Photon of

Structures of Gaseous Ions by Infrared Multiple Photon Dissociation (IRMPD) Spectroscopy

St. John's

by

©

Khaclijeh Rajabi

A thcsi::; submitted to th' School of Graduate Studies

iu partial fulfillmmt of the requirements for the degree of

Doctor of Philosophy

Department of Chemistry Memorial University of Newfoundland

January 2010

Ncwfouwllrltlcl

Abstract

The technique of infrared multiple photon dissociation (IR.MPD) spectroscopy has shown its ability to determine the 3D (dimensional) structure of gaseous io11s. Usiug this technique, the structures of small biologica1ly relevant ions such a.c; <uniuo a.cicls and nucleic acid bases bound with a proton or metal cations in the ga.c;-pbase wC're determined. The necessity of employing computational methods to a.naly:0e and in- terpret the experimental data has been demonstrated. llowevcr, the computational results must be analyzed with extreme caution to prevent a.ny iuconcct condusious. Among the parameters one has to consider when dealing with the COH!plltatiomt.l rc- ::>ults is the experimental method used to obtain the cl~:~ta. For example, the simnlc1t('d lR spectra for only two of the four lowest-energy protona.tecl adenine cliuwr iso111crs were ::>imila.r to the experimental IR.MPD while, ba::>cd on gas-phase calculations, thC' four lowest-energy structures were almost isoenergetic. Since the io11s were prod UC'C'd

by c~lcctrospray iouizatiou (ESI) from the solution phase, thC' effect of water fts a solvent was considered by applying two independent computational approa.cl!cs lo take solvatio11 effects iuto Ftcconut. Polarizable cout.iuumn model (PC~l) cakula.t.ions as well a.-; microsolvation with live explicit water molecules calculaLions showed that.

water only preferentially stabilizes these two observe I isomers, consist.eut with the interpretation of the IRMPD spectra. Th' results suggest great caution is required when using ga.c;-pha.sc ca.lcula.tions to predict the structmes of ga.<;cous ions borne i u

·olution by ESI.

11

To invest iga.Le the influem:e of sol vent. ou the st.rnct.ure of ions, 811 experimcu- taJ method wa1:> developed to produce solvated ions in the gas-phase. These .-olvated iom; then were investigated by IRMPD spectroscopy and blackbody infrared radiative dissociation (BIRD) to obtain kinetk and thermodynamic data. The solvation of clcc- trospraycd ions occurs in the accumulation/collision hexapolc of a. hybrid quadrupole- Fourier transform mass spectrometer (Q-FTMS) by introducing the solvellt iuto the collision cell. The most 1:>ensitive parameter1:> based on our experience were the colli.-iou energy in the hexapole, the pressure of both collision gas and solvent in the hcxapole, ion accumulation time, and the chemical nature of the species. This method wr~s suc- cessfully applied to adenine and thymine cluster ions to produce multi ply hycl rated ious. The 1:>tructmes of singly hydrated ion1:> were determined by lRMPD spectroscopy.

Due to the importance of m-DNA (metala.ted-DNA), an attempt to investigate the structure of singly hydrated thymine zinc ion-bound dimer was initiated. lt was found that thymine loses one proton in the presence of zinc. Therefore, the [(Thy2 -

ll)-Zn-(ll20 )]+ clu ter was singly charged. Solely comparing the IRMPD spcctrulll in the 3100-3850 cm-¹ fwd sinmlftt.c!cl JR spcct.ra, WI-tS !lOt snfficicll t. to assig11 o11ly one structure to the observed spectrum. Based on thermochemical values, the two lowe1:>t energy structures were aH ·igned as possible structures under the experimental conditions. In the most stable structure, the water is directly attached to the :,~,inc in the zinc ion-bound dimer in which the Zn2+ is shared betweeu the two thymines at.

N304 sites. Furthermore, compntational data suggested that recording au InMPD

spectrum in the 1800-2800 cm-¹ region might be useful to distiuguisb between the two lowest energy structures. Therefore, recording the spectrum for the [ (Tbyr 11 )- Zn-(1120)]+ cluster in this lower energy region is part of the fnturc work.

IV

- - - -- - - - - - - -

Acknowledgements

I have greatly enjoyed the yea.rs 1 have spent at Memorial University of ewfound- lancl (MUN) and it will be hard to express my appreciation to the ma.ny people to whom l arn indebted for such unforgettable experiences. This rcsca.rch project would not have been feasible without the support of many people.

First of all, I would like to express my deepest sense of gratitude to my wondcrfnl supervisor, Professor Travis D. }'ridgen, for his patient guidance and cxccllcnt. advice throughout this study. I wa.s very fortunate to have him as my supervisor. Tn1vis hn.s lwcm a.lmnda.nt.ly helpful a.ud off'crcd inva.lua.blc ~tssist.;:wcc, support and Cll<'OIH-

agement in terms of both academic and non-academic matters and becan1e more of a friend than a supervisor. I appreciate his vast knowledge and skill in mauy <Hca.s and his assistance in writing papers. Also I would like to tlwuk Lis~:~, Travis's wife (boss) and a super-mom, for her unrestricted hospitality and all the fnu we had at their place.

Deepest gratitude is also due to the members of the supervisory committee, Prof.

Christopher Flinn and Prof. Robert W. Davis without whose knowledge ~1ucl assis- tance this .study would not have been successful. I would like to thank Dr. Flinn for abo being a teaching .supervisor for the graduate program in teaching (GPT) pro- gram and trusting me to take charge of three sessions of his course. A very special thanks goes to Dr. Davis who has been the head of the dcpartme11t during most of my graduate studies. Ile is probably one of the most phil~:~nthropist l have ever met.

I doubt that l will ever be able to fully convey to him my appreciation.

Also, l would like to thank my internal and external examiners, Prof. Raymond Poirier, Prof. Robert Helleur, and Prof. John Klassen for carefully reading this thesis as well as providing me with the construcbve comments.

Special thanks a.lso to all my friends, especially our research group members:

Os~nna, Mike, Julie, Elizabeth, Chad and Amench for sharing the literature and invaluable assistance. l am not forgetting all my best friends who always been Lllcrc for me.

I would also like to convey thauks to the warm a.ud lovely stFiff at the cht'mist ry department who welcomed me as a frieud during tbe different stages of IllY life a.s it

graduate student. Just to name a few, I would like to thank Professor Peter Pickup, Rosalind, Viola, Teresa, Mary, Gerry, and Marlene for their admini tra ti vc in fonuH- tioll. In addition, I would like to thank them for always positive compliments, funny

little~ storic~s, condolc~nccs wllcll I needed. them, teltchillg me Nc·wfie (sill<'<' T liv<'d ill Newfoundland for four years, I gues · 1 can authorize myself to use this word) c·x- pressions b'y, as well as suggesting inventive methods to survive in the cold s11Ch HS

putting hot water bottles under the coat.

l wish to express my love and gratitude to Saeicl, my beloved husband. llc lws been always there for me with his endless love, encouragement, patience ancl ~mpport.

Of cours<\ I can never cxpn~ss my grc~at apprcciatiou for his sacrifice for giving 11p

his settled life in lra.n and immigrating to another continent for me. lle has bc<'ll

Vl

very sensitive to the quality of my academic work as well and supported me in every

a~pects. He patiently listened to my presentations before l gave them, providing me with excellent comments and feedback. I feel incredibly privileged to ha.vc him as my lmr-;band.

Although Newfoundland weather was not completely in-line with my precli tions, l feel very lucky to have become acquainted with this part of world. The fact tlw.t St.

