 4068
dx.doi.org/10.1021/ac200045n | Anal. Chem. 2011, 83, 4068–4075
Analytical Chemistry
ARTICLE
Table 1. Number of Product Ions Observed Following CID in the LTQ and EID in the LTQFT and a Comparison of Product Ions Common to Both Techniques number of product ions observed > m/z 80 product ions common to EID and CID compound name
EID
CID (MS2)
(number common product ions; m/z of the common product ions)
[haloperidol þ H]þ
5
4
3; (m/z 123, 165, 194)
[sulfamethazine þ H]þ [diphenhydramine þ H]þ
9 6
8 2
6; (m/z 92, 108, 124, 156, 186, 213) 2; (m/z 88, 167)
>7
1
1; (m/z 152)
7
4
[reserpine þ H]þ
13
15
[raffinose þ Na]þ
>21
5
4; (m/z 203, 275, 305, 365)
10
2
2; (m/z 436, 454) 3; (m/z 168, 257, 283)
[diphenhydramine fragment]þ [caffeine þ H]þ
[terfenadine þ H]þ [tazobactam þ H]þ
2; (m/z 110, 138) 6; (m/z 195, 236, 365, 397, 436, 448)
13
6
[AZ_A þ H]þ [AZ_A þ 2H]2þ
>14 15
14 15
[AZ_B(i) þ H]þ
4
4
[AZ_B(ii) þ H]þ
2
2
[AZ_B(ii) þ Na]þ
7
11
4; (m/z 331, 400, 464, 515) 7; (m/z 165, 179, 224, 238, 272, 316, 167*)
1; (m/z 340) 5; (m/z 216, 310, 325, 338, 366) 1; (m/z 399) 2; (m/z 325, 280)
[AZ_B(iii) þ H]þ
18
12
[AZ_B(iii) þ Na]þ
7
5
[AZ_B(iv) þ H]þ
6
5
0
[AZ_C(i) þ H]þ [AZ_C(i) þ NH4]þ
>16 >18
2 3
1; (m/z 471) 2; (m/z 471, 516)
[AZ_C(i) þ Na]þ
>19
12
[AZ_C(ii) þ H]þ
12
4
[AZ_C(iii) þ H]þ
2; (m/z 125, 163)
6; (m/z 250, 263*, 278, 295, 357, 460) 2; (m/z 91, 93, 103, 120)
0
3
[AZ_C(iii) þ NH4]þ
>9
4
4; (m/z 211, 354*, 381, 398)
[AZ_C(iii) þ Na]þ
17
8
6; (m/z 234, 262, 278, 306, 342, 384*)
[AZ_C(iii) þ K]þ
10
7
3; (m/z 147, 154*, 358)
[AZ_C(iv) þ H]þ [AZ_C(iv) þ NH4]þ
10 12
10 2
2; (m/z 120, 164) 2; (m/z 408, 352)
[AZ_C(iv) þ Na]þ
15
4
2; (m/z 330,374)
[AZ_C(iv) þ K]þ
6
3
[AZ_C(v) þ H]þ
3
>7
[AZ_C(v) þ NH4]þ
1; (m/z 390) 3; (m/z 146, 164, 336)
6
4
3; (m/z 164, 336, 392)
[AZ_C(v) þ Na]þ
13
1
3; (m/z 120, 164, 358)
[AZ_C(v) þ K]þ
6
5
2; (m/z 164, 374)
> 330
> 190
Total Number of precursor ions studied 33
Average 3 peaks in common * suspected corresponding EID fragment 1 m/z higher than CID
the positive charge on a solvated carbonyl group of a peptide and inducing hydrogen atom transfer which results in cleavage of the NCR bond.23,24 More recently, the second mechanism, proposed by Turecek et al., describes electron capture that does not result in a mobile hydrogen atom.2527 Here, electron attachment to a peptide amide π* orbital, which is viable as such orbitals can be stabilized by Coulomb interaction with a positive charge, results in a strong base that undergoes proton transfer leading to bond cleavage within the amide (NCR). Both of these mechanisms only concern initial electron capture and primary fragmentation. Secondary fragmentation has been reported whereby nonergodic electron capture generates a radical species that, following radical migration, can lead to multiple radical rearrangements resulting in numerous cleavage sites.2830
It is, however, important to note that most mechanistic studies have involved peptides. Investigation into the electron interactions with single positively charged molecules in an FTICR MS has not been thoroughly examined as it has been assumed that charge neutralization will occur when a single positively charged ion captures an electron, resulting in a neutral species,31 although studies as long ago as 1979 clearly show this not to be the case if electron energy is sufficiently high.32 In this research, and subsequent studies, Cody and Freiser showed that interaction between radical cations and electrons can be used to provide valuable structural information.32,33 Budnik and Zubarev later showed that a singly protonated peptide could be further ionized by “fast” electrons and that subsequent capture of “slow” 4069
dx.doi.org/10.1021/ac200045n |Anal. Chem. 2011, 83, 4068–4075
Analytical Chemistry
ARTICLE
Figure 1. MS2 spectra for protonated diphenhydramine. (a) LTQ CID spectrum obtained with normalized collision energy of 20 and (b) LTQFT EID spectrum obtained with electron energy set to 20%. Collision energy and electron energy values were set as per manufacturers’ software.
Scheme 1. Dissociation of Diphenhydraminea
a
(i) Product ions common to CID and EID of 256 m/z. (ii) Additional fragments produced from EID of 256 m/z. (iii) CID of 167 m/z. (iv) EID of 167 m/z. All EID mass measurements were performed in the FTICR MS and are accurate to