John's, the provincial capital of Newfoundland, ha~ the thir I mildest winter i11 CamHla because of its humid weather is in agreement with my geographical expcctatious bnt it also ha~ the most snowy, windy and cloudy clays in Canada. St. John's, as the oldet>t (----.~1500s) English-founded city in North America, has an cxccptiow-tlly W<ll'lll

and frien lly population. I take this opportunity to pass my special tlnnks to the Mouland family. In particular, my landla.dy Wilma, who sadly passed awHy from cancer a.lmost a year ago, was a very plea~ant and loving person and treated me as one of her family members.

Of course, none of this would have happened if uot for my parents, Agltclas and Ali, especially my mom. They always provided me with the freedom of dwicc and supported my interest in s ience, even when that interest led nw to distaJJt shores.

- To Saeid, The love of my life -

Vlll

"A scientific truth docs Hot triumph by convincing its opponents a.nd lllctking Lhclll sec the light, but rather because its opponents cventua.lly die and a new gcncra.Lion grows up that is familiar with it."

- Max Planck -

Contents

Abstract ¹¹

Acknowledgements ^v

List of Tables ^XVI

List of Figures ^XIX

Abbreviations ^XXX

Co-authorship Statement ^XXXI

1 Introduction ¹

1.1 Structmal Determination of Gas-phase Ions by Mass Spectrometry 1 1.2 Mass Spectrometric-based Techniques for Structural Studies of Gaseous

Ions

1.2.1 Collision-Induced/ Activated Dissociation (ClD /CAD) Icthods ~1

1.2.1.1 Low-energy CID . . . . . . . . . . . . . ^G

X

1.2.1.2 lligh-encrgy ClD 1.2.2 Thermochemical Methods

10 1 1.2.2.1 Threshold Collision Induced Dissociati011 (TClD) 18 1.2.2.2 Ion-Molecule (I-M) Reacbons . . . . . . . 19 1.2.2.3 Pulsed-Electron Iligh Pressure Matis Spectrometry (PU-

PMS) . . . .. . . . . 1.2.2.4 Blackbody Infrared Dissociation (BlilD) 1.2.3 Spectroscopic Methods . . . .

1.2.3.1 Matrix h;olation (MI) Spectroscopy . 1.2.3.2 Vibrational Pre-Dissociation (VPD)

22 2G 31 31

35

1.2.3.3 Infrared Multiple Photon Dissociation (lR.t\IPD) 38 1.3 Computational Chemistry to Determine Ion Structmc . . . . ,11

Bibliography 45

2 Methods 52

2.1 Experimental Method: Infrared Multiple Photon Dissociation (lnMPD) G2 2.1.1 Mechanism of IRMPD

2.1.2 Experiments Set Up 2.1.3 Ion Traps . . . .

2.1.3.1 Quadrupole or RF Ion tra.ps (QIT)

G9

62 63

63

2.1.3.2 Fomicr Transform Ion Cyclotron Resonance (FT-lCH.) 70

2.1.4 lR Laser Sources . . . . 2.1.11.1 lR Free Electron Laser (FEL)

2.1.4.2 Optical Parametric Oscillator (OPO) 2.2 Computational Methods

2.2.1 Ab initio Method

2.2.2 Density Functional Theory (DFT) . 2.2.3 Ba!::iis Set!::> . . . .

2.2.4 Calculation Procedure

2.2.5 Comparison of Experimental IRMPD and Calculated

m .

^Bancl

76 77

79 86 87

91 93 91

Intensities and Frequencies . . . . . . . . . . . . . . . . . . . . 96

Bibliography 101

3 Structures of Aliphatic Amino Acid Proton-Bound Dimers by IRMPD Spectroscopy in the 700 to 2000 cm- ¹ Region

3.1 Introduction 3.2 Methods . .

3. 2.1 Experimental 3.2.2 Computational 3.3 Re!::iults and Discu!::lsion

108 108

112 112

113 JH 3.3.1 Elcctrospray and IRMPD of Protonatcd Amino Acid Dimcrs 11¹1 3.3.2 Glycine Proton-Bound Dimcr . . . . . . . . . . . . . . . 115

Xll

3.1

3.3.3 3.3.4 3.3.5

Alanine Proton-Bound Dimer Valine Proton-Bound Dimer

Alanine/glycine mixed proton-bound dimer .

3.3.6 Comparison of the Aliphatic Amino Acid ProtoJJ-J3ound Dimcr Spectra

Conclusions . .

Bibliography

120 123 126

131 133

137

4 The Structure of the Protonated Adenine Dimer by IRMPD Spec-

troscopy and Electronic Structure Calculations 1.1

4.2

4.3

Introduction

Methods . . 4.2.1

4.2.2

Experimental Computational Rcsul ts and Discussion

4.3.1 Computed Structures and Thermochemistry of Adenine Prot()ll-

4.3.2

1.: . t:s

Bound Dimcrs . . . . . IRMPD Spectroscopy .

Acpwons Solvation EfFects ⁰¹¹ Dinwr St.ahilit.ics .

142 112

1 J H8 H9 1G1

151 155 161 4.3.1 Comparisoll of IRMPD Spectrum With lligher-Encrgy lsomC'rs l69 4.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

4.5 Supporting Information . . . . . . . . . . . . . . . . . . . . 172

Bibliography 173

5 Solvation of Electrosprayed Ions m the Accumulation/ Collision Hexapole of a Hybrid Q-FTMS

5.1 Introduction . 5.2 Experimcuta.l

5.3 Results and Discussion 5.1 Conclusions . . . . ..

Bibliography

179

179 182

1 11 197

198

6 Structures of Alkali Metal Ion-Adenine Complexes and Hydrated Complexes by IRMPD Spectroscopy and Electronic Structure Cal- culations

6.1 Introduction . G.2 Methods . . .

6.3

6.2.1 Experimental

6.2.2 Computational Results and Discussion

200

201 205 205 206 207 6.3.1 A Comparison of the ExperimentallRMPD Spectra. of (C5ll5N5)2Li+,

6.3.2

XIV

208 211

6.3.3 6.3.4

6.4 Conclut>ions . . . . .. . 6.5 Supporting Information .

Bibliography

7 Mctalatcd Thymine 7.1 Introduction .

7.2 Methods

7.2.1 Experimental 7.2.2 Computational 7.3 Results and Discussion 7.4 Conclut>ions

Bibliography

8 Conclusions and Future Aspects

Appendices

VITA

. . . 219

. . . . 221[

22 230

231

236

. . 236

2tl¹1 2/11[

2116 2tl7 2G,J

256

259

266

275

List of Tables

2.1 A comparison of the ratio of _IJ3(Ih0): 1J1(Il20) intensities from a.ct.ion spectra. and the enthalpy change for the lowest dissociation dt<umcl for cation water complexes (reference 71 and references herein).

3.1 The experimental and calculated IR absorption wa.venumbcrs in cut-¹ for glycine proton-bound dimers. sh=shoulder, gl=glyciue labelled 1

99

in Figure 3.1, g2=glycine labelled 2 in Figure 3.1. . . . . . . . . . . . 118 3.2 The experimental and calculatedlR wavenumber positions iu cm- ¹ for

ala.nine proton-bound dimers. sh=shoulder, al=a.lanine labclkcl l iu Figure 3.3, a2=ala.nine labelled 2 in Figure 3.3. 122 3.3 The experimental and calculated lR absorption frequencies in cw-¹

for valine proton-bound dimers. vl=va.linc labelled 1 in Figure 3.11, v2=valine labelled 2 in Figure 3.4. . . . . . . . . . . . . . . . 12r:

XVI

,---~-----

4.1 Binding energict> (kJ mol-¹) of the four lowct>t-energy isomers of the proton-bound adenine dimer relative to N9H adenine and N1-protouatcd- N9H adenine. The experimental value from reference 58 is 61-1(500 l\:)

= 127±4 kJ mol ¹. . . . . . . . . . . . . . • . . . . . . . . . . . . . . 15,J 1.2 Computed relative cuthalpics for various adenine proton-bound climer

structures. . . . . . . . . . . . . . . . . . . . . . . . . . 156 4.3 Table of assignments for experimental IRMPD bands for the adenine

proton-bonnd dimer and predicted bands for structures A tlmmgh D. 1G3 4.1 Relative enthalpies (relative free energies in parentheses) of ::;olvatcd

and unsolva.tcd adenine proton-bound dimers A, B, C, D. . . . . . . . 16r:

1.5 298 K relative enthalpies (and free energies) for singly solvated a.clcninc proton-bound dimers. Sec Figure 4.6 for structure::;. . . . . . . . . . . 16

5.1 Summary of BIRD rate constants for solvent los::; for (Adc)2(1120)71K+

and (Ade)2(Cil:.lOil)₇₁K+ clusters at 298 K . . . . . . . . . 192

6.1 Comparison of 298 K MP2/6-31l++G(2d,p)/ /B3LYP /6-31+G(cl,p) relative Gibbs energies (in kJ mol-¹) for the lowest-energy amino" aud imino⁰ structures of (C5115N5)M+ and (C5ll5N5)M+(Il₂0). . . . . . . 211 6.2 Bands (in cm-¹) observed for (C5ll5N5)2Li+, (C5115N5)Li+(lh0) aud

(C5l15Ns)2Li+(l-f20) compared with neutral adenine and the a.clcuinc proton-bound dimer. . . . . . . . . . . . . . . . . . . . . . . . . . . . 229

7.1 Vibrational frequency as~ignmcntr:; for [(Thy2-ll)-Zn-(ll20)]+. Observed frcqucncier:; arc obtained form the IRMPD r:;pcctrum and the calculatccl valuer:; for r:;tructurc A aucl B arc extracted from the predicted frcqucn- cicr:; a.t B3LYP/G-31+G(d,p)/ /LANL2DZ r:;caled a.t 0.956. . . . . . . . 253

XV Ill

~-~---~~~---~---~

List of Figures

1.1 ClD spectra of a) rotaxane and b) axle-wheel complex when collided with argon. Note that no signal is observed for the dcprotonnt.cd axle in the rotaxa.ne spectrum while it is prominent among the products generated from the axle-wheel complex. Figure reprocl uced from lnt.

J. Mass Spectrom. 2004, 232, 249 with permission from Elsevier. . . . 8 1.2 ESI-MS² of a tetra(aryl)benzidine derivative aJoug with fraguJcnt.nLion

reactions of the precursor ion [M]+• at m/z 884.5. Fignrc rcprocluccd from Eur. J. Org. Chem. 2007, 5162 with permission from John Wiley a.nd Sons Inc. . . . . . . . . . . . . . . . . . . . . . .

1.3 Schematic of a typical thr ^'C sector instrument of BEE geometry. 11 1.1 Metastable ion (Ml) mass spectrum of ionized 2,3-pcnta.ncdione. Fig-

ure reproduced from lnt. J. Mass Spectrom. 2006, 249-250, 222 with

permission from Elsevier. 13

1.5 CID mass spectra of: (a) the middle, (b) the edge component of tlJC composite (M-28)+• peak and (c) ionized 2-butanone. Figure repro- duced from Int. J. Mass Spectrom. 2006, 249-250, 222 with permissio11 from Elsevier. . . . . . . . . . . .

1.6 Schematic overview of the guided ion beam ta.11dem mass spectrolllct.er.

FS represents focusing stages. Figure reproduced from J. Chcm. Phys.

1985, 3, 166 with permission from American Institute of Physics. . . 18 1.7 Proposed structure of (M+ll+Na)²+ of gramicidin S (cyclo[-Pro-VHl-

Orn-Leu- D-Phe-]2) using molecular mechanics in which the Na+ i::>

attached to the exterior surface of the peptide. Figmc rerroclllccd from J. Am. Chern. Soc. 1996, ll8, 202 with permission from tlte American Chemical Society. . . . . . . . . . . . . . .

1.8 Structures and geometric properties of FllF-... (ll20h cluster. Oxygc11 in red, F in Blue and II in gray color . . . ⁰

1.9 !late constant for loss of a water molecule from S042-(lb0),, n=G-17, with BIRD at 21

oc

as a function of n. The error bars represent 1

22

25

standard deviation of the mea.c::>ured rate constant. ^{0 • • • • • • • • • 0} 2 1.10 Molecular mechanics lowest energy structures of (a) S04 2

- (lhO) ¹² <1 ncl (b) S0₄²- (Ib0)₆. Figure reproduced from J. Phys. Cllem. A 2003, 107, 10976 with permission from Amcricctn Chemical Society. . . . . . 30

XX

1.11 Isolation of molecules of a reactive species (black circles) by a rigid host lattice (shown as open circles). . . .

1.12 Schematic layout of MS-Ml instrument. Figure reproduced frotu J.

Phys. Chem. A 199 , 102, 3162 with permission from American Ch('lll- ical Society. . . . .

1.13 Vibrational prcdissociation spectra of water proton-bonnd dil!ler tagged with Ar (a) and Nc (h) messenger atoms. NB corresponds to non- bonded. Figure is reproduced from Science 2005, 308, 1765 with per- mission from The American Association for the Advanccmcut. of Sci- cnce and J. Cbem. Phys. 2005, 122, 244301 with permission frolll American Institute of Physics . . .

l.ltl An ion mollility spectrometer with a traditional t.inJC-of-flight. Fig- ure is reproduced from Appl. Spectrosc. Rev. 2006, tll, 323 wiLlt permission from Taylor and .Francis . . .

1.15 Energy profilcfor int('rconwrsion of the CH₃0ll +• (1) and ·c112 +oJ

h

⁽²⁾

showing t.bcir rt'lnJ.ivc st.n.hili t.i<~s (OK. modi fled G2). Figar<' is r<'pro- duced from J. Am. Chern. Soc. 1996, 118, 6299 with permission from

32

37

10

American Chemica.l Society. . . . . . . . . . . . . . . . . . . . t12

2.1 Schematic view of the experimental apparatus for low-intensity C\11/

IR laser radiation. Figure reproduced from J. Am. Chem. Soc. 1979, 101, 5503 with permission from Americau Chemical Society. 53

2.2 (a) Photodissociation spectra of (C2ll5)20ll+ and [(C2115)20]21J+ and (b) of (C₂D₅)20D+ over the C02 laser spectral range. Dotted line is the infrared absorption spectrum of (C2D5)20 at 1G Torr. Figure reproduced from J. Am. Chern. Soc. 1979, 101, 5503 vvith permission from American Chemical Society. . . ^{0 0 • • • • • • • • • • • • • • • •} 55 2.3 (a) Schematic representation of the lRMPD mechanism and (h) vibm-

tiona! state density in a molecular ion. . ^{0 • • • • • • • • • • • • • • •} GO 2.1 Quadrupole ion trap. (a) an open array of the three clectroclcs. (h)

cut in half along the axis of cylindrical .symmetry (c) schematic of the 3D ideal ion trap representing the asymptotes, r₀ ancl llo. Figure reprodncecl from J. Mass Spectrom. 1997, 32, 351 with permission from John Wiley and Sons lnc.

2.5 Stability diagram. Ions wirh a" and q, valnes that fall in the region

2.6

2.7 2.8

A and D will have a stable trajectory and be trapped within the QlT

ma~s spectrometer. Figure reproduced from J. Mass Spect.rom. 1!)97, 32, 351 with permission from John Wiley and Sonl::i Inc.

Mass analysis by ion ejection at the ::;ta.bility limit.

Ion motion in the ICR cell. . . . ^{0 • • 0 • 0 0 • • •} Schematic reprc.-entation of a cylindrical lCR. cell.

XXll

G8

G9 71 72

2.9 A schematic of the instrumentation for ESI-ICR along with pressure at different stages. TP and RP arc abbreviations for turbo pump and rough pump, respectively. The pressures shown in each stage ^<He iu mbar.

2.10 Steps in ion detection in a FT-ICR MS. B rcprcscuL the lllfl.guctic fidel. 76 2.11 A schematic of a FEL. . . . . . . . .

2.12 A schematic of the YAG:OPO laser. .

2.13 The four-level system energy diagram of the d:YAG atomic transition priuciplcs at 1064 nm.

2.14 Simplified diagram of a Nd:YAG Q-switched laser.

2.15 Schematic of a Brukcr Apcx-Qe 7.0 T FT-ICR mass spectrometer, laser

78

80

2 83

photons arc shown with an arrow entering the ICR cell. . . . . . . . . 8·1 2.16 A schematic of a Brukcr Esquire 3000 QIT ma~s spectrometer cottplcd

with FEL; la~er photons arc shown with <tn arrow entering the QlT.

Figure reproduced from l11t. J. Mass Spectrom. 2006, 2511, 1 wi t.b

permission from Elsevier. . .

8G

2.17 electrons with the same spin (a) and electrons with opposite spins (b). 90

3.1 IRMPD spectrum of the glycine proton-bound dimcr as well a~ the B3LYP /6-31 +G ( d,p) predicted spectra for the four lowest-energy stntc- tures. B3LYP/6-311+G(2df,p)/ /B3LYP/6-31+G(cl,p) relative free eu- crgies compared to A arc in parentheses. . . . . . . . . . . . . 116

3.2 Cctkulat.cd infrared spectrc:l for the three lowest-energy structures for the glycine proton-bound dinwr at B3LYP /6-31 + , (d,p) also ~howing the 2000-3 00 cm-¹ region where it may be more likely t.o be able to distinguish between isomeric structure·. In the inset the cxpcrimcutal JRMPD spectrum (in black) from Oh et alY⁸ is cowparecl wit.l! the

colllpll ted ::>pcctra. 120

3.3 lR IPD spectrum of the alanine proton-bound climcr as well as the B3LYP /6-31 +G ( d,p) prcclietcd spectra for the four lowc~t -<'JWrgy st rue-

tun~~. . . . . . . . . . . . . . . . . . .. . . . 121 3.1 lRMPD spcctnun of the valine proton-bound climcr HS well as t.!Ie

B LYP /6-31 +G ( d,p) predicted spectra for the th rce lowest-energy st rue- tun's. . . . . . . . . . . . . . . . . . . . . . . 121 3 ^{.;) r:} lR 1 PD spectrum of the bet erogeneou::> alanine/ glycine prot 011-bcmud

di mer aucl the B3LYP /G-31 +G ( cl ,p) predicted spectra for t.!Ic two lowest.- energy structures. B3LYP /6-3ll+G(2clf,p)/ /B3LYP /G-3l+G(cl,p) rei- aLive free energy compare I to A i::> in pmcnthescs . . .

3.G B3LYP/G-31+G(d,p) calculate I infrared ::>pectra. for the two lowest- cnngy structures of the heterogeneous alanine/ glycine proton-bound climcr iu t.hc 2200-3800 em ¹ region.

3.7 lRMPD SJWCtra. of all four proton-bound dimers in tl1e 700-2000 CL11- 1

127

129

region which show:> the ·imilarity of the spectra in this rcgiOJI. . . . . 132

xxiv

3.8 Calculated IR ::;pectra for glycine, alanine, alanine/glycine, and valine proton-bound dimcr::; at B3LYP/6-31+G(d,p) ::;bowing that they arc predicted to ha.ve very ::;imilar ::;pectra in the 700-2000 em- ¹ rcgiou

(a)

but that they might be di::;ceruiblc in the higher-cucrgy region (b). . . 13·1

1.1 B3LYP /6-31 +G ( d,p) ::;tructmes of the four lowe::;t-encrgy protou-bonncl adenine dimers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 1. 2 B3LYP /G-31 +G ( d,p). tructurc::; of eight high-energy proton-bound ade-

nine dimer::;. . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 4.3 Comparisou of the experimentallRMPD spectrum of the a.dcniuc proLon-

bound dimcrs with the B3LYP /6-31 +G ( d,p) predicted IR t>pcetrn fm ::;tructures A-D (structuret> in Figure 4.1). . . . . . . . . . . 15 4.4 Comparisou of the experimental IRMPD spectrum of the adenine proton-

bound dimer (the black in the bottom) with the gas-phase spccLrnm of neutral adenine (the black on the top). Also shown arc prcdict.cd ::;pcctra for the N9Il tautomer of neutral adenine. . . . . . . . . 161

11.5 B3LYP /6-31 +G ( d ,p) ::;tructures of the four lowest-energy adenine proton- bound. clin1crs Jnicrosolw1tc'd wit.h five w~1t.E-~r molccnlcs. . . . . . . . . 166 4.6 The four B3LYP /6-31+G( d,p) singly microsolvatccl structure::; each for

A and C. Microsolvatcd ::;tructurcs for B and D arc in Appendix 1. . . 170

4. 7 Comparison of the experimeutallRMPD spectrum of the a.cleuiue proton- bound dimers with the B3LYP /6-31+G(cl,p) predicted IR. spcctn-1 for structures E-L (structures in Figure 4.2). . . . . . . . . . . . . . J71

5.1 Schematic showing the ::;omce, Qh region, ion transfer optics, and ICR cell for the Bruker Apex Qe 70 FTMS.

5.2 The fraction of solvated Thy2Li+ a.nd total ion intensity versus (a) collision voltage and (b) hexapole accumulation tinte. The .. fi tt cd"

lines arc merely to guide the eye. Arrows indicate the ordinate to which the data belong. . . . . .

183

18G 5.3 (a) ESI mass spectra of a 0.1 mM CuCb solution containing rv0.1m~J

thymine and a few drops of 0.1 ml\I adenine in 18 MD water. (lJ) ^Scl.llH' experiment except with water vapor in the hexapole accumnlHtion cell. 188 5.4 (a) BIRD plots for sequential loss of solvent hom (Acle)2(XhK+, ((a)

X=lhO, (b) X=Cll₃0ll). Rate constants for sequential solvent. loss a.re in Table 5.1. . . . . . . . . . . . . . . . . . . . . . . 191 5.5 (a) ESI mass spectra showing the formation of (Acle)2Li+(lJ20) by

reactiug (Ade)2Li+ with water vapor in the accumulation hcxapolc.

(b) Mass spectrum showing the effect of a.lJsorptiou of the OPO lasn tuned to 3350 cnc ¹ for 0.5 s.

XXVI

19 J

5.6 IRMPD spectra of (Aclc)2Li+ and (Acle)2Li+(II₂0) in 2500-1000 cm-¹ region. Also show11 arc B3LYP /6-31+G(d,p) computed structmes and infrared spectra. of the lowest energy structure for (Adc)2Li+ and two lowest energy structures for (Ade)2Li+(Il₂0). . . 106

G.l Comparison of the experimental IRMPD spectra. of (C₅ll5Ns)2Li+,

6.2 ExperimenLa.l lRMPD specLra of(C₅H5N5)M+ (where M=K,Cs) as wl'll as B::!LYP /G- 31+G(d,p)-computecl lR spe<.:Lra (a) for Lhe six-lowesL energy sLrucLutTS (h). The

298 K B3LYP /6-3l+G(d,p) as well a,; MP2/6-3ll++G(2d,p)/ /B3LYP /G-3l+G(d,p) (in parenLheses) relative Gibb~; energies (ld mol ¹) are provided for each ](+-bound

structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 6.3 Comparison of the experimental lRMPD spectrum of (C5115N,,):zLi-¹,

with B3LYP /6-31+G(d,p) calculated spectra. Letter designations cor- rei:ipond to structures in Figure 6.4. . . . . . . . . . . . . . . . . 21G 6.1 B3LYP /6-31+G (d,p) ca.lcula.ted structures for (C5Il5N5)2Li+. B3LYP /G-

31+G(d,p) and MP2/G-31l++G(2d,p)/ /I33LYP /G-31+G(cl,p) (in parcn- theses) Gibbs cncrgic::; (kJ mol-¹) arc also provided . . . . 217 G.5 lRMPD spectra of (C₅H₅N₅)2M+, M=H, Li, Na, a.ncl K. Also shown

arc the B3LYP /6-31 +G ( d,p) predicted spectra. for the lowest-energy structmc for each dimcr. . . . . . . . . . . . . 218

6.6 The cxpcrimcntallRMPD spcctmrn for (C5115N5)Li (lhO) iu the 3300- 3800 em ¹ range along with the five lowest-energy structures and their computed IR spectra .. The 298 K B3LYP/6-31+G(cl,p) ^ciS well m;

MP2/6-3ll++G(2d,p)/ /B3LYP /6-31+G(d,p) (in parentheses) relative Gibbs energies (kJ mol-¹⁾ arc provided for each structure. . . . . 220 G. 7 Comparison of the experimental lRMPD spectra and the predicted

IR spectra for the B3LYP/6-31+G(d,p) lowest-energy structun·s In spectra for (C₅11₅N₅)M+(lb0) where M=Li, Na., and K. . . . 223 6.8 The InMPD spectrum for (C5ll5N5)2Li+(Il20) in the 2500-¹1000 cm-¹

region along with the predicted IR spectra for the eight lowest-eucrgy strnctures composed of A 7 A 7 and A 7 A9 adenine tautomcrs (for strnc-

tun~s and energies sec Figure 6.9) . . . .

6.9 The lowest-energy calculated structures for (C5H5 5)2Li+(ll:z0). The 298 K B3LYP/6-31+G(d,p) and MP2/6-31l++G(2d,p)/ /l33LYP/6- 31+G(d,p) (in parentheses) relative Gibbs cuergics (k.J mol-¹⁾ arc pro-

225

vided for each structure. . . . . . . . . . . . . . . . . . . . . . . 227

7.1 Comparing the experimental (a) and calculated (b)

m .

spectra for structure l3 of [(Phc-11) Zn (Phe)]+. Figure reproduced from J. Am.

Chern. Soc. 2006, 128, 517 with permission from American Chemical Society.

XXV Ill

239

7.2 The proposed structure for dACGT and three Zn²+ species. Figure reproduced from Int. J. Mass Spectrorn. 2001, 201, 55 with permission from Elsevier. . . . . . . . . . . . . . . . . . . . . . 2110 7.3 Proposed structure of [Zn(9-EtA-N7)Cl3](9-EtAll). 211 7.1 The three lowest-energy structures of thymine Zn(ll) calculated at.

B3LYP /6-311 +G(2df,2p). Sec text for the relative energetics. . . . . 2,12 7.5 FT-ICR MS instrument coupled with an OPO/ A laser at Memorial

University. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2,15 7.6 IRMPD spectra for [(Thy₂-ll)-Zn-(lb0)]+ (in black) and [(Thy3-ll)-

Zn]+ (in red) in the 3100-3850 cm-¹ region. . . . . . . . . . . . . 218 7.7 A comparison of the experimental IRMPD (in solid black) and cHktt-

latecl lR spectra for the [ (Thy2-Il )-Zn-( lbO) ]+ four lowest eilcrgy. TIH' relative free energies (kJ mol-¹⁾ at 298 K compared to structnrc A arc provided for each structure. . . . . . . . . . . . . . . . . . . . . 251 7.8 The predicted lR ·pectra for structure· A, B, C, and D (a::; shown i11

Figure 7.7) in the 1 00-3000 cm-¹ region. . . . . . . . . . . . . . . 25~1

Abbrev i ations

BIRD blackbody infrared dissociation ClD collision-induced dissociation

CLIO FEL centre la.ser infrarouge cl'Orsa.y free electron la:-;cr DFT density functional theory

ECP effective core potC!ltial ESI elcctrospray ionization

FELIX FEL (free electron laser) for infrared eXperiments

FT-lCR MS Fomier transform ion cyclotron resonance mass spectrometry lRMPD infrared multiple photon dissociation

l VR intramolecular vibrational energy redistributiou KTP pota.ssium titauyl pho ·pha.te, KTiOP01

MALDl matrix-as::;isted la,<;cr desorption/ionization tn-DNA mctala.ted-DNA

MS-Ml ma.ss spectrometry-matrix isolation OPO /A optical parametric oscillator /amplifier Ql T q uaclru pole iou Lrap

RF radio frequency

TCID threshold collision-induced dis ·ocia.tion YAG yttrium aluminum garnet

XXX

--- ---

Co-authorship Statement

Mot:>t of the experiments were performed at Univert:>ite de Paris Smlll where there arc facilities for lRMPD spectroscopy studies. Recently at Memorial UJJivcrsity of Newfoundland (MUN) we have built a laboratory to study the structures aucl ener- getics of gaseous ions in an FTICR mated to a tnna.blc IR la~er. All <'xpcrimcnts arc conducted by the principal author and co-authors. The experimental and conlpllLa- t.iowtl daJil. was analysed by the principal author ~~Jl(l t.hc first draft of pn.p<'rs wa.s written lJy the principal author. The fiua.l vcr:-;ions of the papers were writL<'ll witll the help of Dr. T. Fridgen. lu Chapter 4, the micro-solvated model calculn.tim1:-; at the donllc-hybri i B2P3LYP density functional were performed lJy Ms. K. Theel and Dr. G. I3cran at the University of California, Riverside. Most part of experiments in Chapter 5 were conducted at Bruker (at Billerica) where the principal Httthor was trained by Dr. C. Berg and Dr. l\1. Easterling to perform BIRD expcrimcJJt.s.

Chapter 1

Introduction

1.1 Structural D etermination of Gas-phase Ions by Mass Spectrometry

Determining three dimensional (3D) structures can lead to a fullunclerstaudiug of a molecule's chemical properties as in chemistry it is believed that "structure controls function". One or two dimensional information gained from chemical formula and molecular drawing, respectively, may not provide all the necessary information to be able to characterize chemical behaviours whereas by obtaining 3D ::;tnl('tmcs om' can fully describe a molecule in terms of its atoms connectivity ancl their oricnLHt.ion in space as well as its bond length and dihedral augles. Once the tructmc is knowu, the chemical behavior of the species can be understood. Protein sccomlH.ry (2° regularly repeating local structures stabilized by hydrogen bonds) and tertiary (3" t.l1c spatial

1

relationship of the secondary structures to one another) structmc bas an impact on nutritional quality, fermentation and degradation behavior in botb llllman ancl mammals. ¹ For example, a high percentage of ,8-sheets usually reduces the access of digestive enzymes to the protein, causing poor digestibility and, as a result, low protein value. Such information is only available from a precise description of the structure.

Mass spectrometry is the leading technique for determining gaseous ion st.ruc-

tun~s. l\Iass spcctrolllct.ry has significantly cOJltribut.ed to determining ··prinHtry'' and "secondary" structures for molecular ions- primary refers to the connectivity of the atoms within the different building blocks of a biological nJoleculc, and secoll(]nry refers to the position of the difrerent components in a complex in the genen11 3D form, rc::;pcctively.

The development of clcctro::;pray ionization (ESI)² and matrix-assisted laser des- orption/ionization (MALDI)³ lm~ led to an extensive invesLigation of biomokcular io11s. Noncovalently bound biological complexes such as protein-receptor, cn:t,ylll<'- snbstratc and multiprotcin complexes can be transferred from <1qncous solutions to the gas phase by ES1.⁴ Noncovalent interactions arc generally much weaker than co- valent bonds and control the biochemical function of biopolymcrs by providing them more flc~xibility to vary their structure. The study of these systems i11 the gas phas<' is a promising experimental approach to investigate directly key intl'nnolcclllnr iutcr- actions in the absence of solvent perturbation. This is in coutrast to the COll(lcllscd

phase where solvent cffect.s can afl'ect. the int.enwtions within a system. Fmtll('rmorc, the interactions of the ion with solvent in the gas-phase provide informa.tioll about the preferred coordination of ions by solvent molecules and their conformational changes of geometry upon t>olvation.⁵ ln contrast to the solution phase, numemns conformers and isomert> can be present in the gas phase.⁶ These solvent. effects can also lw studied through mass spectrometric (MS)-based methods by adding solvent n10lcculcs in H se- lective and stepwise manner (sec Chapter 5). The ultimate purpose and challenge for studying the physical properties of sequentially solvated ions is to extrapolate to or contrHst these properties with those in the solution phase, especially for !Jiologic<dly relevant molecular systems.

Different mass spectrometric-based methods have been d<'veloped for selectively cleaving linkages in biopolymer::; for sequence information a:-; well as for t>Lructural and for thermodynamic determinations. These methods arc based 011 collisional <utd photochemical activation methocb, as well as the employment of neutral rcH.gents to dlcct dwmical transformations in low-energy collisions?

This dissc~rtation is focussed 011 t.hwc go~ds. The first is clucidnJ.iott of tltc~ stnw- tures of small biologically relevant ions such as proton-bound aliphatic amino a.cicl complexes and metal ion-bound/proton-bound D A 11uclcic acid bat>e contplt'xcs in the ga.-; phase using infrared multiple photon dissociation spectroscopy (lR.t--lPD) Hud density functional theory (DFT) calculations (Chapters 3 a.nc1 4). A second import aut goal (Chapter 5) is the development of a MS-based tcch11iquc to produce solvated g<ls-

3

phase ions. Finally, in Chapter G and 7 this method of solvating ions was successfully applied to produce singly solvated metal ion-bound a.clenine aml [Zn (Thyminc-11)]¹ systems. lRMPD spectroscopy was used to study the structure of these solvated io11::;.

The conclu::;ions and propo::;als for future work will be summarized in Chapter . ln !.he followiug sect ious of !.his chapter, different. MS-ba.'ied t.eclllliqttes for s\.ntc- tnra.l st ndics of gaseous ions including collision-induced dissociation ( ClD), tlwnno- chcmical methods, and spectroscopic methods will be introclucccl. lRl\lPD spec- troscopy is the technique used extensively in this resea1·ch. Experimc1tt.al all(! tltco- rctica.l details of IRMPD will be discussed in more detail in Chapter 2 as well as t.hc computational methods used to assist in analy~ing the experimental cia ta.

1.2 Mass Spectrometric-based Techniques for Struc- tural Studie s of Gaseous Ions

1.2.1 Collision-Induced/ Activated Dissociation (CID /CAD) Methods

Collision-induced dissociation (CID) is a common method nsccl to obtain stmc- tnral connectivity information by dissociation of the ana.lyte and investigating t.hc resulting ionic fragments.⁸-10 In general, mass-selected ions, which arc accelerated,

undergo colli ·ions with a noble gas (typically a.rgon) at low pressure. Jn this method, each ion collides with an inert collision gas atom and, in the ca~c of very low prcssmcs, the collision happens only once, i.e. single-collision conditions. During a collisioll, a portion of the ion's kinetic energy can be converted into internal cuergy which cause's dissociation. According to the law of conservation of momentum, the kill(•Lic energy of a rapidly moving particle (analytc ions) colliding wit!J a static target (inert col- lision ga~) cannot be entirely converted into intcmal energy. The nwxillltllll ki11ct.ic energy converted to internal energy of the ion clnring a collision with a backgrouud gas molecules is given by Equation 1.1,

( l.l)

where E('()w is the maximum energy fraction converted into internal cucrgy (center r f mHss collision energy), Ela.b is the laboratory-frame kinetic energy, My and ~II' arc t.hc ma8s of the neutral collision ga~ and the parent ion respectively. ¹¹

ClD is highly dependent on the conditions of the experiment such H~ iustniJucut configuration, the uaturc of the inert collision gas and til' parent iou, the ion ki11d ic energy, and the number of collisions occurring during the activation period. ¹² In prac- ticc, two commonly used regiments for CID of gas-phase ions arc based on the ion kinetic energy. These two regiments arc low-energy ClD with a collision ell('rgy of up to 100 cV (1-100 cV) and high-energy CID in the range of several tliollsancl <•lcct.rou volts (kcV) of collision energy.

1. 2 .1.1 Low-energy CID

Low-energy CID is typically conducted in quadrupole/multipole collision cells, ion tra.ps, or ion cyclotron resonance (lCR) rnal::>s spectrometers. The collision energy i 11

these instruments is low (under 100 c V) in order to 1 c able to trap and detect the lllass-sclccted ions inside these traps. Due to the low collision energy, tlw exci ta- tiou usually occurs in the vibrational levels and leads to the weakest bond cleavctge pathways. For instance, in the low-energy ClD of bradykinin, l:J side-chaiJt cbwagc peaks arc absent but they were observed a.t the keV dissociation ClD.¹~ One of the adv;wtages of low-energy CID is its high collision yields. i.e. the llUlllhcr of effective collisions leading to dissociation of an ion. This is partly because the selected collision gas pressure allows multiple collisions (tens to hundreds) to occur.

ClD can be performed inside the lCR cell by introducing the collision gas into the cell through a pubc valve and at a pressure on the order of

w-s

^mbar. Trmlition- aly, fragmentation of ions in the lCR cell was achieved by exciting the ions with a short radio frequency (RF) pulse ( -<500 J.Ls) in which the frequency of the l\F pulse is on-resonance with the cyclotron frequency of ions(scc Chapter 2). Ions a.rc a.ccder- a.tccl to the desired kinetic energy and activated by collisions with ncu trHI gas. Si uce ions collide with the ncu tral target gas, they lose kinetic energy. Therefore, 1111 tl t i pic- mllision ::t.ct.ivaJ.ion with on-n•soHFtncc~ cxcit.ationis relatively inefficient. Seveml ~tltc·r-

native rnclhodR of cxcit.at.ion such ^ElS Rust.aiuecl off-resomwce irradiation collisionally iuclucccl dissociation (SORl- ClD)have been developed.¹⁵,JG ln this techniqttc, the RF pulse is ofi:-resonmH.:c. slightly above or below, with the r<.'sona11t. frequency of t.lH' pn'- CHrsor ion, causing ion kinetic energy to oscillate with time. Fragments arc produced by increasing the kinetic energy of ions at the same time as the collision gas is pulsed into the lCR cell. Frciscr et al.¹⁷ first conducted CID of small molcC'ules in a Fouri('r tra.nsform ion cyclotron resonance (FT-ICR) spectrometer. A valua.blc review on t.ltc ClD in FT-ICR mass spectrometry ha.':l been published by Lifshitzi H

ClD has been widely used for structural determination allCI to distiuguish I >ctwccn isomers . .!:<or example, two isomers of intertwined rotaxanc and the JJOn-iutcrtwincd (axle-wheel) complex shown in Figure 1.1, arc distinguished using ClD by tll('ir dif- ferent fragmentation pattcmsHl,²⁰ In the ClD spectrum of the complex, the nwjor fragmentation pathway is loss of the axle which is not observed i11 the roh·lxanc ClD spectrum. This can be explained based on their topologies. In rotaxane, the axle is mechanically bonded to the wheel and cleavage of this covalent bond is a !Jig!J energy pathway. llowcvcr, in the complex, the axle is connected to the wlH'el througlt four hydrogen bonds which can be easily cleaved. Therefore, the inta ·t axle cmt be r<'-

lccu.;ed and the evidence is provided by the CID spectrum. These types of cxpcrimcuts arc highly interesting from both a fundamental point of view and thH.t of synthetic chemists dealing with these species.

ClD is a fast and reliable method for structure elucidation in nmlti-cornponcnt

7

N'V"'N"'"'- • OH N..._..N

H

~0 ON~H HN~O

^{H ; t g}

H H 0

' NH HN •

0 0

mil 243

mil 518 mil 505

(rotaxane-H)- mll 1767

a

300 600 900 1200 1500 1800 mlz

(axle-H)- mil 805

[complex H)- mil 1767

b

300 600 900 1200 1500 1800 miZ

Figure 1.1: CID spectra of a) rotaxanc all() b) axle-wheel ('omplcx when col- tided with argon. ote that no signal is observed for the deprotom1lcd axle in tlw rotaxane spectrum while it is prominent among the products generated from the axlc-whe I complex. Figure rcprodnccd from Int. J. lass pcclrolll. 2004, 232, 249 with permission from Elsevier.

mixtures wi t.hou t. an extcm;ivc t;ample cleanup or scpara.tioll process. Tltc CJ D 1 )('- havior or tctra(aryl)lwnzidinc derivatives is intercstiug because they ('011( radict the

''even electron rule·' ^{21 23} Ba~ccl on this rule, lD of closecl-shcll ev<'u-clectrou (EE)

00

0 :!SO

'-l?

mlz 699.4

No o >_; A

o~ o

600.3

I

300 350 400 450 500 sso bOO

699.4

854 5

6SO 700 750 8CO 850 90 m/z

'f!

^M•·

l

mlz

884.5 ~\

'\

tf1

0 QE+·

I -30u ^{I '}

m/z854.5 s

~

(, _{,. m>5}

^-CH20

^z ~"") ^{< 5-}

N

0 ⁽⁾ 0

r I'

l /;'

I.

Figure 1.2: ESI-MS² of a tetra(aryl)bcnzidinc derivative along wiLh frHg;JneJitation reactions of the prccmsor iou [ I]+• at m/z 4.5. Figure reproduced fro1n Eur .. J.

Org. 'hem. 2007, 5162 with permission from John Wiley and OilS Inc.

9

catious preferentially produce dosed-shell product ions and llcHtral frngments. These compounds arc easily oxidizable with E1;2(o.r) values di tinctively below 1 \ (vs. kr- rocene). However, in Schafer ct al.²¹ lD experiments, transient odd-electron reaction intermediates were detected which were explained based on the stroup, <lhility of com- poHnds wi til low oxidation potentials to sta bili~c the correspouding open-shell ca I ions.

The ESl-l\1 j 1 ( 1 ²) spectrum of the radical precursor of one of til<' derivatives is shown in Figure 1.2. ln this Figure, OE repre ·cnts the odd-elect ron-m1mlwr internlt'- cliatcs \\ hik EE reprc1-;ents ions with ('\ <'11 numbers of electrons.

1.2.1.2 High-energy CID

ln high-cncrg_ lD, the precursor ion kinetic cncrgie1; arc on the orcin of s<'vcnd t housaud ckd rou volts (lee V). Only sector instruments, time-of-fligh I . or hybrid in- strnmcnts are practical for . ·uch high-energy stnclies. In this tccllll iq II<', the ion · <m' directed through a higher pressure rcgiou of the mass spcctromC'Ier ndkd the collision cell. .!\ sclwmahc of a typical three sect.or in1-;trument of BEE geonwt ry in which B and E refer to magnetic and electrostatic 1-;ectors respectively, is .·howu in t igure L.3.

The desired ion ('C\n be sepam Led frmn all other ions by !.he magnetic field ( 13) and directed int) the collision cells which am locat<'d iu the fielcl-fr<'<' rt'gions to pcrfonn ClD. The analysis of all ionic dissociation products of the <'Xpcrinwnt is p<'rf'ormed in the <'l<'ctrostatic sector.

Magnet (B)

F1eld-free reg1on (1)

Figure 1.3:

F1eld-free reg1on (2)

<(- - - --- - - ----+

Electrostatic Analyzer (E) I

F1eld-free reg1on (3)

chematic of a typical three sector instrulllent of BEE gcotnctry.

Wang and llolme~²⁵ u~cd CID with an ion accelerating voltage of 8kcV and be- lium colli~ion ga~ to invc~tigatc an unexpected composite metastable ion (l\IJ) peHk, rcsultiug from the isobaric lo~ses of 0 and C2IL, ( r+·-2 ) from 2,3-pcntanediOIH' radical cation. lctal-\lablc ions arc formed with sufficient excitation to dissociate spont.ar1cously and without collision during its flight from the ion source to the de- tcctor. ln the l\111 time frame (10-⁴ to 10-o s), a NkLaffcrty rcarrangcment²G-2

' (sec Scheme 1.1) leads to an cthcnc loss in the 2 3-pcntancdionc raclical cation. However.

the presence of a composite peak at ( r+•-2 ) shows that there must cdso be another fragmentation channel (Fig. 1.4). Since this peak consists of a large kinetic en erg.\'

11

0

0

Scheme 1.1

+·

[CH3C(OH)=C=O] + C2H4

rd<'n.se· (T<t~R) (til<' broad ba.:-;e•) :-i<'C'tion and ft Slllall l.(ER s<'ction (at til<' ce·nte·r) .. the·

possibility of two competiug dis:o ·iRtion pathways was iuvcstigat.t'd hy ide·ntifying til<' :·d.mct.m<' of <'Hch cornpone·nt of the• e on1posit.<' pcftk using a n1odific•d VC:-ZJ\

n

tandc•m mAss spectrometer (Fig. 1.3).²u l'vlett~stablc iou t~ncl ClD lllFISS spectra of th<' ionizC'd dikctonc· mokculc•s we•rc• p<'rformc·d in th<' s<'cond fie•Jd-fr<'<' r<'gion (FFn) of the ^111r ss sp<'drouwter. sing the Pkdrostatic analyzer. c•ach cmnponc•nt of the•

peak wa:-; tmusmit.tcJ into the third FFH where aCID :-;pcdnuu of the frnguwut ious gen<'ratcd iu tht' second FFR was rec rded. In the CID spcdmu1 of th<' hugt' KEH C"Olll])Ollent. (Fig. 1.51>). t.wo major prodnct:-; wert' ob:-;crved at 111/z 13 awl G7 which is similar to that of tlH' ionized ~-lmtanone (Fig. l.5c). This similarity in tlH' fraguJc•nt nt iou spec! rum proves that the largt' KER compoucut nrisl's from t hl' los:-;

of CO from t1 tTRrraugcd, ionized 2.3-p('}}tau ·d.ione. The CID spedru111 of the SJJJall KEH con1poucut (Fig 1.5a) shows auot.her major fragment. nt. 111/z IH which is nhscnt in the CID sp<'ctrum of the 2-butauou<' iou. This fragmeut.ation wFls explain<'d ht~sc•d ou t hc• lc Laffer( y rearrangement. The carbouy I group at at om 2 of 2,3- pe•n t au<'dion<'

Broad base

Narrow base

57 H H..,

(

.1

41

Figure 1.4: l\Ictastablc iou (l\11) mas1; , p trum of ionized 2,3-pcnt a ned ionc. Fip,- ure reproduced from Int. J. l\Iass pectrom. 2006, 249-250, 222 with pcnuission frolll Eb vi r.

methylketcnC' ion with an m/z of 72 from the intcrmccliatC' shown in 'chC'mc 1.1.

13ascd on computational data, a small harrier for thC' clcavag<' of t hC' n-/1 bond in t h<' i ¹¹ t crmccl ia tC' ion which lead::; to t h<' forma Lion of cthcn<' is prcd ict cd. Therefor<', ba:>C'd ou the ID ::;pcctra, it was concludccl that the CO loss rC'sult.s in the larg<' KFR part while th(' small KER ::;lice ari::;es from the los::; of C211.1.

ClD can provide the conncctivit. of atoms in an ion, bnt. it is not obvious how

13

a

b

c

43

27 29

43

C H, C (O) OCH 3 +•

43

27 29

44

57

57

57

Figure 1.5: CID mass spcctra of: (a) the middle, (b) the edge couJpOIJC!lt. of the composite (M-28)+• peak aud (c) ionized 2-buta.nOIJC. Figure reproduced from Jut..

one would obtain higher order ~tructure information directly from CID clata.. One of

the main limitations of ClD is that the energy cannot be completely transferred to the ion. This limits the degree of fragmentation. As well, for large ion~ with b-trgc numbers of degree of freedom, the energy is distributed over the higher ntunbcr of bonds which slows fragmentation rates. ClD of proteins provides information a bon t the primary structures to ~omc extent but it docs uot necessarily contain iufornwt.ion

about 2° and 3o structures. In practice, the ClD fragmentation technique ~ccms to be insufficient iu protcmuiu;ao The main rcellion is that this technique docs not m'ccssar- ily lead to the cleavage of all required bonds to provide full information on the primctry

~tr11cturc of a polypcpticlc.'l¹ For instam:e, a variety of factors ~uclt ;.:ts tltc difFcrelln's in the proton affinities of backbmle amides, stcric cfrccts, aud iutcractiom; bct.wc<'ll ccrtaiu amino acid side chains and the backbone amide cau~e::; a. large prdcn'ucc for cleavage of certain sites while cleavages at other sites arc supprcsscrl.³² Evc11 in cHscs

wlH'rc all bonds an~ clc~wc~cl, id<'Jtt.ifyiug the ~C(jtWltCc~ aud it.s tuodificntions is diffk11lt ba.scd on information obtained on the fragment ma.sscs.³:.1 An overlap of the m;.-t.~scs of N-tcrminal and C-termina.l fragments in sequencing of polypeptides cau ca.11sc <Hl l-

biguity in ma.ss spectral interpretation and peak assignment. Furthermore, pos~iblc

isomerization and/ or rearrangement of molecular ions during ClD is another problclll that one ha.s to take into consideration when cletermiuing ion stmcturc.¹⁰ One cxmit- plc that CID experiments could not address properly is the keto-enol i~omcrizntio11

of methyl acetate ion. Based on the early cxpcrimeuts, both keto Cll;3C(O)OCll;/•

15

A A'

Scheme 1.2

ami enol Clb=C(OH)OCI13 +• isomers fragment by loss of a [C,Ii:!,OJ- (metltoxy radical) exclusively to produce Cll₃CO+. The kinetic energy release for these dis- sociations were found to be identical for the two isomers, suggesting tlwt tlJC two isomers cnn int.crchnugc wi t.hont. n signific1wt. c~ncrgy ha.rricra⁴ A 1,3-hydrogcJJ shi fL was then believed to take place in the tautomerization. However, fmther mct.lwxy deuterium-labeled (CI1₃COOCD:₃₎ experimentf:l revealed Il/D f:lcrambling since loss of [C,D3,0J• waf:l not the only fragmentation channel. Instead ct 1,11-!Jyclrogeu shift was proposed in which an intermediate ref:lultf:l in enol-if:lomer forlllation, followed by an 11-atom abstraction from the methyl group of the enol-intcnnecliatc (A) as s!Jowu ill Scheme 1.2.:{⁴•35 WHh thif:l mechani ·m the ll/D scrambling was explained. The case was not closed as more invef:ltigations through collif:lionally indu ·eel dis::;ociative ion- ization (ClDl) showed loss of •ClbOll radicals in addition to CI-1₃0• occnrs clming the ionization of methyl acctatcY^{6- 39} In CIDI method, the structure of neutral procl- ucts from the unimolecular dissociative ionizationf:l of mw:;s selected ious arc st udiccl by mea.nf:l of the collisiona.lly induced dissociative ionization of the 11eutral spcc·ics t!Jcmsclvef:l. The neutral f:lJ)CCies, with kilovolt translational energies, e11Lcr <-1 posi-

B

Scheme 1.3

t.i vcly charged collision cell local ed iu Uw second held free region of a stall< lard ZAB mass spectrometer. Dissociative ionization of the neutrals results therein froll1 t!Jeir collisions with lle target ga~. The resulting ions arc analysed by means of the electric sector and the relative ion abundances arc shown to be structure characteristic. Durg- en; ct al.⁴⁰ proposed another intermediate, shown in Scheme 1.3, to be involved in

· cn

20H production. The problem with this new intermediate was that. it is diff-icult

to produce this distonic ion independently. Therefore, the ion chemistry experiments conlcl not comprehensively clarify the tautomerization mechanism. Fimdly, compu!a- tional chemistry techniques were used to fully understand the mcchanislll and these will be explained in section 1.3.⁴¹,~²

17

1.2.2 Thermochemical M e thods

1.2.2.1 Threshold Collision Induced Dissociation (TCID)

Threshold collisiou induced dissociation (TClD) is a modified ClD t.cdllliquc. A schematic overview of a guided ion beam tandem mass spectrometer used for TClD

experiments is shown in Figure 1.6.⁴³ Sample vapours arc introduced iut.o tile ion source at pressures between 0.5-0.8 Torr of hclitun and arc iouizccl by ckct.ron iou- ization. Ions arc mass selected by a magnetic sector and focussed into the octopolc ion guide region. CID occurs with an inert gas (usually Xe bcc<wsc of its large l\ly in Eq. 1.1) in a collision cell at prcssmcs ra.nging from 0.05 to 0.2 mTon surrounded

ION SOURCE

REACTANT MASS SELECTION

FS2

10 c ..

INTERACTION REGION

PRODUCT MASS ION ANALYSIS DETECTOR

~ PMT

Figure 1.6: Schematic overview of the guided ion beam tandeut mass spcc:t.rotttc- ter. FS represents focusing stages. Figure reproduced from J. Chcm. Phys. 1985, 83, 166 with permission from Americau Institute of Physics.